Follow TV Tropes

Following

Troubled Production / Video Games (N to Z)

Go To


    open/close all folders 

    N-P 
  • Neverwinter Nights 2's premium module Mysteries of Westgate didn't take long to develop. However, Atari wanted the DLC shipped with a DRM scheme. This delayed the release of the game. Adding to this, Obsidian patches constantly broke down the Adventure Pack, pushing the game's release even further, much to gamers' frustration.
  • After it was funded via Kickstarter, OMORI quickly fell into such a production, lasting 6 whole years; enough time for the proposed Nintendo 3DS version to be scrapped in favor of a port to the Nintendo Switch. "Highlights" included a port from RPG Maker VX to RPG Maker MV, which resulted in such catastrophic plugin compatibility issues that a programmer from Enterbrain (the developers of RPG Maker) had to step in to help the team fix the issues, and coupled with outright radio silence on the side of the developers, the prospects appeared grim. Thankfully, OMORI turned out to be worth the wait, as once it was released on Christmas Day 2020, it earned rave reviews from both backers and players alike.
  • According to several Starbreeze and Overkill developers, Overkill's The Walking Dead had massive troubles behind the scenes.
    • Development had begun on the Valhalla game engine Starbreeze had purchased in 2015, departing from the aging Diesel game engine previously used in PAYDAY: The Heist and PAYDAY 2. Unfortunately, Valhalla was in a very early stage of development and in no shape to be used, lacking even a function to open files. Those who worked with Valhalla had nothing good to say about the engine, calling it "a piece of shit", "unworkable" and "an insane waste of money". Developers begged management to move the game to the proven Unreal game engine, which was finally granted after two years of development on Valhalla. The game was pushed back to 2018, but that only left the team a single year to build the game, many of whom had no prior experience working with Unreal.
    • Mismanagement plagued the game and the company at large. The release of RAID: World War II while Payday 2 was still receiving regular content updates, left many at Starbreeze confused as to why they were competing against their own game, which helped RAID drop to a mere 40 concurrent players within three months of release. Starbreeze was also spending money on ill-advised projects, with tens of millions dumped into failed Virtual Reality projects and on promotional crossovers with various Hollywood films.
    • On the game itself, producers would make wild demands for features based on the latest games they had seen, such as Far Cry and Dying Light, only to change their mind within weeks. The showcase at E3 2018 had management assuring the team that reactions to the game were glowingly positive, even as team members saw news and videos showing that actual impressions had been lukewarm at best. The team was forced to work up to 100 hours a week during the last months of development to get the game done by the deadline, severely affecting the mental and physical well-being of many working on the game. By the end of development, the team had been thoroughly disillusioned about the quality of the game.
    • When the game released in November 2018, it staggered out as an Obvious Beta that was panned for being riddled with glitches and various online issues, and others criticized the game's pacing and gunplay. Expected to sell millions of copies, it fell disastrously short of Starbreeze's expectations and Starbreeze's stock price went into freefall. The company would continue to be rocked by bad news throughout the year as the beginning of December saw Starbreeze CEO Bo Andersson leave his post as the company announced it was filing for reconstruction. Two days later, Starbreeze would have their offices raided by police, who arrested at least one person and confiscated multiple computers on suspicion of insider trading. In February 2019, Starbreeze had the license to The Walking Dead ripped away, removing the game from digital storefronts and cancelling planned releases on the Playstation 4 and Xbox One. Starbreeze came dangerously close to shutting its doors entirely, and would only regain stability after a year of restructuring, selling off most of its assets, and resuming active development for PAYDAY 2.
  • The production and eventual success of Overwatch was well-documented as being the result of the production and stillbirth of a previous game known as Titan. In 2007, Blizzard Entertainment had begun production on Titan, a second MMO following in the monolithic wake of World of Warcraft that would also contain many class-based first-person shooter elements. After many years of production, Titan was officially cancelled in 2013 as a result of multiple factors, from the lack of focus and control of scope resulting in countless balance problems, to the MMO market becoming increasingly saturated as development chugged along, resulting in what was speculated to have been a $50 million loss. As Blizzard developers began restructuring to new projects from Titan's failure, Overwatch was formed as a massive retool that recycled many ideas from Titan (Series Mascot Tracer was reportedly based off a "Jumper" class), with Blizzard's clearer vision and tighter control on scope allowing the game to be announced in 2014 and officially launched in 2016.
    • The quasi-sequel/expansion, Overwatch 2, was reported to have had a rougher development time than expected following its announcement in 2019. Shortly after it was announced Activision Blizzard was being bought by Microsoft in early 2022, Overwatch producer Tracy Kennedy accused Activision-Blizzard CEO Bobby Kotick of serious pipeline mismanagement; for the past few years, it had become increasingly common for the Overwatch team to be forced to halt progress on 2 in order to work on other large side projects at Kotick's request, only for them to suddenly be axed, wasting months of their development time. Kennedy has also suggested that the team had experienced a major turnover rate, including the losses of director Jeff Kaplan and executive producer Chacko Sonny, as a direct result of conflicts with Kotick and other "corporate BS" (the massively-publicized sexual harassment scandals from mid-2021 and the subsequent incendiary backlash against Activision-Blizzard also really didn't help with developer morale). 2 would eventually find stable enough ground for a release date in October 2022 and ongoing roadmaps, but only for its PvP — by Blizzard's admission, they required more time to properly complete the major PvE expansions.
  • Owlboy is notorious for its lengthy development, beginning in 2007 and releasing in November 2016 nearly a decade later, with planning going back to at least 2005, before the release of the Nintendo Wii. The director, Simon Stafsnes Andersen, has suffered from depression since childhood, and the team were so worried about expectations from fans that they cut and restarted the entire game several times. The game became acclaimed for its gorgeous pixel art, story, and characters, but a Broken Base emerged around its controls; some have no problem with the controls while others think the controls are awkward.
  • The production of the third installment in the Pac-Man World series was pretty chaotic. Namco commissioned Blitz Games to develop the title after a previous game, Pac-Man Adventures, was cancelled, but financial problems kept causing hurdles to the point Namco wanted to cancel the game entirely. They eventually allowed the game to continue, but Blitz Games had to pay for it out of their own pockets and had to take the loss. On top of that, they still had a tight deadline to finish the game in order to tie it to Pac-Man's 25th anniversary. The result was met with mixed reviews, and a rather divisive reception among fans. That, combined with poor sales, led to the game becoming a Franchise Killer for Pac-Man Adventure-platform games outside of a duology of tie-in games for Pac-Man and the Ghostly Adventures.
  • Palworld was an Affectionate Parody of Pokémon fueled by Refuge in Audacity that became a smash hit literally overnight, selling a million copies in eight hours and eventually six million in four days. When people questioned the dev company Pocketpair's founder & CEO Takuro Mizobe the secret to their success, in his own words it was nothing short of a miracle that the game even came to exist as it was "the antithesis of proper game development".
    • The game cost almost seven million dollars to develop, with no budget planning whatsoever. Takuro preferred the Ignorance Is Bliss approach and actively avoided looking at the development costs, while being completely willing to risk bankruptcy. When he finally saw near the release date that they were running low on funds but morale around the office was high, he doubled down and hired even more people to help finish it in time. A week before the game's Early Access release they realized that they were finally out of money.
    • Many praised the game's high level of polish by Early Access standards, which made it all the more shocking that the entire team was a Ragtag Bunch of Misfits with basically no idea what they were doing. Their code had no naming conventions or version control, relying on "buckets of flash drives" to keep it all straight. The models had no rigs, and no one knew how to CG model or animate as their previous games used bought assets. The lead engineer didn't know how to use git, the lead artist was a new graduate who was rejected by many other companies for having No Social Skills, and the lead weapon animator was a dropout who was completely self-taught through YouTube tutorials and playing FPS games. All their network infrastructure was designed by one 24-year-old, and many new recruits were Game Modders Takuro messaged through Twitter.
    • They were rejected by every publisher they sent it to, and had to self-publish. The game was initially made in Unity, but halfway through development the decision was made to scrap the whole thing and remake it in Unreal Engine. At one point when they were low on manpower, Takuro pulled an Iwata and started optimizing memory bottlenecks himself despite having never touched Unreal before.
  • Paranoia: Happiness Is Mandatory initially released in 2019 before being quickly delisted for sale with no explanation. The creators of the original Tabletop Game Paranoia, Costikyan and Goldberg, later explained that they filed legal action against the publishers for trying to release an unfinished product, until the game suddenly reappeared in late December 2023. Upon its release, gamers quickly became aware that no progress on it was made whatsoever throughout the four years it was in limbo and the game remained a jumbled mess.
  • Parasol Stars, a spin-off from the Bubble Bobble series, was ported to several systems in the early nineties. However, the Commodore 64 version was cancelled because of a relatively minor Troubled Production. The port was assigned to a programmer whose marriage was falling apart. One night, the programmer got into a heated argument with his wife, so she got on his computer and erased all his work on the port. You're probably thinking "But why didn't he make backups?" Actually, he did, but his wife found those and destroyed them as well. At first, Ocean Software tried to cover for the programmer by claiming his computer had been stolen, but later came clean.
  • Perfect Dark was one of the most problematic productions for Rare.
    • Wanting to make the game a Spiritual Successor to their previous hit GoldenEye (1997), Rare quickly encountered troubles as their ambitions ran into the limitations of the hardware of the Nintendo 64, which eventually necessitated the usage of the Expansion Pack and shipping the game on a 32 megabyte cartridge. Fourteen months in to the nearly three-year-long development cycle, with the development team growing frustrated at the lack of progress, workplace atmosphere, and their increasingly stressed relationship with Nintendo, lead designer Martin Hollis left the company and was followed by a number of GoldenEye veterans to form Free Radical Design.
    • Despite this setback, the remaining team was able to continue on, bolstered by members from other Rare teams joining the project. Late in development, real-world events also impacted the game, as the Columbine High School massacre drew unwanted attention to the family-friendly image of Nintendo and the violent, M-rated shooter that was to be released as a heavily-hyped Nintendo 64 exclusive. This forced Rare to cut some minor content from the game, most notably a feature that would have let players upload pictures of people's faces from the Game Boy Camera to put on enemies in the game, which Rare quickly realized would lead to no shortage of salacious headlines about teenagers giving enemies the faces of their classmates, teachers, and parents in order to kill them in-game.
    • The game was released in 2000 to critical acclaim, but only sold roughly one quarter as many copies as GoldenEye due to being one of the last games released on the Nintendo 64. Moreover, one of the most common criticisms concerned the game's sluggish framerate, a sign that Rare had pushed the Nintendo 64 to its limits. Tellingly, when the game received an HD remaster for the Xbox 360, it was hailed as a Polished Port that fixed the original's biggest problem.
  • The prequel, Perfect Dark Zero, managed to be even more troubled, with a development cycle spanning five years and three platforms.
    • With initial development starting on the Nintendo GameCube, the increasingly fractured relationship between Rare and Nintendo along with the stresses of making Star Fox Adventures kept Rare from making much progress. With Rare bought out by Microsoft and made a first-party developer for the Xbox in 2002, development on Zero picked up, now on the Xbox platform. Despite plans for an Animesque art style being scuttled after fan backlash proved too intense, development on the Xbox was fairly smooth until 2004.
    • With the Xbox 360 launching next year and the expected Killer App titles like Halo 3 and Gears of War unable to make the launch window, Microsoft pivoted Zero to be an Xbox 360 launch title, giving Rare a year to rework the game to be a showcase for the new hardware. This required brutal crunch development that forced several features to be cut from the game, and development was so rushed that Microsoft waived the final certification requirements for Zero, a move that could have been disastrous had a Game-Breaking Bug been discovered in the final build. The game was released along with the Xbox 360 in November 2005 to good reviews and profitable sales, although the later runaway successes of Halo 3 and Gears of War painted Zero in a more negative light in retrospect.
  • Given the history of its predecessors, production on the reboot of Perfect Dark being just as troubled was only appropriate. This article by Rebekah Valentine for IGN goes into more detail on a game that, several years after its announcement, still remains far from the light of day with little information released.
    • In 2018, Microsoft Studios (now Xbox Game Studios) sought to beef up their lineup of exclusive first-party games in the face of stiffening competition from Sony Interactive Entertainment, whose acclaimed titles for the PlayStation 4 played a key role in that console decisively beating the Xbox One. As such, not only did they purchase a number of third-party developers, they created a new in-house developer called The Initiative, founded by veteran employees of Crystal Dynamics, Naughty Dog, Rockstar Games, and Sony's Santa Monica Studio, among others. Studio head Darrell Gallagher and game director Dan Neuburger (both formerly of Crystal Dynamics) decided that their first game at The Initiative would be a reboot of Perfect Dark, a famed IP that Microsoft owned which had laid fallow for over a decade, and they were not shy about their ambitions with the project. In a development video released alongside the first teaser trailer at The Game Awards in 2020, they said that they wanted to make the ultimate Spy Fiction game, set in a False Utopia where a Post-Cyberpunk veneer masks an underworld of Corporate Warfare with action gameplay that would feel more like playing as a secret agent than a soldier like in other shooters. The term "AAAA" (in reference to big-budget "AAA" game development) and comparisons to Game of Thrones and Westworld were thrown around to describe just how big the game would be.
    • Preproduction, where ideas are thrown around as to what kind of game The Initiative would actually be making, started optimistically. The problem was that this process, which normally takes about a year for games of this size, kept going into the second year of development... and the third. Ideas were constantly pitched, scrapped, and changed with little to show for them beyond gameplay prototypes and builds that never went anywhere, and employees at both The Initiative and Certain Affinity, their partner on development, grew frustrated by Gallagher and Neuburger's refusal to commit to a firm vision for the game. What's more, the two studios themselves also saw clashes between them on the game's direction, with The Initiative seeing Certain Affinity as a creative partner to bounce ideas off of but Certain Affinity seeing itself as a studio that made games to specific instructions. The COVID-19 Pandemic only made matters worse as both studios went remote. One former Initiative employee compared communications within the project to a game of telephone.
    • In early 2021, design director Drew Murray left The Initiative to return to Insomniac Games, the start of a flood of talent leaving the studio, with some estimates claiming that half their employees quit. More importantly, Certain Affinity, growing increasingly frustrated working with The Initiative, also pulled out of development that year. Many former Initiative employees cited a frustration with the studio's culture and development's lack of progress, especially as the departures both sapped morale and left growing holes in the development team, with one estimating that development was stalled for nine months as a result of a lack of people in key departments. One female employee also felt that the studio was male-dominated to the point that ideas for the game that she felt were sexist were routinely discussed with little pushback.
    • September 2021 saw Gallagher and Neuburger's old studio Crystal Dynamics step in to replace Certain Affinity. While they were initially hired to merely support development, they quickly realized that The Initiative was in over its head and started taking a far greater creative role, especially after they saw the vertical slice of gameplay that The Initiative and Certain Affinity had put together before they came aboard, which they decided needed to be completely reworked. With that, in 2022 production restarted from scratch in Unreal Engine 5, this time with Crystal Dynamics increasingly taking a leadership role, much to the frustration of some veteran Initiative employees who felt that Crystal Dynamics was stepping on their turf. By the end of the year, it was generally agreed that the reboot of Perfect Dark was a Crystal Dynamics game in all but name, with half the studio working on it.
    • The fact that the game had been in development for four years and had so little to show for it left both studios' management deeply skeptical, leading to a push for deadlines and progress that led to significant stress on the team even if the project never faced serious crunch. While Crystal Dynamics hired many people to work on the game, development was still frequently understaffed. That said, their involvement saw actual progress made on the game, and while, as of this writing, it is still far away from completion, they actually have an end goal in sight.
  • Iwakura Productions' Fan Translation of the PSP remake of Persona 2: Eternal Punishment was in development for years, but development fell apart at the last hurdle due to several factors, eventually leading to it being canned after it was preempted by a similar patch that came out of nowhere in July 2022. Family deaths and the COVID-19 Pandemic ate into team members' free time, but that was surmountable; what was not surmountable was the Russian main programmer losing access to his programming tools as a result of the invasion of Ukraine, causing beta testing to stall. Team leader CJ Iwakura claims one of Sayucchin's team ghosted Iwakura Productions several weeks prior, but found nothing plagiarized from his project despite the inherent similarities in content and noted that the only difference script-wise was that this project used Japanese Honorifics, which official Atlus localizations use but Iwakura dislikes.
  • As shown in the final part of the "Making Of" documentary, The 8-Bit Guy's Planet X3, the MS-DOS sequel to his Commodore 64 strategy game Planet X2, went through trials and tribulations during the stressful 14 month development cycle.
    • The game was originally envisioned as an enhanced port of Planet X2 for DOS, with extra features that weren't in the C64 version due to RAM limitations. No longer being limited by computer RAM of DOS machines, a new limitation came in the form of disk space. Like with Planet X2, David wanted to fit the game onto a single disk, shipping it on both a 360K 5.25 inch disk and a 720K 3.5 inch disk. David had three graphics modes for the game in mind: CGA, Tandy and VGA. The maps were larger in file size than the C64 prequel and took up a huge chunk of space. Furthermore, a planned intro sequence was cut from the game due to the numerous graphics modes and the fact that it would not fit on a 360K disk. The intro sequence was uploaded as a standalone YouTube video. The large map file sizes also meant that certain maps from the 720K disk version had to be removed just so the game could fit on a single 360K disk, and the VGA graphics mode was cut altogether on the 360K disk version.
    • David had plenty of experience on the 6502 assembly language, but he had no knowledge on assembly language programming for x86 processors. It took him a full month before he got to know how to program assembly language on x86 processors and he found a good assembler to use for coding the game. When David was adding extra features, this made the game even more complex than the C64 prequel, but this came at a certain cost. As he kept adding more code, the game had difficulty running properly on IBM XT machines running at 4.77 MHz, David's primary goal to make the game run on all MS-DOS computers, so he had to leave off features he wanted to add to keep the game stable on old hardware. David and beta testers discovered several bugs, some of which David considered almost un-fixable as they seemed random and impossible to reproduce. At two points of development, David almost considered quitting and he even went almost a month without any coding.
