News for hackers. Stuff that matters.
Re:How Karma Should Be Measured by onedognight on Tuesday August 23, 2011 @11:26PM (#2919319)
As far as I can see, the formula gives no penalty for commenting infrequently. There is a penalty for commenting highly once and then poorly many times as compared to what you would get with an average.
Re:How Karma Should Be Measured by eric-hu on Wednesday August 24, 2011 @12:07AM (#2919421)
I believe there is under "A Better Metric"

dx is the number of days the user is registered. If the user made one post and then stopped, his karma score would shrink over time as dx approaches infinity while the numerator remains constant.

The numerator technically has an extra term for each successive day, but it looks like that additional term is 0 if no comments were made on that day.

Re:How Karma Should Be Measured by gjm11 on Wednesday August 24, 2011 @05:02AM (#2920039)
It's worse. The additional term is undefined (it's 0/0-1) for days with no comments.
Re:How Karma Should Be Measured by eric-hu on Thursday August 25, 2011 @03:00AM (#2924040)
I figured that was just a special case they didn't explain directly. The equation follows their description if you treat the 0-comment days as a 0 for the term in the parenthesis: people who post infrequently will see a shrinking karma.
Re:How Karma Should Be Measured by Udo on Wednesday August 24, 2011 @08:02AM (#2920344)
Yes, it does. At the end, the score is divided by the age of the account in days. Essentially, this formula is a treadmill forcing users post as much as they can just to keep their karma. It's a very bad idea that has no upside.