Re: [Bug-gnubg] Snowie error rates versus gnubg error rates

[Joern Thyssen]

On Sun, Apr 06, 2003 at 12:10:03PM +0000, Joern Thyssen wrote

> ad (a): I've recieved a sample of 50 matches from Albert. I can use them
> to deduce an empirical conversion factor between gnubg and snowie error
> rates. This can be used to adjust the threshold for assigning the
> ratings "supernatural", "world class" etc, so gnubg and snowie are more
> aligned.

I've analysed close to 300 matches (50 from Albert and 250 from my own
archive), and the average conversion factor is 1.38 (std.dev. 0.37).

Snowie's thresholds are:

0.0 - 1.2: Extra-terrestrial
1.2 - 4.4: World class
4.4-5.9: Expert
5.9-8.8: Advanced
8.8-12.6: Intermediate
12.6-18.5: Beginner
18.5 - ??: Novice

I suggest the following change (pending the discussion):

                gnubg's old               gnubg's new 
                                          (based on a factor of ~1.4)
Supernatural    0.000 - 0.005             0.0000 - 0.0017
World Class     0.005 - 0.010             0.0017 - 0.0062
Expert          0.010 - 0.015             0.0062 - 0.0083
Advanced        0.015 - 0.020             0.0083 - 0.012
Intermediate    0.020 - 0.025             0.012  - 0.018
Casual Player   0.025 - 0.030             0.018  - 0.026
Beginner        0.030 - 0.060             0.026  - 0.035
Awful!          0.060 - oo                0.035  - oo

Jørn