Hello and welcome to our community! Is this your first visit?
Register
Enjoy an ad free experience by logging in. Not a member yet? Register.
Results 1 to 13 of 13
  1. #1
    Gütkodierer
    Join Date
    Apr 2009
    Posts
    2,127
    Thanks
    1
    Thanked 426 Times in 424 Posts

    signal / noise ratio user stats

    I was wondering if there were any signal to noise ratio user stats anywhere on this site, or, if not, whether anyone else would be interested in that sort of thing.

    This would probably mean number of thanked posts / number of total posts (for users with, say, more than 100 posts total).

    I believe the 'thank user' system has not always been in place, so obviously posts before that would have to be disregarded.

    I myself try to keep the noise as low as possible (or practical, while still having fun here) and be helpful as much as I can, and I'd be very interested to see how I compare to other users in that regard.
    Last edited by venegal; 01-15-2011 at 11:30 PM.

  • #2
    The fat guy next door VIPStephan's Avatar
    Join Date
    Jan 2006
    Location
    Halle (Saale), Germany
    Posts
    8,927
    Thanks
    6
    Thanked 1,041 Times in 1,014 Posts
    You mean stats that show how many crap posts a user has made and how many useful ones?

  • #3
    Gütkodierer
    Join Date
    Apr 2009
    Posts
    2,127
    Thanks
    1
    Thanked 426 Times in 424 Posts
    Quote Originally Posted by VIPStephan View Post
    You mean stats that show how many crap posts a user has made and how many useful ones?
    Well, "crap posts" sounds a bit harsh, since many posts that have not been "thanked" will in fact be quite helpful.

    We all know that sometimes a simple link to a manual page will get thanked, while a very time consuming response won't, and some users just prefer to show their gratitude verbally, which of course is fine too.

    All in all, though, I think that once a user has reached a certain number of total posts, the ratio of total posts to thanked posts is a very good indication of helpfulness, and that's what I'm interested in.

    Of course, I can check that out myself on a per user basis, but I'd very much like to see some sort of comparison between the top "helpful" (using this ratio as definition) users.

    There's already top thanked and top posted stats, but I'd find the one I'm proposing much more interesting than either one of those (with the top thanked one still being more interesting than the top posted one).

  • #4
    The Apostate Apostropartheid's Avatar
    Join Date
    Oct 2007
    Posts
    3,215
    Thanks
    16
    Thanked 265 Times in 263 Posts
    As you pointed out, there is no even vaguely accurate way to measure this. Sometimes a user will make a very helpful post and the OP will not return; sometimes they are not thanked; sometimes they're just not recognised. As Thanked posts are the only statistic we really have, it won't work; using Stephan as an example, he has a thanked/normal ratio of about 1:10. The accuracy is dubious at best and farcical at worst.

  • #5
    Gütkodierer
    Join Date
    Apr 2009
    Posts
    2,127
    Thanks
    1
    Thanked 426 Times in 424 Posts
    Quote Originally Posted by Apostropartheid View Post
    As you pointed out, there is no even vaguely accurate way to measure this. Sometimes a user will make a very helpful post and the OP will not return; sometimes they are not thanked; sometimes they're just not recognised. As Thanked posts are the only statistic we really have, it won't work; using Stephan as an example, he has a thanked/normal ratio of about 1:10. The accuracy is dubious at best and farcical at worst.
    Sure, it would be silly to assume that 9 out of 10 of Stephan's posts aren't helpful, and interpreting it that way is not at all what I'm suggesting.

    What I'm saying is that all those different thanking behaviours of different user types seeking help here apply to all of us in exactly the same way and ratios, and once the sample is large enough (let's say 500 posts, to make it a bit larger than my previous suggestion), this will be a very comparable measure, and a very intersting one at that.

    If two users both have 500 posts, and one has been thanked 20% more often than the other one, sure you could say the other one just had the bad luck of perfectly helping lots of people that didn't appreciate it. But it's 500 posts. I tend to believe that having reached such a number, they both had roughly the same share of unappreciative users, and one of them is just more helpful than the other.

    It's not my intention to conceal what this is about for me either — as mentioned in the first post, I believe that I am a very helpful member, and I want to see where I stand in comparison; I'm competitive that way. And I absolutely don't think that my suggestion is just a random number that doesn't say anything.

  • #6
    Rockstar Coder
    Join Date
    Jun 2002
    Location
    USA
    Posts
    9,074
    Thanks
    1
    Thanked 328 Times in 324 Posts
    Quote Originally Posted by venegal View Post
    What I'm saying is that all those different thanking behaviours of different user types seeking help here apply to all of us in exactly the same way and ratios, and once the sample is large enough (let's say 500 posts, to make it a bit larger than my previous suggestion), this will be a very comparable measure, and a very intersting one at that.

    If two users both have 500 posts, and one has been thanked 20% more often than the other one, sure you could say the other one just had the bad luck of perfectly helping lots of people that didn't appreciate it. But it's 500 posts. I tend to believe that having reached such a number, they both had roughly the same share of unappreciative users, and one of them is just more helpful than the other.
    The problem with that is not all problems and solutions are equal in the sense that user A might spend 15 posts helping the a user address and issue and user B might spend 3 posts helping another user and they both might only get one or two thanks.

