Re: [Bug-gnubg] gnubg-020719 error ?

[Joern Thyssen]

On Wed, Jul 24, 2002 at 07:03:31AM +0200, max-d wrote
> hi
> using gnubg-020719 i have had this strange Match winning chance report
> am i missing something ?
> 
> Overall statistics:
> 
> Error rate (total)  -0,147 ( -6,491%) -0,438 ( -4,382%)
> Error rate (per decision) -0,005 ( -0,232%) -0,017 ( -0,169%)
> Overall rating:   World class          Advanced
> 
> Match winning chance
> against opponent:    47,89%    52,11%
> Guestimated abs. rating: +2045,29  +1812,93

For some reason player 0 gave up more *total* MWC than player 1 (6.491%
vs. 4.382%). The match winning chance against opponent is calculated as
50% - 6.491 + 4.382 = 47.891%.

However, player 0 has a lower error rate pr. move than player 1 (-0.005
vs. -0.017), hence he has a higher rating.

Typically errors become more expensive in terms of total MWC towards the
end of the match. However, error rate pr. move reported in equity is
relatively independent of match length. So the above indicates that
player 0 made his errors towards the end of the match, whereas player 1
made his errors in the beginning of the match.

Jørn

-- 
Joern Thyssen, PhD
Vendsysselgade 3, 3., DK-9000 Aalborg, Denmark
+45 9813 2791 (private) / +45 2077 2689 (mobile) / +45 9633 7036 (work)