TLP reform

~Water Blap. That's freaking awesome . That's great.

I can see about sorting the data through Excel. It might take a little work to get in in there, but it's doable to make some good reports. Go ahead and paste it to the sandbox or PM it, whichever works for you.

edited 5th Apr '17 7:48:06 PM by pokedude10

Okay, I made Sandbox.TLP Reform Statistics. I included the tropes from the "Unresolved Issues" folder, which is from the tally I recently posted in the crash rescue thread, in order to make sure I didn't miss any when going back through the comments. I won't be able to do more tonight, but I can work more on it Sunday (I can chip away at it before then, but RL stuff is important).

edited 5th Apr '17 8:12:54 PM by WaterBlap

Look at all that shiny stuff ain't they pretty

But does that "2 months and Up For Grabs" doctrine really help?^note If the OP manually declared a proposal UFG, then there would've been a post reflecting that. Proposals are shunted off the front page if it doesn't have a reply in the past six-to-twelve hours. One month from last reply gets us to Page 18, Two months is Page 24 right now.

I do admit that my timeframe expectation was grossly inaccurate. (I'm more used to how the TRS was 2 years or so ago when any debate that stalled for 2 months or more was Morgued.) But if "aging and refining the proposal to a workable state" is the priority here, than having the proposals displayed in order of most recent reply seems counterintuitive, as it pushes items that have had their debates stall off of the first few pages. Granted, there are also options for Most Hats and Most Replies, but Most Recent Reply is the default.

I have no problem with a Proposal that is 5 years old still having replies within the past 2 months. That is fine, the debate is still fresh, the "kid" is being fed. But I see a problem when a proposal isn't commented on at all in over a year. Although I now realize that the current focus is preventing things from launching too quickly, I fear that any restrictions we place to prevent that problem would EDIT: expand the number of proposals that spend more than 3 months without a response. The question is: what's the acceptable trade off? What can be done to minimize the casualties in the effort to stave off Sturgeon's Horde? The comment requirements are a worthy measure, but I fear that anything that would make launching more restrictive (yeah, including my prior proposals,) would make it easier for worthy Tropes to fall into the mists of time and obscurity.

edited 5th Apr '17 9:36:41 PM by DonaldthePotholer

Yes, the Up For Grabs doctrine does help. However it only helps by putting drafts up for sponsorships to tropers who are interested in helping it get launch-ready. If no one, OP or otherwise, is interested in sponsoring it, then Up For Grabs doesn't help the draft. Sure, it might be a good draft, and a worthy idea to be on the wiki, but unless there's someone willing to sponsor its creation, it won't happen.

However... I see your point and concerns. The backlog is large (granted, it's magnitudes smaller than it used to be), and the activity does stretch back when sorting on "most recent." I agree, I don't want good tropes to be lost to time. Yet if there are drafts that haven't been commented in over a year (thus are Up For Grabs), and no one has tried to sponsor it, then there's not really anything we can do.

I think our goal right now is just to prevent "toddlers" from being launched before they're ready. If there's a trope that has passed kindergarten, and there are not issues that might pop up after launch, theoretically the trope can launch. If it could launch, but it was abandoned, then someone has to dust it off to launch it.

Does this help answer what you're looking for?

edited 5th Apr '17 10:04:19 PM by pokedude10

The backlog isn't causing any harm in TLP, unlike what it was doing to TRS, where newly made threads were keeping older, decided ones from being finished, because it was much more interesting to talk about a new one than to clean or move the wicks on an old one.

...if you don’t love you’re dead, and if you do, they’ll kill you for it.

I agree that the backlog doesn't cause harm, unlike on the TRS. We don't need a draft to launch, there's no urgency. However, I think I understand Donald's concern that increasing the time it takes to launch might lead to more drafts being abandoned.

This is going to sound callous, but abandoned TL Ps may not be much of a problem. I say "May not", because we need to see how often tropes that come to Crash Rescue had more than a couple of months between last comment and launch to know whether they are or not.

...if you don’t love you’re dead, and if you do, they’ll kill you for it.

No, that doesn't sound callous. I completely agree with you, an abandoned draft doesn't harm TLP or the wiki. The only way it might harm the wiki is if it's launched before it's ready, thus ending up in crash rescue. I think Donald's concern is that there could be a correlation between "time spent in tlp" and "probability of a draft being abandoned" . Again, I don't think abandoned drafts harm TLP, but its worth considering if a higher "abandonment rate" has other effects.

