i hate to complain because i really like this website and all the features, but here are some things that might or might not improve it.
i donno if these have been requested or not.
here they are in order of importance (to me):
1. in calculating the top favorite videos/anime, use sum(1/rank). the current formula, sum = sum(((101-anime.position)/100)+1), gives popular videos too much of an advantage. Example: Ten people watch Video A and think its decent, but nothing great. Nine of these ten people then watch Video B but think its the greatest thing they've seen. If the ten people gave Video A a rank of 10, and the nine people give Video B a rank of 1, Video A would still beat Video B, which in my opinion, isn't right. by using sum(1/rank) it would take 10 people voting a 10 for a video for it to match 1 person voting number 1.
2. a link from the forum profile to the a-m-v.org profile. i know people can add org profile links in their signatures, but most people don't and sometimes when your reading someone's post, you want to see their org profile, and its a real hassle searching for their profile on the org.
3. merge "escaflowne" with "vision of escaflowne" on the favorite anime list. i complain only because its my favorite anime, and it would be ranked much higher if they were combined. u don't have to delete one of the titles, because i know in the future people are going to enter both names, but if u can add up the two scores, and list them under the title "escaflowne/vision of escaflowne," it'd be a better representation of the anime's popularity.
4. have main page announcements also go on member page announcements. i have the member page bookmarked so there were a few times i was clueless about some change in the website until i heard someone talk about the announcement on the main page.
5. this is a small complaint and i dont even know if this will work well, but use the median for usefulness instead of the mean. seeing how the mean is at 162 (at the time of writing this) and the most useful reviewers are in the thousands, and having taken a basic statistics course, i would have to presume that the curve of usefulness is highly skewed. in which case, the median would be a better representation of the typical user's usefulness. but i dont know how much harder it is to calculate the median than the mean (u would have to pool all the opinions and the number of words, then order them and find the middle one or something).
