Categories
games industry programming

GameDev.StackExchange: don’t use it for game-dev info

I’ve given that site two serious attempts – I had nothing to gain from it, I was just trying to share best practice and info from within the commercial side of the industry. I wanted it to work. StackOverflow (which it’s cloned from) has been a huge success, and the nearest equivalent for games industry – the GameDev.net forums – is very weak by comparison; lots of people doing their best, but often dominated by those who have time, rather than those who know what they’re talking about. And very few professional members.

Attempt 2 has crashed and burned. And there I’d walk away silently. But … while I was there, I noticed how much misinformation flows around that site, and I can’t keep quiet about that. It’s doubly depressing that it trades off the reputation of StackOverflow – a site that works many times better, and has a many times higher signal:noise ratio.

Here’s a favour to anyone tempted by that site: don’t. If you need answers to the questions you might ask there, there are much better places to get them (if this is tl;dr – just scroll to the bottom).

What’s so bad?

Bad enough that:

  1. many (more than half of the 50+ I read in the last week) of the answers are significantly or fully wrong, but casually upvoted or selected as correct (seems too few people on the site for the voting to “fix” this problem naturally)
  2. most of the community has no idea what they’re talking about (Vague questions, vaguer answers, and STUPID SHIT like “I was talking with one of my professor and we couldn’t figure out why all game engines (that I know of) convert to triangles.”. Really? You’re that dumb / lazy? Incidentally: “professor” of what? If it’s a course involving programming, that professor is a fraud. Personally, I think it’s a homework question. Should have been shutdown immediately.)
  3. many questions (around 20% of those I’ve read) would be solved trivially by typing them directly into google and clicking the first link

Worse that:

  1. without lots of reputation-whoring, you are “not allowed” to comment on a wrong answer (NB: I get the impression that the rep-limit for this is much higher than on SO; certainly, it’s unrealistically high). Your only option is to write a separate “answer” that explains why the first is wrong. This is sadly common on the site. Readers have to read EVERY answer (including the ones at bottom of screen) just to find the corrections to the “top-most” answers
  2. without even MORE reputation-whoring, you are “not allowed to answer more than 1 question every 3 minutes”. You’re a professional game developer, right? You’ve been doing this stuff for years? You see a couple of related questions you can answer quickly (stackexchange explicitly GIVES YOU the list of “related questions” and invites you to answer them) … well, so far as this site is concerned: “Oh no you don’t!”

Ask any game designer: the rep-limits on that side are somewhat FUBAR. They positively encourage people to reduce the quality of information / organization. Any sane designer would have *at the very least*, said:

“If this site is new, and small, with a small community, you need to set all limits low to start with – there’s just not a large enough pool of campable spawns / experience points (*ahem* reputation sources) to support those limits from day one”

Anyway. I tried it. It was disappointing. Unlike StackOverflow, it isn’t (currently) working well – many in the community are full of their own ignorance and don’t want to read anything that’s based on knowledge or experience – they want only things that support their private theorising. Actually trying this stuff in practice? “Whoa! That sounds like actual WORK! Don’t go there, man!”

IMHO, if equivalent questions to many you see on gamedev.stackexchange appeared on StackOverflow, they would get shutdown quickly with some variant of:

  • “this question is irrelevant and trivial”
  • ” this is not a website for getting your homework done for you”
  • “this is not a website for “tutorials””
  • “please just google it – the answer is the number 1 link”

…etc

Revisiting gamedev.stackexchange, Spring 2011

At the gentle urging of friends and ex-colleagues, all of them professional game devs, I tried again. Their winning argument? “Give them time, the community will improve; it’s better now – it’s getting there” and, most key: “if we don’t help, it’ll never go beyond blind leading the blind”.

I quickly found several trails of different people asking THE SAME QUESTIONS over and over again, and getting different incorrect answers. Zero effort to read what was already there. Zero effort from the community in marking duplicates. I found some related to entity systems, and thought “Oh, FFS, this is easy for me – I’ll answer these”.

SMACK! “You can only answer 1 question every 3 minutes”.

Annoyed, but undaunted, I kept going. Just to be clear: my answers were all different, but the questions themselves were close duplicates; obviously, they’d been asked without bothering to check if they’d already been asked. And then a few hours later, I started getting emails about my answers being downvoted. No explanation, just … someone didn’t like them. WTF?

