"If a statistician hands you a die insisting that 'any given roll has the same odds of rolling a one or a twenty', it means he's handing you a depleted die in the hopes of taking advantage of you. Don't fall for it!"
A statistician can tell you that properly generated random numbers will follow a discrete uniform distribution, giving you a roughly equal chance of rolling very good or very bad numbers. They may wax eloquent about how pseudo-random numbers are generated in computers, and how dice are wonderful randomizers. A good statistician will even disabuse you of the notion that a six sided die has a 1/6 chance to land on any one side; most dice are not built to that kind of precision. (Casino craps table dice are. They're also more expensive.) But it's going to be close to even distribution, even with a cheap die.
A gamer, either of the tabletop or video variety, will tell you that this is all a load of dingoes' kidneys. The characters that gamers play live and die on good or bad rolls, and even if your dice are perfectly square and uniform, even if your game uses cryptographically strong random numbers, these numbers do not follow "distributions" or "probabilities." Instead, they usually produce whatever number you really didn't want. A tabletop gamer may say that "the dice are trying to kill" him if he encounters a long set of bad rolls (e.g. broke every weapon he was carrying). Unless, of course, you know the proper way to placate the Random Number God: then the dice will smile upon you. Usually.
The name of this trope comes from Angband's and NetHack's fanbases (coined in the Angband fanbase, spread by Nethack's), as a fanciful expansion of "RNG", for "Random Number Generator"; both games are partially Luck Based Missions, considering how many Instant Deaths there are, and the players of both games have been known to build altars to the "Random Number God" or curse his/her/its name. Or both.
Some games try to cut out the nonsense by supplying their own Luck Manipulation Mechanic. This never works — the Random Number God is not mocked, and it'll find a way to mess with you regardless. In Real Life or certain other situations, a Two-Headed Coin can masquerade as a mere minion of the Random Number God up until The Reveal.
open/close all folders
Anime & Manga
Saki proves that this trope applies to mahjong tiles as much as it does to dice. Nodoka refuses to believe in players being "loved by the tiles" and such, but other players' freakish abilities to defy all probability when playing (as someone mentioned on the page, "how much can we make a statistician somewhere cry?") pushes her into Flat Earth Atheist territory.
In Quick Start!!, a 4koma about tabletop gaming group, a resident powergamer Karasuyama Sachi is on RNG's bad side, despite buying dice with equal distribution.
In Knights of the Dinner Table, related to Hackmaster, a character was once beaten to a pulp for touching another gamer's dice. He was blinded by a cupful of soda and then was on the receiving end of a flying tackle. No one (except Sara) thought that this was uncalled for, as "that's one dice squirrel who'll think twice before touching another man's dice!"
The attacker in that scenario was the 5'3" pencil-necked Bob Herzog, and the man who touched his dice was the 6'11" ex-marine Nitro Ferguson, and every time the event is referenced (the fight itself was never shown) it is strongly implied that Nitro got hurt at least as bad as Bob did.
In the strip, they even outlined a "dice cleansing" ritual, used to banish bad luck from the character's dice.
Many dice superstitions are explored in this strip. Bob refers to his dice by name, keeps them segregated according to purpose and game genre. Dave refers frequently to having his section of the table "trained" the way he likes it for dice rolling. All of the Knights, even the sensible Sara, have dice rolling styles. But the most infamous incident had Brian and Bob "fame-rubbing" their dice on Gary Jackson's corpse (a "legendary RPG creator" pastiche of Gary Gygax and Steve Jackson) to charge them with good luck. This backfired because Gary, being dead, was obviously out of good luck (as Sara pointed out, someone who'd died in a plane crash couldn't have been very lucky to begin with). All the fame rubbed dice performed poorly in game play. (Or possibly because that corpse wasn't Gary Jackson's.)
The actual pen-and-paper role-playing game, Hack Master (written by the authors of the comic, naturally) enshrines a lot of these dice superstitions in its rules. Included on one page are prints of the signatures of the game's (fictional) creators, complete with instructions for how to empower dice with them.
Lone Wolf has the dreaded Random Number Table, a grid in the back of each book numbered 0-9, that you have to use whenever you were in combat or were just trying not to get killed by some random. A few gave you bonuses based on Disciplines and equipment, but many more would gladly kill you if you didn't get a seven or higher. In many of those cases, not having that discipline would be even worse off as Lone Wolf would gleefully leap into deadly ambushes, activate ancient traps, fail to notice certain death curses or swallow poison whole. Giving you a 7-or-higher chance to live was actually the better path to follow in those events.
The thing is, the Random Number Table really WAS very much skewed and not really random, considering that the numbers in the grid are in a fixed position and you could more or less "aim" for sections of the RNT which were more likely to grant a favorable result. With practice, you could pull out a 0,1 or 9 anytime you wanted.
Midkemia has two gods worshipped by gamblers: lady luck Ruthia, and mischief god Banath. They know to never invoke Ruthia's name to get good luck, because, the fickle lady that she is, might just give you the opposite of what you wanted instead.
Discworld has an actual Random Number Goddess, but trying to curry her favour is not a smart move. The Lady (Always "The Lady", she will never come to those who use her namenote Strongly implied to be Luck) does not appreciate being invoked — gamblers who try to worship her directly always seem to die in strange and unlikely (not to mention unlucky) ways. It is said that she has a soft spot for hopeless cases, but relying on this would be a spectacularly bad idea.
