Over on the nominations page, one of the nominations has a large comment discussion about this. Some of us have been hoping that the person who originally brought up the concern would bring the discussion to meta, but it doesn't look like that's going to be happening, and it seems worth discussing not only on its own merit, but because one particular nominee has received some negative attention over the issue.

The question is, are Googled answers a problem for StackExchange? The issues I've seen listed are:

  1. There is no guarantee the Googled answer is right. I've seen one of the major Google answer people give a wrong answer before.

  2. There is no guarantee that the site the answer was found on has a license permitting the content to be used here. People who write FAQs for GameFAQs tends to explicitly forbid this sort of thing.

  3. "Google answers" suffer from "answer rot." Similar to "link rot," "answer rot" is when the answer becomes wrong but is never fixed. A Google answer to a question could be wrong as of any update. Every question asking about quests or leveling paths in WoW were made wrong by the releast of the Cataclysm expansion in December 2010, due to major changes to the game's zones and game mechanics. I understand something like this also happened in the more recent versions of . Since people who post Google answers don't know the games, they also don't know when the game changes, let alone if those changes invalidate their answers.

Points 1-3 originally written by Powerlord.

share
22  
I'm not going to write up an answer, the ones that exist are fine, but I want to protest the misconception that an answer that written based on information gained through a search or a wiki lookup is somehow not an 'expert answer'. There are a variety of characteristics that differentiate an expert in a field from someone who isn't, and one of those is not just having more knowledge, but, perhaps more importantly, having the skillset and baseline knowledge needed to fill in any gaps as they come up. Being better at searching is an expert skill that will help provide expert answers. –  LessPop_MoreFizz Feb 7 '12 at 3:37
add comment

7 Answers

up vote 26 down vote accepted
  1. Any answer can be wrong. This is what downvotes are for. The bigger concern is users who blindly vote and/or accept without checking the answer.

  2. I'm no legal expert, but anyone should be able to rewrite the relevant information and cite the site(ha!). If this is done in a manner that makes the answer bad, that is what downvotes are for.

  3. Games are updated on a regular basis. Answer rot is a valid concern, but "Google answers" are not the only source. My answer to a Terraria question was rendered incorrect by a recent patch. I play Terraria (ask John and Tristan) and wrote that answer from experience. By the time this had changed, I had completely forgotten that I answered that question. It wasn't until someone pointed out to me that the answer was outdated that I did anything about it. I saw that there were already other correct answers, so I deleted my answer.

Therefore, (correct) "Google answers" are not a problem. They fall victim to the same pitfalls that any source can. If anything, it is the way inexperienced (or sometimes experienced) users react to answers that is the problem.

share
    
Point #1 all the way. Users voting on answers should be familiar with the subject at hand, unless the answer is exceptional -- either obviously terrible, or obviously amazing (such as an SC2 post that others have verified from tzenes, who I know has played the game a lot). –  Matthew Read Feb 6 '12 at 16:53
    
Plagiarism—copying content without even attributing the source—is a little more serious than adding the source back in later: treating it as such aligns us pretty close with content farms. There is an insistence that content be original and well-cited: it's even against the terms of service to provide content for which you cannot yourself license as CC-BY-SA. –  user3389 Feb 7 '12 at 11:06
    
@MarkTrapp - I didn't mean edit the source in later. I meant starting with an answer that has its source cited. I said rewrite to emphasize not just posting a link to the content. –  user9983 Feb 7 '12 at 14:03
add comment

There is no guarantee the Googled answer is right. I've seen one of the major Google answer people give a wrong answer before.

From a moderator's perspective, incorrect answers are not subject of moderation, typically. So long as it is an answer, all that we really can do is downvote like any other human being.

If you do find a question with an accepted but incorrect answer, just post a new one! It happens all the time and having Google as your source of information does not matter here.

I do have one exception on record, but that was due to an incorrect, "joke" answer with a massive amount of votes that couldn't be fixed while keeping the joke and that the author was not interested in fixing, and well, exception handling is the core of what moderators do.

There is no guarantee that the site the answer was found on has a license permitting the content to be used here. People who write FAQs for GameFAQs tends to explicitly forbid this sort of thing.

If you're copying content, you're copying content and there's nothing googling has to do with it. If you find plagarized content, please flag it for moderator attention and we'll take care of it.

Don't plagarize. If you find something amazing on the internet, consider writing a summary in your own words explaining the core of the matter and then link to it. If you must, quote, but keep it short and sweet.

Since people who post Google answers don't know the games, they also don't know when the game changes, let alone if those changes invalidate their answers.

