When computers are losing they make weird moves sometimes, also it was Fritz 11 not 12. The dude that put the video up isin't advertizing for Ubisoft so I have no reason to believe that that game was not accurate. Also we don't know the time controlls it was played in. It looked pretty fast to me, and in closed games with short time controls computers usually perform poorly.
Chessmaster 11 vs Fritz 11,12

Chessmaster is a CPU hog. If both engines were running on the same box, Fritz was running at half or less of its usual search depth. Even weakened thus, it created a winning position that it failed to exploit.
Given the technical limitations of the Chessmaster software for such engine-engine battles, and the clear evidence that this is a low budget test (unregistered screen capture software), we have no reason to trust that this test was fair.
http://www.youtube.com/watch?v=CTQi2e7vt4s - first game
http://www.youtube.com/watch?v=lVnkbhoSHQI&feature=related - 2nd game
http://www.youtube.com/watch?v=kUR43fOLJfw&feature=related - 3rd game
Chessmaster 11 vs Fritz 12???
-> http://www.youtube.com/watch?v=mjA58LwBQFE&feature=related
but there is no way this is true.
At 1:58, Fritz moves up his pawn instead of moving his bishop to G3 which would have paralyzed white's king. epic fail.
Then black king moves three spaces and decides it's a losing battle. Fritz 12 should be able to calculate that.
Then when he realizes it's a losing battle he makes two incredibly bizarre moves with his bishop instead of retreating his king or advancing his pawns.
There is no way this was Fritz 12.
Please post your games between the two engines :)