This Bugzilla instance is a read-only archive of historic NetBeans bug reports. To report a bug in NetBeans please follow the project's instructions for reporting issues.
I've tried to vote for the "Geronimo 2" plugin. As it seems not to be actively developed (other than e.g. some api-related upgrades) and it seems not to support Geronimo server versions other than 2.0.2, I wanted to vote for it as "bad", so clicked on the left star. Unexpectedly, the voting result is 3.3, though the statistic said, no user has voted before.
BTW, the "rating" comment is: Rating: 3.3, by 1 users You already rated: 1
Yes, this is as designed Peter. When the number of votes is small (no matter good or bad) the rating is normalized towards the average = 3 although I would expect 2.7 in your case and not 3.3. Honzo, can you please clarify this? Thanks!
(In reply to comment #2) > Yes, this is as designed Peter. When the number of votes is small (no matter > good or bad) the rating is normalized towards the average = 3 although I would > expect 2.7 in your case and not 3.3. > Then there should be a hint, at least, as this mechanism makes a rating looking better, though I do understand the intent, that one user's rating should not have too much impact.
We use Bayes average for calculating overall average. It corrects plain average depending on number of votes plugin has and what is overall average in ratings and overal number of ratings etc.. http://leedumond.com/blog/the-wisdom-of-crowds-implementing-a-smart-ranking-algorithm/