Interestingly enough, only 498 users out of ~32K have 2000 or more votes. Whereas 4877 users only have one vote. Someone feel free to double check this.
as mentioned in the reddit thread, one coud assume that power users would allow public voting and therefore would be voting more. I of course, just looked at the extremes and do not have a timeframe
> one coud assume that power users would allow public voting and therefore would be voting more
I don't think that's a good assumption. I surely have more than 2000 votes on reddit, but some of them are on political stories and I don't particularly want a public record of that.
> Due to the way that I pulled the voting information (I actually pulled it from the cache that we use to show you liked and disliked pages, which is in Cassandra and turns out to be cheap to query), you won't get more than 1k upvotes or downvotes per user, no matter how many votes they've made, so that so many have 2k isn't surprising.
Comments
Interestingly enough, only 498 users out of ~32K have 2000 or more votes. Whereas 4877 users only have one vote. Someone feel free to double check this.
The data was opt-in (redditors had to specifically elect to enter this dataset), so it's probably biased somewhat.
as mentioned in the reddit thread, one coud assume that power users would allow public voting and therefore would be voting more. I of course, just looked at the extremes and do not have a timeframe
> one coud assume that power users would allow public voting and therefore would be voting more
I don't think that's a good assumption. I surely have more than 2000 votes on reddit, but some of them are on political stories and I don't particularly want a public record of that.
> Due to the way that I pulled the voting information (I actually pulled it from the cache that we use to show you liked and disliked pages, which is in Cassandra and turns out to be cheap to query), you won't get more than 1k upvotes or downvotes per user, no matter how many votes they've made, so that so many have 2k isn't surprising.
Somewhat related: I've been archiving the stories (and vote/comment counts over time) via the JSON API for the last few months or so. Post is here if anyone is interested: http://www.reddit.com/r/datasets/comments/bttjw/top_1000_ral...