The strange thing about this is that the StackOverflow team seems to think that people will use these reputation points for hiring decisions. They have little widgets you can embed into your site to display your reputation, and I've even heard them suggest that developers put it on their resumes.
I can't think of a better indicator of somebody not to hire than a guy with 10,000+ reputation on any community site. You can only get that sort of rep with a 4hr+/day commitment over a period of several months.
That time is pretty much guaranteed to happen at the office. I'd hate to have to take up the slack for one of those guys. Worse, I'd hate to have a room full of employees spending their days gaming StackOverflow on my dime.
I have 16K and I haven't spent 4 hours a day on it. I do spend some time at work on it, usually while I'm trying to distract myself from a problem so I can get a fresh perspective on it. I just answer a couple questions a day and eventually you start to stockpile rep. Old questions/answers get you rep like interest on a savings account.
I have the rep on my blog, because it has some perceived value amongst my professional peers (the .NET community) and there was recently a job I applied for where it actually was a useful number. That's the exception though since it's a unique opportunity. I would never put it directly on my resume, but I wouldn't hide it either.
Ha, great response Jason. Maybe it'll feed back to SO devs. These quality scoring crowd sourced systems need refinement. Although I'm a believer in emergent properties of large systems, the initial reputation should be tied to domain expertise. Let a group of really smart folks set the standard and promote intelligent answers.
Out of the 6 tips provided, "write relevant, detailed answers" wasn't one of them. Quite the contrary, in fact:
"1. Be the First to Answer. Even at the cost of quality."
Gaming a reputation system just means you figured out how to maximize a number determined by some arbitrary formula. Real life reputation works differently. People aren't going to remember the dude with 10k points. They will remember the person who guided them through one of their problems. Or the person who consistently shows up with quality feedback under a certain topic.
By gaming the system you're only making it harder for others to use the mechanisms the site provides for signalling quality. Fortunately most humans have the insight to not rely solely on a number.
Tip No. 1 is a bit similar to Jeopardy strategy of ringing in the buzzer as fast as possible and figuring out the answer afterwards.
Sadly, people will respect someone more with 10k gamed points than someone with 1000 honest points, because we have placed implicit trust in SO ranking system.
Looks like Stack Overflow risks the same descent in to mediocrity that plagued Yahoo! Answers. This is the problem with points systems: people will game them, often to the detriment of the site. None of these tips involve actually providing a good, useful answers. Compare with Ask MetaFilter, which provides excellent quality with hardly any game mechanic - instead using community norms to keep the quality high.
It is inevitable that any Q&A system managed by the masses versus a knowledgeable few will descend into mediocrity. The harder the question is, the fewer people will vote on it, and even fewer would be qualified to vote on it.
Having said that, I think Stackoverflow does a reasonable job of coping with it, but abuse still goes on. I only first noticed these tactics when they were used against me. Without naming any names, I can tell you that many of the people with 15k+ rep have used these tactics.
The only advice that really worries me is the advice to downvote competing answers. Getting votes for providing quick answers that you then refine is not a bad thing. But I think Jeff and the team may need to prevent people who've answered already from downvoting other answers, otherwise you could be right.
No, it IS a bad thing. What you get (and what happens on SO) is that certain people will post something, anything, to new posts so they have the first comment. They then spend the next five minutes gradually expanding their answer either based on their own knowledge or worse of all, by cherry picking info from other peoples answers.
This is certainly not the case with every answer, and it it may be more prevalent under some tags than others. However it happens enough that I've become fed-up of crafting well thought out replies just to see them copy+pasted, with some minor word changes, into answers from first-posters.
To be honest I don't see that as a problem; first, the people answering a question are the people who are more likely to come back, read other answers and possibly downvote the very bad ones. Leaving negative comments is very obvious too, you won't get too far with that. Second, it never happened to me (and I've spent way too much time on SO), never seen it happening so I suppose this can't be too widespread.
