Originally Posted by Analise
But the rating isn't to please one person; isn't it supposed to help others wade through the enormous # of threads? That is what I am reading. Hence, you just contradicted yourself. If a negative rating is strictly a personal problem of the rater, it is thus meaningless. If you disagree with that, then you are a victim of the emotionalism you accuse others of having.
Low ratings which are in the minority and are outliers in the statistical distribution of all the ratings given to a thread will have minimal effect on the average rating given a thread. What the meaning of a thread rating is depends on so many factors, not restricted to the content of the post, or whether or not somebody likes the fact that I posted such-and-such a thread, etc. The average rating on a thread, however, starts to reveal what people feel about the thread in aggregate. This was precisely why I advocated (in the thread in the suggestions forum) that the "minimum number of votes before rating is displayed" setting be increased to something like 10, 15, or even 20; that way, a statistically significant number of votes are counted before the ratings are even displayed at all. There is no threat from people who choose to vote low when everyone else votes high or from people who choose to vote high when everyone else votes low.
IMHO, to base arguments on speculations of people's motivations as to why they vote the way they vote is dragging the issue off-topic. The discussion centres about what the meaning of the overall, average thread rating is, not the meaning of individual votes. Again, all that needs to be done is to choose an appropriate setting as I stated above.
Thanks again for reinforcing more than ever that your interest in ratings
includes expessing your personal feelings about the person. You right there give a solid example of how ratings can be abused. To rank a low "Happy Birthday" thread on some HB threads because you don't like the people???
Yup, thanks for that example. You did more than I could have without that stellar example. ^
Your sarcastic tone only detracts from the credibility of your argument. If you feel that people giving low ratings
to a thread because of personal feelings about
a person is a form of abuse, then simply back this up with logic. I ask you this:
- How do you know what a person who gave a low rating to a thread is trying to do? How do you know if he/she is rating the thread low because of the dislike for the content, poster, opening poster, or some other reason?
- Same situation reversed: How do you know what a person who gave a high rating to a thread is trying to do? How do you know if he/she is rating the thread high because he/she likes the content, a particular poster, the opening poster, or because of some other reason?
By your arguments should we say that someone who votes a HB thread 5 stars "just because they like the person" is doing so based on personal feelings, and hence it's also abuse? I would think not and I hope not. People may vote as they please (and yes, it's one vote per person only; please, no red herring arguments about multiple handle voting). I don't see the guy at the ballot box asking you why you voted Democratic or Republican and that you have to explain yourself. In the end, the outcome of the vote depends on the overall vote of many people. If you are not in the majority; then huffing and puffing about it doesn't change the outcome. Just campaign harder next time, perhaps.
In the end, you may argue that the thread ratings (and cynically that democratic elections) are meaningless. Well, I'd agree insofar that one should not attach too heavy a meaning to them (as in speculating about people's motivation, claiming "voter fraud" when a thread gets higher ratings than you expect, or "abuse" when a thread gets lower ratings than you expect). These are voting systems based on statistical aggregation. There's a fascinating aspect in that statistical analysis can reveal trends that are not apparent in individual votes, but one also must be careful not to "read too deeply" into statistics.
Once again, I maintain the solution is simply that the thread rating display needs to be based on an increased sample size, such that it is more statistically significant.
FewMiles..