-2

It was made evident on the Stack Overflow question Search algorithm but for functions and many others that people with famous names in their fields answer questions and (granted are usually correct) get an insane number of upvotes for their answer that other people with lower reputation points also post.

I can't help that feel that a well known name basically means extra reputation points earned for every answer. Part of me wants to say that there should be some form of bottleneck on reputation: as you gain more you gain less. Then you can have people on 5k answering questions and getting good reputation from it while big names with 50-100k can answer a question, still get reputation point, but the mental number of upvotes for simply being a big name won't tip the scales too much.

Edit:

So for all the people that decided to downvote my discussion on this point here is Sven Marnach talking about it:

Jakob, I actually do think you have a point.

There is a slight bias towards high-reputation users (I wouldn't really call them "famous").

In my experience, however, the most important factors are speed and correctness, sometimes with far too much emphasis on speed.

See Fastest Gun in the West Problem for a discussion of this phenomenon, which I consider far more serious than the slight bias towards high-reputation users. – Sven Marnach 2 hours ago

24
  • 2
    People tend to upvote on anwers from these very trusted users because their experience told them that these users know what they are talking about.
    – juergen d
    May 24, 2012 at 13:37
  • 16
    8 up-votes qualifies as an insane amount these days? May 24, 2012 at 13:49
  • 7
    Perhaps slightly OT and personal, but did you notice that the answer you're referring to also simply looks good? Without looking at the content, the formatting of that answer stood out to me and looked clear and deliberate. Which is often the case with the well-informed and well-thought-out answers of experts. There is a thing to be said for the clarity that comes with experience. Which in turn might explain some of the upvotes (in combination with the content being correct of course).
    – Bart
    May 24, 2012 at 13:58
  • 5
    Heh, I don't recognize any of the users involved in that question.
    – yannis
    May 24, 2012 at 14:00
  • 12
    Jakob, I actually do think you have a point. There is a slight bias towards high-reputation users (I wouldn't really call them "famous"). In my experience, however, the most important factors are speed and correctness, sometimes with far too much emphasis on speed. See Fastest Gun in the West Problem for a discussion of this phenomenon, which I consider far more serious than the slight bias towards high-reputation users. May 24, 2012 at 14:15
  • 1
    There was a question somewhere on the SE network recently asking about something John Carmack (of id games) said recently, and it was answered by Carmack himself. He got hundreds of upvotes; within an hour I think there were over 400. Wish I could find the link but Google is failing me. May 24, 2012 at 14:26
  • 2
    @ErnestFriedman-Hill That was on Super User. superuser.com/q/419070/1170 May 24, 2012 at 14:29
  • 5
    Well, Carmack is a special case. That's not bias because of the famous name, it was about the question's subject answering the question
    – Pekka
    May 24, 2012 at 15:03
  • 12
    There was actually a blog post on the statistics SE site about this very topic. The analysis indicates that high-rep users that get a lot of upvotes always got a lot of upvotes (even when their rep was low). This is likely due to their expertise and the quality of their posts, and not an effect of their high reputation. May 24, 2012 at 16:49
  • 7
    "So for all the people that decided to downvote my discussion on this point here is Sven Marnach talking about it:" So you found one guy who agrees with you. Slightly. May 24, 2012 at 17:16
  • 2
    I just think its a very harsh attitude to have towards beginners, "Oh he made a mistake or a bad comment DOWNVOTE HIM!" May 24, 2012 at 19:48
  • 4
    @JakobBowyer: Says the person who downvoted an answer which was quick and accurate? Wow.
    – Jon Skeet
    May 24, 2012 at 20:47
  • 1
    Okay okay it was a silly response! but I still think the reception here has been less than friendly May 24, 2012 at 21:15
  • 4
    @JakobBowyer: Yes, that is indeed harsh. The downvotes on this question aren't harsh though - downvotes on meta are around disagreement.
    – Jon Skeet
    May 24, 2012 at 21:25
  • 3
    @JakobBowyer: Usually this link accompanies such discussions: meta.stackoverflow.com/faq#vote-differences
    – Jon Skeet
    May 24, 2012 at 22:57

5 Answers 5

47

Here's the reasoning posted as a comment to that answer:

Its not that I didn't get your point, it just feels like you camp on stack overflow and answer quickly (and yes always perfectly correctly) but give others no chance to grow in rep. I mean you have 66K. The other people answering will get 10x less rep on their answer than yours. I don't think that's fair

Frankly, that's stupid (no offense to you personally, but the idea itself...). If the guy's answers are quick and always perfectly correct then why would anyone think it's a good idea to discourage that?

To answer your question, no, I don't think famous names get more upvotes because they are famous. It's the other way around. You get a famous name on Stack Overflow by providing lots of great answers that get lots of upvotes.