As for the data analysis, I'm waiting for Water Blap to post the full dataset on the sandbox. I can create a preliminary analysis if you're interested, but 11 isn't a full sample.

edited 6th Apr '17 11:11:51 AM by pokedude10

It is callous and a form of omission bias: Not having a particular trope can hurt the wiki just as much as having a poor page.

"For a successful technology, reality must take precedence over public relations, for Nature cannot be fooled." - Richard Feynman

I didn't collect that specific data, so I need to comb through the thread.^note The lists were made to help come to actionable resolutions for crashed pages, not for reforming TLP, so it was kind of expected. Luckily, though, the lists that I did compile help check to see if anything was missed through the combing.

The relevant data that I did compile includes what we decided to do (resolution) and where the sent back pages are (in the "resolutions" folder). But yeah, I have to look through for the dates and such, which I don't mind doing, actually.

Update on the sandbox: I added a folder with links to the comments that brought up the crashed tropes. This is very much because I don't have access to my file at all hours of the day, but I do have some time for this project, so... compromise.

Look at all that shiny stuff ain't they pretty

It depends though. As I mentioned a few pages back, TLP is where new ideas come into the wiki. We need fresh input otherwise we become stale, and the status quo is unchanged. I don't want to see more abandoned drafts, but overall, we don't want broken launches either. ^edit  Looking at my post again, I think we were referring to the backlog not hurting tlp, not necessarily abandoned drafts hurting the wiki.

And really, I don't think either "hurts" the wiki in a long-term sense. A crash rescue just takes manpower, and if left alone, may eventually result in a TRS and more manpower (things such as content violations, flame bait, and spam DO hurt the wiki). An abandoned draft is a trope that doesn't get to exist yet, the new idea can't reach the wiki. However. Abandoned is still better than Discarded. It has the potential to launch. Again, I don't want to increase the abandonment rate, it's hard enough for tropers to keep interest.

Honestly, I'm very aware that tlp tends to attract new users with new ideas. I don't want to scare them away or worse, discourage them. We need them. ^note 

That's why I would be behind a clear multi-stage system where the new ideas aren't limited, and it would be straightforward to get it launch ready. Even if the stages were short, it would be doable for new or old users to create a trope.

edited 6th Apr '17 12:09:57 PM by pokedude10

If it's a trope we're really missing, another troper will either find the abandoned draft or make a new one on their own.

I guess what I'm concerned with is diluting our focus from "Make the ones that Launch better" to "Get everything that should be launched, launched." They both need attention but the bad launches actively and immediately harm the wiki, the non-Launches don't.

...if you don’t love you’re dead, and if you do, they’ll kill you for it.

It's not that we're "missing" a specific trope, it's that the flow of ideas into the wiki needs to stay open. We don't think we "need" a specific trope, it's not like we're asking the expected long-term value for each trope. No, that concern is about tropes becoming more difficult to launch, or simply not launching at all. There could be other side-effects that come with that.

We're not trying to change the focus, Donald brought up a side-effect to keep in mind. Yes, we all want to increase the launch quality of works so they can survive on their own. We want them to not need Crash Rescue. But he made a good point, we don't want to overly restrict the flow of new tropes. We all want healthy tropes to launch from TLP that have the ability to grow, but there might be a real trade-off. We should also think about other additions to the process that would counteract the tradeoff.

The blocking the flow of ideas is why I was against anything that would outright block launches.

I don't see how any of the ideas on the table would block the flow of new drafts. They're to help better tropes come out, not to keep new ideas from coming in.

But, a new troper needs to understand how things work here, or they'll be in hot water no matter how great their ideas are. New tropers have been called to the edit banned/suspended thread for launching junk tropes, stubs, duplicates, ZCE-magnet appearance tropes with no understanding of the necessary symbology. There's some big and ongoing reasons the Main namespace is locked to average tropers.

New ideas are useless if not expressed properly. Isn't that a trope itself for a reason?

Coming back to where you started is not the same as never leaving. -Terry Pratchett

~Pokedude10 and everybody wondering about the TLP Reform Statistics. There's enough data on "resolutions" to check that variable, and there should be enough for the other variables to check that, too.