Information rot is something that does concern me, and I have an outstanding feature request that should help in this regard.

That said, I do have a number of rotten Minecraft answers. I've played the game, I've occasionally sourced my answers with other resourced (Minecraft wiki on top), I try to keep up to date with what's changing, but... I have written, what, 70-odd answers on this game in the course of the last year and a half?

I can't remember each and every one of them, and going back to look at every answer of mine to see if it's aged well is... well... it's not the kind of "fun-sized unit of work" this site works on. It's capital-W work, and one I've been slacking off from. Bad badp, bad!

I see this as a problem in the system, and certainly it worries me. I believe anonymous feedback can be an important signal for identifying rotten content, hence my feature request.

At the end of the day, though. Just because you googled your answer, it doesn't mean it's more likely to rot. This depends entirely on the game at hand.


In general, information rot is perhaps the biggest challenge Stack Exchange as a whole faces right now. In my humble opinion, it's a system-level issue in need, right now, of a system-level answer.

Using Google to source your answers is both good and does not directly lead to information rot. Certainly, it doesn't help however. So here's my advice:

If a question can be easily answered by Google, don't just stop at the first hit. Dig deeper, look into the matter, do some science. Write something original, show us the money.

If a question can't be easily answered by Google, yet you managed to (likely, the asker is using the wrong terms), there's really no shame in answering with search guidance. You're genuinely helping the community, you're genuinely helping Google and you're genuinely making the internet a better place to be.

So no, I really can't bring myself to say "stop Googling your answers dammit!" Typically, we can tell when an answer is a cheap rip off from Google (if you have suspicions, just select the text, right click and search it on Google itself!) and the answer will be voted accordingly.

Generally, sourcing your answers is a great thing to do.

That said:

  • If you feel the community is upvoting a specific incorrect answer, post your own, talk about it on chat and let us know. If this answer has too many votes to be brought back in its place, poke us and we'll try to right the wrong, somehow, hopefully.
  • If you feel an user is engaging in copyright infringment, bring proof. What did he copy from what? We take this kind of stuff pretty seriously.
  • If you feel a contest is bringing people to post about games they haven't played, the issue probably is more with the contest rather than the user. Talk about this specific thing on meta.
share
9  
If you feel a contest is bringing people to post about games they haven't played, the issue probably is more with the contest rather than the user. Talk about this specific thing on meta. +1 million for this. –  z - Feb 6 '12 at 15:32
add comment

Mark Trapp and tzenes both appear frequently in the query Raven links (or by total) in his answer. Both are known for in-depth and high-quality answers, including much research and expert knowledge. It doesn't make sense to say that this query implies something negative about those (in this case, agent86, not that I have any problems with Raven, LP_MF, yx, or Kevin's answers) who share their company.

Saying, without any evidence, that agent86 hit the sweet spot and is getting blind upvotes is honestly kind of offensive, and I understand why he's taking issue with Raven's answer. Jon Skeet's old answers (from before he was recognizable by rep) got just as many votes as his newer answers, for example. The SE model tends to work, I think. Let's take a look at the actual data for agent86:

graph

I cut out one answer with 137 score to prevent compression of the graph, but its affect on the trend line was not noticeable. As you can see, the slope of the trend line is very slight and essentially meaningless. His first answers are just as recognized via votes as his new ones.

You can argue that old answers have had more time to get votes, but think about how long questions usually remain active and visible. Regardless, jumping from that argument to "he's getting blind upvotes" is completely ridiculous. It might be true, but it's completely unsupportable with fact, and going by the facts we do have (actually looking at his posts and evaluating them) I think I can safely say that by and large he earns his votes.

I don't really understand what Powerlord's problem was, and it's not fun to see an active user leave, but I think we ought to stop singling agent86 out immediately. It's simply inappropriate, and I think especially so given the election (users who can't follow all of this may just see controversy and get a negative impression). I applaud Sterno for making the question neutral and focusing on site issues instead of Powerlord and agent86; let's follow that example, please.

Blind upvotes might be an issue for SE and Googled answers may have issues; I don't dispute that. I'm all for being circumspect with votes and properly evaluating and writing answers. I think that's all we need to do, though.

share
    
Wow. I am not worthy of this much SCIENCE. Thank you. –  agent86 Feb 7 '12 at 19:59
6  
I wish I could give you all the upvotes. Thank you for bringing science in to show that people are being silly about this –  Dave McClelland Feb 7 '12 at 20:02
add comment

I don't pay much heed to the argument that there is no guarantee that Googled answers are right, simply because there is no guarantee that any answer, Googled or otherwise, is right. Incorrect answers that receive upvotes are a problem, but I don't think Google is the cause.

