The black king is the sole defender of f6. This suggests 32.Rxf6, threatening Qe5:

A) 32... Kxf6 33.Qe5+ Kf7 34.Bd5+ and mate in two.

B) 32... Qe8 33.Re6, with the double threat Rxe8 and Qe5+, looks winning. For example, 33... Qf7 34.Qe5+ Kg8 (34... Kh6 35.Qg5+ Kg7 36.Re7 wins) 35.Bd5 Bf3 36.Re8+ Rxe8 37.Bxf7+ Kxf7 38.Qf4+ Kg8 39.Qxf3 wins decisive material.

C) 32... Re8 33.Rxg6+

C.1) 33... hxg6 34.Be5+ followed by Qxd7 wins decisive material.

C.2) 33... Kxg6 34.Qg5+ Kf7 35.Bd5+ Re6 36.Qf5+ followed by Bxe6 wins.

C.3) 33... Kh8 34.Be5+ and mate in two.

D) 32... h6 33.Bh3

D.1) 33... Qxh3 34.Rf7+ and mate soon.

D.2) 33... Kxf6 34.Bxd7 Rxd7 35.Be5+ Ke7 36.Bf6+ wins decisive material.

Gone for 32.R:f6 Bb3
(33...K:f6 34.Qe5+ Kf7 35.Bd5+ Qe6 36.B:e6+ Ke8 37.Qh8# )

This doesn't work if black plays <35...Kg8>. In fact it's surprising just how much more effective this is than 35...Kh8. After <35...Kg8 36.Rf7> the threat of Rxh7 no longer comes with check and with the K on g8 black is able to defend with <36...Rf8>.

<trnbg: Simply 32.Qxb7 (instead of 32.Rxf6) should win as well. The threat is 33.Re7+, so Black has to exchange Queens, but after 32...Qxb7 33.Bxb7 Ra7 34.c6 it's over.>

The instructive part is here (White to play)

click for larger view

29.Bd6 allowing the Queen to Rook Skewer and judging 29...Ba4 30 Qxd5 to be strongly in White's favour.

Too many replies, including Black not taking on d1 to through so 29.Bd6 was intuition. 'Suck and see what happens. it looks very good.'

29...Ba4 30.Qxd5 Bxd1. 31.Rxe6 Got as far as here with 29.Bd6 now waiting to see the reply. 31...Kg7 (the POTD position).

Now with the choice of replies narrowed right down to a human level White got down to pure calculation digging out 32.Rxf6, a good chance he saw it whilst waiting for Black to make their 31st move.

I'm sure less experienced players would see the idea of 29. Bd6 and maybe 'feel' it was the correct move but then try and go through the myriad of positions and replies to 29.Bd6 before committing themselves and possibly talking themselves out of it.

We are asking them to allow a simple Skewer, a bread and butter tactic at their level. (very often a game winner) and jump into what to them is at best a very unclear position.

It is a fact that less experienced player think more during a game than IMs'/GM's. They think at all the wrong moments and do not yet trust their instincts.

Good Game.

32. Qxb7 Qxb7 33. Bxb7

a. 33.. Re8 34. d5 Rxe6 35. dxe6 Re8 36. Bd5 Bg4 37. e7

32. Rxf6 Kxf6
33. Qe5+ Kf7
34. Bd5+ Qe6
35. Bxe6+ Ke8
36. Qh8#

this is an interesting comment.

the lore from cognitive research is that experienced players can reconstruct a glimpsed position better than less experienced players.

the interpretation usually offered is that the masters visually "chunk" positions into piece groupings (pawn chains, castling behind a fianchetto, doubled R's, etc.) that lets them "encode" a position more efficiently.

note how the cognitive strategy is assumed a priori to be visual rather than, say, dynamic. note also that the computer strategy of grinding out and scoring all possible n-deep permutations is effectively a visual, not a dynamic, procedure. possibly, what alphazero discovered with its neural nets, praised as "morphy like" by the cognoscenti, may be the cues that lead to dynamical considerations rather than the uniform reliance on exhaustive analyses. like morphy, alphazero seems crazy happy to throw away material for room to maneuver, which is hardly a crunchy scoring style of play.

so it's not accurately a matter of instinct (inborn stimulus-response link) but ingrained experience that guides a player. and the novel suggestion in the comment is that experienced players can "chunk" positions in several different dimensions that let them efficiently judge tactics, strategies and cues for time investment, clock management, opponent intimidation, whatever. none of it visual in the literal sense.

Hi Drollere,

<"... experienced players can reconstruct a glimpsed position better than less experienced players.>

This is true, that is experience. Things tend to even up with random positions, pieces just placed haphazard on the board with no rhyme or reason.

get a grip. you don't judge a troll by how they play but by what they post.

if <chrisowen> is a strong player, then he might provide insight. if he feels insight is fruitless, then he might say so in plain text, or otherwise abstain from participation.

yes, but even there one can find ambiguity. "random" strictly means grammatically nonsensical (both B's on dark squares, a white p on the eighth rank, white K next to a black R, etc.), whereas "disordered" might mean something dynamically possible if highly improbable to develop in play.

Hi drollere,

Some of the positions may have been problem mates in two, they can be very difficult to solve and quite hard to recall with a glance at the board.

There are still chunks but not regular patterns.

click for larger view

White to play and mate in two. (first look at it for 15 seconds and then recreate it on a board....then solve it.)

Regarding Chris,

He pays a sub, over the years his posts have slipped further and further into what we see now, the feeling is he is not well (or he has solved chess and is trying to find a language with which to tell us) and from what I can see he is not insulting or hurting anyone. Just tap the space bar, it takes you to the next post.