Malazan Book of the Fallen has Oponn, the twins of chance (one increases your luck, and the other decreases it, but you don't know which you'll get if they bless you). One character unthinkingly named his sword Chance (and as a result has it blessed by them), and most every intelligent character he ever meets advises him to either break it or pass it to his worst enemy if his luck ever turns.
In God's Debris, God Is Dead, having killed Himself at the moment of the Big Bang. However, He left behind matter ("God-Dust") and probability. Said probability determines how matter moves and evolves, effectively being the Random Number God. Despite its seeming randomness, the probability's final goal is God's complete reassembly.
In a season 5 episode of Big Bang Theory, Sheldon makes all of his decisions on a 2d20 roll.
Almost always averted in modern Pinball games. The end-of-game Match Sequence, where a player gets a free game if his score ends in the same digits as a randomly-generated number, is seldom truly random. Instead, most tables are set to distribute freebies based on various criteria, such as the number of players, a particularly low score, or occasional generosity (such as matching 5% of the time). The "random" nature of the Match is merely for show.
Played straight in Bugs Bunny's Birthday Ball, where the game-breaking 50 Million shot is randomly enabled after the player blows out enough candles. And then there's the "surprise present" feature; at the end of a game, the winning player can have his highest score taken away from him and awarded to someone else.
In the back of the Hackmaster 2nd Edition rulebook, there's actually a list of various dice rituals that are prescribed for the game, including rubbing the dice clockwise for higher rolls and counter clockwise for lower rolls.
A number of bizarre good luck superstitions have arisen in the Warhammer 40,000 community, such as never calling missile launchers by their proper name (it has the word "miss" in it), the idea that painted models are luckier than unpainted modelsnote Ork players don't count; their paint actually does change their units' abilities, by the rules., and the practice of occasionally muttering prayers to the Emperor. Never taken seriously, but often endearing.
Don't ever say "anything but an X." The more important the roll is, the higher the chance that you will get that X. Most players have lost more games by saying "anything but an X" than by bad strategy.
Also worth mentioning are the Chaos gods' chosen numbers. Chaos players used to get bonuses for fielding units of a certain size based on which god you were using. Slaanesh was 6, Nurgle 7, Khorne 8, Tzeentch 9, Malal 11, the idea being that the Chaos god makes the characters more lucky as long as they spend time in their chosen number. Modern versions of the game, for the sake of simplicity, have removed this rule.
The W40K rule book advises players who are rolling large numbers of dice to take out the dice that failed and continue rolling the successful onesnote each bullet is determined by 3 dice: whether it hit, how much damage, and whether the target's armor withstood, so squads with automatic weapons can expect to roll 30+ dice up to three times each. Players will tell you the exact opposite: NEVER reroll the successful ones immediately afterwards, because they have just used up a good roll.
A player will say, "I never seem to roll as well as I just did." A statistician will say, "Well, duh. Regression to the mean. You just had a lucky break, so the same entirely random process is very likely to produce a lower value simply because most possible values are lower." Both are correct. The player will usually retort, "Whatever," and produce another fifty dice. Warhammer40000 players always have enough dice, just never enough dakka.
Ork players are advised to use green dice for standard rolls ("green iz best"), red dice for moving through terrain ("coz da red ones go fasta"), and ("lucky") blue dice for critical rolls. Buy in bulk.
It is also recommended, as the melta is one of the most powerful anti-armor weapons a troop can carry, that it never be referred to as such, as firing a melta weapon is an exercise in comedic inaccuracy. Simply referring to which model is firing should suffice. And never use Meltaguns against infantry. They might be great for blowing up tanks, but they against infantry you're guaranteed to roll a 1.
For the uninitiated, the game also sports something called Mathhammer, whereby anyone who can perform simple probability calculations about six-sided dice can figure out the expected results of any particular attack offhand. Naturally, anyone who engages in a quick Mathhammer calculation shouldn't do it aloud; you can't believe the superior chuckling of those who don't understand probability if you're even one off in your calculations.
Some Eldar players mutter prayers to Khaine, or to Cegorach if using Harlequins. Chaos players, meanwhile, invoke the Dark Gods, and Ork players just shout "WAAAAAGH" when shooting. Or charging. Or at random intervals.
In the Wuxia RPG Weapons Of The Gods, observed good or bad luck with the dice can be utilized as a game effect, discovering that the character is under a curse or blessing which can then be either increased to add actual bonuses or used to create a balancing effect of the opposite type.
The system itself also features the River mechanic, designed to mitigate the influence of the RNG. You can take dice results out of a role and store them for later use, so if you make an awesome roll on a trivial task, or roll two sets when you only need one, you can set some aside and break them out later when you screw up something important.
In Call of Duty: World at War (as well as Black Ops), the zombie killing "Survival" mode often involves a fair amount of spending credits earned by killing the aforementioned undead to take a spin on the weapon box, which awards a random weapon to the player. Since every gun in the mode can spawn there and the money is paid regardless of if you TAKE the gun, one must pray to the RNG.
Black Ops has an equal amount of RNG'ing, both due to the Mystery Box, randomly dropped power-ups and a glitch that can randomly give a zombie insta-kill hits. Which still remains unfixed.