    • Producing physical copies of the game wasn't an easy task, either. David decided to start a Kickstarter campaign for the game, earning over $110,000, most of which was spent on materials like box design and disks. Unlike Planet X2 boxes, which came pre-assembled, the Planet X3 boxes were flat, and so a lot of time was spent folding over 2000 boxes of the game, some of which were defective. Duplicating disks was also a challenge, due to the fact that the 720K disks had no write protect notches, forcing the developers to stick paper over the holes to write the disks. Writing 360K disks also required vintage hardware with 360K drives, as 360K disks that were written on 1.2 MB disk drives cannot be read on 360K drives. Even with over 2000 disks for each format produced, that only covered about 20% of the Kickstarter fulfillment, and David realized that he couldn't do all of those disks and ship the product on time.
    • When the month of February of 2019 came along, David started shipping the game to the backers who funded the game. He had estimated that it would take 2-3 weeks to complete the shipping process, but in reality, it took 3 months. David spent over $22,000 shipping the packages, and most of them came back as undeliverable (most of them being from the UK) because either people never responded or the post offices never informed them of the packages. Thus, he asked the backers to update their shipping addresses and pay for the shipping, as he can't afford to pay refunds for all of the packages that returned to him undelivered.
    • Even though David still has plenty of leftover Planet X3 materials, which take up a lot of storage space, the game was successfully shipped and received positive reviews, with plans for a shareware demo, opening the game's source code, and a possible Sega Genesis port.
  • PlayStation Home, a virtual world created by Sony for the PlayStation 3, effectively existed in a state of limbo for its entire lifespan. This article on Kotaku UK lays out the whole story of "Sony's most successful failure".
    • Home began life as an online mode for the PlayStation 2 game The Getaway: Black Monday, but soon expanded once Phil Harrison, vice president of Sony Europe, had a look at it and decided that it could be something far more. Harrison envisioned the project, then known internally as 'the Hub', as a 'space between games' that would function as a lobby system of sorts for a whole slew of games. While Harrison was able to secure funding for what was becoming his dream project, many of Sony's other executives didn't understand his vision for the Hub, a problem that would plague Home for its entire lifespan. The Japanese executives especially couldn't get their heads around it — to them, multiplayer gaming was a social activity where friends get together in the same room, the antithesis of the Hub's Western model of playing with strangers through online matchmaking. (A similar philosophy was visible in Nintendo's much-maligned "friend code" system.)
    • With the announcement of the PS3, what was now being called Home was soon positioned as a flagship title for the fledgling console. This led to a number of problematic design changes. The original animesque character style, for instance, was thrown out in favor of a more photo-realistic one designed to showcase the new console's power, but it soon turned out that having hundreds of such highly detailed, player-controlled characters in a virtual space (especially an online one) was incredibly taxing on the PS3's hardware; the number of avatars in any given space had to be capped at fifty. Furthermore, these more realistic models fell straight into the Uncanny Valley.
    • The greater world's introduction to Home came at E3 2007, in the form of an incredibly awkward presentation featuring the digital avatars of Sony executives Jack Tretton and Kaz Hirai. It was a poor first impression, and it added to the woes that the PS3 suffered early in its life cycle. Furthermore, Home missed its planned autumn 2007 release date, being pushed back into the following year.
    • Phil Harrison's departure from Sony at the beginning of 2008 produced a revolving door of producers, many of whom also didn't understand the concept of Home and whose expertise came in widely disparate fields. Oscar Clark, the man brought in to sort out the disorganized project, remarked that, in early 2008, there essentially wasn't a Home, the project having been filed with all manner of half-formed ideas.
    • When Home finally launched in December 2008 after more than a year of delays, it was an Obvious Beta. The massive amount of detail on the avatars and the worlds they inhabited produced outrageous load times of up to ten minutes, and there was a good chance players couldn't get into the areas they were trying to enter due to the strict, resource-mandated caps on the number of avatars in any given space. The team could do little to fix these problems, as they were small and underfunded and the bugs so numerous and deep-rooted; there's a reason the game never left its Perpetual Beta. There was little content to see either, and a number of features planned for later updates were scrapped. One such feature was the Hall of Fame, a personalized room where players could see physical, three-dimensional trophies corresponding to their achievements in various games; most studios, even Sony's second-party developers, saw designing dozens of unique trophies as a waste of time.
    • Developers' skepticism extended to Home in general. Few of them saw any point in creating virtual spaces and other content in Home for their games, especially given how difficult the process of creating such content was, and as a result, most of Home's actual content had little to do with video games. While a few companies like nDreams and Veemee did create a number of unique original games for Home, shopping for clothes, houses, yachts, and other items for players to customize their avatars with became the main activity. Home had gone from a gaming hub to a bizarro version of Second Life...
    • ...which actually allowed it to start turning a profit and develop a passionate fanbase. Commercially, it was a huge success in the long run, even if it was swiftly forgotten outside its cult following. In fact, it was precisely this financial success that caused Sony to wait until 2015, after the PS3 had become a Daddy System, to finally shut down its servers.
  • Pokémon:
    • Pokémon Red and Blue was a massive Sleeper Hit. So creating a successor in Pokémon Gold and Silver ended up being very complicated for Game Freak.
      • For starters, Gold and Silver were originally developed as Game Boy games, with Super Game Boy compatibility and a set release date of "late 1997." Enough of the game was completed for a demo at Spaceworld 1997, and from hacking the leaked ROM of the demo, one can observe how much of a mess things were in this state. Hardly any of the gym leaders were programmed into the game, the Kanto region had been heavily distilled into a single map, and many other features were not anywhere close to completion. This, combined with the impending release of the Game Boy Color in 1998, forced Game Freak to delay the games to June 1999, something they didn't announce until March 1998, three months after the original release year lapsed.
      • Despite this, Game Freak still faced massive difficulties with getting the game properly programmed, due to them having a lot of ideas that they didn't really know what to do with, and due to them having to reprogram much of the graphics to be compatible with the Game Boy Color in addition to the Super Game Boy, with both of these factors pushing the release date further back to November 1999. Even then, the game was in such a sloppy state that it seemed unlikely that it could be released at all, but eventually, Game Freak was finally able to release Gold and Silver on November 21, 1999 in Japan.
      • The sheer chaos of the games' development is highly unusual and a much-talked-about subject among analysts, as not only did it mark the only time Game Freak delayed a mainline Pokémon game, but it also marked one of the most dramatic revamps of a Pokémon game from the initial drafts to the final release. The delays in development are also notable in that they forced the staff behind Pokémon: The Series to improvise around the games' constantly-shifting release date, creating a Filler Arc with the Orange Islands to ensure that the show could still keep running during the roughly two-year gap between the intended and actual release dates.
    • Pokémon Sleep took four years of development before it was finally released in 2023 because it was difficult to playtest sleep tracking, its core mechanic.
  • Postal III was a nightmare in development according to developer Running With Scissors. As Matt McMuscles chronicled on an episode of Wha Happun?:
    • After the release of Postal 2, RWS lost their main publisher Whiptail Interactive, who bailed on them after the game was released. Seeking a new publisher for the next Postal game, RWS decided to make Russian publishing company Akella develop it instead through their in-house studio Trashmasters, since Akella had handled publishing for P2's Russian release, promising to still design and write the game themselves. Akella assigned their A Team to work on the game, planning on releasing the game on the Xbox 360 and PS3 alongside the PC port with a 2008 launch date. So far so good...
    • ...until the 2008 financial crisis happened. Russia was hit hard by the crisis, especially Akella. So to cut costs, Akella fired their A Team. Then their B Team. Next their C Team. And finally their D Team. All of whom never received payment for their prior work on the project. Akella also had to make downgrades to the core game in order to stay in business, like replacing the promised open world with a more linear level based structure. RWS suggested cancelling the game to put it out of its misery, but because Akella had sunk so much money into the game, they had no choice but to complete whatever they could and ship it after numerous delays. The console versions were cancelled as well, and promised patches were never released.
    • The game finally saw release in 2011, where it received incredibly poor reviews from not only critics, but fans as well, who balked at the game for its weak graphics, numerous glitches, poor controls, and its obnoxious sense of humor. RWS would completely disown the game on their website, to the point where it was retconned in the Postal 2 Paradise Lost Expansion as a coma-induced fever dream by the Postal Dude. The game would also serve as the killing blow to Akella, who ultimately went defunct a year after the game's launch, no thanks to the game's abysmal sales, plus a massive lawsuit filed by the fired staff for unpaid royalties. RWS would eventually work on their own in-house Postal sequel, Postal 4: No Regerts, without any third party assistance. As for outsourcing, they wouldn't attempt that until 2021, when they contacted Hyperstrange to make Postal: Brain Damaged as a spin-off as opposed to a mainline title.
  • Another 3D Realms game with development issues was Prey. The first attempt (1995-1996) of the developers failed because the creative director left the company to join his friend's new business and the rest of the team also did not stay in the project, either. The second attempt (1996-1998) failed due to technological problems the team could not solve. At the third time (circa 1999), there was only one person working on the game, the tech programmer. This attempt, too, was unsuccessful because of technology-related issues. The project was eventually transferred in 2001 to Human Head Studios, which then developed the game. It was finally released in 2006.
  • The sequel Prey 2 suffered from a similarly chaotic production to the original game; unlike the first game, it only made it to alpha before Bethesda unceremoniously canceled it in 2014.
    • Originally the game was going to be a direct sequel announced in 2006 and made by 3D Realms. However, nothing came out of it, and the trademark was bought by Bethesda in 2009 before being given to Human Head Studios for development with a 2012 release date.
    • As the release date came and went, virtually no information came out from either Bethesda or Human Head about the game. It turned out that for then-unstated reasons, Human Head had quietly stopped development of Prey 2 near the end of 2011, after having spent about two years of development on the title, throwing the game into Development Hell. At this point, they had only managed to create an alpha version of the game; said alpha wasn't shown publicly until it was leaked by several former Human Head employees in March 2018.
    • According to an article by IGN, Bethesda had agreed to give Human Head an additional six months or more on the project to complete the game and still make the planned 2012 release; however, this extension was not written into their contract. Bethesda was disappointed by the state of the game and also wanted to buy out the studio, so they used this as an excuse to start cracking the whip on the devs to make development milestones, causing Human Head to stop development completely in protest.
    • Frustrated by the issues with Human Head, Bethesda took the IP back when their contract with Human Head ended, then instantly moved development of the game over to Arkane Studios. Arkane was actually chosen among a number of other developers Bethesda wanted to take over the game (some of their other choices were Obsidian Entertainment and Rebellion Developments). Arkane promptly scrapped everything Human Head had made and started the game over completely for the second time.
    • After about a year of further rumors, Bethesda officially cancelled Prey 2 and Arkane set the IP on the backburner so they could finish work on Dishonored. Eventually they went on to dust off the project and "reboot" the entire IP with Prey (2017).
  • Princess Maker 2 was slated for an English release in 1995, but publisher Intracorp went bankrupt before it could be released. MS-DOS becoming obsolete, and the rising dominance of first-person shooters in the PC market, did it no favors. The English beta of the game eventually leaked to the internet several years later... but thankfully, a fully-translated version of the Updated Re-release Princess Maker 2 Refine was released in September 2016 by CFK on Steam.
  • Project: Eden's Garden, a Danganronpa Fan Game, started off on an incredibly rocky note due to the antics of its creator and former director Sean "Obelusk" Lusk. As indicated by an October 2021 statement by the head writers after they quit in protest, after already having to shift from a live action series due to the COVID-19 Pandemic and monetizing promotional streams that could've caused Spike Chunsoft to sue, Lusk subsequently repeatedly rewrote the entire plot several times without warning, alongside frequent unprofessional conduct towards staff members and volunteers. Things significantly improved after Lusk was forced to resign — enough for the prologue to be released in December 2022.
  • Psychotoxic, a First-Person Shooter released in 2004, took six years to make and shipped in an Obvious Beta state, as elaborated upon in this video from DXFan619, who refers to it as "The Craziest Game Ever Made".
    • The game got its start in 1998 when Frank Fitzner, a German game designer who had made a number of edutainment games before then, recruited a number of people from the Half-Life modding community to form Nuclear Vision Entertainment and begin work on what was then called Psychotoxic: The 4th Horseman. The first levels they made for the game were originally created as tech demos for Vulpine 3D Technologies' Vision game engine, with an eye towards later using them for their game to attract the attention of publishers. It worked, and they were picked up by CDV Software, one of Germany's largest game publishers.
    • CDV, unfortunately, would only spend 1.5 million Deutsch Marks on the game, enough to recruit a team of six people and one part-time freelancer, far less than the 2.6 million Marks (enough to hire eleven people) that Fitzner felt was necessary to bring the project to fruition. That said, Fitzner took the deal, and sure enough, the lack of budget forced Nuclear Vision to cut many corners, leaving multiplayer, a number of the protagonist's angelic powers, and numerous other features on the cutting room floor. Furthermore, Vulpine's planned update for the Vision engine fell far behind, delaying production to such a degree that CDV at one point asked Nuclear Vision to inquire about what it would cost to license the Unreal Engine.
    • Work started in earnest in late 2001, and went smoothly through 2002 and early 2003 despite the prior problems. In August 2003, however, things began to fall apart virtually overnight as Nuclear Vision realized that CDV, having expanded far too quickly in the last several years, was in a perilous financial state. Psychotoxic only avoided the waves of cancellation and firings that befell CDV's other games and staff by virtue of the fact that Nuclear Vision had met all of its milestones for the game's development, but that didn't stop CDV from cutting the game's budget in order to stave off bankruptcy.
    • By September, Fitzner, seeing the writing on the wall, was searching for a new publisher, and seemed to have found one in Ubisoft. However, this led to a breach-of-contract lawsuit from CDV, forcing Nuclear Vision to pay 30,000 Marks to get out of its old contract, a bill that was paid for by Fitzner and other project leaders going for three months without pay while borrowing from family and friends in order to keep the studio running. Worse, the lawsuit also killed the deal with Ubisoft due to how long it dragged on for, causing Ubisoft to lose interest and move on. The only thing that stopped the game from getting canceled outright was when Vidis, a gaming accessories company that mostly published smaller titles before then, agreed to publish the game.
    • Psychotoxic finally staggered into stores on September 3, 2004... just in time to get run over by Doom³, which had come out a month prior. Its behind-the-scenes woes and shoestring budget were evident in the amount of Game Breaking Bugs that it shipped with, the worst of the bunch caused by a faulty batch of DVDs that made the Starforce copy protection system go haywire. Fitzner was willing to make a patch to disable Starforce, but Vidis, afraid of piracy, wouldn't let him do so. Furthermore, due to the game's lousy sales, Vidis cut the price to 30 euros (which Fitzner had originally wanted) — and used that as an excuse to deny Nuclear Vision the rest of the money promised to them in their contract, as Nuclear Vision had agreed to the price change. This led to another lawsuit.
    • Nuclear Vision closed its doors the following year. Fitzner left video game development to become a graphic designer before suddenly passing away, while the other developers went to work for other companies.
  • Pump It Up Infinity, a spin-off of Andamiro's Pump It Up arcade rhythm game series, struggled due to a disorganized, unpaid and unmotivated team; deadlines that were missed or delayed; a head producer who was hard to contact; and one developer's family who attempted to wipe out his hard drive which contained work on the game.

    Q-S 
  • The Ratchet & Clank series is no stranger to this.
    • According to a podcast with Mike Stout and Tony Garcia, the development of Ratchet & Clank: Up Your Arsenal was very turbulent, hence the heavily scaled down and rushed feel of the game's content. The dev team was forced to split up programmers and designers in order to focus on the game's new multiplayer mode, forcing them to cut numerous corners in the single player modenote . There was also getting Insomniac's new propietery Phoenix level editor for the game functioning, as well as getting the Galactic Rangers' A.I. to work, which took the entire development timeline to get right and it was reportedly a nightmare the whole time. On top of that, the game had a tight development timeline of a year (3 months preproduction, 9 months full production). There was a genuine fear that the game was going to be a flop, with one of them going as far as saying the game was seen as a disaster until mere weeks before it went gold, with the dev team scrambling before then to even make the game presentable for release.
    • According to Mike Stout, Ratchet: Deadlocked was an even harder game to make than Up Your Arsenal (itself having a turbulent production) and broke the dev team behind it. They were operating on the same period of time to make it as the previous two Ratchet games (three months preproduction, nine months full production) but the first few months were spent on three attempts to retool the game (one was the proposed but aborted Nexus gamenote , an R&C racing game that was scrapped to avoid competition with Jak X: Combat Racing and a failed attempt to expand the mission segments of UYA into a full but non-linear game), leaving only 9 months left to get a finished game ready for the 2005 holiday season. Making matters worse was that the Deadlocked engine was built almost entirely from scratch as opposed to reusing the engine of the previous games and, like UYA, had to be built with multiplayer in mind, forcing the main gameplay variety, enemy lineup and length to be even more scaled down than the previous game.
    • The 2016 game also had its own fair share of issues. Despite being announced in 2014 and coming out in 2016, the game only had about 10 months of development to coincide with the release of the film, along with having a smaller than usual budget. Not only did this lead to the developers being unable to hire more people to help out, they also had to cut corners, which included dropping levels from the original game, recycling weapons from previous games and omitting Skill Points altogether. They even had to resort to using data from the original PlayStation 2 game to save time, which was easier said than done due to them having lost the source code to the original games. Not helping matters was having to work off a movie that they had little involvement in with the main restriction being that they could only use a limited amount of footage from the film itself. Thankfully for everyone involved, the game went on to earn critical praise and became the fastest selling title in the series.
  • The development of Redfall was practically a repeat of the development of Anthem (2019), with Arkane Austin being ordered by Bethesda to produce a game that could leap on the live service boom that was kicked off by Destiny, despite the studio's experience being mostly with "immersive sim" games such as Prey (2017). Development lacked a clear goal throughout, with the game directors being accused of seeming content to make things up as they went along, and the development team was sorely under-staffed for a game of the scope that Bethesda wanted. It got to the point where the developers were actively hoping that Microsoft would order the game scrapped or development to be rebooted when they bought out Bethesda, only for them to not only not order any real course-correction on the game, but to start hyping it up as the next big exclusive for the Xbox Series X|S, which backfired big-time when the game released to heavily underwhelming reviews.
