5 ms·
The problem with this system is in the sorting. The list of "Highest Rated" teas is dominated by results where 1 person rated the tea 100. Steepster should use
by lonestar 17y ago
The problem with this system is in the sorting. The list of "Highest Rated" teas is dominated by results where 1 person rated the tea 100.
Steepster should use a Bayesian average (http://en.wikipedia.org/wiki/Bayesian_average http://en.wikipedia.org/wiki/Bayesian_average) so that the uncertainty of a small number of ratings is reflected in the sorting.
- mpotter 17y agoYeah, sorting is an issue we're still looking at (and is still very much in transition considering the new rating system). Appreciate the suggestion! We'll add it to our list of potential solutions.
- selven 17y agoStart everything off with a single 50-point score. That way one person ranking it 100 will bump it up to 75, the next to 83, and so on. Such a system would cause teas that have more people upvoting them to rank higher than those that just happen to have one or two good opinions.
- Eliezer 17y agoThat's a special case of a particular sort of Bayesian average.