This concept is taken to extremes in Kingdom of Loathing, where the RNG is a conscious entity that has its own account and frequents the various chat rooms. Those who please the RNG in some way may find themselves "Blessed by the RNG", whereas those who annoy it (especially by begging for a blessing) may find themselves "Cursed by the RNG" (both of which are active character effects). Because KoL game mechanics rely heavily on random number generation, and because KoL effects are rarely explicitly defined, there is still a significant debate over whether or not the Blessing and/or Curse actually affect a player's RNG-based "luck" in the game, or if it's just a red herring.
A player may also be "Blessed by The RNG" by sending a gift package with an 8-ball, and a look in The RNG's display case may reveal other effective sacrifices.
One particular area that makes players rage to the RNG: the F'c'le. In order to earn Pirate Fledges and be able to access the Obligatory Pirate Cove without wearing fairly weak equipment (the Fledges themselves are actually pretty strong), you must defeat three specific enemy types in this area, pick up their random drops, and use them. Every KoL player can regale you with stories of spending hundreds of adventures praying for the right enemy to show up, or, alternatively, having all three enemies show up in turn and give up their items in a row. The RNG in the F'c'le is an extremely fickle godling, and no one can predict its mood on a given day.
In the Roguelike communities where the term originated, finding a very good item early in the game is a sure sign that the rest of it is going to be a hellish struggle against the wrath of the Random Number God. It's given you your one good thing, now it's going to do its damnedest to kill you.
The ultimate item in this is an Amulet of Life Saving, which effectively gives you an extra life (in Roguelikes, if you die, that's normally it — your (one) save game is erased). If you find one of these early, rest assured the game is going to maneuver you into a situation where even instant resurrection will not help you. After all, if something is tough enough to kill you once, it can probably do it a second time... Paradoxically, "lucky finds" like this that should improve your chances of survival will just make players incredibly paranoid and even more cautious.
Added as a character in TOME (and its parent game, Z Angband). Random Number Gods are weak, annoying, fast multiplying monsters that drop decent loot — and cause confusion.
In some roguelikes, the random number generator is known to get locked into generating the same number repeatedly for an extended period of time more often than probability would suggest it would happen, so seeing a 1.7% chance happen 10 times in a row is not unheard of, especially when you are on the wrong end of it. Ouch. Of course, this is normally because, when you're on the right end of it, you don't need a second chance. Your critical hits are basically kills, but when Everything Is Trying ToNibble You To Death...
Starting characters in Angband and its offshoots are very dependent on the Random Number God smiling favorably due to the lack of abilities and equipment. Character deaths on the first few floors happen very often, requiring a few do-overs before being able to finally survive.
We of the Final Fantasy Tactics online simulator Super Tact pay our sacrificed goats to her most revered Random Number Goddess.
Also from the Final Fantasy Tactics main series, it shows you the odds that your attacks will hit before you make them. However, anyone who has played the franchise will tell you that these odds are not true, because anything less than an 85 will not hit. Ever.
Anyone who has ever played Fire Emblem will curse his name. No exceptions. Just for those who don't know, Final Death is in place for all units. Think you can take that guy who has only a 1% chance of landing a critical hit? Think again. In some games, the sequence of random numbers generated after a save is always the same, allowing for a degree of predictability. Proper manipulation of the RNG is key to any Fire EmblemSpeedrun.
Various theories point to the RNG being a Random Number Goddess, specifically Anna, an NPC (one who has appeared in every game to date note except for Gaiden, sometimes with her husband Jake... who always has above-average stats.) She's something of a Series Mascot.
It's probably vindicated by Fire Emblem Awakening as she appears as the final boss of the secret path in the hardest DLC chapter, Apotheosis. She comes with stats that no class could ever achieve (even with Limit Breaker), a forged Spear and Brave Lance, and the skills Dragonskin, Aether, Vantage+, Counter, and RIGHTFUL GOD.Pick a god and pray indeed.
Path of Radiance actually features an instance where the RNG can be made to work for you, instead of on his own whims, with the Bonus Experience. Got a character that really needs consistent growths in certain stats? Simply save your game right before giving them the bonus experience, the only time when you can consistently A) save your game, and B) have a character gain a level without any other factors at play; if you didn't like the stats they gained, no problem! You can reset the game and try again as many times as you want, until the character gets the desired stats. And then the sequel, Radiant Dawn, allowed quick saving (except on Hard), so you could do this same thing whenever you wanted, in battle! Foolish, Random Number God. Foolish.
Quick saving was also present in Genealogy of the Holy War, though it had the "Use Exact Same Strings" method, as well as a bug would cause the game to run out of Numbers during Arena combat, resulting in level-ups with no stat gains.
Players who have become aware of higher-dimensional spaces may start to briefly encounter the RNG in other locations, and one of the more traumatic experiences during a quest is briefly seeing it in its true form.
In World of Warcraft, it's customary to use the /roll command a few times before rolling for a specific piece of equipment to please the Random Number God and to get the low numbers "out of the way". In addition, there are persistent (but frequently debunked) rumors that the game seeds the random number generator according to specific criteria, including, but not limited to, the raid leader, the first person entering the dungeon and/or the number of damage-over-time spells (more dots!) on the boss.
