There have been many discussions about counting votes with various weighting schemes: self nominated expert (bad), rep (bad bad), rep (again, but not bad?), and several down-vote specific schemes (some good).

What I'm proposing is a flexible and complementary display of a second, optional, weighted vote count besides the current raw "every vote counts" count.

The weighted count is NOT displayed by default for visitors. You have to opt-in from your user prefs, and either choose from one of SO-provided, community-approved schemes, or design your own scheme with some kind of scripting language or raw SQL. (Perhaps then allow multiple weighted counts, since sometimes you're not sure which scheme is the best, or there's simply not a "best" scheme.)

This can also be very valuable to employers as they can design a weighting scheme that fits their criteria for a specific position they're trying to fill.

UPDATE I didn't specify any particular scheme on purpose, but please don't assume I only meant to use rep, which I would oppose myself. Up-vote count on a tag or a "guru" badge, for example, can be a quite reliable measure.

  • Welcome to MSO, @Geoffrey! Based on the failure of previous alternate vote system proposals, I think you'd be better off suggesting a specific scheme. At this point, most people, including me, are of the opinion the there is no good alternate scheme until proven otherwise. As an aside, you can get a one-time +100 rep bonus here by going to your profile's accounts tab and disassociating/reassociating with your SO account, since you have >200 rep there. – Pops Sep 22 '10 at 16:15
  • I was wondering what would happen if I get negative rep (would my monitor explode?) and Holy Skeet I got +100 here AND SO by asssociating! (Up-vote meta.stackexchange.com/questions/65292/…, please!) I'll come up with a specific scheme that I myself would like to use. – Geoffrey Zheng Sep 22 '10 at 17:21
  • you can't get negative rep. Someone has tried; you can see his reputation graph here. Even though it goes into the negatives on the graph, his display rep is never lower than 1. But, to make things more complicated, I should point out that his current rep is artificially set at 1 because he's been suspended. His "normal" rep is actually greater than 1, because the order of votes matters. – Pops Sep 22 '10 at 18:17
  • (Same comment for my bad proposal to scrub +100 for association) Sorry I hallucinated about seeing +100 on SO. I did gain some points that pushed me over from 500 to 600. I guess I'm too hung up on rep, an early and sure sign of SoAS (SO Addiction Syndrome). What's the deal with that Evan Carroll guy? I'm curious to learn how he got so negative. – Geoffrey Zheng Sep 22 '10 at 19:28

I can't speak for the team, but I don't think it's a good idea:

  • It would require the team to have code from users running on the same server as StackOverflow.
  • It soak up a lot their development time for something that probably won't generate a lot of revenue.
  • Since vote-counts couldn't be cached, it would require said code to run the first time each user using this feature looks at a post.

You could always do this yourself using the API or data dump (maybe as a Greasemonkey script).

  • #1: if Google Apps can do it, surely you can, no? #2: I don't know about the development time part, but the revenue part can be justified by charging employers. #3: agree. As for using API, it would have the same performance issue unless you throttle/limit API usage. – Geoffrey Zheng Sep 22 '10 at 17:08
  • #1: I doubt SO has the same amount of resources. I said that based on how often the data explorer goes down (as I said in my post, I'm not part of the team btw). #2: If they're prepared to pay for it, surely they'd also be prepared to use a custom solution? If they're going to spend money developing something custom, they might as well use the API/data dump. #3: For posts the API can only retrieve upvotes and downvotes, and it's harder consume a large amount of resources than the data explorer (which I assume this would be similar to). – Gelatin Sep 22 '10 at 17:19

As Jeff so frequently says: Any "feature" that works by making the system several times more complicated to use or understand is broken by definition.

I vote an emphatic "no" for this one.

  • Sometimes complication is necessary for greater utility. We never want confusion, that's why I suggest an optional, second measure. – Geoffrey Zheng Sep 22 '10 at 15:31
  • 1
    @Geoffrey: "Make it an option" is never an answer. The past 15 years of software development and UI research have proven that beyond a shadow of a doubt. This is not a decision you should ever ask a user to make, and giving wildly different experiences to different users is never a good idea. – Aarobot Sep 22 '10 at 16:06
  • Options are only bad when you stick them to a user's face. Can you find a desktop app that you use every day (I'd guess Visual Studio from your SO profile) with no or few options? And why would different (or rather, customized) experiences be bad if that's what a user is specifically looking for? – Geoffrey Zheng Sep 22 '10 at 17:15
  • @Geoffrey: Visual Studio is a morass of useless options, starting with the front page "recent files" setting. A "feature" is something that makes a specific use case easier to accomplish. What benefits does this provide? – Aarobot Sep 22 '10 at 18:50
  • I didn't mean that VS is a good UI example (I'm glad I don't have to use it). How about Firefox, then? Anyway I think @Simon made it pretty clear mostly from a pragmatic perspective. Thanks for the discussion! – Geoffrey Zheng Sep 22 '10 at 19:07

You must log in to answer this question.

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