Just to reduce uncertainty, what Granny Magda described is how it works. In case it helps make it more clear, here's a simple example.
Let's say I rate five of Buzsaw's messages, giving them these ratings: 5, 2, 3, 4, 3. The average of these five message ratings is 3.4, so I contribute a rating to 3.4 to Buzsaw's member rating. If I'm the only person who has rated any of Buzsaw's messages then 3.4 would be his member rating.
Now let's say someone else rates five of Buzsaw's messages as follows: 5, 4, 5, 5, 3. The average of these five message ratings is 4.4. If you average 3.4 and 4.4 you get 3.9, which would be Buzsaw's new member rating.
As quickly became apparent when I added message ratings, the current approach is inherently both unfair and ambiguous. The ratings are based upon opinion, which is highly subjective, and every person will judge according to their own individually selected criteria.
While I wouldn't consider the current rating system a very reliable parameter upon which to judge the quality of a member's contributions, there does seem to be a correlation of higher ratings with those who discuss things they know something about or who take the time to investigate. Those who tend to put fingers in gear without taking the time to develop an appropriate level of understanding of the topic seem in general to have the lower ratings.
-- | Percy |
| EvC Forum Director |