Re: [Bug-gnubg] Luck analysis, round-off error

[Joern Thyssen]

On Tue, Aug 19, 2003 at 08:37:54AM -0400, Christopher D. Yep wrote
[snip]
> 1. I then analysed the match at 0-ply checker, 0-ply cube.
> Luck adjusted result: Advanced: +17.63%
> 
> If I save the .sgf file, then re-open it, the luck adjusted result is then 
> reported as +17.66%.  If I then Analyse - Clear Analysis - Match and redo 
> the analysis, it's back to 17.63% again.  Is this round-off error?

Actually yes! 

Although gnubg uses 4 byte floats (7-8 digits) internally it writes the
SGF file with 4 digits for equities and 3 digits for luck. So the
difference between 0.1766 and 0.1763 is a rounding error!

I've commited a fix so gnubg uses 5 digits for the luck. With this fix
you should get the luck to 4 digits after loading the files -- you need
to resave your existing files to enjoy this feature :-)

Perhaps we should consider 5 digits for equities as well.

Jørn

pgp8LL8fuWjFD.pgp
Description: PGP signature