This will flag comments for moderators to take action.
A place for Round 3 Raters (and writers!) to discuss their experiences
Posted: 8/19/2011 7:58 AM PDT
I'm not familiar with the background to this problem. I gained expert status only yesterday but, when coming to the forum was surprised to discover no recent posts, yours being the most recent in any thread. The idea behind the site is that we are all readers as well as (aspiring) writers. Some have studied or are studying literature others are just people who love books. These factors should make us broadly representative of "the reading public" and the theory behind the site is that if we like it so will they and, hence, it will be of interest to agents and/or publishers.
I do wonder, though, from the perdominance of Sci-Fi, Fantasy and Teen fiction in the submissions we are asked to rate, if that is really true. What is your take on this?
Posted: 7/22/2011 5:36 AM PDT
I'm more intersted in focusing on the bigger issue. Round 3 raters aren't rating anymore. Now that the thrill is gone, the ratings have nearly stopped coming in. IMO, this is because the WeBook staff isn't generating excitement nor letting anyone know what's going on.
I'm on another site that's a lot more vibrant than this one, and part of the reason is the staff is involved. They don't have contests, but if they did, I know the staff would be involved.
Posted: 7/20/2011 7:41 PM PDT
Simpler and better might be to give a 5 to works that are in the top 5% of a raters ratings and a 4 to a score in the top 15 to 10% of a raters ratings. If a rater gives everybody 4 and 5, his ratings would never count. If he gives almost everyone 2 and 3, his 4 might count as a 5. I assume when you mentioned stdev you would give a 5 to works given 3 x stdev above the upper quartile (a statistical outlier), and perhaps a 4 to works 2 standard deviations above it.
I'd rather stick with the top 5-10% of a raters ratings.
Posted: 6/20/2011 4:58 PM PDT
I realize that this will probably be to difficult to implement, but a lot of the complaints and problems with the rating system could be fixed with a couple of calculations:
1) subtract the raters average rating from each subsequent rating. Thus a rather who always gave a five would, in effect, do nothing. The rather, on the other hand, who gave ratings of all types would have each count except for his 'middle of the roads'.
2) Divide this rating by the standard deviation of his ratings (this is fancy stuff, the mathmaticians will understand what this does. Basically it evens out the person who rates with all fives and ones with the person who rates with all threes and fours)
3) Allow the readers to rate the raters (this would only work if the two things above were done), and have those raitings 'weight' the raters ratings.