First thing to notice is that most of the crashed or crashing tropes are under two years old. One outlier was about seven years old. I haven't looked at the counts for the variable of "reason brought up," but I think the name, description, and example section were the primary reasons. Three times (exactly), "en masse launching" was an issue (and by its nature, it looks like it's a more prevalent issue than it is), though on a per-trope basis it's much more than that.

I'm actually going to do a statistical test on the age variable. I looked through the launched pages list and found that a reasonable hypothetical average age for all launched pages is about 10 months and two days, so I'm going to use that for the test I'll do within the next day or two.

edited 9th Apr '17 1:08:07 PM by WaterBlap

Look at all that shiny stuff ain't they pretty

I've seen mass launchings before over the last few years. The biggest problem is when one person does it, one or two others think it's okay to launch a few drafts themselves that aren't out of swaddling bands yet. The tropers themselves may not even know better.

So while they're infrequent, they tend to cause messes beyond the initial mass launcher.

But the answer isn't to block launches, since the problems resulting are greater then the problem resolved. Education on site norms is really the only long-term answer. (But isn't good education always that?)

Coming back to where you started is not the same as never leaving. -Terry Pratchett

+1 on educating new tropers (including nominal "old timers" who are still as ignorant on the topic as actual newbies are) on site norms. I've time and again suggested that newly made accounts should get a PM that strongly "advises" the new member to read all of the site guidelines and rules, clarifying that a lot of them go beyond "generic netiquette", and the member will be held wholly responsible for any infraction that results from him/her not even giving a cursory reading of said guidelines/rules.

edited 10th Apr '17 4:26:59 AM by MarqFJA

Fiat iustitia, et pereat mundus.

I agree, education should be favored over punishment. I also think that's why having official medics/mentors/heralds would help guide new tropers to the appropriate resource as well as catching clear issues early.

---

On the note of the data analysis, I've run more numbers over on the sandbox using the 55 with full info.

In terms of age, 36% of drafts were launched less than a month before reaching Crash repair. That's not very long before someone considers a trope "crashed."

Of those younger than a month, 27% had issues with examples (too few), 18% had issues with the description, and 14% had issues with the name. (Another 14% were Not A Trope, while another 14% were not cross-wicked).

So... If 27% of examples younger than a month are launching without enough examples, does that mean the drafts aren't following the Three Rules Of Three? Or is it just they're not getting new examples after launch? The lack of crosswicking seems odd, I think that means someone launched the page and didn't follow up with the necessary work.

On a positive note, 54% of drafts launch within 3 days of the last comment (78% less than a month). While that last comment could just be "launching...", that does indicate that it's not just someone hitting the launch button. There is participation in the draft before it is launched.

Yes, there's more that can be done, but this is a good start.

EDIT: Misread the meaning of age.

edited 10th Apr '17 2:36:29 PM by pokedude10

~Pokedude10 I'm not sure where you're getting that, unless it's what you had already. I didn't report that statistic. The "age" refers to how long the draft was in TLP prior to launch — not how long the page was up before being reported. It's the "age of the draft at launch" rather than "age of the page at being reported." Maybe I'm just misreading you, though...? I just want to be as clear as can be

Essentially, most of the poorly written tropes are either really young or really old.

Honestly, it's kind of a toss up. Some comments were simply "Is this ready?" or "This is ready" and then they didn't wait for a resonse. Note that these are all poorly written pages, given the source of the sample. I think it's more of a negative note that so many drafts are launching within 3 days of the last comment.

edited 10th Apr '17 1:37:51 PM by WaterBlap

Look at all that shiny stuff ain't they pretty

~Water Blap Oh, It looks like I misunderstood your format for the age. I assumed that since these drafts were on the crash rescue thread and that you called them "pages" not drafts, that age referred to the time between proper launch and crash rescue. I retracted the analysis on my last post, and I'll retract it on the sandbox as well. However, I think the "time between last reply and launch" is still correct, right?

Fair point about the last reply though, it is a toss up. While those may be poor quality, a reply before launch is marginally better than no reply before launch. At least a bump opens the door to more replies, hats, or bombs.

