|
|
|
Topic: Too many protected ratings
| |
|
|
Author
| Message |
|---|
|
|
Mario!
As this is shading off into another kind of discussion, I have posted some other thoughts, in the general chess forum, under the heading 'FUN, HOBBY, ADDICT'.
Best wishes,
Clive
|
|
|
My rating has climbed to 1700 Players rated 2100 and above are willing to play me. that may be due to their unusual rating rule regarding 400 points. My rating doesnt matter as long as masters are willing to play me!
|
|
|
I believe my present rating 1988 is near my true strength, though my history graph has a large positive slope. I ignored beco's advice becuase I wanted to play strong players. I wonder how players felt losing to someone 500 points lower. Some seemed bewiledered.
In math dicontinuos functions are not well behaved. You gain 32 points for a win against someone 400 points higher and 0 for a win against someone 401 points higher. That seems bizare and definitely not well behaved.
|
|
|
The problem is the bizarreness, not the alleged mathematical ill behaviour. (I'm not convinced there's any mathematical problem here at all, in fact.)
|
|
|
My math comment was intended to be somewhat amusing. However, many fundamental statements in math assume continuity. Any measuement has uncertainty. Ratings are supposed to measure a players strength. The bizarreness could be restated as a surprisingly large uncertainty. The cuase of this large uncertainty is the aforementioned discontinuity.
There was also a secondary effect. When I played someone within 400 points of my rating thier rating change was based not on my true strength but my rating at the time. Instead of losing 2 points they lost 30 points.
|
| Previous 1 2 3 |
|
|
|
