1... Nd2+
2. BxN Qe2+
3. Kg2 Qf1+
4. Kf3

Now the repetition-draw is obvious, but instead...

4... Be2+
5. Ke3 Bc4

getting the bishop out of the way with tempo. Seems to me that White cannot both save his queen and thwart a queen checkmate on e3. Is that right?

White threatens 32.Qxb5.

You see - chess is purely a game of luck. No skill at all is required.

Let's suppose that I play a weaker opponent. I calculated a deep mating variation and had to see everything. My opponent needs to see nothing... it's enough that he doesn't resign because of my obvious confidence in playing the moves, and then after I deliver the killing knock-out punch, it's enough that he sees, or just blindly goes 6. Bd3, enjoying the hole in my computation.

It's a little bit like in table-tennis when you win a point because your opponent, who is obviously stronger, does a smash, or spiking shot into your field which then goes out.

click for larger view

31...Ne5+! is not only best, it seems to be the only way to win. Thus my first suggestion 31...Bb4 is met by 32 Bd1; <solskytz>'s suggestion <1... Nd2+ 2. BxN Qe2+ 3. Kg2 Qf1+ 4. Kf3 4... Be2+ 5. Ke3 Bc4> is met by 6. Bd3; and <LoveThatJoker>'s 31...Ba4 is met by 32 Qb7! with perpetual check threats.

I actually don't understand why the endgame trainer doesn't take my loose B on a4 but rather proceeds 34. Bb3. The correct move is certainly 34. Bxa4 a5. I'm still better. My pieces are more effective in the attack than his - but I'm sure we've fallen into a bug in the endgame trainer. 34. Bb3 is crazy and easily squashed.

I often wonder how many times during a game we think, "Gee, I'm now at the "puzzle" part of my game." Of course, nobody really thinks like that.

Seems too simple, but I think it covers everything.

31 ... Nd2+ leapt out at me (3 seconds maybe?). But after about another half an hour of looking at it, I was still no further! I then I looked at Bb4, but finally, I went back to my original plan, but this time with a new checking square for the knight!

31 ... Ne5+
32 fxe5 Bd2+ (Kg2 Qf1# or dxe5 Bd2+ Kg2 Qf1#)
33 Kf4 g5#

<<solskytz>: oops.... looks like I just threw a piece away. Now I see 6. Bd3 at the end of my variation. Frustration!! You see - chess is purely a game of luck. No skill at all is required.>

There's no doubt chess does include luck. It is there because of the "human element". If you didn't get quite enough sleep, or your blood sugar goes too high or too low, or you had a disagreement with your spouse, etc. etc. it all impacts analysis. Generally the better your analysis becomes, the stronger the player you will be so it doesn't completely involve luck. If it was all luck, I assure you my rating would not have gone up and you wouldn't see such consistency in strong players.

It was a case of zig-zag and I picked zig.

My own mistake in analysis, as shown in this puzzle, was what Mr. Heisman calls "Quiescence error" if I remember correctly. It is actually "Stopping calculations when the situation is still 'hot', as in - rife with tactical threats and counter-threats".

I said - well, black surely can't defend against both threats... they are both very powerful - a checkmate, a queen... they are in (relatively) removed areas of the board... just too much.

Thanks! I did a lot of work to accomplish that but Dan has given me a lot of great advice and pointed me in the right direction. I try to provide some of the same advice on here, trying to be helpful, but seldomly get comments so I'm not sure if it does any good or if anyone reads it sometimes.

You're right about quiescence errors. But not seeing the Bd3 move could also be a board vision error, which is a completely different issue. It could also be that you "played (solved) too fast", in which case it becomes a time management issue. I never considered your variation at all but looked at other lines that made me suspicious of Nd2+. So then I looked at another check (Ne5+) and came across a forced mate and never went back to Nd2+.

<My own mistake in analysis, as shown in this puzzle, was what Mr. Heisman calls "Quiescence error" if I remember correctly. It is actually "Stopping calculations when the situation is still 'hot', as in - rife with tactical threats and counter-threats".> Very well stated.

My friend, yours are some of the most lucid and helpful comments on this site. I certainly look forward to what you have to say each day.

I get the impression that this site has lots of lurkers who read, enjoy and learn but don't necessarily comment. I wouldn't worry too much about the occasional silence - it hides a lot of watching eyes.