As far as the specific example of copying content and GameFAQs in particular, my experience as a reader and contributor to GameFAQs is exactly the opposite... most authors don't mind if you copy their work so long as you aren't trying to make money off it. And most FAQs I've read have specifically stated if someone had a problem with it. But that specific example aside, I don't know enough about this issue to know if legally it's a real problem or not.

When it comes to answer rot, every answer risks suffering from answer rot. Googling aside, does this mean I shouldn't leave an answer on the site if I don't think I'll be around a month, year, or decade from now to correct it myself? Am I responsible for maintaining and updating every answer I leave, for all of time? There are already tools in place to combat this problem.

If someone literally knows nothing about the game and slaps the first thing they find on Google as an answer, then yes, that can be a problem, particularly if what they cut & paste looks good, which will tend to give it upvotes, correct or not. But how often is this really the case? I've answered quite a few SWTOR questions that were a combination of my personal experience with the game along with researched (a.k.a., Googled) knowledge. So I think this is definitely a problem if the answerer doesn't have enough experience to even know if the answer is correct or not, but again, that's what downvotes are for. What other alternative do we have? Reject correct answers because we don't like their source?.

For me, the issue really comes down to whether or not the answer is correct and well-written, not where it came from. The community has tools in place to deal with out of date answers and incorrect answers, regardless of their source. And lastly, if SE has a policy against Googled answers, how exactly does one police that? Short of an extremely obvious cut & paste job, there's no way to know whether an answer was Googled or not.

share
3  
+1 I agree with everything you've said. Most of the issues raised are just symptoms of a bad answer, regardless of the source. Most of the "issues" are solved by proper voting. If anything this issue is more of an indicating of users up voting answers without knowing if it is correct. –  Wipqozn Feb 6 '12 at 14:31
4  
I personally see no issues with googling answers as long as the answerer is responsible. Don't just quote something from a yahoo answers that's over a year old for a MMO. But @Wipqozn is correct, proper voting will naturally solve those issues. Dealing with outdated answers is a problem for every game that's not static even if the original answer wasn't googled in the first place. –  z - Feb 6 '12 at 14:43
5  
Also, whats wrong with obvious cut/paste answers? Some questions naturally trend towards c/p answers since its obvious the question asker could've found out if he bothered to try googling it in the first place. I see that as more of a symptom of the question than the answer. –  z - Feb 6 '12 at 14:45
    
@yx. IMO, it's only a problem when the cut & paste answer is WRONG (so, just like any other wrong answer). I mentioned it simply because well-written answers, regardless of correctness, tend to get upvotes, so if you copy and paste something that is well-written but wrong, it'll probably attract upvotes. But no, it's not inherently any worse than writing an incorrect answer all by yourself. –  Sterno Feb 6 '12 at 14:48
add comment

I think the core of this issue is a different question altogether. If you peel this back to its bone, the real question is:

What are the qualifications required to post an answer on Gaming.SE?

In my mind, there are two major ones:

  • A reasonable belief that what you are posting is correct and complete
  • A willingness to accept the consequences if you are incorrect.

There are probably some others that are secondarily important:

  • A willingness to revisit answers over time so that they don't become stale.
  • A willingness to accept that a better answer may come along later.

These are important - no doubt. But I feel like if we made them "primary" we'd exclude many "drive-by answerers" with little to no rep from the site. However, as "power contributors" we acknowledge these things are part of our responsibility.

Plagiarism or copyright violations are a different topic. I don't condone copyright violations in any form, and I encourage everyone to be extra careful about this sort of thing.

We seem to be discussing adding yet another qualification:

Your knowledge or expertise should (must?) contain some (or all?) of the following:

  • Come from first hand sources or original research
  • Be acquired by experiencing the same problem or question the asker has
  • You must be active in experiencing the latest changes that might effect the answer

Don't get me wrong - these things can improve answers. They are not guaranteed to, but in most cases they do. However, as a criteria for judging answers, I find this problematic. How do we evaluate this?

  • Should each answerer be asked to prove that they've played the game in question?
  • How long ago should they have played it?
  • What if they played it for a while, but don't anymore?
  • What if they haven't bought all the DLC, and maybe one of the DLC packs makes a difference?
  • What if they just played the demo, or read a review that covered this topic?
  • Is watching a 2-hour Let's Play or a Speedrun enough?
  • What if they played the game, but had to look this particular question up?
  • What if they've played a bunch of games in the series, but not this one, is that enough to form a basis to provide an answer?
  • What did they know before they read the question, versus what they had to go and do research to determine?