At this point, I realised the futility: it’s actually quite *hard* to get down-rated on StackOverflow. The attitude of SO users is “help, explain, and educate” (shaped by the SO penalities for downvoting). By contrast, it appears easy on gamedev.stackeschange: and it comes with no explanation, no commentary. Drawing from my experience in MMO development, these are the characteristics of a negatively charged online community: it’s not heading towards a flourishing, happy family. Time to get out while the getting’s good.

So, I’ve killed the account. I want nothing to do with it. At least, not until a good community-moderator takes things in hand and changes the culture of the site. Hard job.

Want answers to game dev questions?

So, what are you to do instead?

My advice: send twitter messages @ game designers, artists, programmes, etc. Maybe half of them are on twitter these days, and generally very easy to find. Most of them (including all the non-twitter ones) have blogs, and are happy to help and give copious free advice.

As a secondary source, read a lot of blogs. Blogs are better than following people on twitter – game-dev is too deep and nuanced for twitter, and *nearly* all the good stuff is in blog posts. This may change, but for now it’s the case.

Finally, try emailing people who’s blogs and tweets you read. But this is a last-resort: most people are very busy, and will struggle to respond to your emails with the depth they would LIKE to. Most will – therefore – not answer at all. They’re not ignoring you, they’re just waiting until the get time. After years of doing that myself, these days I try to reply to people quickly, even if all I can say (in 99% of cases) is: “I’m sorry, I don’t have time to answer this properly.”. Where I can, I’ll add pointers to other people to ask – but usually they’re just as busy, so it doesn’t much help.

27 replies on “GameDev.StackExchange: don’t use it for game-dev info”

Yes, the signal:noise is not super great on the site. But some of what you say is just wrong.

I did not need much karma to be able to comment on a wrong answer. I received that privilege within about a week’s time, having answered about three questions.

I have plenty of privileges on the site, and if you check out my user page, I have an “accepted answer” for exactly one question. I have only written answers for 7 questions total, and none of them have been upvoted more than 4 times.

http://gamedev.stackexchange.com/users/378/dariusk

I mean, I like to think that I am increasing the relevance of the HTML5 tag on the game dev stack exchange with answers like this:

http://gamedev.stackexchange.com/questions/7498/techniques-for-creating-sound-effects-in-a-browser-based-game/7506#7506

I don’t spend an awful lot of time there, but if I’m bored I check for recent HTML5 questions and do my best to answer them.

I had an accepted answer, it wasn’t enough.

You need *popular* answers. Yes, you can get lucky, but most people have to reputation whore.

I checked, and the karma levels are the same for SO. that’s insane – SO is an almost infinitely broad remit with masses of members. Gd.stackexchange is a small niche industry and a site with very few members.

Or just read the TCE thread – my experience of being effecirvely gagged out of the site is not unique.

Oh, and NB: the few times I was personally attacked, people apologized. I’m not saying TCE is full of monsters; what I am saying is that I don’t have room in my life to opt in to a community where that happens at all, even with a sincere apology after the fact.

I’d urge you to keep posting Adam, if only because of the quoted comment “if [you] don’t help, it’ll never go beyond blind leading the blind”.

And my eyesight was never too great to start with!

Just noticed – it doesn’t look like you’ve associated your SE accounts with each other (Profile/Accounts).

If you do so you can get an immediate 100 rep boost on all your SE accounts provided your main SE/SO account meets a certain score (100 or 250 has been bandied around).

I was going there regularly to test the waters of the community and I noticed the same issues. Many of the answers that, to me, seemed the most knowledgeable were the ones being voted down.

I use Unity Answers( Unity3d’s Stack Overflow equivalent) since I’m mostly experimenting with Unity anyways, and have been pretty happy with the quality of help.

I feel like one of the major issues is that the percentage of professional developers on the site is extremely low. So, there are an inordinate amount of entry-level/beginner questions that are definitely duplicates and never get flagged as such or removed.

I find about 1 or 2 questions a week that are worth my time to answer; but, they’re typically not all that technical. It’s the same as most IGDA events I go to–more students and wanna-be devs than actual devs.

That’s not to say it won’t improve. That’s why I’m sticking around.

@darius

interesting point. I don’t think of TCE as an online community, not before wondered why. I guess:

– it’s all private. Not even visbile on the web
– it’s tiny. it’s so small that most people know most people after a shor while of Reading
– it’s specifically founded for saying the things you can’t or won’t say in public. Originally it was enitrely anonymous
– it’s heavily gatekeeper / curated: you know that everyone there has been filtered by hand. You know because how many false negatives there are of people getting kept out wrongly

just like a normal online forum changes the nature of what’s considered “normal” etiquette compared to face to face, I believe TCEs unique nature bends the standards too.