  • Return to Krondor, a long-awaited sequel to Betrayal at Krondor, suffered from this:
    • The poor sales of the floppy disk version of Betrayal At Krondor caused Sierra to hastily scrap their plans for the original Betrayal At Krondor sequel, going as far as to sell the game rights to The Riftwar Cycle back to Raymond E. Feist and restructuring the game's developer, Dynamix, toward making simulation games... only for the CD-Rom sales of Betrayal to be exceptionally good. This blunder not only delayed development of a proper sequel, but spurred the development of Betrayal in Antara out of Sierra's regret.
    • 7th Level got hold of the license and announced their Krondor game in 1995. Their lack of experience in developing role-playing games along with financial issues led to a protracted development where PyroTechnix, who developed the game engine used for EverQuest, was called in to assist on the game's technical aspects. By late 1997, 7th Level was on the verge of collapse and sold their unfinished games to others, leaving PyroTechnix to finish development of Return. Despite reports of strong sales, lukewarm reviews and stiff competition against Baldur's Gate and Fallout 2 ended any further attempts at continuing the Krondor series outside of novels.
  • Scalebound, a game about a badass smart-aleck who could fight and control dragons, was stuck in Development Hell for seven years, then finally looked like it was going to get out of it in 2015 as a collaboration between PlatinumGames (specifically Hideki Kamiya) and Microsoft Studios. Unfortunately, the game ended up being cancelled in early 2017 due to the two companies' different creative ideas and design philosophies clashing together, resulting in a negative impact on the morale of the development team. Platinum wanted a large world that had a consistently great frame-rate, while Microsoft wanted a more scaled down game that showed off the graphical capabilities of the console, additionally demanding it be produced within the Unreal Engine and contain substantial online components, neither of which Platinum had experience with. All of these disagreements caused the game to remain in limbo until its cancellation, and for Kamiya and producer J.P. Kellams to take a month-long absence from their respective companies to recuperate their mental health.
  • The production and design of the Sega CD add-on for the Sega Genesis proved to be quite difficult for Sega. Technical hurdles to getting the Sega CD to work with the Genesis, as well as paranoia about the capabilities of competing hardware, pressed the development team into continually beefing up the add-on's specs to the point where it became ridiculously expensive to manufacture. Adding to the mess was the lack of cooperation and coordination between Sega's Japanese and American branches (which would go on to derail the Sega Saturn), with Sega of Japan refusing to send prototypes to the West, rendering Sega of America and Sega of Europe unable to promote the add-on to consumers and developers before its release. When systems finally arrived at Sega of America, they found them plagued with manufacturing defects up to and including spontaneous combustion. All of this scramble to simply get the hardware out the door meant that Sega couldn't properly solicit game developers for the system, causing the Sega CD's library to consist mainly (though not entirely) of FMV games, with only a few or so games that actually made (comparatively) more innovative use of the system's specsnote .
  • Secret of Mana's development is an example that had profound ramifications for Nintendo, Sony and Squaresoft, as its fate is closely tied to the collapse of the SNES CD-ROM.
    • The SNES CD-ROM was a joint project between Nintendo and Sony with Secret of Mana being developed as a Killer App for the new add-on/hybrid console, but the CD-ROM project collapsed due to a bitter falling out between Nintendo and Sony; Sony's contract terms insulted Nintendo and Nintendo publicly snubbed Sony to partner with Philips at the 1991 Consumer Electronics Show (which itself would fall through). The game fell on the brink of cancellation as a result of the drama, but Executive Meddling and contractual obligations forced the developers to continue working the game as a standard title for the SNES.
    • The game went through massive cuts and reworking as a result of being forced onto the limited storage space of a cartridge. Director Koichi Ishii estimated that 40 percent of the game's content was axed in the transition, though most of the content and concepts cut from the game would find use in other Square RPGs such as Chrono Trigger and Trials of Mana. While the game was ultimately a success in both reviews and sales, the SNES CD-ROM incident and its impact on the game's development was a major catalyst behind Square's decision to sever its ties with Nintendo and partner with Sony, who would turn the CD-ROM console idea into the wildly successful PlayStation. To add insult to injury, many of Squaresoft's games would prove to be major hits on Sony's system, most notably Final Fantasy VII.
    • The English localization deserves special mention: Ted Woolsey had only one month to localize the game and was given the script out of order. Due to issues compressing the game's text space and the limited time given to localize the game, the English script was cut by about 40 percent with conversations chopped down to the base elements.
  • Shaw's Nightmare's development had a long and troubled history as detailed here:
    • The project began in 2007 when the developer got the idea to make the game when his sister got an idea for a series. The series idea didn't get very far. Due to lack of a programmer the project was cancelled.
    • The project was revived a year later and the Build engine was chosen to be the engine that powered the game. A programmer joined early on but he left when the author discussed commercial licensing with Ken Silverman.
    • The developer tried to program it himself but he ran into issues and he tried to ask for help in the Allegro forums but he wasn't able to find anyone to help him.
    • The developer continued to make the game but he eventually lost interest. The project was revived in 2011 but he lost his work due to hard drive formatting and he had to restart the project.
    • Another developer joined the project but he left without warning and did not do any work on the project.
    • The sequel was also delayed due to level designers joining and leaving the project without any warning. OBLIGE (a random level generator) was used to fill in the missing levels.
  • Sherlock Holmes Chapter One was initially planned for release on PC, PS4, PS5, Xbox One, and Xbox Series X|S platforms, but due to production limitations (presumably related to Coronavirus) the last-gen versions were delayed and so didn't get released alongside PC and current-gen upon the game's initial launch in November 2021. Kyiv-based developer Frogwares later revealed that unsurprisingly the Russian invasion of Ukraine had further impacted production, and while the PS4 port eventually made it to launch in April 2022, the Xbox One version has been indefinitely delayed, if not cancelled altogether.
  • Sierra Ops had a lengthy development with some hiccups, as detailed on the developer's website. They had to find a new lead artist in 2015 after the original one left the project, and progress was slow from 2016 to 2018 for various reasons. Things sped up mid-2019, at which point InnoMen decided to release it as an Episodic Game so that the people who backed Sierra Ops on Indiegogo would have something to play while the team finished development, only for the COVID-19 pandemic to slow things down.
  • Silent Hills immediately emerged as one of the most anticipated horror games on the horizon after a demo called P.T. (for "Playable Teaser") was released at Gamescom 2014 and scared the pants off nearly everybody who played it. It was to be a Silent Hill game made by a team consisting of Hideo Kojima and Guillermo del Toro, with Norman Reedus voicing the protagonist. Unfortunately, P.T. was all that ever came of the project, as production came to a halt several months later thanks to a bitter feud and falling out between Kojima and publisher Konami, with both Del Toro and Reedus confirming that the game was delayed indefinitely, if not outright canceled. Rumors briefly swirled that Microsoft was looking to buy the rights and restart production as an Xbox One exclusive, but those hopes turned out to be little more than wishful thinking. The experience (together with that of inSANE, a previous video game project of Del Toro's that became vaporware when THQ went bankrupt) caused Del Toro to swear off working on video games. Fortunately, some good did come of the experience, as Kojima and Reedus would later collaborate on Death Stranding, with Del Toro making a cameo appearance within that game.
  • The Simpsons Hit & Run, as described in this video by James Tyler of Cleanprincegaming, faced massive headwinds. While it was a smash hit that became a Cult Classic acclaimed as one of the best licensed games ever made, it did not have an easy time getting there.
    • During production, Radical Entertainment was facing a lawsuit by Sega arguing that their previous Simpsons game, Road Rage, was such a blatant ripoff of Crazy Taxi as to constitute copyright infringement. While the lawsuit was settled out of court, it was a close enough call that Radical decided to retool their next Simpsons game into a Wide-Open Sandbox game instead, shoehorning in on-foot segments to a game that had already gone into production as a vehicle-focused sequel to Road Rage. It is a testament to the quality of the work done by Radical that the nature of the shoehorning doesn't stand out too much and hurt the actual gameplay, but it is still visible in things like the lack of on-foot enemies and the use of fade-to-black wipes whenever characters get into or out of most vehicles (since they didn't have time to create the animations for such).
    • Radical first sent Fox a ludicrously optimistic schedule (16 months from concept to completion for an open-world game, with no assets to base it off), likely in an attempt to get the license. Radical were already behind schedule before the first playable prototype, but didn't inform Fox.
    • Due to the delayed schedule, the game failed to meet several deadlines, resulting in months of crunch time and a tense work environment. Poor communication and dysfunction between teams resulted in the game being unstable to the point of unplayable for weeks at a time. Radical and Fox had conflict over the game's content and design for the duration of the development period.
    • Morale on the team was low, with many thinking that the game would bomb. Two entire levels were apparently cut from the game due to time constraints, the code still present in the game files. Delays resulted in hasty design changes being made in October 2002 (a food and turbo meter were cut, Levels 2 and 3 were cut by a third of their previous size, two planned boss fights were cut, and bonus missions and collector cards were added) and additional time was needed to polish the game. These design changes continued to be made almost until the game was done.
    • Furthermore, the game was slated to come out on the Nintendo GameCube, but publisher Vivendi didn't want to spend much money on porting the game to a system with a limited install base, so they gave the job to just one person, lead programmer Cary Brisebois. Amazingly, he managed to convert all of the game's code to work on the GameCube in the few weeks he was given to port the game, though the fact that the port was entirely the work of one man shows in its Obvious Beta nature compared to the PlayStation 2 and Xbox versions.
  • Though not as extreme as its successor, The Simpsons: Road Rage also faced its share of development troubles, most of which clearly shows in the final product:
    • Development began smoothly, as Radical Entertainment's pitch for a Simpsons-themed Crazy Taxi game was accepted thanks to standing out at a time when every other licensed driving game was a Mascot Racer. The team was given a short 11-month deadline to complete the game, which limited the scope of the project but was not an issue by itself.
    • The game was shown at E3, where it displayed an art style very faithful to the cartoon and a large rendition of Springfield to explore. When Matt Groening was shown the game however, he believed that the game didn't look video game-y enough and advised that Radical not try to emulate the show too closely. This resulted in much of the game's textures needing to be remade.
    • When it came time to port the game to various platforms, the Xbox and GameCube versions worked fine, but the PlayStation 2 version showed serious frame-rate issues, primarily caused by the new, more detailed textures. The programmers eventually fixed this issue, but it created a new one regarding the game's streaming ability: if the player moved fast enough, they could outrun the game's ability to load new areas. With the strict deadline looming, the developers made the decision to slice the game's map up into six different areas independent of each other, an obvious downgrade that nobody was happy about. This decision is very obvious in the final game; the transition zones between areas are blocked by quickly-placed props, and the mini-map clearly shows that the areas are supposed to be seamlessly connected. This extra work also meant a few additional features had to be cut, including side-missions during the main taxi mode.
    • The final result was a game seen as So Okay, It's Average at best, and severely hurt by the string of compromises made to deliver the game on time. Some good did come from this game however, as Radical's experience in making the game would be put to good use with the far more polished Hit & Run, and helped to smooth out that game's rough development too.
  • Rounding out the trifecta of Simpsons games with tumultuous development cycles is The Simpsons Wrestling, the story of which senior programmer with the game's developer Big Ape, Robert Leyland, goes over in this Matt McMuscles "What Happened?" video:
    • Despite the much more powerful PlayStation 2 being the hot new thing at the time and the Sega Dreamcast also having been out for a year, executives at both Fox Interactive and publisher Activision insisted on only releasing the game on the by then long-in-tooth PS1, believing that they would have more market share among children if the game were released on the already established PS1 instead of the PS2. The executives also asked for rather arbitrary cuts: An animated character select screen where the characters could heckle each other and trade blows before the start of the match was forced to be cut in favor of a much simpler select screen with no animations or banter between the characters. It was also apparently a delicate dance to even have the characters be shown exchanging blows (even in a wacky, cartoonishly exaggerated way) in the first place without causing the executives to demand changes to characters so they could not attack others.
    • By Leyland's own admission, Big Ape, despite being a reputable name in the game industry, were out of their league trying to create a fighting game when they had previously largely developed 3D platformers and Zelda-esque top-down adventure games beforehand. Working with the weaker PS1 quickly proved to be a chore for Big Ape as well: It was clear that the PS1's hardware was not powerful enough to properly emulate the show's style and aesthetic without causing the game's performance to tank, especially since the outlines for the game's cel-shaded appearance were generated in real time (and real-time cel-shading on a PS1 + the fast-paced gameplay and visually intensive stages = D'OH!!). Failing to help matters was development time being only a single year (so the game could be released in time for the 2000 holiday season) along with the game's inexperienced producer suddenly quitting in the middle of development, throwing the studio into chaos as the devs tried to restructure without them and causing the game's release date to be pushed back by four months. On top of this, the devs experienced communication issues with the show's writers and voice-actors when going to them for dialogue. The writers and VAs, knowing they were writing and recording for a wrestling game and nothing more, repeatedly turned in scripts and dialogue that, according to both Leyland and level designer Michael Ebert, apparently would have been too graphic and vulgar for the TV show itself, let alone a video game based on it that was supposedly aimed at children. It took repeated attempts to tone the dialogue down to something that would be presentable in a video game that children were expected to be playing.
    • Despite selling relatively well, the game was critically ravaged and was left in the dust by its competition on other consoles (including the PS2). Activision subsequently dropped the Simpsons license like a hot potato and the one-two punch of both this game and the similarly poorly-received game based on Celebrity Deathmatch ended up killing Big Ape, whose staff migrated over to Toys For Bob (and, to their credit, went on to more successful projects there like Skylanders).
  • According to Raven Software employee Keith Fuller, Singularity's development was a mess. After two years, the game was nowhere near complete despite repeated delays and siphoning staff from other teams within Raven Software, and a good chunk of the game's maps could not load on either of the console version. The game was as good as cancelled when Activision's then-VP visited the studio and saw what a sorry state the game was in, but a deal was made at the last minute where the team behind the X-Men Origins: Wolverine licensed game would take over the project and get a 10 month extension to ship (not finish, ship) Singularity in a presentable form - a very short development for an AAA game. Managing this required taking an hacksaw to every aspect of the game, with the TMD mechanic mainly serving for repetitive box-dragging puzzles and a heavy reliance on Bioshock-style audio logs to accommodate the plot rewrites. Despite the rough development, however, Fuller goes on to conclude the team was proud to be able to make a solid game out of a disaster.
  • According to a Kotaku expose, almost no progress has been made on the Skull & Bones during its eight years in development. With the core premise of the game changing at least four times and development costs stated to already have hit $120 million, it took eleven years for it to finally see its release in February of 2024.
    • Development began in 2013, originally as a multiplayer expansion for Assassin's Creed IV: Black Flag; before it expanded into a full online game. Development is currently headed by Ubisoft Singapore, known for working as a support studio on multiple games, as well as having created the free-to-play shooter Ghost Recon Phantoms.
    • The central reason as to why development has taken so long is that beyond being a multiplayer pirate game with ships, the developers cannot decide what kind of game Skull and Bones is meant to be. Does the player play as an individual pirate captain; or as the entire ship? Is there a large overworld similar to the Assassin's Creed games; or does the game use individual sessions? Can the player do missions on land? Are crafting mechanics involved? Fundamental and foundational design like these were constantly being opened, closed and reopened, with no end in sight.
    • Part of this is due to the revolving door of senior developers. Three creative directors have taken the position, and each time a change occurred, the new managers would try and change and shift the design in their own manner.
    • Conditions at Ubisoft Singapore were also described as being poor, with management often described as being surrounded by yes-men who refused to listen to developer feedback and try to sell the Paris headquarters a vision of the game that was impractical. Many developers grew fed up with the dysfunction and lack of progress; jumping ship for better paying jobs elsewhere.
    • In normal circumstances, the game would have been cancelled outright long ago. However, Ubisoft has a special arrangement with the Singaporean Government in which the studio would have significant tax credits in the country, so long as original IP games are being made there. In effect, Skull and Bones is too big to fail.
  • Skullgirls was hit with a litany of production problems, mostly after its release in April 2012:
    • In May 2012, Autumn Games and Konami, the publishers of Skullgirls at the time, got hit with one heck of a lawsuit involving the game Def Jam Rapstar. This led to the development staff for the game being laid off by the original developer Reverge Labs. The game's development team was able to form a new studio called Lab Zero, and were allowed to maintain the game and release the planned Downloadable Content because Autumn Games owned the Skullgirls IP. The legal and financial hurdles in the formation of the new team is what caused the six month delay.
    • The first patch for the game arrived in November 2012, but only for PlayStation users (and from a different developer) — Xbox Live Arcade size limits prevented the patch from being released on the Xbox 360.
    • After a Japanese localization was released and the developers used crowdfunding campaigns to raise enough money to realize the first four DLC characters (as well as the Xbox patch), Konami wouldn't respond to any form of contact from Lab Zero, meaning the DLC couldn't be released on consoles. Lab Zero decided to cut their ties to Konami and seek publishing from Marvelous AQL, who published the PC version. In response, Konami requested the game be delisted from both Xbox Live and PSN in December 2013. Lab Zero was only notified of this after Microsoft and Sony approved the delisting. Thankfully, due to the transfer of publishing rights from Konami to Marvelous AQL, the game was re-released as Skullgirls Encore in January 2014, and included the long-awaited console release of DLC character Squigly (for free, no less!).
    • Not long after the announcement of Encore, Cyberfront, the game's Japanese publisher, closed up shop, and the game's future in Japan looked uncertain. Thankfully, some time later Arc System Works stepped up to publish the game's 2nd Encore version there.
    • And just when everything seemed like it would be going smoothly, the Red Cross requested that Valentine's crosses have their color changed (the symbol of the Red Cross is protected by the Geneva Convention; this is why you'll never see a Nurse Joy with a red cross on her hat). Lab Zero at least took advantage of the necessary patch to add a new character (Fukua) to the game.
    • Even after many years of relative peace, problems still seem to haunt this game. In late 2020, following a series of allegations of harassment and poor working conditions against Lab Zero Games's owner, Mike Z, employees left the company in droves and their departure forced the closure of Lab Zero entirely, making it seem like Skullgirls was about to meet its final end at last (a problem considering at the time, they were working on adding a new character, Annie, for the first time in five years). Fortunately, this was averted as the IP ownership was able to be transferred to Autumn Games, the developers of the mobile version of Skullgirls, while said employees that left Lab Zero Games went to found a new company, Future Club, thus ensuring continued development. Unfortunately, this shift would lead to Lab Zero's other game, Indivisible, abandoning all plans for future content and updates as a result.