In short, how are we to evaluate this person's qualifications to answer, if we're going to filter 'good answers' from 'bad answers' and vote based on what we believe their level of personal expertise is?

What I see happening now is, if a post cites sources, especially multiple sources, or it answers the question but doesn't provide tons of specific/irrelevant details, people are assuming a lack of personal experience, and they're downvoting answers and taking the answerer to task over this.

The problem is, citing references should be a way to improve an answer. Giving as much information as the asker needs should form the basis for a "good" answer. There is certainly a continuum of answers possible to a question, and some are better than others. The core systems of the SE network are designed to handle this situation.

share
2  
I agree with most of your post, but am not so sure about calling a willingness to revise your answer as a required qualification to post. I mean, that willingness is a great quality in a user, and it would be a nice bonus. But if someone is encouraged not to post a great answer simply because they don't have the willingness to stick around and maintain that answer in the future, I think that would be a shame. –  Sterno Feb 6 '12 at 15:54
    
@Sterno, yes, that's why I left it as a "secondary requirement" - I was probably editing the exact clarification you're mentioning in as you were reading it :) –  agent86 Feb 6 '12 at 15:56
3  
I think a willingness to revise answers is a subset of willingness to accept consequences of being wrong. Anyone answering should be aware that games get updated and answers can become incorrect. If they don't want to update their answer that's fine, but they must accept the downvotes they receive when it does –  Dave McClelland Feb 6 '12 at 16:11
    
I think we all agree that it's important to the site that we keep answers fresh, and as regular participants we have some level of responsibility in this regard. I like to think that when something goes out of date, I'll know because someone will comment on it or similar, so that I have a chance to fix it. I do occasionally audit my own answers though. –  agent86 Feb 6 '12 at 16:29
add comment

I wanted to highlight a part of badp's answer, as I think it's really important:

If a question can be easily answered by Google, don't just stop at the first hit. Dig deeper, look into the matter, do some science. Write something original, show us the money.

Not doing this, I think, is why people get annoyed with the "Googled" answers: they're kind of thoughtless and in some ways condescending: "eh, you could've spent 30 seconds Googling it yourself like I did and answered your own question. Dummy."

I wish the people who do go for the Googled answers assumed good faith more often, or assumed the original poster isn't completely incompetent and already considered the obvious answer. If the answer was so easy to find, why was the question asked at all?

I don't really know why people don't do that more often: I think maybe—given some of the contests we've had recently—people think a fast answer is better than a thoughtful one. For urgent situations, yeah, I suppose that's true.

But gaming is a recreational activity: it's not mission critical for most people. And what sets Gaming.SE apart from our competitors—stuff like GameFAQs's answers board and Yahoo! Answers—is the level of thought and science we put into our best answers. We should be emphasizing—and insisting—on that level of quality for all our answers, not just a small subset.

So right isn't the bar we want to hit: useful is. Did the internet just get a bit better with an answer? Will someone see that answer, having read the content on GameFAQs or Yahoo! Answers, and say "wow, that's interesting!"?

If not, and all we're doing is shuffling bits around between the dozen or so gaming help sites by copying answers, is all this really worth it?

share
add comment

I don't see any fundamental problem with answers that rely on external sources instead of personal experience. That's probably not a big surprise as I'm moderating an SE site that is all about finding reliable external sources.

The important part when your answer is not based on direct personal experience is to judge the reliability of your sources. An official answer from a developer is obviously a very good source, a random forum post might be accurate, but it's not reliable by any means. A wiki that is used by a large number of players tends to be pretty accurate, though there is still the danger that it has become outdated.

One very important aspect of creating a trustworthy answer is making it easy to follow or reproduce the answer. Answers that explain how they got to their conclusions are just far more useful than answers that are just statements without any explanations how the user arrived at that knowledge. The exception are answers that are directly and easily verifiable ingame, voting works very well on those and they get quickly corrected if they're wrong.

It is very unfortunate that this issue was brought up tied to one specific user, relying on external sources is really a widespread practise on this site. My subjective impression is that a large number of answers is directly or indirectly based on external sources, not direct personal experience. Many answers quote some wiki directly, or the users are writing answers based on something they read a while ago, but never independently verified themselves.

What I expect from an answer based on external sources is the following:

  • The answer should indicate where the information is from, this allows readers to judge the reliability themselves by examining the sources.
  • It should not be based on the first result you can find, but if possible on multiple independent sources.
share
add comment

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .