28. Bd2 Bxh3 29. Rd8 Rxd8 30. Qxd8+ Kg7 31. Qf6+ Kg8 32. e6 fxe6 33. Qd8+ Kf7 34. Qxc7+ Ke8 35. Qc8+ Kf7 36. Qd7+ Kg8 37. Kxh3

click for larger view

The checks get the white queen onto the d file with tempo so that the Bd2 is protected white can grab the Bh3. During this sequence, black has to avoid parking his king on a black square or the Bd2 will move away from the black queen with check.

Does black have a perpetual here? Fritz doesn't think so, with an eval of +2.7.

I did find an interesting wrinkle if black tries to exchange queens after 28 Bd2 Bxh3 29 Rd8 with 29…Qf5?! (hoping for 30 Qxf5 Bxf5).

click for larger view

That line immediately loses to 30 Bh6!, with mate in 3 or less.

click for larger view

You wrote: "If only black had played 31...Kg7 instead of Kh7. Then it looks like whites bishop could not have stepped in and white could have lost... Comments?"

How does White lose? He's a piece up, with all of Black's compensating Pawns far away from their Queening squares! After 32. ... Kg7; 33. Qd4 White is threatening a discovered check, Black has no checks and almost no moves for his Queen.

Hi Playgroundplayer;

To your defense, we must admit that each and every puzzles this week were a bit tricky.

For example, the monday one featured a position where White, already down a piece, could 'only' achieve re-balancing material + 1 pawn + the better position, as soon as one sees that Black could not avoid that from happening because of mate.

Then, we had the tuesday puzzle where one had to catch that there was no win but instead a drab draw for White, achievable through stalemate.

The puzzle we had been submitted yesterday was also uncommun in the sense that all White could really achieve was the gain of a pawn. The very reason why one might call that one a 'puzzle' (as for the one we had monday) is that for White to 'only' win a pawn, Black had to refuse some kind of a poisonnous gift, here, in the form of the WQ.

As for the one we have been strugling on today, there was that thin and subtle relation between 4 elements, namely, the a1/h8 and c1/h6 lines, the threat of advance of the e5-pawn, and the pin of the g2-pawn, that White had to find in order to really understand the nature of the position. And also, the fact that the 'combination' had to start with a rather quiet move (28.Bd2) was nothing to help.

So do not be too depressed (!) with your recent results. We' re all in a learning process. And sometimes, the misunderstanding may result in a better understanding, at some point or the other.

Personnaly, I found 28.Bd2 after spending some time over a1/h8, but totally missed 28. ...Bxh3(!), as I could not realise the importance of the WK to be unsafe. Thus, I learned that when one' s king is not safe, nothing holds.

Peace !

<28.?>

Material: =

Candidates: <[Bd2]>

-ML-
This seems a no-brainer... I think White mates in an extremely recognizable pattern with:

<28.Bd2!>

/A\
<28....Qxd2> the 'sacrifice', but neither move of Black seems to work.

<29.Bh6 Qg3† 30.Kxg3 Rd8 31.Qg7‡ 1-0>

/B\
<28....g5> the ONLY move to avoid this checkmate at all.

<29.Qxg5† Kh8 30.Qf6† Kg8 31.Rg3† Qg6 32.Rxg6 hxg6 > and White should win easily. Badly enough, upon opening the game I saw the last move was Bg5. Not that strange, because the game probably ended in the forced 33.Bd2 Kh7 34.Qh4† Kg8 35.Bg5, but I looked at this specifically due to the move shown. However, White is clearly winning after 32....hxg6. Time to check.

Black's surprise defensive reply 28...Bxh3! complicates things, as noted by <Once> in his amusing Star Wars analogy, because White must now proceed accurately to avoid losing.

Tarrasch's heroic followup 29. Rd8! puts him back on the winning track. With a little help from his opponent, Tarrasch manages to find a long and difficult winning continuation after 29...Bxg2 30. Kxg2. However, he missed the quick mate-in-three with 30. e6!

White to play (28.?) "Medium" (2.5 stars)

This was fairly easy, once you see it!

28. Bd2! Threatening 29. Bh6 and 30. Qg7 mate, and there is no defence:

28...Qxd3 29. Bh6 and there is no defence against the mate.

Tarrasch missed 30.e6!, with a mate in three. Instead, he made a serious error by playing 30.Kxg2??.

After 30.Kxg2??, Baird would have had very good drawing chances by playing 30...Qxb2!.

click for larger view

Additional analysis is needed to better determine the drawing possibilities for this position, but here is Fritz's preferred line after 30...Qxb2!: (.33) (24 ply) 31.Rxf8+ Kxf8 32.Qd8+ Kg7 33.e6 fxe6 34.Qe7+ Kg8 35.Qxe6+ Kf8 36.Qc8+ Kg8, (.22) (21 ply) 38.Qd8+ Kf7 39.Qd7+ Kg8.