[Moved by wahaha, originally posted in Feature Requests - AniDB Clients]
Title says it basically, just think that this would make the list a little more helpful: by rating only those who liked a series enough to download the whole thing, rather than every five minute fan who heard Cowboy Bebop was cool so got the first episode.
Rar
to count only those with the anime completed [tracked]
Moderator: AniDB
I think this kinda got overlooked in the muddle. It is a serious request, and I think it would be a more representative indicator of which are the most prevalent animes in the community.
For instance, Naruto and Koukaku Kidoutai STAND ALONE COMPLEX are 1 and 2 at the moment, but the number of users drops off rather heavily, both to under 600 by later episodes. The more established favourites Cowboy Bebop, Full Metal Panic! and Shinseiki Evangelion at 3, 4 and 5 all maintain over 700 users all the way to the end, Bebop currently having 831 users for the last episode, compared to GITS:SAC (such a cool abreviation) having only 567.
If a new user looks at the to ten stats on the front page at the moment, the are given a rather odd impression of what the series have the 'most users', favouring the current crazes over the long term favourites - and series people have downloaded one episode to see what it was like over those they liked enough to get it all.
For the record here, this is not to favour animes I prefer myself, (look at my votes to see the only one I really rate is Evangelion), but to give viewers of the front page a clearer impression of what series have the most users all the way to the end.
Rar
[Edit: Playing with formating]
For instance, Naruto and Koukaku Kidoutai STAND ALONE COMPLEX are 1 and 2 at the moment, but the number of users drops off rather heavily, both to under 600 by later episodes. The more established favourites Cowboy Bebop, Full Metal Panic! and Shinseiki Evangelion at 3, 4 and 5 all maintain over 700 users all the way to the end, Bebop currently having 831 users for the last episode, compared to GITS:SAC (such a cool abreviation) having only 567.
If a new user looks at the to ten stats on the front page at the moment, the are given a rather odd impression of what the series have the 'most users', favouring the current crazes over the long term favourites - and series people have downloaded one episode to see what it was like over those they liked enough to get it all.
For the record here, this is not to favour animes I prefer myself, (look at my votes to see the only one I really rate is Evangelion), but to give viewers of the front page a clearer impression of what series have the most users all the way to the end.
Rar
[Edit: Playing with formating]
This again depends on what you think the statistic is for - whether it is for giving other users hints on what to watch next, or what many users tried out 
Since I also think it should give hints about what most users watched to the end and kept in their mylist (I sincerely hope that "deleted" does not count towards this statistic!) I too think it _could_ be changed.
Problem is: some series have not ended yet. How are people supposed to get Narturdo as a hint as long as this series is going on and on and on, like Inuyasha?
Not all users have broadband and maybe it just takes a little longer to get all files of Narturdo than the 26 of GitS:SAC
Since I also think it should give hints about what most users watched to the end and kept in their mylist (I sincerely hope that "deleted" does not count towards this statistic!) I too think it _could_ be changed.
Problem is: some series have not ended yet. How are people supposed to get Narturdo as a hint as long as this series is going on and on and on, like Inuyasha?
Not all users have broadband and maybe it just takes a little longer to get all files of Narturdo than the 26 of GitS:SAC
I see some problems here:
1. If after watching few episodes you think that some title is rather bad, then why you should watch the rest of anime, you didnt like only to have possibility to rate it low?
2. If you have watched all episodes, but you dont have files for them (or you have, but they are not good quality and dont wat to add them to anidb). Then you will have to add to mylist some files (files you dont have) for all episodes only to make a vote for anime?
3. You have watched allmost all episodes (even including first and last) but some of them waiting to download (and we know it can time long time).
Why isnt possible to rate title, when have you watched more thane 90% of episodes? Will really the missing episode change you thinking about title so much? And even so, there is 'revoke vote' for changing vote.
4. Problem with still ongoing long-lasting series (Inuyasha, Naruto, ...). Have you really watch all episodes to rate it?
1. If after watching few episodes you think that some title is rather bad, then why you should watch the rest of anime, you didnt like only to have possibility to rate it low?
2. If you have watched all episodes, but you dont have files for them (or you have, but they are not good quality and dont wat to add them to anidb). Then you will have to add to mylist some files (files you dont have) for all episodes only to make a vote for anime?
3. You have watched allmost all episodes (even including first and last) but some of them waiting to download (and we know it can time long time).
Why isnt possible to rate title, when have you watched more thane 90% of episodes? Will really the missing episode change you thinking about title so much? And even so, there is 'revoke vote' for changing vote.
4. Problem with still ongoing long-lasting series (Inuyasha, Naruto, ...). Have you really watch all episodes to rate it?
OOps, but some points (2 and 3)is the same: i dont think having all episodes is required to be listen in 'Most users'. It is not called 'Most users having complete title' - when completing a title is in progress why not be counted as already having it.
And as i see those current numbers, i didnt see a rule to counting them, are there already not counted some users? It looks alike (when number showed here is lower than number of users having 1st, 2nd, 3rd episode for title).
Some user filtering could be usable. I vould set it to 10 or 20% of episodes for title (not 'Episodes' field in title, but number of episodes having already registered at least one file in anidb).
And as i see those current numbers, i didnt see a rule to counting them, are there already not counted some users? It looks alike (when number showed here is lower than number of users having 1st, 2nd, 3rd episode for title).
Some user filtering could be usable. I vould set it to 10 or 20% of episodes for title (not 'Episodes' field in title, but number of episodes having already registered at least one file in anidb).