  • Soul Saga, which was initially pitched as a classic turn-based JRPG, had a very difficult development. Originally conceived in 2008 with 2D graphics, Mike "Disastercake" Gale was dissatisfied with what he had and made a Kickstarter campaign in 2013 to make the game 3D, which raised over 3 times its 60k goal. Intended for a 2014 release, it faced several delays due to Mike's ever-changing vision of what the game should be like, leading to things like his wife divorcing him due to his obsession with working on the game and him becoming homeless and having to work from his car. Not helping was a Game-Breaking Bug that was completely the fault of the Unity game engine, meaning Mike could do nothing but wait until a new version of the engine came out without that issue. Episode 1 of Soul Saga would only see release (in Early Access form) in 2020, where it ended up getting negative reviews, mainly due to being too different from what was initially promised, and eventually stopped getting any updates, seemingly being abandoned.
  • South Park: The Stick of Truth began development in 2010, with a release date slated for circa 2012. The game originally was going to be published by Viacom themselves but they decided to offload the publishing duties to THQ, which had been financially ailing for some time. In 2012, Obsidian's Project North Carolina was cancelled by Microsoft, causing a portion of the game's development team to be laid off. Then THQ went defunct and ultimately sold the game to Ubisoft. The game then underwent a near-wholesale rework from its THQ-era build after Ubisoft saw it as being too bloated in scope, delaying the game until its eventual release date of March 2014.
  • The 2010 Splatterhouse reboot had a very rocky development. Originally greenlit in 2007, the game went through multiple dev studios before settling on BottleRocket Entertainment, with them announcing a 2009 release date. Unfortunately, BottleRocket head developer Jay Beard was a severe case of Small Name, Big Ego due to his success with The Mark of Kri, ignoring Namco’s wishes to do whatever he wanted with the property, going so far as to lie to their faces when they came by for brief visits. Namco, naturally, ended up not being satisfied with the final product, so they chose to scrap everything BottleRocket did and develop the game in-house, with only a year left to go before its new 2010 release, which helped lead to the closure of BottleRocket (though Namco was gracious enough to hire most of the former staff minus Jay Beard). As a result of the rushed development time, Namco wasn't able to iron out the game's problems, such as the long loading times, some graphical issues, and clunky platforming bits. The game was released to lukewarm reviews and poor sales, but has now become a Cult Classic.
  • Splinter Cell: Conviction: It took almost four years from the time the game was announced (via an internal leak of images from the game in mid-2006) to its release because of several major gameplay shifts, including a halfway-finished product that was essentially thrown out midway through production. The original game, helmed by Ubisoft Montreal, featured Sam Fisher (now on the run from Third Echelon) as some type of homeless drifter sporting a beard, hoodie and makeshift weapons and devices, and the gameplay was intended to be a sandbox-type shooter where Sam would investigate various locales to get information (and memories) about his daughter. The game was seen as a serious departure from the franchise, and Ubisoft canned it midway through development over negative fan reaction and claims that its gameplay was too similar to the original Assassins Creed (also made by Ubisoft Montreal). Several features were unceremoniously thrown out (including several abilities that enabled Sam to blend into his environment, move objects around and fight hand-to-hand against enemies), and the game's entire structure was revamped. Conviction would eventually be released in early 2010.
  • SpongeBob SquarePants: Revenge of the Flying Dutchman, as covered by Who Needs Normal? in this this video.
    • Original developer Kalisto Entertainment USA run into technical problems regarding the internal engine developed by their parent company in France, who had to constantly update the code in order for the American development team to properly do anything at all. The financial turmoil brought upon Kalisto France from the failures of Nightmare Creatures 2 and NYR: New York Race led to them desperatly trying every tactic to stay afloat; producer Billy Joe Cain described in a 2019 interview (around 11:43 in the video) instances where, upon calling the French office, he and the rest of the development team found out that they were taking the entirety of August off, and in another instance, Kalisto France outright begged the USA team just for a piece of the money they were making on development. Kalisto France would shut down in April 2002, leaving the USA team with two unfinished products that they needed and wanted to complete; they scrambled to form BigSky Interactive and reprogram Kalisto's engine to something more user-friendly in order to prevent publisher THQ from canceling the project. Because of this, the team had to scrap a number of intended features in order to get the game out on time. Notably, the GameCube version was to have functionality with the GBA version, where the Flying Dutchman's treasures would be found with the GBA serving as a radar; the final version uses the GameCube's rumble function for this, with the remnants being left in the code. Development was so rushed, there was an infamous glitch that occurred in the PS2 version where the game would be stuck on the loading screen, which would lead to the game saves being corrupted or even outright deleted; the result of a soft reset tactic that BigSky had to work around due to Sony's draconian policy about loading screens being longer than six seconds.
    • THQ and BigSky's relationship became strained during development. Initally, they were kind to let them continue working on the project, since they were so deep into development they couldn't hand it off to a different developer. THQ began to pay them less and less, to the point where it would come down to the last hour before the bank closed. This greatly stressed management to the point where they were begging THQ for funds until the last possible minute. The entire development team became tense due to crunch time, and issues would rear their ugly heads. One of the lead designers lost his grandfather, who lived in Japan, and had to be gone for a whole week, a move which didn't sit well with one of the producers at THQ, who went ballistic.
    • During development, BigSky contracted a number of publishers in order to secure funds for their future projects. Towards the end of development, they realized that none of the publishers they contacted were calling back; Cain would find out from a friend that Jeff Lapin (who at that point had transferred to Take-Two Interactive shortly before the game was released) had blacklisted the entire company to spite them for rejecting THQ's offer to buy them. The rest of development was spent laying off most of the staff just to get the game finished on time; what few staff remained could only work on the game on an on-and-off basis.
  • The production history of Spyro: Enter the Dragonfly could very well be a horror story. A large part of it came from the outlandish demands of Universal Entertainment, often demanding certain changes and retracting them at a drop of a hat, and shooting down ideas for the game if they didn't consider it "Spyro enough", the definition of which seemingly changed from meeting to meeting. Crunch time was in full effect, the game would be completely rewritten multiple times, and the pressure resulted in hostility between the employees, including one who would snap at random and would often spiral into violent behavior, to the point of choking a programmer for disagreeing with one of his suggestions. The end result ended up being Christmas Rushed, with many obvious cut corners and having so many bugs and glitches that it would make Sonic Boom: Rise Of Lyric blush, and Enter the Dragonfly is commonly the butt of jokes among the Spyro the Dragon fandom to this day.
  • Following the dissolution of LucasArts, Bioware being purchased by Electronic Arts, and Obsidian's bankruptcy and acquisition by Microsoft, the Video Game Remake of Star Wars: Knights of the Old Republic fell to, of all people, Aspyr Media who are primarily known for porting games. The KOTOR remake had been rumored for years but was officially announced in late 2021 with plans to release the game by the end of 2022. However, in July 2022 it was announced that Aspyr Media had fired the design and art directors almost immediately after they presented a working demo, and the game would be delayed indefinitely with a release date in 2025 at the earliest as Aspyr's parent company Saber Interactive took over development.
  • S.T.A.L.K.E.R.: Shadow of Chernobyl suffered from almost a decade of Development Hell due to the developer GSC's grand ambitions battling with their own X-Ray Engine, along with near-endless feuding between publisher THQ and GSC on the road to becoming known as the Eastern European analogue to Duke Nukem Forever. Originally announced in 2001 and not released until 2007, the result was an amateurish Obvious Beta. It eventually became Vindicated by History as a huge Cult Classic, but it had quite a hard time getting there. This video chronicles its development through its known prototypes.
    • What became Shadow of Chernobyl started as a completely different game called Oblivion Lost, and was originally intended to be a sci-fi game in a futuristic setting akin to Quake before GSC opted to put the game in a more grounded setting after their similarly-themed shooter, Codename: Outbreak, flopped. Some on the team were apprehensive about the game being set at and around Chernobyl, as the effects from the Chernobyl disaster was very much still in recent memory and they didn't want to make light of it by using it as a setting.
    • The team's wild ambitions constantly ran afoul of reality. Initially planned as a realistic and massive open world game with cutting edge AI and photorealistic graphics, the team regularly fought against the X-ray game engine, which was less than well suited for the monumental task asked of it. Many of their gameplay ideas proved to be unworkable as various mechanics fought against each other, such as drivable vehicles clashing with the deadly anomalies in levels. In late December 2003, a pre-alpha build of the game was leaked to peer-to-peer file sharing networks. This build, marked as version 1096, inadvertently acted as a fully functional tech demo of S.T.A.L.K.E.R.'s engine, despite its lack of NPC enemies and fauna.
    • A version of the game dubbed Build 1935 was the closest the game got to its original ambitions. Despite being in a near feature-complete (albeit very buggy) state, much of this build's content was axed due to issues with the engine, negative feedback from playtests and THQ demanding a more story-driven game upon being shown the build. This forced GSC to hastily cut the game's story to pieces, and thanks to the issues with the engine and amount of cut content, the game was delayed multiple times for almost two years by THQ, while GSC frantically tried to make the messy remains of the project into a playable game. This build would later be released in 2009 by GSC to the community.
    • Additional problems followed in 2005, when some members of the development team, most notably Oles' Shiskovtsov and Aleksandr Maksimchuk who had worked on the X-ray engine, were frustrated with the constant delays, workplace conflicts and abysmally low pay. THQ's representative at GSC, Dean Sharpe, faced language barrier issues and open hostility from the development team, with one incident leading to a physical altercation between himself and a developer. His attempts to get the game focused and cut down to something releasable also made him a target of scorn and death threats from the community. These factors ultimately resulted in a mass exodus from the company on October of that year, with many of the former GSC developers forming 4A Games to begin work on Metro 2033. Sharpe, despite his icy relationship with the GSC development team, would himself move to 4A and eventually become its CEO. Sergei Grigorovich would later accuse 4A of stealing assets and engine code from S.T.A.L.K.E.R. for Metro 2033, though no legal action was taken.
    • THQ ran a competition in January 2007 offering the winners the chance to play the beta in a 24-hour marathon session. The event, scheduled to take place on 24 January 2007, was subsequently changed to a 12-hour session merely days before it was supposed to occur. On the morning of the event, the winners were met at the venue by the THQ staff that had organized the event... who told the winners that they couldn't get any copies of the game, because it was "unplayable" according to GSC and the developers refused to give THQ any beta copies.
    • The standalone expansion Clear Sky suffered from this too, as the entire X-Ray Engine was retooled for DirectX 10 support halfway through development with only a year of development time. The final result was a somehow even worse Obvious Beta than Shadow of Chernobyl; the game shipped as a near-unplayable mess of glitches, awful optimization/performance, insanely unbalanced Nintendo Hard gameplay that flew headfirst into Fake Difficulty, and bugs - including several bugs that made the game impossible to finish. Thankfully, Call of Pripyat was much smoother and the result was the most polished game in the series.
  • The development of S.T.A.L.K.E.R. 2: Heart of Chernobyl managed to surpass both Shadow of Chernobyl and Clear Sky in the sheer scale of troubles that the development team faced.
    • According to different sources, the first incarnation of the game was in development after Call of Pripyat, but GSC Game World was facing serious financial difficulties as well as having trouble finding a publisher for the game. Internal conflicts within GSC were becoming increasingly severe. GSC's CEO, Sergei Grigorovich, was becoming dissatisfied with the direction that the game was taking, ultimately leading to the studio shutting down in December 2011. What little remained of the game's staff desperately continued their work without receiving any financial compensation or pay for several months. They tried to find new investors and a publisher to finance the project, but Grigorovich still wanted to hold onto the S.T.A.L.K.E.R. license and forbid the development team to continue work on the game, which forced the remaining staff to finally kill off the project in April 2012. Most of the remaining development team would thereafter go on to found a new company called Vostok Games who made a multiplayer free-to-play Spiritual Successor called Survarium, while Grigorovich and some other old GSC workers went on to re-establish GSC in December 2014 and announced in May 2018 that the game was back in development. In the meantime, they released Cossacks 3.
    • Even then, it has been far from an easy road. Initially given a release date of 2021, it was pushed back to April 2022 and then December 2022. In December 2021, GSC announced plans for in-game Non-Fungible Tokens, only to be met with such backlash that they retracted such plans with an apology. In February 2022, development was suspended due to the Russian military invasion of Ukraine while much of the team relocated to the Czech Republic to resume development. One of the developers, Volodymyr Yezhov, was confirmed to have been killed in battle in December 2022.
  • Star Fox Adventures started life on the Nintendo 64 as a stand-alone title by Rare named Dinosaur Planet. Considerable technical wizardry was required to make a game of the scale that the development team were aiming for feasible on the Nintendo 64, but the team nonetheless got the game engine working and had the product about 90% complete... when Shigeru Miyamoto gave them the idea to make it an entry in the Star Fox series. The development team had grown attached to the work they had created and were hesitant, but ultimately decided to go with Miyamoto's idea due to the increased sales potential of the Star Fox brand. Rare also decided to move the game over to the GameCube because of the N64 reaching the end of its life cycle and to take advantage of the tech. However, due to Rare's inexperience with disc-based media, a lot of content was lost during the transition from N64 to Gamecube, and because of the rebranding, they had to add content that was never originally intended to be in the game and that it wasn't built for like the Arwing segments. To make matters worse, Microsoft bought Rare out mid-production. This caused Rare to have to rush to get the game out before the deadline of them officially going to Microsoft and scrap more than a third of the content that was planned for the game, including having Krystal as a playable Deuteragonist, an actual boss fight with General Scales, a Desert Force Point Temple, six Spellstones instead of four and eight Krazoa Spirits instead of six, and more. Even then, they just barely made it, with Microsoft acquiring Rare the day after the game released.
  • The development of Star Trek: Klingon Academy embodied the rampant Head-in-the-Sand Management of Interplay Entertainment at the time, and it is regarded as a small miracle it became a Cult Classic space sim at all.
    • As told in this interview and a post-mortem with designer Ron Hodge, the game encountered troubles from the offset. Interplay wanted the game to be an Mission-Pack Sequel produced on a limited budget, ideally avoiding the live-action footage and costly Trek actors of Star Trek: Starfleet Academy. As Starfleet Academy was greeted by mixed reviews in the US press, Interplay also wanted Klingon Academy to address its issues but were reluctant to give the team the resources needed to do so. Producer Raphael Hernandez lied to Interplay, agreeing with their demands before going behind their back with far more ambitious plans.
    • Problems were immediately encountered with reusing the Starfleet Academy code, much of which was in a messy state due to the lead programmer being inexperienced and not saving the source code of key features. Hernandez made the call to focus on the graphics engine above all else. This had ripple effects through the entire production; While the graphically impressive internal demos pleased the executives at Interplay, gameplay systems were not implemented until late in development, resulting in unfinished and underdeveloped game mechanics.
    • Hernandez was also blamed for the combative atmosphere at the studio. He insisted on keeping the different departments from communicating with each other in favour of routing communication through himself, which only worked until Hernandez became detached from the game's development and rarely spoke to the team at all. Changes from one department would often require another department to react accordingly, creating a great deal of resentment between members of the team and particularly the scripters from the lack of communication, who vented their anger to 14 Degrees East producer Brian Christian, who would then make summary judgments without consulting the rest of the team.
    • As the scope of the game grew to include new live-action cutscenes for the story, Interplay co-founder Brian Fargo and Development VP Trish Wright grew hostile toward Klingon Academy, blaming it for the rapidly declining finances of Interplay. Christopher Plummer, asked to reprise his role from Star Trek VI: The Undiscovered Country, demanded enough money that Wright wanted his his character cut from the game, only for Paramount to step in and threaten to pull the license if Plummer was cut. Even after all this, designer Brent Kollmansberger had to repeatedly argue with Fargo against canceling the game during the final months of development.
    • Though it was eventually released to a decent critical and commercial response, the game's chaotic development likely played a major role in Paramount pulling Interplay's Star Trek license and consolidating all video game development on the franchise under Activision, who had just released two of the franchise's most critically and commercially successful games, Star Trek: Elite Force and Star Trek: Armada.
  • Star Trek Online started out being owned by Perpetual Entertainment and was given a set schedule for launch. However, two years into development, they had plenty of pretty conception work, but no actual gameplay and it's been said that Perpetual was only doing this for shadier reasons. Either way, CBS was angry at the waste and gave the license to Atari and Cryptic, who forced the companies to rush out the game (though it was thankful that Cryptic had the easily compatible engine from Champions Online to use). However, even that wasn't enough as the game was bleeding players as Atari used the money the game brought to pay off their massive debts. It wasn't until Perfect World Entertainment bought Cryptic that the game would flourish.
    • A good example of Cryptic's forced rush was the Klingon faction. Compare the group at launch and compare the group during the first expansion, "Legacy of Romulus": Klingons were not chooseable at the start, they were unlocked when a Federation player reached Level 20. The Klingons started out at Level 20, but had no extra places to actively gain levels like the Federation did and had very little ships to play with. As well, the Gorn race was laughably bad in design, looking almost as bad as the Star Trek: The Original Series versions. Nowadays, the Klingons are chooseable at the start, the Gorn look much better and there's a starting story for the Klingons.
  • Star Trek: Starfleet Academy hit trouble with the live-action cutscenes on the PC version, which were hurriedly completed in just a couple of days. Complicating matters was the fact that someone used red dots on the green screen as a tracking method, believing that it would help make combinining the actors with the CG backgrounds more accurate. It did nothing of the sort and as it was a near-match for the colour palette used by the Trek uniforms it created all sorts of errors. Unable to reshoot the scenes, the team had to spend over half a year fixing the live-action footage frame-by-frame. On the plus side, the extra time allowed them to add more multiplayer modes to the final release.
  • State of Decay 3, as revealed in this article by Ethan Gach for Kotaku, spent several years mired in pre-production, all while an incredibly toxic culture festered at its Seattle-based developer Undead Labs.