That's ridiculous. Everyone knows that the dice will only favor you if you A) contributed the least of anyone to the party/raid, B) can't use or shouldn't be using the item in question, or C) have someone else present who has waited for the random drop for months. These effects are cumulative, and if all three are true you're almost guaranteed to win. The Random Number God of World of Warcraft is a sadistic bastard.
It is possible to increase the odds of a desired gear piece to drop by spending a lot of gold on ridiculously expensive gear enchantments or gems to enhance the piece you desperately want to replace, or to buy a lesser temporary substitute at the auction house (or the crafting materials to have one made by a blacksmith). Ask any player, everyone has a tale of spending a fortune on a piece of gear only to have a better one drop magically in their hands not one day later.
It's also not unheard of for someone to get an upgrade on their first try as long as another member of the party has been attempting to get a drop from the same boss for a long amount of time. It's not just sadistic, it's pure evil.
A needed upgrade for a raid member will often drop week after week in their absence, only to never be seen when they do show up.
Or worse, an item that no one can use after the first week will drop every. Single. Week.
God help you if your prep roll yields a 100 or similar high value; that was the only one gifted to you by the RNG this month.
As a tank, God help you if your prep roll was over 15. Tanks have by far the worst luck rolling on anything, and you aren't likely to top 15.
Doubly so for warrior tanks, who are notorious for losing a roll on the same piece of gear multiple consecutive times, by one point.
Part of that is also that tank specific gear is also the least likely to drop in any circumstance, and with the random dungeon finder, they have the easiest time getting into a dungeon.
The developers are equally sadistic with this. Two of the Achievements involved having to get 100 on rolls to Need or Greed loot (once the system was redesigned to distribute loot drops automatically). It can take as long to get those to come up as specific, desperately wanted gear.
Suikoden has trouble with this at times. There's a cup shuffling game where the dealer always follows a specific pattern every time you quit and play. However, there's another game that involves throwing three die. Despite the fact that the odds of getting very good rolls (three of the same number other than one, or a straight of 4-5-6) or very bad rolls (all 1's or a straight of 1-2-3) should be low, you'll find that you and the computer will get these combinations very frequently. Murphy's Law of course will turn what should be a 1/216 chance into 1/3. Also, a ground rule is that you must roll something valid in three turns (i.e., you automatically lose if you roll three "no scores"), the computer will always roll something valid on the third turn (even if it's one of the dice going out of bounds).
The poker minigame in Tales of Vesperia has a rigged RNG. It's very easy to win once, but if you try to double your earnings, the RNG makes your life horrible. The higher the value of your hand is (i.e. straights, flushes, full houses, etc.), then the more the RNG will screw you. If you bet 'Low' on a King or Queen, then there's a very high chance you'll get an Ace or Joker. Vice versa applies to betting 'High' on threes or fours. And even if you have a relatively low value hand (Three of a kind, two pairs), the RNG screw up kicks in at about the fifth time you want to double your earnings. Roughly 90% of the time, you'll lose by betting low on Kings and Queens, while you'll have a 50% chance if you bet high.
Anyone who has ever played EverQuest and attempted to level tradeskills beyond a certain point knows that the RNG is a spiteful beast that hates you. There is a reason the premier crafting forum has the appropriately named Primal Scream Room...
Players of City of Heroes will notice that, after extended periods of time, you will miss with several attacks in a row despite your accuracy hovering from anywhere between 75% to 95% (the highest possible). Clearly the Random Number God at work. As opposed to, say, you just having an unlucky streak of rolls which is nearly guaranteed to happen the longer you play the game. Nope, the Random Number God definitely woke up with a hangover that morning and decided to take it out on you. It ended up being such that the developers turned the RNG into a quasi-RNG: If the random distribution did not meet the ideal after a certain amount of time (generally meaning that a player missed a lot), it would force favorable results until it did.
The Monster Hunter fandom has come to the conclusion that the series has a sadistic, psychic, "desire-based" RNG for determining how often a really rare item Randomly Drops, also known as the desire sensor. It will skew the odds out of your favor if you really want something, and the only way to end your curse is to not want it. You can't act like you don't want it, you have to actually not need it; the system is immune to reverse psychology. So if you spend hours farming the same giant monster for its armor plating, you may go through fifty corpses without seeing one, but you'll end up swimming in the stuff once you give up and move on to a different goal. The only other way to appease it may be petting the pig.
FTL: Faster Than Light and it's RNG can result in an unwinnable game if you are unlucky enough to go through the entire game without getting a set of good weapons.
Given that the engine knows the characters' profiles, quests, recipes and any discovered players' paths, the implementation of a sadistic "you'll always get stuck with only nineteen bear asses" drop routine is easy. In fact, it would be a usual semi-random drop with reversed adjustments. So the question is whether it's really that random, or not.
Due to the nature of the makers (and fanbase) of Dragon Fable the RNG is nicknamed the Rude (or Reall) Nasty Gnome. He sits on a floating pile of all of the gear in the game and throws items at heroes who finish quests. Those heroes that 'amuse' him, get a better item thrown at them (albeit faster).
