2 puzzle questions:

Firstly, I'm confused: I pooched Wed, but I found Thur withs a lot of work, Fri in about 5 secs, and I found e4 today after some light work in the Knight move dead-ends along with h3, f3, and e3. This weekly result is not uncommon for me. Next week I'm likely to ace the ez's and not hit on the hard ones. What gives?

Secondly, although this game is not an example, am I the only one who gets bent when the text of the puzzle does indeed demonstrate the best move, but it isn't clearly winning, and is only so because the opponent fails to continue with best moves? I mean best move is best move, but we're to look for clear wins in these positions, and if it's only clear after the opponent boots a continuation, isn't that somewhat counter-productive to the exercise?

And 2 Engine/Evaluation questions (I use freebie versions of Spike and Rybka on a crap laptop):

1.) Why is it that when I'm evaluating a position of a game, and don't see the subsequent text move in the best lines, and then I advance to the next text move, evaluate a little, and then return to the original position, that next move now often shows up in the best lines, often as the very best line? Is this the whole "backsliding" deal I hear about? Is it a function of the weakish versions of the engines I'm using? Or is it my a function of my crap hardware coupled with the horizon effect? Or something else?

2.) I have noticed that when evaluating a position of say, approximately .85 to 1.75 (greater advantages answer their own questions, generally), and the side with the advantage makes a less than best, but still advantageous move (say .25), that even with best play by both sides to follow, that advantage will creep back up, often to clearly winning. I thought that a less than best play would allow the other side to equalize or reverse. This happens with significant, but not common, frequency, in my experience. Is this a function of engines not being able to *truly* evaluate long-term positions properly as dominating almost regardless of the next move? Or is it again my crap engine and laptop and ol' horizon effect?

It is good that sometimes that isn't the case here. For me I don't mind getting them or not, probably because I more often than not don't get em.

However, these puzzles do seem to open up more discussion. You read a lot of what if's and people looking for lines that are better. Anyhoo, I think that CG puzzles are no so much puzzles as an opertunity to discuss a games that are often flawed. Makes ya feel like a chess journo trying to annotate a game :)

I am toffee at the ol' chess but often find stuff I shouldn't. Like today I did consider the e4. However, I could not really follow it up with any good juice and it looked flimsy to me.

I think the reason I sometimes see things I shouldn't is because I do really know chess motiffs and, therefore, any move becomes a possibility to me. Some stonger players maybe are blinkered by looking for certain lines or what not. It's tough to think of a dog as anything other than a dog unless you've never seen a dog, if you get my drift.

31 e4 Qe8
32 Nxf5+ Rxf5
33 Qxe8 Rxe8
34 exf5 Nxc2
35 Rxc2 Rc8
36 Nxb7 ...

White has two pawns and black has little to say about it.

I spoiled the problem for myself, so I don't know if I would have found e4 on my own, but I can say that the move I was considering at the time of accidental revelation of the correct answer, is worth talking about. I was thinking about 31 Rd2. The line

31 Rd2 e6
32 e3 Nb3
33 Nxb3 axb3
34 e4 Qd7
35 e5 Rf7
36 d4 Nd5
37 Rd3 Ne7

At this point it becomes clear that white is attacking the pawns on the e, f, and g files. So my thinking about how to get the solution was that the better move would lead to a better attack. 31 h3?? is sure the wrong end of the stick, but 31 e4 would be the obvious other choice.

Because nobody else is stepping up to the plate, I will offer my two cents worth.

Question 1:

You have the basic imagination to see key moves, but you are not building up a routine for analysis. The routine should be thorough enough to get the elements of the position into your head, but its exact details are unimportant. My routine is the four basic chess elements (material, position - space and time, pawn structure, and King safety), the defensive burdens and positional vulnerabilities (e.g., to discovered attacks) carried by the opposite pieces, and the available moves (the "2-ply brain" - what can your pieces do?). By the time you have run through your routine, you should find the elements of the position in your head, and you can start to weave them together. When I am stumped, as I was today, I "go back to basics" (the best and most general advice I had from any teacher); I run through my routine. Eventually with enough practise, I expect to go through my routine unconsciously. Voila, my "intuition" is then doing all the work.

I found <UdayanOwen>'s posts this week very helpful. In particular, they explain that the routine for analysis terminate when a tactical stroke fulfills the strategical aims of the position. This wonderful harmony of tactics with position is something that I have not appreciated properly. I also recommend <dzechiel>'s analyses (who doesn't?) because they indicate his routine for analysis so transparently.