    In addition one post can receive multiple thanks.

    Quote Originally Posted by venegal View Post
    It's not my intention to conceal what this is about for me either — as mentioned in the first post, I believe that I am a very helpful member, and I want to see where I stand in comparison; I'm competitive that way. And I absolutely don't think that my suggestion is just a random number that doesn't say anything.
    Doing so might provide statistically interesting information but it would be factually worthless. I don't think your suggestion is a random number, it would say something but because of the wildly unreliable ways people are thanked the number wouldn't mean much.

    I think the reputation system is a much more reliable indicator since it allows for negative feedback. But even fewer people use it than the thanks system.
    OracleGuy

  • #7
    Gütkodierer
    Join Date
    Apr 2009
    Posts
    2,127
    Thanks
    1
    Thanked 426 Times in 424 Posts
    Alright, I see that it's not a statistic people are eager to see on the front page. Would it be possible, though, to run a quick query against the user base and post such a statistic — just once, for the fun of it? Who would have the power to do that? It would sure help me with not starting to think about doing something really stupid like screen scraping the whole thing and making the stat myself.

  • #8
    Rockstar Coder
    Join Date
    Jun 2002
    Location
    USA
    Posts
    9,074
    Thanks
    1
    Thanked 328 Times in 324 Posts
    Quote Originally Posted by venegal View Post
    Alright, I see that it's not a statistic people are eager to see on the front page. Would it be possible, though, to run a quick query against the user base and post such a statistic — just once, for the fun of it? Who would have the power to do that? It would sure help me with not starting to think about doing something really stupid like screen scraping the whole thing and making the stat myself.
    You would want to talk to WA.
    OracleGuy

  • #9
    WA
    WA is offline
    Administrator
    Join Date
    Mar 2002
    Posts
    2,596
    Thanks
    2
    Thanked 19 Times in 18 Posts
    Not quite the top 100 thanked users, but here's a more comprehensive list: http://www.codingforums.com/misc.php?do=cybstats
    - George
    - JavaScript Kit- JavaScript tutorials and 400+ scripts!
    - JavaScript Reference- JavaScript reference you can relate to.

  • #10
    Supreme Master coder! Philip M's Avatar
    Join Date
    Jun 2002
    Location
    London, England
    Posts
    18,310
    Thanks
    203
    Thanked 2,563 Times in 2,541 Posts
    “Get your facts first, and then you can distort them as much as you please: facts are stubborn, but statistics are more pliable”. - Mark Twain

    “I can prove anything by statistics except the truth” - George Canning

    84.7% of all statistics are simply made up.

    I have to say that I do not see what the practical value of such an index might be. I think that both the number of thanks and the rep points should be treated with a certain reserve. There are far too many variables and subjective influences affecting both values.

    And how do you interpret:-

    No of posts 1 Thanked posts 1 Index 100%

    Another pointless fact: There are more Turks living in Berlin than in any Turkish city except for Istanbul and Ankara, making Berlin Turkey's third largest city.

  • #11
    Super Moderator
    Join Date
    May 2002
    Location
    Perth Australia
    Posts
    4,108
    Thanks
    11
    Thanked 101 Times in 99 Posts
    Quote Originally Posted by iBall View Post

    However, I think an improvement on the thanked system would be to also publish the percentage of thanked posts to the total number of posts.
    shame for those who made most of their posts before the thanks system was even introduced

    ..please feel free to thank me for that snippet of information
    resistance is...

    MVC is the current buzz in web application architectures. It comes from event-driven desktop application design and doesn't fit into web application design very well. But luckily nobody really knows what MVC means, so we can call our presentation layer separation mechanism MVC and move on. (Rasmus Lerdorf)

  • #12
    Supreme Master coder! Philip M's Avatar
    Join Date
    Jun 2002
    Location
    London, England
    Posts
    18,310
    Thanks
    203
    Thanked 2,563 Times in 2,541 Posts
    Another statistic of how valuable a member is might be the number of times that a user has been banned from this forum (under different user names). iBall would probably qualify for the world's record here. It is an achievement that for some reason he does not mention.

    All the code given in this post has been tested and is intended to address the question asked.
    Unless stated otherwise it is not just a demonstration.

  • #13
    Regular Coder
    Join Date
    Apr 2011
    Posts
    286
    Thanks
    2
    Thanked 39 Times in 39 Posts
    The only way such a statistic would be helpful is if you implemented a "Not helpful" option and somehow use that with the Thanks.

    Gauging people by thanks/total posts would cause people that cared about this statistic to avoid certain users/posters to keep a relatively high ratio, or not bother to follow up on the thread if it appears the user is not getting it and requires many follow ups to get a single thank you. Then on the other hand you could do it based on threads participated in, then it wouldn't be so bad.


  •  

    Posting Permissions

    • You may not post new threads
    • You may not post replies
    • You may not post attachments
    • You may not edit your posts
    •