Yeah, I do the downvoting thing from time to time. If there are few answers, and I think mine is slightly better, I will downvote them so mine goes to the top. This usually results in lots of points, since people see mine at the top and upvote, and see the others at the bottom, and downvote them. (The same thing works here, but I don't do it. My comments usually get upvoted anyway, for some reason.)
The funny part is that someone "noticed" this and complained to me about it. (There was a whole meta thread about it, which I thought was hilarious.) In that case, though, his answers were just plain bad, and I had to downmod his answer and answer the question myself just so that people would not be horribly misguided by his high reputation but wrong answers.
Anyway, I hate to "game" sites like StackOverflow, but there is no incentive not to. If I get banned, it just means people won't get my knowledge for free anymore. I have never needed to ask a question there; I have a social network for that (which can't be "gamed", btw).
(As to why I care about rep, I don't know. It just feels good to get a lot of points, I guess. It's like winning the lottery, except you can't cash in reputation for goods and services...)
Hmmm... I dont do this... perhaps thats why I'm still only around 3500 points :(
I rarely downvote an answer preferring to just upvote correct & good answers. So unless answer is a real "stinker" then I just reserve downvotes for trolls & language flamers
It is a secondary issue if individual users can hack the system to gain somewhat undeserved reputation.
The primary issue for the quality of Stackoverflow is whether good answers are generated, and whether the best (most correct, relevant and informative) answers float to the top. This seems to be the case AFAICT, so I don't think this reputation gaming is a big problem for the site. (It will be a big problem for somebody (theoretical) who hired people based on Stackoverflow reputation. But that would be stupid anyway.)
You've exposed some systematic weaknesses of the SO quality score. If I worked for SO I'd thank you.
My proposed solution is to manually identify X judges who represent a nicecross section of programming backgrounds. These judges are selected by their continued support and passion for helping programmers. Over time their upvotes would create new quality judges in an intelligent manner.
I used a system like this for characterizing 3D shapes in the protein database (kinase). The templates were created from known kinase chains, we called them motifs. Thanks to Tony Sommese, and Jim Dwulit for really making it happen (I just helped out on that project).
Another case for this model is crystal formation. Seed crystals allow for fractal like formation of patterns in supporting media dishes.
Googles voting system also relies on "reliable" sources to help judge the value of links.
Reference papers use citations to determine how close an author is to a renowned scientist.
People that game the system tend to get bored after awhile and go away. Only those that truly like answering questions and helping people stick around. Overall, most of the time questions are getting answered and that's the primary goal of the site.
If you're getting sniped and then having your content copied, post first or post faster. I have never had my content stolen for another answer.
Also if you want more rep with less games, stay away from the high-profile low-hanging fruit. Avoid basic language syntax lessons that are easy for people to answer quickly. Pick a few tags that you have more specific knowledge in and follow those. That's an easy way to get rep without as much competition.
What's missing from this piece is an answer to why I should care about getting Stackoverflow reputation? Besides access to the "special moderation tools", is there any benefit to justify the time I would spend to amass more than 10,000 points? I think that the only reason someone should use a site like Stackoverflow, or HN for that matter, is that he enjoys it. So from that perspective, there is little incentive to "game" the system.
The more interesting question, for me at least, is to consider what kind of tweaks StackOverflow could make to the reputation system in order to improve the quality of the content.
I'm a bit surprised that Atwood & Co. don't seem to be more interested in this. (I opened a question on meta.stackoverflow.com to discuss precisely this point, and got almost no response whatsoever....)
Huhn. I go there to ask questions, and hope people answer them. Helps me learn. I'm not in school anymore, I don't need 'good grades.' I need to know how to do x with y.
These tips describe a fairly "premeditated" approach to gaining reputation, but like anything that yields a reward, it's just reality that people look for optimizations.
I can't think of a better indicator of somebody not to hire than a guy with 10,000+ reputation on any community site. You can only get that sort of rep with a 4hr+/day commitment over a period of several months.
That time is pretty much guaranteed to happen at the office. I'd hate to have to take up the slack for one of those guys. Worse, I'd hate to have a room full of employees spending their days gaming StackOverflow on my dime.