What is meta? ×
Meta Stack Exchange is where users like you discuss bugs, features, and support issues that affect the software powering all 131 Stack Exchange communities.

When SO was new, users seemed to vote a lot more than they do today.

Since reputation is intended to reward valued participation in the sites, is there perhaps a need to encourage users to vote more?

Furthermore, if this is implemented, should a reputation recalc be done to add existing user's voting to their reputation?

share|improve this question
    
Out of curiosity, how should downvotes be handled in this system? Are they unaffected (-1), just incremented (0), or set to be positive (+1)? –  Grace Note Sep 30 '10 at 12:59

3 Answers 3

up vote 16 down vote accepted

So all you would need to do to get a lot of reputation is to click the upvote button at random, without contributing anything useful to the site? Sounds like a great way to completely break the system.

share|improve this answer
    
Well, at a maximum of 30 points per day, it would depend on how you define "a lot"; I suppose you could diligently amass about a 1000 points a year. The point is that reading and voting takes effort, which effort is more likely to be expended if there is some reward for it. The same counter-argument can be made for giving report for accepting, which is done - it's much easier and quicker to amass rep by asking tons of simple and obvious questions, and accepting each one. You'll even garner some upvotes on the questions. –  Lawrence Dol Sep 30 '10 at 0:22
5  
@Software 30 * 365 = 10k. And your proposal would encourage people to vote regardless of whether they have read and considered or not. –  dmckee Sep 30 '10 at 0:31
    
@dmckee: Crap - sorry about the math error. Yes, that is significant. –  Lawrence Dol Sep 30 '10 at 0:35
2  
@Software Reading and voting consciously takes effort. Serial upvoting for points doesn't. –  NullUserException อ_อ Sep 30 '10 at 1:06

When SO was new, users seemed to vote a lot more than they do today.

A quick query in the Data Explorer reveals:

year   month   number_of_votes_per_month
----   -----   -------------------------
2010   8       504959          
2010   7       486036          
2010   6       456449          
2010   5       456407          
2010   4       430347          
2010   3       470651          
2010   2       421871          
2010   1       436296          
2009   12      361790          
2009   11      364010          
2009   10      346814          
2009   9       321173          
2009   8       333831          
2009   7       335970          
2009   6       317373          
2009   5       294468          
2009   4       266940          
2009   3       269150          
2009   2       261838          
2009   1       257096          
2008   12      183431          
2008   11      180981          
2008   10      270977          
2008   9       235731          
2008   8       61799           
2008   7       6       

UPDATE:

However, the average number of votes on answers do appear to be in a steady decline: about 13% lower in August 2010 than the all-time high in January 2009:

year   month   votes_per_answer 
----   -----   ---------------- 
2010   8       2.21991530084699 
2010   7       2.24435692744368 
2010   6       2.26759340378189 
2010   5       2.33139620180944 
2010   4       2.26774346471624 
2010   3       2.27444972927229 
2010   2       2.2805144275629  
2010   1       2.355767310253   
2009   12      2.34382948701287 
2009   11      2.32638737758433 
2009   10      2.39763986100524 
2009   9       2.44056752609626 
2009   8       2.44243704544499 
2009   7       2.32734172158977 
2009   6       2.43334320922628 
2009   5       2.42150861663057 
2009   4       2.53494671463945 
2009   3       2.55219082832073 
2009   2       2.57934649910233 
2009   1       2.67963604466184 
2008   12      2.36511564836698 
2008   11      2.32409950864704 
2008   10      2.5333967996192  
2008   9       1.87002470171514 
2008   8       1.404326054835   
2008   7       0.5
share|improve this answer
3  
Can you get that relative to the number of users? –  deceze Sep 30 '10 at 0:20
1  
...it would have to be versus the number of page-views or something rather than raw number of registered users if it is to mean anything. –  dmckee Sep 30 '10 at 0:29
    
@dmckee: I could, but because of the increasing number of inactive users that wouldn't make sense... Number of page views might be interesting. –  Daniel Vassallo Sep 30 '10 at 0:36
    
@Dan, but what about the increasing number of active users? –  jjnguy Sep 30 '10 at 0:46
    
@jjnguy: Posted data on average votes per answer per month. I can't get meaningful results when comparing the data to the number of registered users. There would be a very big drop in votes per registered user, which I don't think is representative of the issue. Mine was a very crude attempt to see if I were likely to get more votes on an answer last year, than if I were to post the exact same answer today. –  Daniel Vassallo Sep 30 '10 at 1:05
3  
@dan, I think the update is a much better dataset. –  jjnguy Sep 30 '10 at 1:28

You're not supposed to be able to award yourself reputation1. You earn reputation... by posting great content.

The theory is similar to how Google ranking is earned by how many people link to your site. You can't give yourself a better Google ranking.


1 The only exception is the minimal points awarded to you for accepting an answer.

share|improve this answer
1  
Actually, according to Jeff the idea has always been more broadly that you earn rep by participating in ways that benefit the site. Otherwise, questions would have no rep or only net-positive rep associated with them. –  Lawrence Dol Sep 30 '10 at 3:34
1  
Simply voting to earn rep (i.e. voting arbitrarily because it is easier) does not help the site. Otherwise, just offer rep for other good activities... like editing. I'll "fix" (read: change) every post I can find and be the site's top user. –  Robert Cartaino Sep 30 '10 at 13:45

You must log in to answer this question.

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