10

I noticed that many older posts used to receive dozens of upvotes. (I’m pretty sure it’s rare to see a post with more than twenty votes nowadays, I think Alex’s Shulchan Aruch count was the last one.) I wasn’t a part of the site for all of its history, so I don’t know what iterations its to through, really. However, I do notice that considerably older posts contain more votes. I was wondering if there is a definitive answer to this question, or something close to it. At this time I have thought of several possible reasons, please contribute if you feel necessary.

  • they’ve been around longer, more votes
  • there used to be more active users
  • users used to be more up-vote friendly
  • posts used to be more primitive and simple in their nature due to the site’s young age, therefore more deserving of an upvote
16
  • 5
    The first and most obvious alternative hypothesis to rule out is that what you're seeing is simply that older posts have had more time to accumulate votes. It should be possible to test this hypothesis statistically using data.stackexchange.com
    – Isaac Moses Mod
    Feb 24, 2019 at 14:27
  • 2
    I’ve been wondering this for a while as well.
    – Alex
    Feb 24, 2019 at 15:24
  • 3
    @Isaac, I think that's the first bullet point. No?
    – msh210 Mod
    Feb 24, 2019 at 16:31
  • 2
    I have been wondering as well - and believe the standards of the community got much elevated over time. Once in a while, you find answers with 10+ votes that today wouldn't qualify. This is a good thing but makes it harder for late starters. In addition, my sense is there was a wave of excitement at the beginning with more active users contributing (and maybe a sense that any content was good content). But I wasn't there and am sure some of the old-timers will comment with more accuracy
    – mbloch
    Feb 24, 2019 at 17:33
  • 1
    @msh210 I'm suggesting that talking about this in terms of "used to get more upvotes" may be a category error.
    – Isaac Moses Mod
    Feb 24, 2019 at 17:58
  • 1
    Note that the Shulchan Aruch post had a lot going for it to amass a lot of votes. Between that answer and the subsequently posted answers to the question there were 22 revisions. That means that the question got sent to the top of the homepage 22 times, which would give it a lot of exposure. It was also bountied, so it was featured for a week, giving it even more exposure. Then it was nominated in the best answer contest so it got even more exposure from those visiting the contest page. Finally it was featured for another week when the best answer bounties were awarded.
    – Alex
    Feb 24, 2019 at 18:01
  • So it might not always be about the inherent value of the post itself.
    – Alex
    Feb 24, 2019 at 18:01
  • 2
    Here's a query that can get started toward establishing metziut: data.stackexchange.com/judaism/query/127654/…
    – Isaac Moses Mod
    Feb 24, 2019 at 18:07
  • HNQ and bounty also might give a boost on upvotes. However, the upvote rate (total upvotes/post age) on older posts might not be higher than some of the recent posts.
    – Andrew T.
    Feb 24, 2019 at 18:10
  • 1
    Someone asked a similar question in 2013 judaism.meta.stackexchange.com/q/1468/1569
    – b a
    Feb 25, 2019 at 10:05
  • @IsaacMoses If I'm reading that correctly, that just looks for upvotes cast to posts in their first week, correct? Not just votes cast, period?
    – DonielF
    Feb 25, 2019 at 22:21
  • 1
    @DonielF Right, to attempt to parse out the vote-benefit posts may have gotten from when they were posted, as distinct from vote-benefit that comes with longevity.
    – Isaac Moses Mod
    Feb 25, 2019 at 22:23
  • @IsaacMoses We'd also need a graph for how posts in given time ranges average per unit time, as that average changes over time, to do the opposite. Your graph already says a lot about older posts: out of the top 20 highest-scored questions, 14 of them are around the graph's global peak, January 2012 ±8 months or so. Notably, even though PTIJ 2012 fell during that range, only two of those 14 were PTIJ.
    – DonielF
    Feb 25, 2019 at 22:31
  • 1
    My .02 shekel: when you run the cursor over the upvote arrow, it says "This question shows research effort; it is useful and clear". That's a fair but high bar, IMO, for an upvote, at least mine. A awful lot of questions, especially relatively recent ones, IMO, do not contain all three of those qualities. They seem like they were just thrown out there, without even a simple Google or even Mi Yodeya search being done beforehand. "Useful" is extremely subjective, what might be useful for me may not be deemed that by 99+% of the rest of the folks here. "Unclear" the mods call out pretty well.
    – Gary
    Feb 28, 2019 at 5:41
  • case in point: judaism.stackexchange.com/questions/15638/…
    – רבות מחשבות
    Mar 22, 2019 at 12:49

1 Answer 1

Reset to default
7

I think there are multiple factors at work here. Your first suggestion is definitely a factor. Posts continue to accrue upvotes over time. I have only posted a few questions/answers over the past few weeks so I'm not currently receiving lots of upvotes on brand-new posts. As a result, the last 10 upvotes I have received are on posts from 2015 or earlier (and only one of those is from 2015; the rest are from earlier).

Another factor is that earlier in the site's existence, volume was lower. As a result, posts stayed on the front page for a longer period of time (frequently days), resulting in more exposure. I remember a time (and I also have not been around for the site's entire existence) when I could easily see every single question and answer that was posted on Mi Yodeya. So, assuming people spend a similar amount of time on the site compared to what they did say 5 or 6 years ago, people today are seeing a smaller percentage of the posts. If the rate of posting is increasing faster than the rate of increase in the number of voting users, available votes get diluted by the increased number of posts. I have also noticed the phenomenon you're describing and I believe this is a big part of it.

You must log in to answer this question.

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