How are these blunders?

Hi,

I had the computer analyse this game the other day. I'm confused as to how moves 34,35, and 36 are blunders, since I lost a Rook, but my opponent lost a Rook and 2 Bishops.

Thanks

The computer analysis explicitly stated that my moves 34,35, and 36 were blunders. Apparently, on move 34, I should have played Bxe5+, not Qxe5+. And on move 35, I should have played Bd4. And move 36 should have been Qf6+ instead. I fail to see how these moves were better. Also, move 33.Qe3 was only a mistake, not a blunder.

I disagree with your statement that 33. Qe3 was "only" a mistake. Paul already pointed out that you had a forced mate, but let it slip away. Giving anyone a chance to survive after you have forced mate can be considered a blunder.

As for the other three, I would say that the computer is the one that's mistaken here. On move 34, taking with the queen is *definitely* correct as taking with the queen allowed the massive gain in material you got immediately after it. Move 35's analysis is also bad, because you can (and did) win both the bishop and rook, and Move 36's analysis is just puzzling. If you checked on f6, the black king either runs away (and you lose your big chance to win material) or he repeats moves. I also fail to see what's so special about checking on f6.

Actually, I meant the computer analysis said 33.Qe3 was a mistake. Perhaps it thought Qf6+ was better because I had forced mate?:

Or:

However, I mated him in 3 moves

Houdini sez:33.Qe3 white goes from mate in 8(Re7+) and mate in 15(Nxf5) or Re8+(+20.0 atleast) to only +7.00(Qe3)

34.Qxe5+ is mate in 8 while Bxe5+ is 8.0+ advantage to white

35.Qh8+ is mate in seven, multiple other moves given as mate. Bd4 is +17.0 and probably mate aswell if i left it thinking long enough

36.Qxh7 is mate in 6 while Qf6+ is only 17.0+

My conclusion is you played well and the chess.com analysis is nutz 

What strength is Houdini? The chess.com computer is ~2000 for basic members.

Edit; whoops, navigated to the wrong page and deleted my earlier comment

Are you saying that the chess.com computer (at least for basic members) is not a good engine?

Of course it isnt. Or maybe it is, but it is not provided enough CPU time to provide a thorough analysis. There are many free engines that will give better results but of course you will have to do the work yourself instead of hitting ''submit analysis'' button

I just ran positions 33.Bxe5 to 37.Qxb7 in Microsoft Chess, and it agreed with my moves.

Because I didn't see it! That's probably why it suggested Qf6+. Thanks for pointing that out.