the technical measure I would most like to see against mass-downvoting
LW could also employ the technique commonly used to prevent brute-forcing logins—limits on the rate of downvoting. Basically, the forum could allow (the numbers are arbitrary and are just examples here) 10 downvotes within 10 minutes, 15 within an hour, 20 within a day, 25 within two days, etc.
Yup. Though some of the mass-downvoting I’ve had has been gradual—a few points a day for several days—so it seems at least one LW mass-downvoter might not be so badly inconvenienced by this. (Have there been multiple LW mass-downvoters? I don’t know.)
Well, a few points for a few days doesn’t sound like mass-downvoting to me. Downvoting is a useful function, I don’t think we should be heading towards the situation when it would require an advance application for permission to downvote submitted in triplicate to a Very Important Committee.
I don’t think anyone’s suggesting that, or anything sufficiently like it to warrant concern.
The behaviour I’m talking about is: A goes through all B’s old comments, systematically downvoting a few of them every day for some time. Not selecting particularly bad old comments, you understand; the comments themselves are irrelevant. The only goal is to be able to reduce A’s karma by a lot more than a single downvote would, without making it too blatant what you’re doing.
(Why old comments? 1. So it’s less obvious to anyone other than B what’s happening. 2. Because A has already downvoted all B’s recent comments.)
Well, you’re talking about the state of mind of the downvoter. Unfortunately, the technology to detect that isn’t available at the moment. What LW software can detect is that user X downvoted N1 comments by user Y on day 1, N2 comments on day 2, etc. As long as N is “a few”, I would be wary of drawing “mass-downvoting” conclusions from this pattern.
Also, I would recommend not getting into a technological arms race with people want to game the karma system.
This is the advantage of merely publicizing their behaviour rather than attempting to prevent it. Even if you get false positives or false negatives, the adverse consequences aren’t severe.
LW could also employ the technique commonly used to prevent brute-forcing logins—limits on the rate of downvoting. Basically, the forum could allow (the numbers are arbitrary and are just examples here) 10 downvotes within 10 minutes, 15 within an hour, 20 within a day, 25 within two days, etc.
Yup. Though some of the mass-downvoting I’ve had has been gradual—a few points a day for several days—so it seems at least one LW mass-downvoter might not be so badly inconvenienced by this. (Have there been multiple LW mass-downvoters? I don’t know.)
Well, a few points for a few days doesn’t sound like mass-downvoting to me. Downvoting is a useful function, I don’t think we should be heading towards the situation when it would require an advance application for permission to downvote submitted in triplicate to a Very Important Committee.
I don’t think anyone’s suggesting that, or anything sufficiently like it to warrant concern.
The behaviour I’m talking about is: A goes through all B’s old comments, systematically downvoting a few of them every day for some time. Not selecting particularly bad old comments, you understand; the comments themselves are irrelevant. The only goal is to be able to reduce A’s karma by a lot more than a single downvote would, without making it too blatant what you’re doing.
(Why old comments? 1. So it’s less obvious to anyone other than B what’s happening. 2. Because A has already downvoted all B’s recent comments.)
Do you think this is a “useful function”?
Well, you’re talking about the state of mind of the downvoter. Unfortunately, the technology to detect that isn’t available at the moment. What LW software can detect is that user X downvoted N1 comments by user Y on day 1, N2 comments on day 2, etc. As long as N is “a few”, I would be wary of drawing “mass-downvoting” conclusions from this pattern.
Also, I would recommend not getting into a technological arms race with people want to game the karma system.
This is the advantage of merely publicizing their behaviour rather than attempting to prevent it. Even if you get false positives or false negatives, the adverse consequences aren’t severe.