    • Undead Labs, after having officially partnered with Microsoft on the first and second State of Decay games, became an official part of Microsoft Game Studios (now Xbox Game Studios) in 2018. Contrary to the fears of many employees, Microsoft took a very hands-off approach to the company that they initially welcomed, but which they believe in hindsight to be responsible for many of the problems that emerged as they felt that they had nobody higher up to turn to. To quote one employee, "we were afraid they would come in and change our culture but our collapse came from within, and we could have used [Microsoft's] help."
    • There were two big problems going into SoD 3. The first was that nobody was quite sure what kind of game they wanted it to be. Would it be the massive, open-world Zombie Apocalypse MMO that studio founder Jeff Strain had envisioned a decade ago, with the first two games as the proof of concept, or would it be an iterative sequel that was bigger and badder than its predecessors but still more or less offered the same gameplay? The second problem was that the team was still working on new content for SoD 2, which had become a live service that required a constant stream of support, drawing away their focus even as they started work on the sequel.
    • Development wound up scattershot, with employees divided into "strike teams" to work on features with little communication between them, leading to lots of duplicated work and a checklist of features while the core gameplay experience was neglected.
    • What's more, discrimination against female, non-white, and LGBT+ employees ran rampant. While Microsoft, in its communications with Kotaku for the article, proudly cited its diverse hiring record, many of those employees said that Microsoft did little to help them and was just using them for PR. Female employees routinely faced sexist comments, saw their contributions belittled by their male colleagues, and were talked over in meetings even when they held positions of authority. Studio head Jeff Strain, who would leave Undead Labs in late 2019, seemed increasingly apathetic about his job, and when he departed, new studio head Philip Holt was seen by many as a poor replacement, one who hired his buddies from other studios and put them in charge of Undead Labs' satellite studios in Florida and Illinois while further entrenching the sexist culture at the studio by pushing out two female directors and replacing them with men.
    • Holt did, however, hire one important woman: the new human resources head Anne Schlosser, who spent her time siding with the bad apples on Undead Labs' staff and doing nothing about the culture at the company. A staff meeting in August 2021, in which multiple female employees shared their stories with each other and realized that they weren't alone, brought the situation to a head, with Microsoft HR getting involved and starting a process that ended in Schlossler's departure the following month. Holt's continued praise for Schlossler, however, indicated that he cared more about sweeping issues under the rug than actually solving them.
    • Collapsing morale within the studio led to a constant stream of departures. Between that, the scattershot production, and the demos the studio had to make to show Microsoft it was meeting its targets, many people working at Undead Labs didn't seem to know exactly what features were even in the game. The demo that they ultimately put together in September 2021 went terribly, beset as it was by tension between Holt's ambition and developers on the verge of revolt. Those who have stayed around at Undead Labs through 2022 feel that the game has finally turned a corner, though many former employees who quit are skeptical, and feel that nothing will change as long as Holt is in charge.
  • Street Fighter X Tekken seemed like a good idea at the time, crossing over two of the biggest fighting game franchises together. But as Matt McMuscles explains with the help of a former Capcom employee here, a series of short-sighted marketing issues doomed it to failure:
    • Capcom, riding high from the fighting game renaissance they started with Street Fighter IV, negotiated with Namco to license their Tekken characters for a crossover fighting game. Tekken creator Katsuhiro Harada and Street Fighter IV producer Yoshinori Ono announced the game at EVO 2010 to a stunned audience. While fans were generally hyped, there were concerns as to how it would meld the different Street Fighter and Tekken mechanics into one game. The answer was two games; Capcom would produce Street Fighter X Tekken and Namco would produce Tekken X Street Fighter, both games relying on their respective flagship franchises' game engines.
    • In addition to the licensing fee they paid to Namco, Capcom contracted Polygon Pictures to produce CGI trailers to promote the game, with licensed songs from bands such as Rise Against. Capcom expected the game to bring in the combined sales of both franchises, which failed to take into account any overlap between their fanbases and those not ready to switch to the new game.
    • Enter the Gem System, a a pay-to-win powerup mechanic. Retailers offered their own gem packs, nearly all of them better than the subpar pack in the base edition. Players rejected the gem system outright, while tournaments banned it, leaving the system, which was expected to go on for years, to be dropped after a few months. The game also featured another unpopular game mechanic, Pandora, in which a player would sacrifice one of their characters for a seven-second power boost but would cause them to lose the match unless they won in that timeframe. Players were also disinterested in the main gameplay, which, among other things, would cause most matches to end in time over.
    • As part of a marketing deal with Sony, Capcom put in five exclusive characters in the Playstation 3 and PS Vita ports. Of those characters, Bad Box Art Megaman was the most reviled, as this came soon after Capcom cancelled numerous Megaman projects, including the highly anticipated Mega Man Legends 3. This soured fan goodwill even further.
    • And then came the DLC characters. Twelve DLC characters were intended to be released for the console ports to coincide with the PS Vita port release. But hackers discovered them early ... directly on the disc. This was a concern Capcom USA's marketing team brought up with the Japanese branch, who couldn't fathom that anyone would do something as illegal as cracking the games. Their request to remove the DLC characters were also ignored. Needless to say, fan goodwill plummeted further, which resulted in the DLC characters officially being released a month early.
    • All of these factors led to the game falling spectacularly short of sales forecasts. The Tekken X Street Fighter game was quietly cancelled, while Capcom came to regret their DLC practices and made pains to not repeat those mistakes.
  • Steven Seagal is The Final Option was designed around the then-unusual conceit of creating an entirely original game starring a big-name Hollywood actor, rather than basing a licensed game on one of their films or TV shows. And much like many of Seagal's films, the production was severely problematic; in this case, to the point where the game eventually died in Development Hell.
    • Publishers TecMagik hired Riedel Software Productions to develop the game, promising them collaboration and oversight from Seagal himself. As the team at RSP soon discovered however, Seagal wasn't really involved in any capacity — partly because at the time he was on location in Alaska working on his directorial debut, On Deadly Ground, but also because actually hiring him would have cost far beyond what TecMagik were willing or able to afford — and the publishers had merely paid for his name and likeness to be used in the game.
    • Seagal's management did point them in the direction of some of his regular stunt doubles as casting choices for the live-action shoot (the game would have used the same digitized actor technique made popular by Mortal Kombat (1992)), but TecMagik thought even that would be too expensive, and they ended up hiring a martial artist named Greg Goldsholl, who bore a passing at best resemblance to Seagal (though at the resolution the game operated at, this wasn't massively noticeable), and had to shoot all his scenes separately from the rest of the cast because he was scheduled to teach a martial arts class on the day of the main shoot, and the publisher wasn't willing to compensate him for the loss of earnings that would have been caused by cancelling the class. As the kicker, when Goldsholl did meet Seagal in-person a few years later, it turned out that for all of TecMagik's claims about hands-on involvement from Seagal, his management didn't even consult him when they licensed out the rights to his image for the game, and that he had tried to have it cancelled when he found out on returning from the location shooot for On Deadly Ground, but discovered that his hands were tied.
    • RSP were given the source code and assets that a previous developer had been working on — which turned out to be useless as it was for a driving game, much to the RSP team's bewilderment as Seagal's films were known for their martial arts sequences, not vehicle chases. This forced them to junk everything and start over.
    • The producer that TecMagik initially asssigned to the game freely admitted to never having played a game in her life — her previous job had been as a bank manager — and she kept pressuring the developers to simplify the gameplay and just copy Streets of Rage. Eventually she got booted off the project... and was replaced by someone who somehow knew even less about game development, despite his having previously designed a strip poker game for the PC.
    • No sooner had he been installed than the new producer did something that would help send the game's development down in flames — he authorized another, highly costly live-action shoot, thinking that the characters from the original shoot looked too generic and the game needed more outlandish character designs. Unfortunately, no-one involved in this new shoot had any idea what they were doing, resulting in the footage being totally unusable. Though considering how ridiculous the characters ended up looking, with one of them being described as "Bart Simpson wearing Freddy Krueger's glove", this may have been for the best.
    • After an initial prototype of the game — which until this point had been intended as a SNES exclusive — was up and running, TecMagik realized that it would probably be a good idea to have a Genesis port, seeing how the system's target audience was older and likely to be more interested in Seagal's films. Since RSP only had experience developing for Nintendo platforms, TecMagik hired another studio to work on a Genesis port. Instead of porting RSP's code to the Genesis, however, the other studio used the live-action footage to produce a bunch of programming demos that were arcade-quality in terms of visuals, but bore zero resemblance to any actual game that could be run on the system.
    • The final nail in the coffin came when On Deadly Ground proved a major critical and commercial bomb, and sent Seagal's career into a severe downward spiral. Between the money they had already wasted on the second live-action shoot and the Genesis "port", TecMagik decided to cut their losses and abandoned the project, with the staff at RSP viewing the decision as a Mercy Kill. By some accounts the publisher tried to revive the project for the original PlayStation, but gave up due to a combination of financial problems and Seagal's star status having waned even further in the intervening years. The game's troublesome development would ultimately prove a Creator Killer for TecMagik's US branch, who went out of business not long afterwards; their UK branch remains in operation to this day, but purely as a consulting business for developers and publishers.
  • One would think that, as bad as Superman 64 was, it was solely developer Titus' fault. However, in an interview with Eric Caen, one of the founders of Titus, with ProtonJon, it's revealed a lot of the reasons for the game's poorness was office and company politics with Warner Bros. and DC Comics towards Titus (the virtual reality world? They didn't want Superman kicking real people). In fact, Caen mentions that the game was "not even 10% of what they envisioned" and that while it was a money-maker, it hurt them in the long run because they were forced to cancel the PlayStation version of the game.
  • An odd one — Super Robot Wars Alpha 3 was actually held back a few months because of a mundane music problem: JAM Project just couldn't hit a specific high note for the game's theme, GONG. Even now, they'll still sing the song in their shows, but they will not attempt to hit that high note since they simply can't get it to work.

    T-V 
  • Even well before Richard Garriott was Screwed by the Network, Tabula Rasa was beyond just troubled. The game began its creation as a Spiritual Successor to the Ultima series after EA's own failures led to Origin's functional demise. Set in a variant of eastern Asia, though, test audiences quickly explained that it didn't track due to the Uncanny Valley — so the project began anew, this time with more familiar (to the devs) Western fare. This, too, eventually had to be scrapped because the situation didn't quite pan out. Finally, under mantle-dense pressures, the sci-fi incarnation of Tabula Rasa was unveiled, but with NC Soft breathing down Destination Games' necks all the way, a comparatively meager reception, quite a lot of bugs in beta and at launch (in fairness, they only had two years worth of crunch to produce the last variant) and the publisher rapidly becoming known for providing developers with just enough rope to hang themselves, NC finally took it upon themselves to fire Lord British and kill the servers in the most ignominous possible manner, leading to him winning a $28 million lawsuit — and the rest of the world losing out on a promising, if flawed, sci-fi MMORPG.
  • "Troubled" doesn't even begin to describe the genesis of Tattoo Assassins, Data East's Mortal Kombat clone. According to insider accounts, the idea was born when Data East Pinball executive Joe Kaminkow got a script treatment from Back to the Future's Bob Galenote  and decided to turn it into a game — with a development schedule and budget that was less than half of the typical arcade game of the time. It bears repeating that this proposal, tight schedule, and low budget for an arcade game was given to a company focused exclusively on pinball machines, not videogames. The developers were promised lucrative bonuses for making the deadlines, which resulted in 12-hour-a-day, seven-days-a-week grinds. To make matters worse, the team was hamstrung by nonstop demands for additional fatalities, time was spent preparing demo prototypes for trade show demos, and the final hardware was far underpowered for the game's intended design. The playtesters had to be forced to test the game, while the developers simply wanted to finish it and move on — and at the end of it all, the game was never released.
    "We knew the game was crap, and that we were no longer capable of fixing it. After we got back from the show, we were so 'crispy' that we no longer cared about the money — our only true reward for finishing up was that we wouldn't have to work on it any more."
  • Thief (2014) was in various states of Development Hell for nearly six years, and the results showed greatly in the final product.
    • Prior to Eidos Montreal overseeing development of what would eventually become the final game, ION Storm Austin (producers of Thief: Deadly Shadows) apparently planned to create a modern reboot, which would take place in the present day and feature Garrett working in an urban city. The concept was only worked on for a short time before ION Storm closed its doors.
    • Eidos began development on Thief 4 in 2008, and intended for it to be a sequel to the original series. After a teaser website and logo appeared in 2009, there was no word for years on the status of the project or where development was at, and various reports and rumors claimed that it had already been worked on for a long time prior to its announcement, that work was being outsourced to another studio, and that Eidos Montreal still hadn't decided on whether it would be a sequel or not.
    • Work went on for so long that an incomplete leak of a CGI trailer from the game's prologue mission made news in 2012, and after the game's release, it was revealed that Stephen Russell, the original voice actor for Garrett, had provided dialogue for the unused trailer, but he was replaced for the final version. Additionally, an unused (but completed) OST from the game was released by ex-audio director Paul Weir after the game's release, which featured many cues (including the one from the unused trailer) and pointed to an earlier version of the game's plot (suggesting that "Erin" was originally named "Ariel").
    • In March 2013, a damning report came out stating that the development team was completely fractured. Designers were coming and going constantly, and all of them had differing visions on where to take the game. Screenshots released around this time also showed a third-person gameplay perspective. The same month a Game Informer cover story came out, lead designer Dominic Fleury left the team.
    • The dev team then spent nearly a year working on press demos instead of the game proper, with heavily modified code being shown to media outlets (which was not released to the public because the team was unhappy with the end result). That also coincided with Eidos being forced to secure additional funding for the game from a German investment firm.
    • During the development of the press demos, the animators were tasked with animating sex scenes for the House of Blossoms level, which they were reportedly very uncomfortable with.
    • The game was eventually released in February 2014 to middling reviews. Three years later, rumors of a fifth game being in development (in tandem with a feature film supposedly being produced by Straight Up Studios, which was later removed from the studio's site) led Eidos Montreal boss, David Anfossi, to mock the announcement, claiming that fans should just "forget" any chance of seeing a Thief 5.
  • This Is Vegas was envisioned as Midway Games' answer to the Grand Theft Auto series, and spent half a decade in development. Unfortunately, as this video by Liam Robertson lays out, it proved to be a boondoggle that stood as a symbol of Midway's downward spiral in the latter half of the 2000s.
    • Surreal Software, makers of The Suffering, had then-recently been snapped up by Midway, who assigned them a pitch by CEO David Zucker for an open-world crime game set in Las Vegas. After several rejected pitches, they eventually settled on something much like GTA, albeit with the lighter tone of Saints Row owing to the game's Vegas setting, its focus on nightlife and gambling, and its heavily satirical script written by former Cracked editor Jay Pinkerton that sent up the city's real-life '90s Audience-Alienating Era. Surreal took several trips to Vegas for research, though attempts to get the city's actual casinos to appear in the game were shot down by their owners, who objected to having their properties used as the scenes of violence in a GTA-style video game. This led the team to use fictionalized versions of the Strip's casino resorts.
    • The problems started when Midway required Surreal to use the brand-new Unreal Engine 3 to run the game, owing to an expensive licensing agreement that Midway had made with Epic Games in 2005. Many people who worked on the game regard this decision as its death knell in hindsight, as Unreal 3 was still an experimental, untested game engine that few people had worked with and which no games had been released for yet, and furthermore, it was overly specialized for linear action games and was ill-suited to the demands of an open-world game. As a result, production was plagued by technical difficulties from the word "go" as Surreal struggled to build their ambitious game within the limitations of the engine they'd been saddled with, especially given that the naturally flat, open geography of Las Vegas, which they'd hoped to depict in the game, meant that they couldn't easily cover up pop-in and other visual issues.
    • There was also a great deal of internal strife over the game's tone, leaving it with something of an identity crisis. On the subject of guns, some developers felt that they clashed with the game's Lighter and Softer tone and its focus on living the good life, and that requiring players to shoot people (as opposed to get into non-fatal fistfights and Mixed Martial Arts matches) would produce a bad case of Mood Whiplash, while others felt that guns were a perfect fit for an open-world crime game and wanted to add blood on top of it. Similar arguments arose over the game's depiction of women, with some feeling that the focus on wet T-shirt contests in the marketing materials made the game look like fratbro juvenilia.
    • The game was revealed to the press in February 2008, with a promised release date later that year. In truth, it was nowhere near meeting such an optimistic date, and it was pushed back to 2009 just four months later. Midway's ever-growing list of demands for new features consistently led to more work for the team, exacerbating the technical headaches caused by Unreal 3, and Midway's tech-sharing initiative, an attempt to cut costs by recycling assets from other games, backfired as it often turned out that it took significant work to get those assets to work with different games. Worse, Midway at this point was losing money hand over fist; it hadn't reported a profit since 2000, it had lost over $250 million since 2005, and its lifeline from National Amusements CEO Sumner Redstone was cut in December 2008 as the Great Recession hammered his finances and caused him to sell his majority stake in the company. Two months later, Midway filed for bankruptcy.
    • Even after Warner Bros. Interactive Entertainment bought out the remains of Midway in May 2009 and saved This Is Vegas from immediate cancellation, morale at Surreal was low and an aura of doom hung over the project. Several Surreal staffers were laid off that year, and plans for post-launch Downloadable Content were scrapped. While work did continue, few had any illusions that the game would be finished, and indeed, in August 2010 it was finally canceled. A total of $60 million had been spent on the game by Midway and WB, and one developer estimated that WB would've needed to spend another $15 million to finish it. Rather than continue throwing good money after an unproven new IP with a troubled production history and a lot of edgy humor that the higher-ups were uncomfortable with (not least of all because of the pot shots it took at various celebrities, including people who were in working relationships with WB), they pulled the plug.
    • After This Is Vegas was canceled, Surreal was merged with Monolith Productions and worked on Middle-earth: Shadow of Mordor, while Pinkerton went to work for Valve and co-wrote Portal 2 and Half-Life: Alyx.
  • Based on The Digital Antiquarian's coverage of the game (1•2•3), the development of Time Zone, the sixth adventure game by On-Line Systems (later known as Sierra), was among the earliest major examples of this in the computer game industry. To make a long story short:
    • Time Zone was the latest of the Hi-Res Adventures, whose selling point was the colorful artwork accompanying every room at a time when most adventures were text-based. It was not to be just another adventure game, however; Roberta Williams, the designer, envisioned an epic adventure through time and space where the player must travel to different points in history, culminating in a climactic showdown in an alien planet of the future that would itself be bigger than any adventure game of the time.