Question 2:

Who would want to be famous for only beating the fish?>

Kasparov and Fischer? ... but I do take your point :)"

Yes. I was surprised yesterday to find out that a guy that works for me was a serious chess player back in the Ukraine, he got close to becoming a master before he gave up. He beat Kramnik once. He probably beat a lot of fish along the way, but it's that victory over Kramnik which he will be able to tell his grandchildren.

Black might have practial survival chances after 31...Qe8! 32. Nxf5+ , but with best play White seems to come out ahead.

Here's my break-out of the combination with Fritz 8:

<31. e4 !! Nxc2>

Black's best may be 31... Qe8! 32. Nxf5+ Rxf5 33. Qxe8 Rxe8 34. exf5 Kf6 35. Nxb7 Rb8 36. Na5 Nxc2 37. Rxc2 Kxf5 38. Rxc7 Nd5 39. Rc5 Ke6 40. Rc6+ Kf5 41. Rxh6 , when White still appears to have a winning advantage..

<32. e5! Rc6>

Other tries for Black also fail:

32... Rf7 33. Qg6+ Kf8 34. Ne6+ Ke8 35. Qg8+ Kd7 36. Qxf7 ;

32...Rf8 33. Qg6+ Kh8 34. Ne6 ;

32... Nd4 33. exf6+ exf6 34. Qg6+ Kf8 35. Qxh6+
.

<33. Nxf5+ Qxf5>

Equally hopeless for Black are:

33... Kf8 34. Qxg4! Re6 35. Qg7+ Ke8 36. Qg8#

33...Kh8 34. Nxh6 Qf8 35. Nf5+ Kg8 36. Qxg4+ Kf7 37. Qh5+ Kg8 38. Qg5+ Kf7 39. Rxc2 Qg8 40. Qxe7+ Kg6 41. Ne4 Nd7 42. Rxc6+ bxc6 43. Qg5+ Kf7 44. Nh6+ ;

33...Kh7 34. Qf7+ Kh8 35. Qg7#.

Now White initiates a pin to gain decisive material with <34. Qxf5! Rxc5 35. Qxg4+ Kh8 36. Qd1 c6 37. Rxc2 .> With a winning advantage, White makes short work of Black's resistance after <31...Rxe5 38. Qd2 Kh7 39. d4 Rb5 40. Qd3+ Kg7 41. Re2 Re8 42. Re6 Rg5 43. f4! Kf7 44. Rxh6 1-0>

Stockfish 15 says:

31. e4

(31. e4 (e2-e4 ♕c8-e8 ♘h4xf5+ ♖f6xf5 ♕h5xe8 ♖a8xe8 e4xf5 ♔g7-f7 ♘c5xb7 ♘d4xc2 ♖c1xc2 ♖e8-c8 ♘b7-a5 ♖c8-d8 ♘a5-c6 ♖d8-d5 ♘c6-b4 ♖d5xf5 ♖c2xc7 ♖f5-b5 ♖c7-b7 ♔f7-e6 ♔g1-f1 h6-h5 ♔f1-g2 ♔e6-d6 h2-h3 ♔d6-e6 h3xg4 h5xg4 ♔g2-f1 ♔e6-d6 ♖b7-b8 ♔d6-e6 ♖b8-g8 ♔e6-f5 ♖g8-f8+ ♔f5-g5 ♖f8-f7 ♘b6-c8 ♖f7-g7+ ♔g5-f5 ♔f1-e2 ♖b5-b7 ♖g7-g8 ♘c8-b6 ♖g8-f8+ ♔f5-g6 ♖f8-f4 ♔g6-g5 ♔e2-d2) +5.33/42 299)

score for White +5.33 depth 42.

However Black didn't follow the Stockfish reply so Black took that Rook on c2. So - what happens?

