QUOTE In that case, consider it done. Best Seiyuu will be changed to Best Character under the restriction that the character must be fully identifiable. This means that the character must not only be a recurring character in the series (i.e. the character cannot just show up for 5 minutes in one episode and never show up again). The character must serve a primary or secondary role in the main cast. Anything below that will be considered null. This is to avoid random nominations like the ones in Saimoe, e.g. "the girl on was walking in a swimsuit in the background in Code Geass"
From the way you say that I take it people actually tried to nominate minor side characters ><
QUOTE I did this for a couple years when I first started this, but then I had complaints let alone realizations that this is a very poor method in prelims because of how much more easily the results can be skewed.
That there is a difference between the votes of two people voting for the same characters in a different order, is, I think, an advantage, rather than a disadvantage of weighting votes. A huge swing with only two votes, on the other hand, is to be expected. Small sample size normally causes that.
In other words, there are certainly bigger swings, especially with a small sample size, because each vote carries more information about people's preferences than with a non-weighted vote system. Put another way, the fact that weighted votes would treat your hypothetical people's votes differently depending on the order they were put in is a good thing. It shows that ranking is being taken into account.
There is no perfect system to deal with a small sample size (less than 20 votes, as we received in the prelims). With unweighted nominations, you're going to get a lot of ties, and some mediocre shows will be overvalued. With weighted nominations you're going to get some swings.
The best solution is actually just to increase the number of voters. In an unweighted system you will get less ties from more votes, and in a weighted system you will get smaller swings.
Either way, your position is somewhat unenviable. Someone will complain no matter how you run the poll =x
QUOTE
Another category I intend on correcting is Best Soundtrack to simply Best Sound, which would be defined as ranging from appropriate voice casting to background themes, and so on. Thoughts on this?
Its a bit broad, but much more user-friendly than soundtrack. It has some troubles, (comparing apples to oranges), but...I think its better. I got the impression people had a lot of difficulty (I know I had a little, myself) selecting series to vote for in the soundtrack category. In the prelims, 9/17 people selected 1 or 0 shows for soundtrack, and 13/17 picked 3 or less shows.
So, trading one problem for another? Yeah, maybe. But an overbroad category seems better than a specific one that people are having trouble voting for.
From the way you say that I take it people actually tried to nominate minor side characters ><
QUOTE I did this for a couple years when I first started this, but then I had complaints let alone realizations that this is a very poor method in prelims because of how much more easily the results can be skewed.
That there is a difference between the votes of two people voting for the same characters in a different order, is, I think, an advantage, rather than a disadvantage of weighting votes. A huge swing with only two votes, on the other hand, is to be expected. Small sample size normally causes that.
In other words, there are certainly bigger swings, especially with a small sample size, because each vote carries more information about people's preferences than with a non-weighted vote system. Put another way, the fact that weighted votes would treat your hypothetical people's votes differently depending on the order they were put in is a good thing. It shows that ranking is being taken into account.
There is no perfect system to deal with a small sample size (less than 20 votes, as we received in the prelims). With unweighted nominations, you're going to get a lot of ties, and some mediocre shows will be overvalued. With weighted nominations you're going to get some swings.
The best solution is actually just to increase the number of voters. In an unweighted system you will get less ties from more votes, and in a weighted system you will get smaller swings.
Either way, your position is somewhat unenviable. Someone will complain no matter how you run the poll =x
QUOTE
Another category I intend on correcting is Best Soundtrack to simply Best Sound, which would be defined as ranging from appropriate voice casting to background themes, and so on. Thoughts on this?
Its a bit broad, but much more user-friendly than soundtrack. It has some troubles, (comparing apples to oranges), but...I think its better. I got the impression people had a lot of difficulty (I know I had a little, myself) selecting series to vote for in the soundtrack category. In the prelims, 9/17 people selected 1 or 0 shows for soundtrack, and 13/17 picked 3 or less shows.
So, trading one problem for another? Yeah, maybe. But an overbroad category seems better than a specific one that people are having trouble voting for.