    • To this end came the first semblance of modern "AAA" game development where instead of a few jacks-of-all-trades, it would be created by a team of people performing different specialized tasks for the game—the largest ever assembled at the time. Williams designed the game; three people translated her design into game code, with each "time zone" coded as a game in its own right; and three people handled the game's artwork, with one drawing on graph paper and two others digitally tracing the art into the game. All of this was managed by Bob Davis, who himself had just finished and shipped his own game, Ulysses and the Golden Fleece, as the fifth Hi-Res Adventure.
    • Alas, Davis did not have the experience or temperament to be a manager, and almost immediately, the project devolved into a desperate attempt to figure out how to fit all the different sub-games together in time for a Christmas 1981 release. It was not until Jeff Stephenson, an experienced programmer who had last worked for the developers of VisiCalc (the first PC spreadsheet application), joined the company that the team managed to fit the whole thing together. The final result is essentially the bare minimum of Williams' vision, with each sub-game essentially stuck together with duct tape and glue.
    • Worse yet was what the lead artist had to go through. The one person who drew the art, Terry Pierce, had to draw 1,400 images for Time Zone... most of which were empty landscapes, as the Hi-Res Adventures brand dictated that every room of the tediously-oversized game grid needed its own artwork no matter how important or interesting it was. The stress of having to scribble hundreds of featureless "fields," "forests" and "city streets" on a tight schedule eventually drove Pierce into a nervous breakdown that reportedly led him to walk down a freezing highway shirtless and barefoot. Unsurprisingly, Pierce not only never worked on a game again, but never spoke to anyone at On-Line for over two decades.
    • All this, and yet there was no way the team could make the Christmas deadline. Instead, On-Line settled on a March 1982 release, at an ambitious and unheard-of price of $99.95 based on the sheer number of locations. Adjusted for inflation ($239.96 in 2012, when the blog articles were written), this makes Time Zone the most expensive single computer game ever made, and the game indeed became a flop largely because of its price tag. One of the few team members remaining afterward was Stephenson, who became the lead programmer because of his leadership, and thus eventually the head architect of the AGI and SCI engines that would power On-Line/Sierra's future and greatest hits, beginning with Williams' King's Quest two years later. Much of the rest left the industry altogether, happy not to work on another game.
    • Worst of all, however, was what happened to Bob Davis. He had a long past of alcohol and drug abuse, but had become clean by the time he worked for On-Line. Unfortunately, by the end of Time Zone's production, the money he made with Ulysses and the Golden Fleece led him to fall off the wagon big time, and soon he quit On-Line with the ambition of making his own games to sell to publishers. Alas, he could never do this without development tools to work with, even when sober... and even if he hadn't tried working with the Atari 2600, one of the most infamously difficult platforms to code for. He lost his shirt and his marriage when the royalty checks dried up, he was reduced to constantly calling On-Line to try to get hired again (or, increasingly, to try to get money), and he ultimately ended up in jail after burning bridges all over his hometown by writing one bad check after another. His downfall would taint On-Line's memories of Time Zone's production, and haunt the company for many years after.note 
  • The Tomb of the TaskMaker, the sequel to the mid-1990s Apple Macintosh RPG TaskMaker, ended up being the undoing of its publishers, Storm Impact. After the success of Storm Impact's first two products (the first TaskMaker and a skiing game called MacSki), the company's next two products (a debug program called Technical Snapshot and a space game called Asterbamm) both failed to catch on. A publisher who kept losing money orders didn't help, neither did the rush to get Tomb out on time — the game has a huge number of cut corners, and version 1.0 just barely got out before Storm Impact went under. See more information here (Wayback Machine archive). The deciding factors were undercapitalization (Storm Impact was mostly just two guys), an inability to get the product out on time, a declining Mac market at the time, and considerable advances in computer gaming since the first TaskMaker came out.
  • The making of Core Design's Tomb Raider games was a chain of troubled productions, as covered by a number of retrospectives, that turned Core's biggest success story into something many of the minds behind it grew to hate.
    • Tomb Raider I was made by a Ragtag Bunch of Misfits put under the leadership of Toby Gard, most of whom lacked experience in 3D games and had to deal with a bevy of challenges, including level creation software that was prone to crashing and Eidos Interactive pushing against having a female protagonist. They also had to scramble to put together a Sega Saturn port six weeks earlier than planned after Core co-founder Jeremy Heath-Smith inked a timed exclusivity deal with Sega, forcing the team to work fifteen hours a day for seven days a week to get the game done on time.
    • Tomb Raider II had the team immediately sent back in the trenches as executives demanded yearly games to capitalize on the popularity of the franchise. Toby Gard and lead programmer Paul Douglas, already disgusted by the sexualized marketing of Lara Croft, balked at the deadline and left the team, turning down over a million pounds in royalties. The game was built by a mixture of old and new talent in the span of eight months, with brutal crunch from the onset that saw developers sleeping under their desks and breaking relationships with their families. The first batch of demo disks sent to Sony were unplayable due to bad copy-protection, which the team hastily fixed after Sony "went nuclear" over the incident.
    • With Tomb Raider III, the team wanted two years to make the game but Eidos insisted on yearly releases, which nearly caused them to walk out until a compromise was made where they would help train a new team to continue the franchise while the veterans moved on to make Project Eden. While the game had three more months of development compared to II, it was still hectic enough that early copies had a Game-Breaking Bug in the Temple Ruins levels. While the PC version addressed this with a downloadable patch, new disks had to be shipped for the PlayStation version of the game.
    • Tomb Raider: The Last Revelation saw the burnout truly set in, with many who worked on the games admitting that the generous royalty payments were the main motivator at this point. The atmosphere at Core was also growing more toxic, with Core's other games falling well behind the Tomb Raider titles in critical or commercial success, creating resentment toward the Tomb Raider developers. Fed up with the franchise, the team decided to kill off Lara hoping it would get them out of making any more games. Upon seeing the ending, an irate Jeremy Heath-Smith confronted the developers but it was too late to make changes, though Eidos took the decision in stride... as long as more Tomb Raider games continued to be made.
    • Tomb Raider Chronicles continued the burnout as franchise fatigue began taking a toll, with each new game slowly declining in overall sales. Legal troubles hit when the character of Jean-Yves from The Last Revelation got the attention of real-life French archeologist Jean-Yves Empereur, who found the character's resemblance to him to be more than coincidental, forcing Eidos to publicly apologize and Core to hastily cut the character out of this game. By this point, the exhausted team was simply going through the motions, with designer Andy Sandham outright calling the game "a load of old shit" that nobody at Core had passion for. The end result was the worst-selling game in franchise history at that point, and even the positive reviews voiced concerns that the series had become Strictly Formula. Its development also cut into resources available for The Angel Of Darkness.
  • Tomb Raider: The Angel of Darkness was beset by a litany of problems that began the moment the game entered development.
    • With Core Design's main team working on Tomb Raider Chronicles, an underfunded new team was put in charge of the Tomb Raider series' next-gen debut on the PlayStation 2. This team had no experience with the PS2's exotic hardware, and spent a year just trying to figure out how to work with it. When Richard Morton, the lead programmer on Chronicles, finished that game and went on to bring his team to work on The Angel of Darkness, he was mortified at what he saw. He and his team had to dump the entire year's work and start from scratch.
    • The game was essentially made by committee. The senior management at Core was seeking to play Follow the Leader with the big games of the time, and told the team to incorporate Metal Gear Solid's stealth gameplay, Shenmue's character interaction, and RPG Elements, among other features. They also commissioned a Darker and Edgier 'epic' story that would span multiple games and take place mostly in modern city environments, as opposed to the tombs and lost civilizations that were the series' hallmark. None of this went over well with the developers, who had enough trouble getting the game to work properly on the PS2 hardware without an unreasonable list of features on top of it.
    • The public first became aware of these problems at E3 2002, when Eidos and Core showcased 'playable' levels that were little more than tech demos, showing that the game was nowhere near a state where it could be finished by its Fall 2002 release date. Another demonstration at a buyers' conference saw Core co-founder Jeremy Heath-Smith cursing on stage over the buggy state the game was in. The game would be pushed back to June 2003 as Core cut the game to pieces to make their deadline, leaving behind a tangle of plot holes, inconsistencies, and unfinished game mechanics.
    • The final product was strongly criticized for being riddled with glitches and unfinished features. All the involved parties wasted no time in pointing fingers at each other for the fiasco, with even Paramount blaming The Angel of Darkness for the poor box office of Lara Croft: Tomb Raider – The Cradle of Life. Jeremy Heath-Smith was ousted from the company shortly after release, while Eidos would be bought out by SGi Entertainment in 2005note . Core limped onward by developing games for the PlayStation Portable, until Eidos sold off Core Design to Rebellion in 2006 and handed Tomb Raider to Crystal Dynamics, who rebooted the franchise with Tomb Raider: Legend. Despite its reputation, The Angel of Darkness became a Cult Classic within the fandom for its dark story and ambitious design.
  • Silicon Knights hit its creative nadir with CEO Denis Dyack's pet project and self-considered masterpiece Too Human. It took second place to Duke Nukem Forever as the king of Vaporware: announced in 1999 and released in 2008, having effectively existed in three incarnations on three console generations (PlayStation/Sega Saturn, Nintendo GameCube, and the final released product on Xbox 360). It was cancelled the first time around by the original publishers (a partnership between Electronic Arts and Metro-Goldwyn-Mayer) for reasons unknown to even the developers, and sidelined the second time around due to too much workload with Nintendo. The reason why the Xbox 360 was ultimately chosen? Because Dyack, a man of graphics and technology, was utterly disgusted by the Wii hardware's lack of horsepower and immediately burned bridges with Nintendo, believing that the Wii would be a failure.

    This proved to be a fatal mistake, for without Miyamoto and Iwata to keep watch over him and his work, Dyack's ego spiraled out of control and he grew progressively more hostile towards virtually everyone, including his own staff. Development on the game suffered dearly for it. This resulted in mediocre sales and reviews, a bunch of projects left in the garbage bin over the next few years and most importantly of all, a disastrous lawsuit against Epic Games. Silicon Knights paid a license to use the Unreal Engine 3 for Too Human, didn't like it, and began reverse-engineering it until they decided the engine was now 100% Silicon Knights and 0% Epic. They stopped payments to Epic and demanded refunds for the money they already paid to them. Problem was, the "new" engine used in the final product still contained portions of Epic code... and Silicon Knights lost big time and paid the ultimate price in 2013.
  • Turok: Evolution ended up killing that series and contributing to the downfall of Acclaim due to troubles in development. Initially the development team intended to produce a direct follow-up to the events of Turok 3: Shadow of Oblivion, but the underwhelming critical response and poor sales of that game resulted in them deciding to instead make a prequel starring the first game's protagonist, Tal'Set. Development went smoothly for the most part, though the team were placed under the same two-year schedule as with the three prior games, which didn't help when they also had to transition to the new GameCube hardware, and then add in a version for the original Xbox to maximise potential sales. Thing really started going wrong near the end of development, when Sony started making thinly veiled threats towards any publisher who was thinking of releasing a game on the other two consoles, but not the PlayStation 2, resulting in Acclaim grudgingly producing a quick-and-dirty PS2 port, figuring that the franchise's previously Nintendo-exclusive status would mean the game would sell the most on the GameCube. Which ended up going completely pear-shaped when sales of the PS2 version eclipsed those of the other two — the GameCube version actually ending up selling the least of the three by some distance — meaning that the enduring image that most were left with was of an ugly, slowdown-riddled, barely functional mess of a game. The other two versions, while not being reviewed as poorly, were still affected by this mess, as the need to produce a PS2 port cut badly into the time the developers had to test, debug and optimise the game.
  • The Typing of the Dead: Overkill was developed under constant threat of cancellation: it began development with a deadline of four months at Blitz Games Studios, which went bankrupt before the game was completed, taking down Sega's original contract from the game with them. Afterwards, the developers negotiated for a new contract to release the game, which they got... with a measly deadline of six weeks. Due to not being employed, the staff had no choice but to become bedroom coders in a crowded apartment; fortunately for that one, Sega saw the location situation and granted the staff temporary office space at one of their subsidiaries. That it sold well with this history is a complete miracle.
  • Ultima IX: Ascension spent 5 years in development — and it shows. As noted on Hackl's Ultima Page, "The Conquest of Origin" and its Executive Meddling entry, the game was fraught with problems behind-the-scenes:
    • Lead creator Richard Garriott originally planned to have IX pick up exactly where the previous game left off (the Avatar confronting the Guardian on the latter's homeworld). However, after realizing that the fanbase likely wanted a return to the series' roots, he scrapped the idea. Shortly afterwards, in 1997, the team came up with a new concept for the game, using an isometric 3D engine. However, this idea was scrapped when the rise of 3D accelerators forced the team to jettison their existing project and begin another new version of the game.
    • During this time, Don Mattrick (later of Microsoft's Xbox brand), who was president of EA Worldwide Studios at the time, assumed control of the division. Mattrick pushed the development teams to stick to their schedules, and began cancelling Origin's projects for unexplained reasons (or outright killing in-production titles). As a result, Origin began to move into the online gaming sector.
    • Garriot went to EA CEO Lawrence Probst to get funding for Ultima Online, and it was initially treated as near-irrelevant by the publisher. However, after said game's beta test drew 50,000 volunteers, EA insisted that Garriott shelve Ascension and work solely on Online...
    • ...and that's when everything fell apart. The original lead designer, Robert "Bob" White, fed up of EA's meddling by this point, led virtually the entire development team in walking out of Origin in favor of Warren Spector's Ion Storm Austin, where he co-designed Deus Ex, leading the "Western RPG faction" opposite Harvey Smith's "Immersive Sim faction". Garriott, under pressure from EA to produce an Actionized Sequel, hired a veteran of the Command & Conquer series as the new lead designer, but was forced to fire him after six months when it turned out he really didn't understand the series. This left Garriott himself having to take over as lead developer again (a role he hadn't occupied since Ultima VII, six years prior) and hastily having to cobble together the game from whatever bits of code and artwork they could find. In the rush to finish Ascension for retail release (after EA set a firm deadline when Garriott wouldn't acquiesce to their original demand), many shortcuts were taken. As a result, the engine was bug-ridden, the team didn't have the time to fully implement the plot from the isometric version of the game into the 3D version, and cutscenes and dialogue were left unfinished.
    • Things finally came to a head afterwards, during pre-production on what was to become Ultima X: Odyssey. EA cancelled all of Origin's planned projects (including Ultima Online 2) and forced relocation of Odyssey's development from Austin, Texas to California, leaving developers who couldn't make the move due to family issues out of work. This subsequently led to the project being scrapped altogether, and Origin eventually being disbanded a short time later.
  • Uncharted 4: A Thief's End, the second-to-final installment in the critically-acclaimed Uncharted series, wasn't exactly smooth sailing for developer Naughty Dog. In March 2014, it came to light that series writer and creative director Amy Henning, along with game director Justin Richmond, who worked on Uncharted 3: Drake's Deception and — at the time — Uncharted 4, had left Naughty Dog to work at Visceral Games and Riot Games, respectively, citing Creative Differences with Naughty Dog. Later, Neil Druckmann and Bruce Straley, the game directors for The Last of Us , were revealed in June 2014 as taking Henning and Richmond's place. Conversely, this meant that some plot ideas and eight months of shooting were scrapped. Todd Stashwick was also set to play Drake's brother Sam, but he was later replaced by Troy Baker. As a result of this, the game missed its 2015 launch to ensure extra development time, and after several small delays was finally released in May 2016.
  • Unsung Story: Tale of the Guardians was a kickstarter game that had loads of promise, talent, and potential behind it. But instead, what resulted was loads of false promises, miscommunication, and delays.
    • The game was created by Yasumi Matsuno as a spiritual successor to Final Fantasy Tactics as a partnership alongside board game company Playdek with a goal of $600,000 and a release date of July 2015. Problems instantly arose when the game couldn't make its required funding in time despite the big names attached (including composer Hitoshi Sakimoto and artist Akihiko Yoshida). That is until a Kotaku article by Jason Schreier got people's attention and allowed it to reach its goal in time (something Schreier would later come to regret).
    • After a few mundane backer updates, backers were getting concerned when Playdek was hyping up the digital card game rather than the actual game. The company's PR manager had to calm down the fanbase a few weeks later making some promises, such as an update in August 2014 (which didn't happen, and was the first of many broken promises).
    • A year after the game received its funding, development became slow, until an update showed off some work-in-progress screenshots of the game, which were poorly received, looking like something made in a few minutes on the cheap. On September 2015 (a month after the game's expected release date), Playdek CEO Joel Goodman apologized for the slow development (a contradiction to the previous month's update, which talked about how smooth development was), and explained that the game was being delayed due to trying to find an outside publisher for the game, and Playdek suffering from a financial crunch. Out of nowhere, Joel Goodman promised an online PvP multiplayer mode, something that not only wasn't promised, but also wasn't what backers wanted. On an AMA session on Reddit, Goodman confirmed that Matsuno jumped ship, despite the game being far from complete.
    • In October 2015, gameplay footage was finally shown, and, well, to say fans were disappointed would be an understatement. The footage contained slow gameplay, and amateurish graphics that look no better than a mobile game. Info about the gameplay was going to be released immediately after, but Playdek, as always, went silent.
    • Four months later, Playdek delivered some bad news. Development was facing more setbacks such as financial issues and losing key staff members; worse, development on the game would be put on hold so that Playdek can create other games in order to secure a proper budget, even requesting an outside development team for assistance. Backers requested refunds, but to no avail. Playdek later announced that they have gotten a budget for the game, and that they have already spent $1.5 million on development. A playable beta build was promised for October 2016 (which ended up not happening, and would never be brought up again). In September 2016, it was announced that Unsung Story would be turned into a franchise thanks to a partnership with a big development company (sound familiar?).
    • No updates would be announced until August 2017 (the last one being in December 2016), where Playdek announced their biggest bombshell: they would be leaving development of the game, throwing the game in publisher/developer Little Orbit's hands. Little Orbit had to start development of the game from scratch and pay for it out of their own pockets, this time focusing on the single player aspect of the game, and promising to deliver on the backers' interests and requests.
