Just to add to what has already been posted, looking down the first page of the forum, it appears that the typical film poll gets an average of 10-15 votes. The current system ensures that, at worst (meaning that the vote is a perfect spread), each rating will have a minimum of 2-3 votes. Thankfully, there is usually some sort of preference center around ratings points, creating tops and bottoms.
LOVELIFE, I had initially pushed very hard to have a seven-tier half star rating system in this forum. Under ideal conditions, more ratings would give a more accurate read on people's preferences. Eventually I realized that we had to correlate the number of ratings categories according to the number of people we had voting. If there were twice as many people voting, I have no doubt that half stars would be better. If there were three times as many people voting, an 11-point system would be even better. Statistically, having (1) a larger sample coupled with (2) more categories always leads to more accurate results. However, if one of these elements is missing, then the result can be a meaningless analysis. An 11-point rating system for 10-15 people would guarantee a little more than a vote per category.
I hope someday that there are 40-50 people voting for each poll, and that we can revisit the ratings policy. Until that time, we are open to whatever suggestion, keeping in mind that there are consistency and appearance issues that also come into play with these types of decisions. Thanks for the input.
LOVELIFE, I had initially pushed very hard to have a seven-tier half star rating system in this forum. Under ideal conditions, more ratings would give a more accurate read on people's preferences. Eventually I realized that we had to correlate the number of ratings categories according to the number of people we had voting. If there were twice as many people voting, I have no doubt that half stars would be better. If there were three times as many people voting, an 11-point system would be even better. Statistically, having (1) a larger sample coupled with (2) more categories always leads to more accurate results. However, if one of these elements is missing, then the result can be a meaningless analysis. An 11-point rating system for 10-15 people would guarantee a little more than a vote per category.
I hope someday that there are 40-50 people voting for each poll, and that we can revisit the ratings policy. Until that time, we are open to whatever suggestion, keeping in mind that there are consistency and appearance issues that also come into play with these types of decisions. Thanks for the input.