In Pokémon games, there is a chance of encountering a "Shiny" Pokemon. This is actually just a normal Pokemon with a recolor and a sparkle animation when it enters combat, but they are highly sought-after nonetheless, just because they are so extraordinarily rare. How rare? The odds of encountering a shiny Pokemon in a random wild battle is approximately 1/8192. Couple that with the odds of capturing a Pokemon you've just found, and you've got one frustrating challenge. If you ever see a player with a shiny Pokemon, you can be sure they've been blessed by the Random Number God. Or they're cheaters. Or they got it at an event. Or it's a red Gyarados.
Unless you're playing PokemonXAndY, where the rate appears increased and people are getting them all over from Friend Safari,Hordes or chain fishing...
Players have managed to reverse-engineer the random number generator, and by methods that are technically not cheating, including precisely timing button presses and setting the Nintendo DS to a specific date and time, can significantly increase their chances of getting a Shiny Pokemon and/or one with the best possible stats.
However, you would still need to have a few Shinies to make this work, or you can read your save file, because one of the factors required to get a shiny is your "Secret ID," which cannot be found in the game. Thankfully, Shinies are the mainly the only thing that requires that number.
And that is now being exploited by people who use another software to help others find that code. Then, all that's needed is to use that method, determine the shiny code each egg has, and find someone on the 'willing to hatch' list whose game id matches the egg's code,and the result is a guaranteed shiny.
Pokémon is made of this trope. There's the randomness of getting the Pokemon that you want (have fun finding ones with lower than 1% encounter rates), and the randomness of Natures and IVs on top of that (and as mentioned above, the extremely rare chance of getting a Shiny). Once you have your Pokémon and use it in battle, you get to enjoy the RNG of critical hits and secondary effects of moves, like flinches or status effects, as well as the RNG of hitting at all. There's even an RNG for attack damage, which varies between 85% and 100% on a bell curve. Get unlucky and have it hit for the low end, and you might have missed a crucial KO.
High level AI opponents such as the gym leaders are eternally blessed by the RNG. If there is one thing they can't stand, that is you scoring a perfect, flawless win against them. So what if you trained your Pokémons to be ten levels above the next gym leader's average Pokémon level? Once you start to one hit KO the gym leaders Pokémons, suddenly they pull out somebody knowing Metronome, which "randomly" selects Tri-Attack (something that has a 20% chance to freeze) from over 550 moves, which in turn freezes your Pokémon instantly. All this just because they don't have an one-hit KO move or an ice-type move to freeze you but you must lose at least one Pokémon to them. Calculate the odds of this happening exactly when you want it to and you realize who is the cheater here.
Some competitive communities, such as Smogon, try to keep the effects of the Random Number God to a minimum, through means such as banning items like Brightpowder and moves like Double Team and Minimize that rely solely on RNG to work. The mindset is pervasive enough that many battlers in these circles will go so far as to apologize to their opponent if RNG starts to favor them, such as with a crucial Critical Hit or an attack with a small chance at a secondary effect getting that effect (ie. If a Thunderbolt induces paralysis in the target).
The Sinnoh games gave players a possible means to fudge the shiny system— the post-game Pokemon Radar can be used to "chain" repeat encounters with a specific kind of Pokemon. The longer your "chain", the odds of finding a shiny Pokemon slowly increase, until it caps at 1/200 (after about forty Pokemon). The Poke Radar also has a unique animation for detecting a wild shiny Pokemon.
And then, there are reportedly various rituals that can be used to increase the chance of successfully catching a Pokemon after a Poké Ball is thrown, such as timing pressing the A button as the Poké Ball closes and every time it shakes, mashing buttons, specific taps on the DS touch screen, etc. They're found nowhere in the games' code.
To a lesser extent, Voltorb Flip from the non-Japanese versions of HeartGold and SoulSilver. Because it is both skill-based and luck-based, you can win big, but you can (and will) more easily win very little. The game is much like a cross between Picross and Minesweeper: you have a five-by-five grid of tiles, and each row and column contain multiplier cards or Voltorbs. If you get a multiplier card, your current coins are multiplied by that number (If you haven't won any coins yet, you simply get coins). If you win, you advance a level and can win more because there are more multiplier cards. The catch is that you can just as easily land on a Voltorb, which not only makes you lose all the coins you got in that particular game, but also ends the game altogether, and sends you back to an earlier level based on how many multiplier cards you flipped; if you're unlucky enough to lose right off the bat, back to level one you go. It's not entirely luck-based, but it's still far easier to lose your coins and/or end a long streak of wins on an unlucky flip than to actually start off said long streak of wins.
To be fair, most players turn off the items for any competitive match because the RNG loves nothing more than waiting until you and your rival are at one life each, then spawning a Bob-Omb or an exploding crate right in front of you at the very moment you attack.
The 4X game Sword of the Stars was practically built around this trope. In addition to the usual randomly generated maps of most 4X games, the game also semi-randomly spawns various "menaces" unaligned with any race that range in threat level from Goddamn Bats to That One Boss, and even randomizes the tech tree. Some techs are considered "core" and guaranteed to be available for research to all races, most have a random chance of appearing at all for each race (with probabilities ranging from 0% to 100%, depending on the race and tech in question), and some are exclusive to particular races, and some of those aren't guaranteed to be accessible to their race in every game.
God Hand can be bad about this. Low on health and need a pick-me-up fast? The dice often aren't on your side.