  • Uru: Ages Beyond Myst nearly proved to be a Creator Killer for Cyan Worlds. After the success of Riven in 1998, Cyan began work on what was originally called 'Project Mudpie', a Massively Multiplayer Online Puzzle Game allowing players to cooperate to solve puzzles in the cavern of D'ni.
    • The game was originally intended to be multiplayer-only. However, when the game was to be pushed for sale, the multiplayer component wasn't ready, so a handful of the ages were retooled for single-player solving and released under the Ages Beyond Myst title, with the multiplayer subscription component called Uru Live launching later.
    • At the beginning of 2004, Ubisoft decided that the game wasn't selling well enough, so they pulled the plug on the multiplayer servers. This led Cyan to do some more retooling and release some of the locations that were available online only in the expansion pack To D'ni, and some that weren't yet available in The Path of the Shell. More assets intended for Uru Live were later used in Myst V: End of Ages, including the age from which the very first screenshot of Mudpie was taken. Meanwhile, the fandom launched Until Uru, a fan series of online servers, with Cyan's blessing.
    • In 2006, Cyan partnered up with Gametap to provide official multiplayer servers again, and to continue the online story, under the title Myst Online: Uru Live. Since official servers were going to be available, fans were asked to shut down Until Uru. More content was made available than had been previously released. This lasted until February 2007, at which point Gametap also pulled out.
    • After this, Cyan started hosting their own server, donation-run and free-to-play this time, under yet another title, Myst Online: Uru Live Again. However, the damage was done, and besides another remastered version of the original Myst, Cyan was only able to produce a couple of mobile games, and at one point were reduced to a production team of only a handful of people, before risking everything on a crowdfunding scheme for Obduction which was a great success and saved the company from closing altogether.
  • Vampire: The Masquerade - Bloodlines was the swansong for Troika Games, thanks to Troika's ambition, Valve's unfinished technology, and a side of Executive Meddling from Activision Blizzard.
    • Troika and publisher Activision became the first to license the then in-development Source engine from Valve, under the condition that Bloodlines would not release until Half-Life 2 was finished. As Half-Life 2 was expected to release in 2003, Bloodlines was also expected to release that year. But the unfinished state of the engine left Troika on their own for implementing important engine-level features, including AI, particles, and lighting. Troika was new to developing fully 3D games with real-time combat and stuggled with gameplay elements, grossly underestimating the amount of work that a full-3D first-person RPG would entail. Troika also had their small studio split between developing Bloodlines and The Temple of Elemental Evil and lacked a cohesive vision, which further harmed development.
    • By the time the game was officially announced at E3 2003, Activision was concerned about the pace of development; Troika had replaced the writing staff on the game and appeared to be a long way from their expected "late 2003" release date. Activision sent industry veteran David Mullich as a producer to oversee development. Seeing the messy and unfocused design of the game, Mullich worked with the team to focus the project and cut features holding up development. And with Half-Life 2 having its own development troubles that forced it to be pushed back over a year, Bloodlines too was pushed back to a tentative "early 2005" date.
    • But having been in development for over three years with no end in sight, Activision's patience ran out; after giving Troika the funds to complete Temple of Elemental Evil and bring its team onto Bloodlines, they gave Troika a number of must-hit milestones and demanded the game be ready for release no later than September 2004. Troika crunched to get the game presentable by that date and cut several completed features at the last second due to a lack of time to properly test.
    • Activision completed its internal testing after receiving the game and declared it ready for release, and then kept it from being released until Half-Life 2 was released in November 2004 as per Valve's contract. Troika narrowly convinced Activision to allow them to continue work in the meantime, but Activision only allowed a small budget, Troika employees worked, many of which were now going without pay, to patch the most important issues prior to the game's release.
    • While reviews on Bloodlines praised the world, story and characters, the game's many, many bugs led to mixed scores. But any chances of commercial success were instantly killed when Activision released Bloodlines on the same day as Half-Life 2. Lead Writer Brian Mitsoda and Creative Director Jason Anderson would blame Activision of shoving the game out at the worst possible time, and Mitsoda would claim that most comsumers weren't even aware the game had released. Troika would release one more patch for Bloodlines, and shut its doors in early 2005. Bloodlines would by Vindicated by History thanks to unofficial community-made patches that restored cut content and fixed bugs, and became enough of a Cult Classic that Paradox Interactive would greenlight a sequel over a decade later.
  • Vampire: The Masquerade – Bloodlines 2, had its own development hit by major troubles, including repeated delays, a revolving door of creative leads and a change in developers.
    • It was announed early on that Chris Avellone, long respected by western RPG fans for writing games like Planescape: Torment and Fallout: New Vegas, had been brought on by developer Hardsuit Labs. This helped sooth fears that the game's writing would not live up to the original... until 2020, where Avellone was among a slew of industry figures accused of sexual harassment and assault. Paradox wasted no time cutting ties with Avellone, stating that his contributions would be immediately removed. Reports vary about how much he had contributed to the game; Paradox officially stated it was minor, but Avellone himself later released a statement showing that he wrote a significant portion of the game, all of which was reportedly thrown out.
    • Bad news continued through 2020 as Lead Narrative Designer Brian Mitsoda, a lead developer from the first Bloodlines who had pitched Bloodlines 2 in the first place, was abruptly let go (to his own frustration and surprise), along with Creative Director Ka'ai Cluneynote  after putting five years of work into the project. Ironically, Mitsoda had been the central figure throughout 2's entire marketing campaign; Paradox used his person extensively to advertise it at cons and engage consumers. The official word is that his work remains despite his departure, but this disheartening news had a souring effect in the Bloodlines community all the same. Ka'ai was publicly replaced around the same time by Alexandre Mandryka, who worked on projects like Assassin's Creed, Prince of Persia: The Two Thrones and Warhammer 40,000: Space Marine, a resume which some saw as ill-fitting the narrative focused RPG promised.
    • Cara Ellison, who had replaced Mitsoda as Narrative Designer, also suddenly departed the project in October 2020, further fueling fears of development going horribly awry. Ellison was replaced by Mass Effect: Andromeda writer Samantha Wallschlaeger.
    • Then, in February 2021, it was announced that Hardsuit Labs was no longer in charge of the project, and the game was nearly scrapped entirely before being handed to a new developer (later confirmed to be The Chinese Room). Paradox pushed the game back past 2022, and suspended pre-orders until a new release window could be confirmed.
  • Vexx, Acclaim Studios Austin's attempt at making their own 3D mascot, faced numerous tribulations that resulted in a comedy of errors production and the game's relative obscurity, as told by Matt McMuscles with sources from former members of the studio.
    • The game was set to star two characters, Clip—later Jinx, which they couldn't use anyway due to copyright issues—and Mischief, inspired by Banjo-Kazooie. However, when Jak and Daxter was announced at E3, pre-production had to start over because of the extremely coincidental similarities between the two properties. Mischief would be phased out, and Jinx would be renamed to Vexx.
    • The game being multiplatform ended up being a detriment, rather than a blessing. Acclaim Austin had thought of the idea of an interconnected world and free reign traversal between all levels, something that wouldn't have been able to work on all three platforms due to Sony's mandate regarding all multiplatform ports being equal to each other. While the Xbox could handle this function no problem, the PS2 and Gamecube weren't powerful enough to handle it. With not much time to properly program this design into every port, the game had to be scaled down immensely. Six worlds and eighteen levels were cut down to just nine levels, and many more concepts were either cut or simplified.
    • Large chunks of the story had to be cut due to a particularly nasty contract snafu. When Brian Cox was contacted to voice Darby and Dark Yabu, an error in his contract resulted in Cox getting paid the entire voice acting budget, with no way to reverse it.
    • Upon release, Vexx would receive mixed reviews for its undercooked ideas and would prove a flop with over 20,000 sales across all three consoles. A potential sequel was canned, and Acclaim Austin would later be shut down along with its parent company.
  • Development on the Sega Saturn version of Virtua Racing was somewhat difficult due to a lack of support from Sega, as chronicled in a 2021 video by YouTube channel "PandaMonium Reviews Every U.S. Saturn Game". Instead of developing it internally as with the previous Genesis and 32X ports, Sega outsourced the Saturn Virtua Racing to American developer Time Warner Interactive, who were not given the game's source code, or indeed much reference material beside model data for the tracks and a loaned cabinet of the arcade original. Despite aiming it to be a launch title for the system, Sega did not initially provide them a Saturn devkit either, which means that for the first few months of development, Time Warner Interactive had neither resources for the source game or the target hardware, and until the team received its own devkit in December 1994, one employee had to stay at Sega of America's headquarters and work overnight on their spare hardware. When they did receive one, they found that the original "Sophia" kit was an unwieldly beast that lacked a debugger, so the solution was to carefully remove the CPU from it and emulate it by placing it in an Hitachi workstation. Not all was bad, however, as Sega exerted little oversight during development, allowing the developers to add a generous amount of extra tracks, cars and modes (indeed, still the biggest of any incarnation of Virtua Racing). In the end, Time Warner Interactive's VR Virtua Racing got a cool reception at release due to its controls and physics being inaccurate to the arcade game and its graphics being massively outdated compared to the likes of Daytona USA, and still stands as easily the most divisive release of the game.

    W-Z 
  • In July 2021, it was confirmed in a Jason Schreier report that development of Warcraft III: Reforged, a remaster of Warcraft III: Reign of Chaos, had been fraught with difficulties.
    • After its announcement at Blizzcon 2018, Activision Blizzard executives considered Warcraft III: Reforged to be a low priority, believing it wouldn't sell nearly as many copies as the likes of Call of Duty, World of Warcraft or Overwatch, as the Real-Time Strategy genre had steeply declined in popularity from its heyday in the 1990s. The development team was given a cripplingly low budget that wasn't enough to allow them to deliver on the promises madenote , effectively dooming the project before development even began. Further budget cuts were also made during development, forcing the developers to cut features that were present in the original.
    • The development team as a whole was an undersized one, and the overworked team suffered from depression, anxiety, and exhaustion. There were multiple heated disagreements over the remaster's art style and scope, and the team was headed by Rob Bridenbecker, whose management was resented by the other developers as aggressive and inconsistent.
    • In 2019, Activision and Blizzard issued a mass layoff of over 800 employees across the two companies. Despite 2018 being a record year for the two, they had still failed to meet the internal expected earnings for 2018's fiscal year. The layoffs earned widespread condemnationnote , and the layoffs were later confirmed to have included members of the Reforged development team, leaving them even more understaffed.
    • Pre-orders were opened long before the game was finished, increasing pressure to launch the remaster as soon as possible... which it did in January 2020, as an Obvious Beta that was widely panned. Even the original game was affected by the changes of Reforged, as the original and remaster were merged together; the EULA was changed to give Blizzard exclusive ownership over any and all user-created contentnote , and features were retroactively cut from the original 2002 game to make it compatible with the remaster. As of 2021, Reforged has yet to restore the missing features to either Reforged or the original game.
  • According to a YouTube comment from an anonymous Carbine Studios ex-developer on nerdslayer's Death of a Game review of WildStar, the demise of the MMO came down to a hot mess of behind-the-scenes design issue and office politics that ultimately destroyed both the game and Carbine Studios in 2018.
    • Leadership from the top down was reportedly very poor. Most of the first 5+ years of game development were thrown out completely by project lead Tim Cain. He was also quite difficult to work with according to the report, vetoing ideas he didn't like left and right. There was also intense mandatory 10am-3am, six-day-a week crunch time for several months, during which the morale among the staff plummeted and turnover rate skyrocketed.
    • A large number of issues were also caused by the art director, who "couldn't keep his nose out of anything" to the point of driving off several of the company's best artists - including lead Cory Loftis, who originally designed the entire art style for the game. The same art director was also responsible for hijacking the story and gameplay design as well; the reason for the lack of new raids and dungeons was allegedly because he thought that several (already completed and tested) levels didn't fit the story anymore and threw all of them out.
    • The decision to turn the game into a WoW-esque hardcore raiding game came very, very late in development, when the devs only had a few dungeons and a single 40-man raid ready to go on launch, so they were forced to hastily throw in a long attunement grind involving speedrunning dungeons and killing world bosses in order to access the endgame. This ended up backfiring badly; the resulting attunement grind caused a mass exodus of non-raiders and casual players weeks after launch, while the "hardcore" raiders the endgame was explicitly designed for quickly accessed the first raid, beat it and put in on farm, and quickly ran out of content to do. This was further exacerbated by the buggy launch, as development time that could be used to create new content was instead focused on fixing glitches and bugs.
    • Warplots were an "utterly broken mess" on launch because it was rushed out the door in a barely tested state; the only reason it worked at all was because several people on the 3D prop team actually learned to script just so they could try to fix it.
    • All of the game's PVP systems were mostly tossed into the game with very few play tests; the system barely worked on paper let alone in practice. The "telegraphing" mechanic for attacks worked well in 1v1 fights, but quickly proved unworkable in any fights larger than 2v2, as it was impossible to see any attacks over the multitude of telegraph effects on the floor. Every single piece of feedback on this was completely ignored.
    • Any ideas or issues that conflicted with the vision of the higher-ups were shot-down and most of the more well-liked features of the game had to be done in secret. The Hoverboards mechanic was snuck into the game by a bored animator, who eventually got written up for working on it in his spare time. The Double-Jump mechanic was put into the game by one of the artists, who eventually got fired for it and other reasons.
  • Sir-Tech Software were so confident that Wizardry IV: The Return of Werdna would be available by the end of 1984 that they actually told a magazine to announce it was available in its November 1984 issue. The game wouldn't actually see release until 1987; when it was released, a combination of still using tech from 1981 and ludicrous difficulty even by dungeon-crawler standards made it the poorest-selling product in the company's history. The Other Wiki has more on the story.
  • Wizardry 8 was developed as Sir-Tech was crumbling from serious money issues. The game (originally subtitled Stones of Arnhem) was initially given to DirectSoft, but the partnership fell apart and that incarnation of the game was canned, leading to DirectSoft's dissolution as Sir-Tech rebooted development under their Canada division. Later in development, Sir-Tech's situation grew so dire that they were forced to rush the game out and even sold in-game advertising space to companies. While the final game was released in 2001 to critical acclaim and won several awards, it wasn't enough to save Sir-Tech and they would cease operations in 2003.
  • World Cup Carnival for the Amstrad CPC, ZX Spectrum, and Commodore 64 has one of the weirder stories behind its troubled production. US Gold had won the rights to make a licensed game based on the 1986 World Cup. They farmed the game's production out to Ocean Software. As the game neared its release date, however, US Gold checked on how Ocean was doing... and found they hadn't even started on the game. In a panic, US Gold bought the rights to a completely different and terrible game, 1984's World Cup Football from Arctic Computing, reskinned it, and put it out as World Cup Carnival. The game was stuffed with Feelies is a desperate attempt to make it anywhere near worth the ten pounds US Gold was charging. The gaming press naturally brought the hammer down on the game, but it still sold well due to the FIFA license, becoming the biggest selling release of 1986; U.S. Gold even managed to bury the bad press by releasing Leaderboard Golf the same year. As a bizarre coda, Prism Leisure Corporation sued Artic Computing for copyright infringement, as it turned out Artic didn't actually own the rights to World Cup Football, having sold them to Prism in 1985 to keep the company afloat; the cost of both the lawsuit and having to pay Prism the profits they had made from World Cup Carnival brought an end to Artic Computing.
  • The Sega Genesis port of World Heroes was the work of a single, American first-party studio, Sega Midwest, which proceeded to churn out a Porting Disaster. But going by this developer interview with Jim Reichert, it's a wonder the port was finished at all. The original programmer, supposedly a "wunderkind" at his trade, turned out to be a Con Man who defrauded Sega Midwest out of a huge chunk of money before vanishing. Thus the job was handed to a novice programmer, Reichert, who was forced to write the whole thing from scratch almost all by himself in an extremely short time. There was no one to test the game or help him port the Neo Geo graphics to the Genesis, he had to squeeze 82 megabytes into a 16 megabyte cartridge, and the only source code he had was a post-compilation assembly dump with no documentation whatsoever. He admits the final result was a Porting Disaster, but considers the fact that he managed to fit all the characters and their animations into the game, including the final boss, to be an impressive feat considering the circumstances.
  • The World of Darkness MMO, as detailed in this Guardian article, spent nearly a decade in Development Hell before finally being canceled. Those years had all the makings of this trope.
    • Icelandic developer CCP, fresh off the booming success of EVE Online, bought out a troubled White Wolf in November 2006 hoping to get its hands on that company's lucrative tabletop gaming IPs, chief among them being The World of Darkness. Almost immediately, CCP began production on an MMO based on the game, opening a studio in Atlanta to work on it.
    • Troubles began almost immediately. While CCP, to their credit, kept on most of White Wolf's important staff and trained a stable of programmers and artists to work on the game, The World of Darkness was very much of secondary importance compared to the golden goose that was EVE. Developers wound up frequently poached to work on expansions for EVE, often for months at a time, causing constant delays in production that saw features being planned, partially completed, and then scrapped. During production on the Apocrypha expansion for EVE in 2009, production on The World of Darkness halted entirely as the whole team was put to work on that project.
    • Furthermore, the manager in charge of the project had little in the way of a coherent vision for the game beyond "buzzword-laden rambles", exacerbating the delays and the problems of work wasted on various abandoned gameplay mechanics. Much of this was driven by CCP's corporate culture of "the war on the impossible", the idea that they should strive to outdo all of their competitors and deliver things that nobody had ever seen before in an MMO. These grandiose ambitions led to the EVE spinoff DUST 514, which further cannibalized the staff. Thanks to all of this dysfunction, the game reached alpha stage (i.e. with the fundamentals of its core gameplay mechanics all implemented) a total of three times, with the staff going back to the drawing board for each one as all of the disparate game mechanics failed to gel together. As production stretched out, management frequently attempted to deflect blame for the delays onto the programmers. Design meetings were described by one former developer as "whoever shouted longest and hardest would dominate".
