On Thu, Jul 30, 2009 at 10:08 PM, BrianBrian.Mingus@colorado.edu wrote:
The second sentence should read: There is no information in the current heuristic that indicates that editors who are allowed to vote are more or less familiar with the candidates than those who are not.
Who says there needs to be?
The recent edits criteria reduces the incentive to crack or otherwise collect old unused but qualified accounts. For example, I could setup a free watchlist aggregation service and users would give me their passwords. Over time I could obtain many and then wait for accounts to naturally become inactive, then I could vote with them.
It also makes it harder to otherwise obtain votes from accounts whos owners have lost interest in the project and might be willing to part with theirs easily. Recent editing activity also provides more information for analysis in the event that some kind of vote fraud is suspected.
A recent edits criteria is justifiable on this kind of process basis alone.
50 edits can easily be made in a couple of hours, even if you're not making trivial changes. If you're not putting that level of effort it seems somewhat doubtful that you're going to read the >0.5 MBytes of text or so needed to completely and carefully review the provided candidate material from scratch. Like all stereotypes it won't hold true for everyone but if it's true on average then it will produce an average improvement, we just need to be careful not to disenfranchise too many.