I'm not sure if this idea will be popular or not considering it is certain to be a sensitive issue as it concerns the entire model of brainstorm.
The problem is that popular ideas are typically more popular when the level of traffic is up and less popular when the level of traffic is down.
The 13 most popular ideas ever were submitted within less than 24 hours of one another. I am guessing that's the day that Digg was made aware of Brainstorm.
In fact, at least five of the individuals considered to be the top 10 idea contributors per the algorithm, earning that ranking from submitting ideas around that period, have submitted less than 5 ideas a piece, and exactly two, from what I can deduce, have only submitted a single idea.
Yes, these ideas certainly are very popular, and I honestly believe that those ideas should be implemented. What makes me believe there is a problem with current working definition of popularity is its strong relation to overall traffic on Brainstorm.
Reason being: an idea that earns the same percentage of individual +1's and -1's from all users with the lower levels of traffic in relation to February 28th and 29th should be considered just as popular.
In essence:
There are 1000 people browsing Brainstorm today, 300 vote up an idea, 10 vote it down.
There are 10000 people browsing Brainstorm another day, 3000 vote up an idea, 100 vote it down.
Should these ideas be considered to have the same level of popularity? I believe so, you may not. Whether or not this can be considered a day-by-day, month-by-month problem is up for discussion. Whether consistency and simplicity are at issue, that's up for discussion too. This idea is meant to be vague in that sense because any solution would require a fair amount discussion.
I believe the problem is most visible in concern to "Most popular ideas ever."
[....]