    • In 2011, the Incarna expansion for EVE experienced its own troubled production, causing CCP to bring in the Atlanta team to help finish the job that their main team in Reykjavik, Iceland was struggling with. When the Atlanta team got the job done in a fraction of the time that the Reykjavik team took to do just a quarter of the work, their work ethic was vindicated, but the Reykjavik team was left bitter that they'd been shown up so badly.
    • The beginning of the end came in late 2011 when, between the failure of DUST 514 and the backlash against CCP for introducing microtransactions to EVE, a humbled CCP began cutting costs. Morale fell apart as developers saw cuts to their pay, to their free meals, and to their medical benefits, and twenty percent of the workforce was laid off by the end of 2011. A teaser trailer was released at EVE Fanfest in 2012, but while it got the feel and style of the source material down, it notably didn't contain anything even resembling gameplay footage. Resources were stretched even further in 2013 when CCP announced EVE Valkyrie, a Space Fighter game for the Oculus Rift set in the EVE universe. At that point, its cancellation in 2014, the only product of its development being some cool concept art, was a Foregone Conclusion. CCP eventually sold White Wolf to Paradox Interactive the following year, washing their hands of the failed World of Darkness MMO and driving the final nails into its coffin.
  • World of Warcraft's fifth expansion Warlords of Draenor became infamous for its barren content and flawed gameplay systems, largely as a result of several unfortunate factors all converging on the team at the same time.
    • Coming off of Mists of Pandaria, Blizzard didn't have a solid idea of how to follow up for some time. Their plan was to have previous Final Boss Garrosh Hellscream left alive and out for revenge with a new army, but the concept of an Alternate Timeline uncorrupted Horde was the last of several different hooks that were brainstormed after an unknown length of time.
    • As is usual, a great deal of work for the expansion had been completed by the time it was revealed to the public. Fan response was less divisive than Mists, but a large number complained about how the expansion had almost nothing except fighting evil orcs, coming off of an evil orc final raid tier, and distaste with Time Travel plots. Partway through the publicly-visible Alpha stage, Blizzard reworked several mostly-complete zones from scratch to avert this complaint. Not only did the resulting zones have their own issues, but the time required to do this was devastating, and they seemingly had to spend more time altering later content or scrapping it with no replacement. The Time Travel aspects were also excised, including the expansion's introduction sequence, forcing them to rewrite it entirely. Man-hours intended to go towards a proposed max-level zone at launch and patch content later were instead spent here.
    • Blizzard's intended new title, Titan, had been reshuffled, downsized, then cancelled after years of development around the same time. Many of the employees that worked on Titan were reassigned to different teams around the company, including the one making Warlords. On paper a larger team would get more work done faster, but Blizzard underestimated the time it would take to get the new members acclimated and up to speed with the tools, engine and workflow. This slowed down the team's pace of work until it was too late.
    • A much-hyped feature of the expansion was revamped versions of the game's oldest player character models, which were intentionally bare-bones and dated even at the time they originally released. Developing brand new art assets for the expansion, while also creating twenty player models more-or-less from scratch, was too much for the art team to fully handle. The result was a set of models that while functional and customization-complete, were extremely hit-or-miss. Players cite unpolished animation work (many of which were unaltered from their originals), inaccuracy to the looks of the older versions, and more—which could have been refined or fixed with more time but weren't, and years later likely never will be.
    • Systems-wise, the central features of the expansion received heavy iteration, thanks to Blizzard's gradual realization that the proposed systems were either fundamentally flawed or not feasible in the limited time they had. The expansion shipped with stripped-down versions of its original pitched systems, which were often exploitable or underwhelming and sometimes had glaring holes where a cut aspect would have been. It took months in some cases for fixes to be implemented.
    • Demand for the expansion was higher than Blizzard had been remotely prepared for, which itself was higher than they thought they would need. A heavy marketing push, as well as a perceived 'return to form' for players who wrote off Mists's premise, resulted in a huge spike in returning players that overloaded the servers and rendered content unplayable for days at launch, the game's rockiest launch yet. There were also murmurings of a possible DDOS attack... but whatever the case, even the release was troubled.
    • With the expansion out the door, with little in the way of content and janky systems, Blizzard hardly supported it at all. Patch 6.1 invited mockery for being a big "point-number patch" with barely any content but instead Twitter integration and a selfie system. The second half of the expansion's first raid tier, Blackrock Foundry, had opened just before the patch which placated players somewhat - but as a knock-on effect led to the perception that 6.1 had zero new actual content. But seven months into the expansion, Hellfire Citadel dropped—not a middle raid, but the final one, hastily wrapping up the story and leaving players with only two full raid tiers and the game's longest end-of-expansion content drought. It's clear that Blizzard realized it was too late to fully salvage Warlords, and instead of keeping players in a broken expansion for longer, put those man-hours into the next expansion Legion.
  • Following the disasterous launch of their remake of the 2003 first-person shooter XIII, French developer PlayMagic and publisher Microids issued a joint statement blaming the COVID-19 pandemic and at-home working for the game's technical issues. Several anonymous PlayMagic employees thought otherwise.
    • PlayMagic was notorious for regularly crunching on work-for-hire projects. CEO and creative director Giuseppe Crugliano allegedly fought with employees, made unrealistic promises to publishers, and repeatedly fell behind on payments. These problems killed the company's mobile VR adaptation of Arktika.1, a first-person shooter by 4A Games, in June 2018. Many employees left afterwards, resulting in the XIII remake — which began production four months later — having a small team.
    • Crugliano originally wanted to remaster the original game in three months with Unity plug-ins. However, his team had access to only a few assets, so the project became a remake. The sheer amount of work to pull this off required a larger team. Crugliano was allegedly blacklisted from recruitment agencies for non-payment, so hiring new talent was difficult. The team had no character animators until about halfway into development.
    • In December 2019, PlayMagic allegedly took on a remake of Stubbs the Zombie, eating up resources meant for the XIII remake. This project was cancelled after about roughly half a year of work.
    • Programmers had to test their own code because PlayMagic lacked a dedicated QA department.
    • The first teaser trailer had a release date of November 13, 2019. However, the game was delayed to November 10, 2020 to reach "the level of polish [PlayMagic] is working for"—a statement that employees claim was false, as the game wasn't even halfway finished at the time.
    • At launch, the XIII remake was critically savaged and sold worse than the original game. Over a quarter of the credited staff left PlayMagic, with Curgliano allegedly harassing them in retaliation. Two years later, Microids replaced PlayMagic with Tower Five, who released a free update that significantly improved the remake.
  • Two conflicting accounts on the difficult development of X-Men: Destiny exist. Whichever rings more true is, as of the end of 2015, up to the reader's discretion.
    • According to Silicon Knights chief Denis Dyack, licensing problems with Marvel Comics threw a wrench in the game's budget. The game was intended as a major AAA release by then-Marvel license holder Activision, with Silicon Knights pumping even more money into the game's budget than its employees' salary. But no one expected Disney to buy out Marvel during development, and with Disney publishing games themselves, Activision and Disney argued over the former's Marvel game license; the budget kept shrinking, and since Silicon Knights wasn't paid by Activision during negotiations, they started funding development out of their own pockets. Because Activision's Marvel contract was "so complicated and detailed to unravel", Disney couldn't do anything to help the project, and the game was released as it was.
    • Meanwhile, a few anonymous ex-employees claimed that Dyack's ego turned nasty after Silicon Knights broke ties with Nintendo, and insisted on taking absolute control of game development and throwing down with anyone who disagreed with him inside and outside of the company. Dyack, who was said to not be a fan of the X-Men and thus didn't care for Destiny effectively procrastinated on production of the game as much as possible for about two years by stalling communications with Activision and shuffling manpower onto his pet projects, like a sequel to Eternal Darkness. Nothing came of Dyack's projects and it became harder to get more funding and time from Activision. Eventually, Activision got fed up and retaliated by publicly announcing the game about a year before its release date with no possibilities of delays while development was still severely messed up, forcing Silicon Knights to crunch under a 12-month deadline.
    • According to this video, Silicon Knights just didn't want to make an X-Men game because Dyack hated the idea of working on licensed games and he was not a fan of the franchise. However, with SK in the dumps, they were not in the position to negotiate. Even more, Dyack's attitude towards the game, seeing it as more than as a necessity than a labor of love, trickled down into the other developers, who began to feel the same way outside of those who were fans of the franchise. An example of this was given by Assistant Director Julian Spillane, who was asked to preview a boss battle against the Juggernaut, which was famed as a simple beat-em-up boss fight. Spillane, realizing how wrong this was, instead suggested changing it up so that the player character could remove the Juggernaut's helmet and have Emma Frost use her telepathic powers to damage him. He was told he was worried about "little details" and sent the build to Marvel and Activision; Marvel sent them all copies of X-Men Marvel Encyclopedia.
  • Development on Yandere Simulator began in 2014 by its sole programmer, Alex "YandereDev" Mahan, and ever since then has trudged through a very lengthy (and rather stressful) production.
    • YandereDev had always wanted to garner a fanbase from something of his creation. He ended up getting his wish, and it far exceeded his expectations; on a daily basis, he was flooded with emails and comments, all of which he tried reading (which consequently gave himself less time to develop the game). YandereDev quickly became loath to reply to fan emails and comments. Furthermore, he issued constant reminders to discourage his fanbase's misguided efforts to assist (including unhelpful bug reports, offers of help without a proper resume, and suggestions that were impossible, already considered, or outright stupid)note . The fact that YandereDev was forced to do so on the game's two-year anniversary and created an Audience Surrogate character based upon those who annoy him speaks volumes.
    • As of January 2016, Twitch banned the game and threatened to immediately suspend anyone caught streaming it. YandereDev stated in a video that he was provided no explanation for this, arguing that the violent and questionable content in Yandere Simulator can be seen in other games allowed by Twitch. An official response finally came in February 2017, stating that things in early builds like the naked Titans and violence against minorsnote  had violated guidelines. While later versions have rectified most of the complaints, Twitch has nonetheless refused to re-review the game until it is closer to completion.
    • Numerous schedule slippages — all of which documented on the game's Trivia page — plagued development with delays. Among the reasons include YandereDev's hard drive dying on January 10, 2016 (which fortunately didn't delete progress) and a trip to Anime Expo later in July 2016. YandereDev projected that the game would be released in 2015 with a Kickstarter project for higher production values and to get it on Steam. The game is still at its alpha state since 2020.
    • Money troubles were apparent, as the $5,000 a month from Patreon that YandereDev received wasn't enough to hire volunteer animators or riggers who were competent (at least, not while also paying voice actors and 3D modelers). This in turn made implementing the first rival excessively difficult, who was incompatible with the animation rigs YandereDev currently had.
    • Despite his best efforts, YandereDev continued falling behind on updates, and on February 2017 finally had a Creator Breakdown. A video revealed that, for the past two years, YandereDev had sacrificed most of his personal life and health to work on the game 24/7—all while operating under fears of under-delivering fan expectations and the game dying from a subsequent loss of interest. Additionally, he called into question his own programming skills, expressing interest on hiring a more experienced programmer to help refurbish the game engine. Thus, YandereDev ended up partnering with tinyBuild for further development on the game... except that plan ended up getting scrapped a couple of months later due to complications involving rewriting the game's code and adding features at the same time.
    • After the deal with tinyBuild collapsed, the source code for the game was leaked, revealing to many why YandereDev had questioned his programming skills in the first place; a lack of advanced programming methods led to a more complicated than necessary code, thus bogging down the game's framerate and performance. In addition, with the first rival being continously delayed, YandereDev's habit of focusing on relatively minor content began to be more openly scrutinized and regarded as Skewed Priorities.
    • The hectic production resulted in a growing number of irate ex-fans who in mid-2020 responded by hacking the game's official subreddit, revealing extremely unsavory information about YandereDev in regards to his beliefs on sexual consent and mental illness, and endorsing the rival product Love Letter: My True Feelings. The one-two-three punch convinced YandereDev that he would have to deliver fast, and so he finally officially released the first rival. The results showed clearly that it was rushed to avoid the threat of competition: framerate, texture and pathfinding issues were in high supply, and bug fixes were being churned out daily. Love Letter soon crashed and burned itself in the wake of sexual misconduct allegations against its project leader DrApeis, and the alpha build of Yandere Simulator eventually became stable enough to cease daily updates.
    • But the damage had been done, as the Obvious Beta nature of the alpha build ended up with even more jaded fans and volunteers who abandoned the game's development in response to not only the botched release, but also YandereDev's controversies being brought into the spotlight, not helped by more scandals, most surrounding his growing inability to handle criticism and poor showings on social media piling on even after the subreddit was reclaimed (with one such scandal regarding a Discord moderator's dissociative identity disorder causing a prominent volunteer to withdraw support for the game). While YandereDev continued to implement new features and occasional bug fixes, the fear that the game will become Vaporware continued to persist.
    • Yet the problems were not over, as in October 2021 YandereDev attempted to dissuade fears that he was incapable of handling 10 rivals (which he had planned from the beginning of development) by releasing 1980s Mode, complete with its own batch of unique rivals. Once again, the new mode launched with a plethora of severe glitches despite his insistence that he tested the game beforehand, most notoriously save file corruptions. Almost the entirety of October was dedicated to bug-fixes, and many continued to be skeptical regarding Yandere Simulator's future.
    • In September 2023, the game's development hit a significant roadblock for a rather disturbing reason; YandereDev was accused of sending sexual messages to a 16-year old girl, followed by pressuring her into retracting her claims, downplaying the incident, and attempting to prevent the release of a compilation of evidence documenting the accusations that she was initially cooperating with. Things took a turn for the worse when the aforementioned compilation of evidence was quickly copyright claimed by YandereDev — under the alleged victim's real name and doxxing her in the process, a move that increased speculation from the general public that he was indeed guilty of what it accused him of as they questioned why he would go to such extremes to avoid people from seeing the evidence unless it genuinely incriminated him. This merely lead to the compilation being reuploaded by multiple other users and created a massive furor on social media such as Tumblr, Twitter and Reddit, that was enough for a mass exodus of staff members, particularly voice actors such as Michaela Laws and Austin Hively (who voiced the main characters) from Yandere Simulator's development, several YouTubers who had frequently played the game to announce that they would now boycott it, a rather gradual decline of donations to YandereDev's Patreon, Glowstick Entertainment to voluntarily delist crossover DLC for Dark Deception: Monsters & Mortals, Sad Panda Studios to likewise scrub all traces of Yandere Simulator from Crush Crush in favor of a Suspiciously Similar Substitute, and for word of an apparent police investigation into YandereDev to quickly spread. This finally convinced YandereDev to make a blog post as an apology and donate to the anti-sexual violence organization RAINN, which was regarded at best as a token effort to defend himself from the accusations, and at worst a thinly-veiled confession. Afterwards, YandereDev would announce that he would take an indefinite hiatus from development and additionally lower the game's scope once he resumed, only to almost instantly go back on his word and announce a rather large update implementing a custom mode for release in 2024, which did little to dissuade people from predicting (or even hoping) that YandereDev would be outright arrested and/or Yandere Simulator would eventually be cancelled, ending its development on a sordid note.
    • During the final days of December 2023, things went even more downhill once a victim statement purportedly denying the September 2023 allegations was posted — only for evidence to surface that portions of it had been written by both YandereDev and his then-main composer Cameron "CameronF305" Fields. This led to a second round of grooming allegations against both of them, including but not limited to the latter sending the victim explicit imagery, and the former having knowledge of her address and making plans to meet her. On New Year's Day 2024, YandereDev would attempt to dispel the controversy with a video in which his key piece of evidence was the aforementioned victim statement, which met a mixed to negative reception which resulted in him rendering it private a few days later, and the victim statement being outright deleted. Occurring around the same time was the release of the custom mode — which yet again had numerous glitches requiring daily bug-fixing builds, followed by substantial work being done on the second rival, which faced criticism for the high amount of corners cut due to the aforementioned loss of professional voice talent.
    • Around the same time, accusations were made of CameronF305 making legal threats towards people attempting to create videos documenting the existing allegations, before a former voice actress for Yandere Simulator provided screenshots and audio recordings from both herself and the initial victim supporting more abuse and outright rape allegations against him at the beginning of February 2024. He too would make a 3-hour long video attempting to debunk said allegations a few days later, only for it to face scrutiny over accusations of Manipulative Editing, focusing more on negatively portraying his accusers, and filtering comments to silence criticism. In April 2024, however, CameronF305 decided to collaborate with a rival Yandere Simulator subreddit, who promptly cleared up and properly debunked the allegations against him, to a cautiously optimistic response by its users... around the same time that recordings of YandereDev soliciting explicit material came to light. With the game still nowhere near complete and the massive controversy regarding the persistent sexual misconduct allegations, YandereDev's responses to them, and CameronF305 getting caught in the crossfire, outside of its remaining diehard fanbase, Yandere Simulator's cancellation has become not a matter of if, but when.
  • The Yogscast video game Yogsventures! met an untimely end because of this, and has since been regarded as one of the biggest failures in Kickstarter video game projects. Read about it here.
    • Winterkewl Games started a Kickstarter campaign to get support from fans, which resulted in $500,000 dollars in funds raised. However, two weeks into the project, the game's artist abandoned it to work with LucasArts, who didn't give him any accessibility in his contract to spend time on working on Yogsventures!. This was a huge setback for Winterkewl, as they already had a contract that guaranteed each of the main artists a $35,000 lump sum payment and they made no clear clause as to how and why someone could legally stop working on the project.
    • In wake of this, Yogscast co-founder Lewis Brindley grew hesitant to trust Winterkewl leader Kristafer Vale. The development team ended up paying Yogscast $150,000 to create and ship the Kickstarter campaign's physical rewards to Kickstarter backers.
    • The money also went towards hiring a main programmer for Winterkewl, who still didn't have one by this point. When these efforts fails, Vale himself took on the role and worked on the game for 18 months.
    • Various design changes and Vale's inexperience as a project lead and developer ultimately marked the end of the line for Winterkewl, who, by this point, had paid so much money for licensing fees and finding a programmer that they had little left to continue the project. Winterkewl ended up going bankrupt and announced the cancellation of the game. Yogscast also officially announced this bit of news as well, calling the whole thing a failure that was "a deep regret". Out of sympathy for the tens of thousands of people who had funded the game and had no chance of getting any money back, Yogscasters gave out free copies of Landmark, a massively multiplayer online RPG (which would itself shut down in January 2017).

Top