31. e4 Nxc2
32. e5

(32. e5 (e4-e5 ♖f6-c6 ♘h4xf5+ ♕c8xf5 ♕h5xf5 ♖c6xc5 ♕f5xg4+ ♔g7-h8 ♕g4-g6 ♖c5xe5 ♕g6xh6+ ♔h8-g8 ♖c1xc2 c7-c6 ♕h6-g6+ ♔g8-f8 ♕g6-g4 ♔f8-e8 ♕g4-b4 ♖e5-b5 ♕b4-e4 ♔e8-d7 h2-h4 ♖a8-e8 ♖c2-e2 ♔d7-c7 ♕e4-f4+ ♔c7-c8 ♔g1-g2 ♖b5-d5 ♕f4-b4 ♖d5-b5 ♕b4-g4+ ♘b6-d7 ♕g4xa4 ♘d7-f6 ♕a4-c4 ♖b5-d5 ♕c4-f4 ♖d5-d7 ♕f4-f5 ♘f6-d5 h4-h5 ♖e8-h8 ♖e2-e4 ♘d5-f6 ♖e4xe7 ♖h8xh5 ♕f5xf6 ♖d7xe7 ♕f6xe7 ♖h5-d5 g3-g4 b7-b5 g4-g5 ♖d5-f5 ♕e7-e6+ ♔c8-b7) +8.98/45 208)

score for White +8.98 depth 45.

OK - so what about my move 31. Rd2 ?

31. Rd2

(31. .. Nd7 (♘b6-d7 e2-e3 ♘d7xc5 ♖c1xc5
♘d4-b3 ♘h4xf5+ ♔g7-h7 ♕h5xg4 ♕c8-g8 ♕g4-b4 ♘b3xd2 ♕b4xd2 ♕g8-f7 e3-e4 c7-c6 ♕d2-e2 ♖a8-f8 f2-f4 ♖f8-d8 ♔g1-g2 b7-b6 ♖c5-c4 c6-c5 ♖c4xa4 ♕f7-e6 ♕e2-g4 ♖f6-g6 ♕g4-f3 ♕e6-d7 ♖a4-c4 ♕d7xd3 ♕f3xd3 ♖d8xd3 ♘f5xe7 ♖d3-d2+ ♔g2-f3 ♖g6-g7 ♘e7-d5 ♖d2xh2 f4-f5 ♖h2xb2 ♘d5-f6+ ♔h7-h8 e4-e5 ♖b2-b3+ ♔f3-f4 ♖g7xg3 e5-e6 ♖g3-f3+ ♔f4-e5 ♖b3-e3+ ♖c4-e4 ♖e3xe4+ ♘f6xe4 ♔h8-g8 f5-f6 ♖f3-e3 ♔e5-f5 ♖e3xa3) -2.04/43 232))

31.e4!! and 32.e5! reveals how poorly based the scaffold is. All comes crashing down, becoming holes and loose bricks in the wall to pick up.

The pawn on f5 protects g4 and the rook on f6 protects g6. These details suggest 31.e4:

A) 31... Nxc2 32.e5

A.1) 32... Rc6 33.Nxf5+

A.1.a) 33... Kg8 34.Qxg4+ (quicker than 34.Nxe7+) and mate in three.

A.1.b) 33... Kf8 34.Qxg4 Qxf5 (34... Qd8 35.Qg7+ Ke8 36.Qg8#) 35.Qxf5+ wins decisive material.

A.1.c) 33... Kh7 34.Qf7+ Kh8 35.Qg7#.

A.1.d) 33... Kh8 34.Nxh6 should be winning. For example, 34... Qf8 35.Nf5+ Kg8 36.Qxg4+ followed by Rxc2 with attack and two pawns for the exchange.

A.1.e) 33... Qxf5 34.Qxf5 Rxc5 35.Qxg4+ and 36.Qd1 seems to win decisive material.

A.2) 32... Rf7 33.Qg6+ Kf8 34.Ne6+ Ke8 35.Rxc2 c6 36.Nxf5 looks crushing.

B) 31... fxe4 32.Nxe4

B.1) 32... Nxc2 33.Nxf6 must be winning. For example, 33... Kxf6 (33... Nd4 34.Qg6+ and mate next) 34.Qg6+ Ke8 35.Rxc2 and the black king is too exposed.

B.2) 32... Nf3+ 33.Nxf3 Rxf3 34.Rxc7 looks winning.

C) 31... f4 32.e5 as in A.

D) 31... Qe8 32.Nxf5+

D.1) 32... Nxf5 33.Qxg4+ Qg6 34.Qxg6+ Kxg6 35.exf5+ followed by Nxb7 wins two pawns.

D.2) 32... Rxf5 33.Qxe8 Rxe8 34.exf5 Nxc2 35.Rxc2 looks good for White.