Concrete example: very stressed sleep deprived people in fear of their job and afraid for their family will often have tcr as the only safe outlet for work issues and questions. That has leadto some extreme outbursts, and in turn I think has lead to the community norm being set much higher tolerance of “say something stupid then come back and apologize when hove had some sleep” etc

As a special case I’d guess you were also poisoned due to igda associations, unfairly of course. You joined at a time where igda had thoroughly fucked the industry and it’s members

“Many of the answers that, to me, seemed the most knowledgeable were the ones being voted down. ”

Yep. Exactly. I’ve seen this too often now. It doesn’t just lead to a “weak” resource, it actively drives-away all the people who are most needed in order for it to be a good resource. It’s meta-destruction of the community, which is far worse than simply inserting bad content.

PS: …and it’s such a massive difference from SO (where answers are much more stupid, yet down voting is much rarer) that it points to the culture / community of the site as being the problem. I’m open to other explanations, but it matches with my many experiences of small online communities where the “smallness” and an automated rating system make it very easy for a small group of people (or even just individuals) to game the system and take control.

Two things. 1 – apart from a brief stint on a single Android game, I’ve had nothning to do with java game dev for a long time, e.g. I’ve been doing iPhone for last 3 years. 2 – Sun fucked up / made promises and broke them one too many times, and I walked away. Sun’s bad stewardship invalidated much of the value of JGO as a community: if the platform holder won’t support you using their platform to make games, you’re always going to be screwed.

I’ll challenge you to use the GameDev.net forums again, as well as expand beyond the forums and into our development journals (blog) and other areas of the site. As a GameDev.net representative, I believe you’ll get more value out of our site than any other approach, particularly since we work with so many industry partners.

On the forums side … they live or die on the community, and for years the GD forums were hampered by major bugs (like the way that all topics got rapidly locked, causing MASSES of duplicate posts). I think GD still suffers from this – I know plenty of industry people who refuse to go there because of the problems they experienced there in the past.

…and, sadly, the world has moved on: StackOverflow and Quora – both of which work in a different way – are rapidly outpacing / invalidating much of what GD.net forums did.

Maybe GD.net could institute a clone of SO’s system, one which – unlike gamedev.stackexchange – actually worked. Using the existing karma, you could seed the community with members at various levels of points, and maybe sidestep some problems. I expect it would be tough to make work as a community.

But plain forums? I have no idea what you’d use them for…

Just to be clear, I think GD.net is great. And the recent forum re-design is a lot nicer than the old one. The new site design is MUCH easier to understand and navigate.

But … in most cases, I don’t know how GD.net parts would benefit me. In the old design, it was all too overwelming – too much to take in. Now, it’s not overwhelming, but it’s scattered and it’s very hard to understand which parts you are “supposed” to use, and why.

For instance: you have a Calendar. It’s the first link in its section. And yet *off the top of my head* it’s missing at least 95% of the content I expected (conferences, events etc in the next 6 weeks) – and I don’t even keep track of this stuff any more; this is just what I’ve heard about by accident. So … I have no idea why/when I’m supposed to use that calendar?

“without lots of reputation-whoring, you are ‘not allowed’ to comment on a wrong answer”. I don’t understand. I automatically get 101 rep on every SE site for merely linking my new account with my SO account.

I agree that some questions are riddled with ignorance, but you can easily vote the good ones up until you can also vote these down. Some people are actually trying to make this a good place, and you leaving just doesn’t help. You should let the GD.SE community time to grow and mature, especially if your first piece of advice is to send twitter messages to programmers. Seriously?

If you read the comments above, you’ll see that the linking HAD NO EFFECT for my account. Maybe I just got unlucky, and got caught with a bug on the SE backend :(.

But … I hear enough complaints from other people who use SO that it seems I’m not the only one who got no up-kick from the account linking.

NB: I’ve now unlinked the accounts, and that had no negative effect either. Whatever it was didn’t un-trigger :).

re: twitter – um, yes. They love it. At least, the ones who are using Twitter do. That’s rather the point of Twitter – if you don’t want to be messaged, don’t tweet.

The “plain forums” will always be around. There are many advantages of a forum system over a Q&A system like SO – one is that a forum also offers a place for general discussion on the technical topics rather than just posting questions and getting answers. We strongly disagree that Q&A sites are displacing forums and rather believe that Q&A sites are complimentary to forums. It’s like saying twitter will replace chat systems – chat isn’t going to go away because there’s another method of realtime communication.