Want to heal your critically hurt union inThe Last Remnant? Better damn well pray that the AI generates the heal option for you on your list of unit actions for that round.
Get enough assassins in Medieval Total War II or its predecessors, and it's easy to see the chances of success the computer gives you are not very accurate. Having a less than 30% chance of success means, more often than not, even eight assassins with that percentage will fail, even if attacking enemies from two separate nations. There seems to be lucky and unlucky turns, as if one assassin succeeds with a 23% chance, there is a good chance others will too, and the reverse if one with an 80% chance fails.
Civilization uses an RNG to resolve combat, among other things. On account of the way the system worked, this led to the famous Spearman Defeats Tank meme within the community. The problem was fixed in Civ II, but then III and IV screwed it up again (though not quite to Civ I levels).
Civilization IV and V at least give you the option to keep the same seed for the RNG between game loads but only for the entire match. With the same seed, you can effectively predict the future. With different seeds on each load, you can save scum all you like until something good happens.
In Nexus War, "The RNG hate my/your guts" became a bit of a meme on the forum, generally in relation to searches or hiding (i.e if you tried to hide with more than 10 AP left or REALLY needed to find a supposedly common component you would fail) more than combat which was PvP.
The Facebook game Empires & Allies by Zynga has this trope due to how critical hits occur. Critical Hits and more importantly Critical Kills give more loot, exp, give Ore and often Energy. They are thus VERY desirable. Players have come up with a theory that aiming for "sweet spots" on various targets grant a much higher than normal chance for a critical. This theory has not been conclusively proven or disproven.
The X-Universe games rely on an RNG to determine such things as NPC cargo and equipment, what enemies spawn and where they travel, and especially Boarding Party casualties. Praying to the RNG has become a Running Gag on the forums.
In Wizardry Labyrinth Of Lost Souls, you will receive a random number of bonus points to allocate to your stats upon character creation. This can be as low as 6 to (allegedly) as high a number as 47. Exiting, then reentering the character creation screen rerolls this number. This often leads to character creation taking hours, not because of many customization options (you only choose your character's name, race, gender, alignment, stats and class), but because players keep exiting and reentering character creation in hopes of getting a higher amount of bonus points. This is compounded by the fact that lower numbers occur far more often, while numbers get rarer the higher they get, so only the truly blessed ones among us will ever see a number in the forties.
In Urban Dead nearly all actions are luckbased and thus dependant on the RNG. People who have wasted their entire day's worth of AP on missed attacks or uselss searches will understand how frustrating this can be.
In XCOM: Enemy Unknown, the RNG will actually help players of the Easy and Normal difficulties, although He/She only provides subtle assistance in trickier dierolls in Normal. The kid gloves come off in Classic and Impossible, though. Expect to be furious about missing 80% chances to hit, or flubbing that Rocket Launcher shot.
The expansion pack, Enemy Within offers a "Save Scumming" option - the original would have the RNG result occur every time (Missed that 90%? It will always miss when taken in that turn order), while this option rerolls it on the RNG.
RuneScape has its fair share of skewed probabilities, but a game moderator once memorably said "the random number generator gods are fickle and bow not to mortals".
The Sims Medieval gives you "success odds" of certain tasks your Sim can perform, but "medium" is more likely than not to result in failure on difficult things. Never try, for example, picking the lock on the stocks with your Spy when the odds are only medium. You will get caught.
All players of the Mass Effect 3 multi-player know that the items store is governed by a Random Number God. Say for example that you prefer to play snipers, so of course you want to unlock some powerful ultra-rare sniper rifles. Maybe you've really got your sights set on that Black Widow. Of course, you'll get shotguns. You'll get assault rifles. You'll get SMGs. And if you do get a sniper rifle, it'll be the Incisor. And at some point, you'll have all your weapons maxed out but the Black Widow is still only at level 2.
This is lampshaded by one of the random conversations you can overhear in the Citadel DLC, where an N7 Fury (a biotic that favours light weapons) is arguing with a requisitions officer that keeps trying to give her upgrades to a very heavy shotgun.
N7 Fury: You're the reason we're losing this war!
The Rune Factory fandom has been citing the Random Number God quite often with the release of Rune Factory 4. Whether or not you can successfully confess to a person, when you will see all the event you need in order to propose or be proposed to, and when you will get access to the final dungeon of the game are all entirely random. You might be able to clear the entire plot and get married in the first few seasons you play through, and you might spend five years waiting for that last event you need to activate.
The RNG of World of Tanks is frequently cursed by players for things including, but not limited to, map selection* Certain maps being detested by segments of the player base, or very unfavorable to some vehicles, game mode, shot deviation, shot penetration, and damage* The last two frequently resulting in vehicles surviving shots that "should" have destroyed them. The "matchmaker" that assembles random teams comes in for a double round of cursing for assembling teams that are seen as wildly imbalanced, either in player skill or vehicle lineup.
Final Fantasy XIV is BRUTAL with RNG when it comes to crafting. Crafting high quality items can require an insane amount of luck (it's completely random on how often your next move can be used on a turn with an excellent condition), even if you try to set up the odds in your favor as best as you can. If RNG doesn't favor you, you'll fail your crafting and lose all the materials that were used for it.
