Pro ratings and SBC
As you probably noticed, the SBC system * Rating did not justify itself. The reasons are simple and numerous, here are the most obvious, I think:
* The average comment score is highly dependent on their number. So the ideal way to climb to the top of people is to write one comment, but super-popular, then raise yourself a little rating - that's all.
* As I understand it - the evaluation of posts in the blog, groups and columns is not taken into account in calculating the SBC, which also clearly affects the coefficient itself.
* Any user can increase or decrease the comment score by exactly one. I repeat - any.
Actually it is clear that the finished methods, how to deal with the problem there. What seems obvious to me can be described as follows:
')
* Make the amount by which the comment score increases or decreases related to the person changing the score. Thus, comments that like Igor Ashmanov are more important than comments that I like, for example, and this is logical.
* The next logical step would be to change the practice of raising or lowering a person’s rating by exactly one. It seems the rating system, when the voice of a person with high ratings is more important than the voice of a less popular comrade, works almost everywhere.
* Estimates for posts and posts in groups should somehow correlate with the evaluation of comments and also be taken into account. However, this is debatable.
* Ideally, it is necessary to consider not SBC, but a certain coefficient, which is considered as a function of the number of comments and points for them. Because the lamers of course, but the number of comments on the site is also quite important.
* The average comment score is highly dependent on their number. So the ideal way to climb to the top of people is to write one comment, but super-popular, then raise yourself a little rating - that's all.
* As I understand it - the evaluation of posts in the blog, groups and columns is not taken into account in calculating the SBC, which also clearly affects the coefficient itself.
* Any user can increase or decrease the comment score by exactly one. I repeat - any.
Actually it is clear that the finished methods, how to deal with the problem there. What seems obvious to me can be described as follows:
')
* Make the amount by which the comment score increases or decreases related to the person changing the score. Thus, comments that like Igor Ashmanov are more important than comments that I like, for example, and this is logical.
* The next logical step would be to change the practice of raising or lowering a person’s rating by exactly one. It seems the rating system, when the voice of a person with high ratings is more important than the voice of a less popular comrade, works almost everywhere.
* Estimates for posts and posts in groups should somehow correlate with the evaluation of comments and also be taken into account. However, this is debatable.
* Ideally, it is necessary to consider not SBC, but a certain coefficient, which is considered as a function of the number of comments and points for them. Because the lamers of course, but the number of comments on the site is also quite important.
Source: https://habr.com/ru/post/93/
All Articles