Also, we already have a reputation cap of 200 points per day, so the bottleneck is already in place.

6
  • 5
    The funny thing of this answer is that it got a lot of upvotes /and/ is made by a "famous" stackoverflower.
    – akappa
    May 24, 2012 at 13:45
  • 6
    @akappa My sense of irony is usually pretty fine tuned. I really should have seen that coming. May 24, 2012 at 13:47
  • 27
    +1 because you're famous
    – GUI Junkie
    May 24, 2012 at 13:50
  • 12
    -1 Because Bill is so famous May 24, 2012 at 14:21
  • 7
    @ErnestFriedman-Hill Being downvoted and called "so famous" by one of the Sheriffs on Java Ranch is probably going to be the highlight of my day. :) May 24, 2012 at 14:35
  • I upvoted a Jon Skeet post and sent Ernest a message on FaceBook instead of upvoting. Jun 13, 2012 at 15:45
16

First of all, this comment is relevant (emphasis mine):

Its not that I didn't get your point, it just feels like you camp on stack overflow and answer quickly (and yes always perfectly correctly) but give others no chance to grow in rep. I mean you have 66K. The other people answering will get 10x less rep on their answer than yours. I don't think that's fair

The point is that if someone is answering both quickly and perfectly correctly, that's great. That person deserves the rep. (And downvoting out of spite is not a nice thing to do.)

Now, apart from the Jon Skeet (and him getting upvoted is not really debatable), there aren't that many famous names. Really, hundreds of users here have more than 20k in reputation, and I don't believe it influences people's voting decisions that much. If a new user posts a really good answer, it will be treated just as well, and likewise, I've downvoted three-word answers from people who do have a lot of rep.

1
15
+50

Just to throw some data at this I used this query to construct the following table

+--------------------------------------------------+
| Rep  Range | Avg Rep Per Post | Std Rep Per Post |
|------------|------------------|------------------|
| 100K >     |      36.02       |      13.40       |
| 10K - 100K |      33.65       |      33.95       |
| 1K - 10K   |      30.71       |      92.38       |
| 1 -  1K    |       9.20       |      18.10       |
+--------------------------------------------------+

This data shows is that

  • There is a big jump when you go from the 1-1K group to the 1K to 10K group for rep per post

  • There are very small improvements as you go beyond 10K

  • The higher reputation the smaller the variance of the rep per post. This means you can't depend on crazy high upvotes as you go up in rep.

So I contend that higher SO rep does not lead to increasingly higher rep per post

Personally I think that people believe in a strong rep based bias because when you look at a profile its easy to see the hits but ignore the misses.

For example a look at Jon Skeets profile will easily show that he's got nearly 3,000 nice answer badges but they rarely take note of the nearly 4,000 answers that got 1 vote or less.

For a much much more thorough analysis by someone who actually know something about data analysis check out this blog post (thanks to @jadarnel27)

3
  • 1
    There are some more numbers to throw at this in this blog post if you've never seen it. Pretty interesting stuff ^_^ May 24, 2012 at 16:52
  • 1
    @jadarnel27 very very cool, I particularly like the bit about the fixed effects regression model. May 24, 2012 at 16:57
  • I particularly didn't understand those words (not a statistics guy), but I'm glad they made sense to you =) May 24, 2012 at 17:07
3

To answer your title bar question with a purely personal opinion:

Yes, of course it does...sometimes.

That's the very nature of the concept of reputation, which Wikipedia defines as follows:

Reputation is generalized or held view of a person or a group

If you hold with the idea that Stack Overflow reputation is a measure of the trustworthiness of the individual's answers, or the community's generalized view of that person's answers, then it follows that a user with a more formidable reputation will gather a few votes here and there from other users who stop by and pick an answer to upvote without putting too much effort into forensically comparing all of the answers on the page.

In an ideal world, every answer would be evaluated strictly on it's merits - but as long as the answerer's reputation score is visible, a higher-rep user's answer is sometimes going to get the benefit of the odd vote here and there from the casual upvoter. over a lower-rep user's answer.

Edit: See also this related question regarding a tendancy to upvote low-rep users - and my very similar answer to that question.

1

Famous People on SO are famous not because they have done some goofy things like movie stars. Instead they are famous for their brains. They have something in their questions/answers that people take note of them. Definitely if we quote their names or answers, it gives our answer a solid ground.

3
  • Yes and I appreciate this. But the fact that they get all the attention detracts from others who still might be providing the correct answer yet don't have this name for themselves May 24, 2012 at 16:37
  • 5
    "Definitely if we quote their names or answers, that thing adds a star to our answers." What. May 24, 2012 at 16:41
  • @BoltClock'saUnicorn: Now OK? May 24, 2012 at 16:49

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