The Atma quest lines are just as bad. To get your foot in the door for powering up your Infinity+1 Sword, you must hunt down 12 Atma items. Said items have a supposed 2.5% chance of dropping after each event you clear, You can get a few Atmas in a few hours or possibly nothing after several days.
The hit Japanese webgame Kantai Collection has this bad. How bad? The initial access to the game is dealt to you via lottery. After going in, the RNG continues to hinder you as you play, ranging from misses and critical hits* Bad, since even the heaviest ships has a disproportionate ratio of hitpoints versus damage, and even normally-weak enemy destroyers can still land a critical hit on your battleships costing them half their hitpoints, especially in night battles to item-crafting and the constant stream of NAKA-CHAN DAYO~* When you craft ships, the odds are skewed towards certain classes of ships due to the sheer size of variety; Sendai-class light cruisers (Sendai, Jintsuu, and Naka) are particularly detested for being extremely common, although they are by no means weak/useless
Speedrunners endlessly bemoan the nature of random number generators. For longer speed runs, and depending on the game, a lot of RNG rolls can be encountered on an ideal fastest route through a game. Manipulating a game's random number generator (to influence item drops, the movement of NPCs, etc.) is virtually impossible in most games played in real time (with rare exceptions), so your highly optimized, world-record pace run could come crashing down just because the game says so. In a hobby as repetitive as speedrunning, where you could spend hundreds or thousands of attempts trying to nail all the tricks in your route, this can be very demoralizing.
Tool-assisted speedrunners, on the other hand, have the benefit of being able to control inputs frame-by-frame and check memory values. They can reliably influence a game's random number generator, making it look like the game is conspiring to have them go as fast as possible.
This◊ Dork Tower strip was the beginning of an arc about scheming dice that started behaving only when the cats got a hold of them.
A strip in Real Life Comics parodies the idea of THAC0 where the main character misses a stationary enemy that wasn't focused on him because he rolled too low to hit it.
There's a sort of Double Subversion to this concept, detailed in the annotations to thisDarths & Droids strip, which involves "rejecting that superstitious nonsense" and instead using the laws of probability distribution:
Take 1000 or so 20-sided dice.
Roll each and every one of them once.
About one-twentieth of these will have rolled 1's. Take these fifty-odd dice, and roll each of them again, once.
Two or three of these dice will now have rolled 1 twice in a row. Statistically, the odds of rolling the same number three times on a 20-sided die is 1 in 8000, so now these dice have the 1's "rolled out of them"
Place them in a special padded container so that they can't roll around, and you may now safely bring them out in emergencies for use for a die roll in which you really don't want a 1.
This is of course patent nonsense; no matter how many times in a row you get a 1, the odds of the next roll getting a one are always 1 in 20, even if it seems "overdue" for a different one.
It's not just nonsense, it may be totally backwards. If a d20 rolls two 1's, which is a 1 in 400 occurrence, then the dice might be flawed, and rolls 1 more often than it should.
Darths & Droids also goes as far as to Lampshade the trope with a link in the annontations of the strip mentioned above that links to this very page.
The Random Number God will smite thee, unbeliever.
This strip demonstrates this strategy in action: The prerolled die yields a natural 1 but the Random Number God's true believers don't waver. They reason it will simply be even more lucky next time.
Full Frontal Nerdity carries this to a ludicrous extreme when Lewis attempts to dispose of a 'cursed' die that can seemingly only roll 1s. The die rises from the grave and the curse is so strong that every random number generation device in the world becomes incapable of generating any number other than 1.
"Now here's an important thing, though," said Clay, pointing at the die. "Where did those 4s come from?"
Wanda looked down at the little brown enchanted pyramid and shook her head. "From you. You cast upon the die."
Clay grinned slyly. "So you're saying my juice had a physical effect on the way the die rolled."
"Yes and no," said Clay, grinning more broadly now. "It's not a direct effect, you see? The 4 is a Number. It had to come from somewhere."
The Elijah And Azuu sequel presents a character who actually manages to beat the Random Number God. She has a lucky 20-sided die, and due to residual subconscious magic the die is literally unable to roll a 1, even when other characters use it.
A College Humor sketch about Tetris called "The Tetris God" involves the eponymous character manually choosing which piece will be next.
Often invoked on Tabletop. In the Elder Sign'' episode, Wil Wheaton claims that the dice are trying to kill him. They succeed.
Cards. When you've already drawn the ones that are good for you, they're not in the deck anymore. Notions about the Random Number God seem to grow out of expecting dice to behave like cards.
There are some calculations scientists perform that require truly random numbers, such that the slight predictability inherent in a computerized system is too large. For these cases a "True Random Number" generator can be purchased, which is effectively a small radio telescope tuned precisely to the frequency of the Cosmic Microwave Background radiation, as produced by the Big Bang.
Radioactive isotopes can also be used - set up a Geiger counter and a numeric counter, then stop the counter whenever the Geiger meter goes "beep". The time between successive radioactive decays is random, so your numbers are random.note For those wondering about Half-Lives, what's normally done is to measure only whether the time between decays is more or less than the previous one, and use the sign as the resulting bit
Actually getting a random number out of a truly random source is far from trivial in itself - you have to build a device that doesn't introduce any bias.
As this video explains, cheap dice have dull edges because they were polished in a rock tumbler. The resulting imperfections change the die's center of gravity and make it easier to roll over some edges than others. So there is a scientific explanation to this phenomenon, and a way to prevent it.
