Test Match, 1'+1"/40+1'/99+1'/99  0                         
1   Rybka 3 1-cpu 32-bit  +190/-59/=69 70.60%  224.5/318
2   Fritz 11              +59/-190/=69 29.40%   93.5/318
For +150 ELO

This is great.  The odd thing was, that Fritz got better and better as the games went on.  Started out as +400 ELO 10 games, was +250 ELO at 100 games, +200 ELO at 200 games and 150 ELO at +318 games.   I am not sure what is going on here.   First if it is "learning" then Rybka is VERY good for Fritz, and it has gotten over 100 ELO better against Fritz in just 300 games.  I wish I could learn that well.  But if it is not learning, I have no idea why Rybka would become that "fatigued" over time.

I am going to retest for a day at a 5 minute blitz time frame and see what happens.

(This was single CPU engine match in Fritz 11 GUI, running in Parallels on my mac, I am not sure if any of that matters).