Edit: Corrected the sandbox. Knowing the age of the draft is still useful, so I kept it but revised it so it made sense. Also added some more various stats.

edited 10th Apr '17 3:14:42 PM by pokedude10

~Pokedude10: I just used the Created By line in the draft and the Page Created line on the page to find the age of the draft at launch. It's a bit difficult not to call them "pages" because the browser loads "pages" whether it's a published page or pre-published page (so to speak).

I think we're finding different means and std dev's due to conversion of months to days or something. For age, I got a std dev or 629.94 (with the 7 year value) and a mean of 508.62. Very different from 605.45 or 482.75 (respectively).

Also, just in case this also wasn't clear, the number for hats-and-bombs is simply the real count — with a period separating them. It isn't a decimal number (this was done for convenience, using a number keypad on the keyboard). Thus, Actually Real Magic's "19.5" means "19 hats, 5 bombs," which is a ratio of 3.8:1. I would have put up the ratio, but I realized a bit too late that the colon-indentation markup changes all subsequent colons into line breaks.

Sort of like this

yada yada yada

Look at all that shiny stuff ain't they pretty

Blap, the age of pages since creation is going to skew young if you're looking at what gets brought to Crash Rescue, because that thread concentrates on Crashed recent Launches; pages that were made more than a couple of months ago tend to get bounced to TRS is they're brought up in CR.

...if you don’t love you’re dead, and if you do, they’ll kill you for it.

Right. Most of the pages brought up were reported the day of (or the next two days of) being launched. That's one reason why I didn't report that number.

This wasn't meant to be so confusing. "Creation of draft" was found directly on the blue-colored draft that is linked from the Trope Launch List. The "launch of page" was found directly on the trope's Main/ page inside the Page Info panel.

Another way to think of it is to just go to an innocuous example. Uncoffee reports "Page Created: March 20, 2010" and going through its discussion page into the YKTTW archive, you'll see that the draft was initially created "March 17, 2010." So, even though the article on Main/ is about 7 years old (and gets larger every year), the age of the draft at launch will always be 3 days. Those 3 days were what I would report.

I don't think it would skew young if we're looking at the time between the draft being created and proper launch.

edited 10th Apr '17 8:33:16 PM by WaterBlap

Look at all that shiny stuff ain't they pretty

Yeah... thanks for clarifying about the hats/bombs. I thought that when you replaced the colons with periods, I figured you just put the ratios up. Not a big issue though, nothing a little excel magic can't fix.

I added a new table that sorts the issue counts into categories. Focusing on the issue brought to crash rescue may help more at this stage. I think I got off track by trying to analyze the ages. They're still important.

---

Ok, so a few thoughts about current policies that came to me from the reasons. In tl;dr form so they don't leave me before tomorrow...

• What is the current policy about using a crowner in TLP for a name? I always remember it being used only if the OP isn't sure. What if a crowner was mandated?
• Regarding making a better description, I rarely see a lot of direct advice in replies on what to do. I see many telling OP what not to do, but usually just see suggestions to improve the description. Is this something having heralds/medics could help with?
• About examples. Honestly, I think the Three Rules Of Three are a bit outdated. Three examples doesn't really give a new trope great footing once launched. At only three, there would only be those crosswicks, plus any indices linking to the page. I could push for a higher minimum.
• Not A Trope, Chairs, etc. Do we still recommend new trope ideas pass though Trope Finder first?

Again, I'm just leaving these ideas here so I don't forget before tomorrow.

I have been lurking (hey guys), and just thought I'd chime in that I like a lot of the ideas, one that I've brought up in crash rescue before being to bring tags back and to rewire the hats/bombs deal. I also support the reasoning behind "heralds", but am putting in a suggestion to have it so that TLP doesn't require their approval to launch, and that diverse interest tropers are collected for the job: it is true that people are more likely to read and weigh fairly something that they are interested in.

Also, maybe have something like terms and conditions, that new tropers (and everyone if it's brought in) can't launch in TLP until they've read some guidelines (but obviously not going to pop up every time). I could try and find where I did the draft of YKTTW guidelines for that?

EDIT: found it, my updates for the YKTTW Guidelines are a few years old and so outdated, never mind

edited 17th Apr '17 3:09:31 PM by lakingsif

OH MY GOD; MY PARENTS ARE GARDENIIIIINNNNGGGGG!!!!!