It also makes reference to one of the classic ways con artists can rig games: "shaving" the edges on a die (usually a casino die, since they have sharp edges to begin with). You trim off a tiny bit on the edges of whatever's opposite the side you don't want to land on. So for example, if you don't want to roll a six, you'd shave the edges of the one face. The die expends less energy rolling over the shaved edges, which skews the probability way below 1/6. (Incidentally, if you find an actual casino die that's been shaved, you could be rewarded by the state gaming commission. They take fair odds seriously. Outside of a casino, though... just remember the page quote.)
Another common method of messing with die odds is to cook them. You take your plastic die, put it in the microwave with the side you want to show face up, then nuke it for a couple of seconds, no more than five. The nuking causes the plastic to melt downwards a bit(thus making the die more likely to stop on that edge, since it's heavier). Done right, it's almost undetectable at casual inspection.
And knowing that means that your GM is justified in confiscating your dice if they seem to be unusually 'lucky'.
Most computer software use a Pseudo random number generator Often, a programming language's library will provide it, but for early games and applications, bugs in the implementations of pseudo random number generators caused biases. C's built-in random library is actually notable for allowing programmers to use it incorrectly and introduce biases to the numbers generated by it. Problems with how the random number seed is picked can also cause the numbers to be predictable.
Since random number generators are used by many security applications, their predictability can lead to serious security vulnerabilities. Which makes computing with serious security need another area of use for the aforementioned true random number generator devices.
Although implementation issues with random number generation itself have been largely averted with more modern games, it is still difficult to keep correct uniform distribution when picking more complicated random events, even while the actual random number generator has a good distribution. For example, consider picking a random point inside a disk, as is needed in certain games (usually some variation of bullet spread). Programmers without prior knowledge, specially amateur ones coding for a game mod can implement this random point generator in a way that there is a bias towards the center of the circle or towards its limits.
In Data Stage it is possible to randomly generate the same numbers as in the course book for exercises, meaning it is completely predictable.
Some languages, such as Q Basic, allow the programmer to generate pseudo-random numbers using the system's internal clock as the seed, so the seed will be different every time a program is started.
One particularly notable random number generator, known nowdays as RANDU, was noted as being, in the words of Donald Knuth, "truly horrible". It has since become a textbook example of a bad PRNG, due to many, many undesirable properties. note Among the lesser sins: only generating odd numbers if fed an odd seed value, and even numbers if fed an even seed value. Given that most programs feed the result in as the next seed to insure consistency across runs, this is not a trivial problem. (On the plus side, if your system is newer than the second half of the Reagan administration, and doesn't involve FORTRAN, it's highly unlikely you will suffer under this particular misguided PRNG.)
Many versions of Linux have a very good PRNG as a system resource with a seed that depends on many hard-to-access and fast-changing internals of the system, making it difficult to guess what the next value will be, even if you have many recent values and normal programmatic access to the system.
Averted with really cheap CD players. For some "random" is the playlist shuffled differently, and always that same order differently.
Even more expensive ones will leave the user scratching their heads at the chances of two songs of the same artist back to back, or the fact that the system seems to throw up one artist more often that others despite the other 100 artists available.
This got so bad, so widespread, that Apple rewrote the "Shuffle" program on their newer iPods so that no two songs from the same artist are close to each other, if possible. Turns out people want high entropy from a shuffle, not random entropy.
In the same vein, Pandora Online Radio. Sometimes it plays songs by the same artist (or in the case of Disney-based stations, merely the same album). And sometimes, on the Disney stations, it will play songs from things like... Avenue Q, between Do You Want to Build a Snowman and Reflection. Other times, the station will play songs that are wholly irrelevant to the station you started off on, based very loosely on songs you've liked on the station.
A school of thought is that nothing is truly random, but rather, very chaotic. The forces that act on an object at any given point, even if static per se, can be graphed as a fractal (an infinitely complex shape). Since everything is only capable of a certain level precision, it's impossible to get the same exact result every time because you're trying to point a spot on something that's infinitely complex.
It is only a school of thought, not reflecting reality. Actually, radioactive decay is truly random in the sense that nothing in the observable Universe has an effect on when an undisturbed particle will decay, and not just because of our sensors lacking enough complexity.
This however is only a matter of which philosophical viewpoint you're taking, which cannot have an impact of our observation of "randomness". Using all our knowledge about the universe, we still could not predict any non random randomness (if it even existed), since we cannot observate closely enough (see Heisenberg) and we cannot compute all data from the universe without having some computing machine which is larger than the universe. So even if chance did not exist, we still will have to act and calculate as if it did. In other words - the existence non existence of randomness can never be proven since you can never assemble all required data, thus the question cannot be answered scientifically.
Basically: to prove randomness does not exist, you need to know every event that happened and the conditions that acted upon them.
It's also largely a question for the philosophers. Whether "true" randomness ultimately exists or not, the concept has too many practical applications to simply throw out.
Some websites with a "Random Page" button keep sending you to the same article, and might even get stuck on it.
One example is located directly above. Try clicking the 'random page' button a few times, and you'll start to notice that there's a few pages that are noticeably more 'random' than others. note What's actually happening is that the random page is chosen once a minute, to minimize server load.