I've been fooling around with Chessmaster's ranked play feature, testing my play against several "personalities" at rapid time control. My usual gig is to open the program to "ranked play" and select "random opponent" and "random color." If I lose (often), I may play more against the same opponent.
Today the program froze for several minutes, and I lost two minutes on the game clock--kind of frustrating in a game 10. Even so, I was doing okay until I reached this position.
White to move
I played 28.Qd1, opening myself to an unstoppable attack. This error was the game losing move.
The game finished:
28...Bxh3 29.gxh3 Rxh3 30.Rh2 Rxh2+ 31.Kxh2 Qf2+ 32.Kh3 Rh8+ 33.Kh4 Qf4#
Here's my beef: the post-game analysis tells me that my "worst move" was 31.Kxh2--the only legal move in the position. I suppose the software thinks I should have resigned inasmuch as it is mate in five.
Such post-game analysis distracts the chess student from a necessary task: identifying the key turning point in a game. Instead Chessmaster highlights the move that changes by the greatest margin the numerical evaluation in an easily won, or completely lost position. This feature is a disservice to users of Chessmaster software. Bad advice can be worse than none at all.
In contrast, Fritz (Hiarcs, Junior, etc.) does not offer numerical analysis at all from move 29 on. Rather, it is mate in 12, mate in 6, and mate in 1 after White's suboptimal moves.
From the diagram position, 28.Qf3 maintains White's advantage. All other moves shift the advantage to Black, my Qd1 decisively so. If one is using software in training, it should help identify the key position, not offer phantom assessments after the battle has been decided.
2 hours ago