Regarding rating and such, we have plans to revamp ratings and reputation for people as well as content.

I’m also not sure about bugs that prevent people from going to the forums. We never had an issue with threads being locked, so maybe you’re confusing GDNet with something else. Anyway, I also know a lot of industry people who do visit the site on a daily basis. I also know a lot of students, indie developers, hobbyists, and people in non-game-but-related industries who visit the site. So I’m not quite accepting of the argument that people don’t visit GDNet because of problems they’ve had in the past.

One area of the site that has exploded in growth is the developer journals (blogs). One of our members’ blog posts is actually going to be published in GD Magazine. Lots of great information is being generated by the community in that area of the site, and it’s turning into a great compliment to the forums.

We’re actively working to revamp the site into something that better integrates the content and aggregates it in a way that you, the user, are most interested in seeing. So again, I’d invite you to keep an eye on GameDev.net. It’s possible we may not be what you need – maybe you simply want a quick Q&A type site instead of an industry community – but we have improvements coming that will address a lot of our deficiencies.

Yeah I’d noticed the journals had taken off. They’re still not perfect, but it’s the biggest collection of quality ones I know of.

I’d love to see some tools making it easier to browse/subscribe to them. Eg automatic Twitter feeds on a per-journal basis, tight integration with Twitter lists, etc. Maybe that’s alteay live. It’s just very tough managing them as a user – you don’t have time to sub to all, nor to have them in your rss reader permanently.

Re: visitors – sure, there’s a lot of people that use gd. But as a percentage of the industry professionals it’s disproportionately small. It’s too insular and isolated; most people I’ve worked with know of it, but never use it.

(Full Disclosure – One of the community moderators on the GD.SE)

Like Sam said, there are people who are working to make this a positive place. It would have been awesome if (in the past, or if you ever decide to come back) you would flag questions that you believe are problematic and/or join in our Meta Discussion (http://meta.gamedev.stackexchange.com) or in our chatroom (http://chat.stackexchange.com/rooms/19/game-development).

Unfortunately, using both of those also has a bit of a rep requirement. It has been argued that for the younger SE sites this would be a bit of a problem. But everything is handled in the context of the trilogy* (outside of the sites that are still in beta at least).

Either way, we welcome the assistance of users that want to get involved in making the site a better place.

*(In fact, if GD.Net wanted to use the SE system, they would have the same problems. They would need to use one of the open source SO clones to do it their way)

Since I’m using Unity I go to answers.unity3d.com though not all the time, but I’d say it does well for itself. Sometimes when I’m bored, I do their homework for them and actually post code, but I make it a point to explain the concepts behind them.

Seems to me from what I’ve read of your post, the community of gamedev.stackexchange are either amateurs or people who do not want it to be successful.

you’re overreacting, or angry about something you’re not telling us about, the site’s moderated, any open questions are being closed, there’s not much difference between it and stackoverflow, you’re just picking the worse examples, you could go to stackoverflow and find same questions, stackoverflow is more general than gamedev and i’m sure you don’t understand 100% of the topics that are posted there, so the overall feel is that it’s “smarter”, while gamedev is near and dear to your heart, so you get angry everytime somebody’s wrong or asks a bad question

Your comment is a year after the original post, so … I can imagine (I’d hope) the site has changed a lot in that time.

As it happens, I was on there the other day accidentally through a google link (looking for a ref to mist/smoke algorithms), and was disappointed – there was a perfect question about it, but no especially good answers.

I found that going back to Google, there were better links on the first page of results … than in the answers on gamedev.SE. So, my guess is: most of the experienced gamedev people are still avoiding gamedev.SE?

lol, a lot of your complaints under your “bad enough that” listing also exist in stack overflow. It’s just that the community is bigger and eventually the right answers pop out. I ended up here cuz I searched how bad/good the gamedev stack exchange was after seeing how very few viewed my question and yes, the only guy to answer me did not have a full understanding of the framework (even though i admit my question was poorly worded).

i guess the only way for now is to talk directly to the people who created the framework still (duh, im stupid).

UPDATE: February 2014

Looking for an answer to a relatively common OpenGL issue, one of the top 2 links was a stackexchange question.

The highest voted answer (with 3 upvotes) was half right – and then horribly wrong and misleading.

But theres nothing I can do about it – the account linking stuff now “works” but even with 14,000+ rep on SO.com, I can’t even downvote a terrible gamedev.stackexchange answer.

…my original complaints stand: gamedev.stackexchange is a travesty (and a pain in the *** for clogging-up Google with crappy results)

Comments are closed.