1 sac the queen

2 dounle check

3 mate with the bishop

Keres adopts an unusual line in the French with 3. Be3. There are only 38 other games (not counting both of the Keres vs Verbac games) in the Chessgames database with this move order. Chessgames refers to it as the Alapin Gambit while ECO refers to it as the Alapin Variation. It is also known as the Alapin-Diemar Gambit since f3 is typically played in similar fashion to the Blackmar-Diemer Gambit.

You are right, it should be winning. This shows how much I avoid quiet moves, because they usually give the other player more options and more ways to refute a sacrifice so I tend to look for lines more forcing. This helps me to prune my tree of analysis and instead look for something more reliable (like Qh6+). But if a move like Qh6+ didn't exist here, then I am forced to look at Rxf6 more carefully and then considering a move like Raf1 as in your line.

So basically when I said that after Kg6 "it didn't look good for white", I'm not really concluding that white loses. I was pointing out that it looks suspicious and therefore I'm not spending any more time on that line just yet until I examine other possible candidates. This follows a principle my coach taught me--"Look wide before you look deep."

In essence my written analysis attempts to demonstrate what I was thinking and how I concluded with the correct (or incorrect) solution to the puzzle.

Of course I saw the potential mate threat of Qh6# if only my rook were out of the way of my DSB.

Naturally, I started looking at forcing rook moves. Although there were none as forcing as I would like, it did seem that <14.Rxf6> was intriguing, since the rook can't be taken (14...Kxf6 15.Qg5# or 14...Qxf6 15.Qh6#).

Black's obvious reply is <14...gxh5>, and I lose my queen, but have a strong attack underway (perfect for a puzzle, right?)

I continue <15.Rf7+ Kg6> (not 15...Rg8 16.Rd7#)

click for larger view

I was kinda stuck for a moment, until I noticed that moving my LSB to d3 would be mate if only my LSB weren't busy guarding my rook on f7.

So, I relieve the bishop from defending the rook with <16.Raf1>, now threatening Bd3#. With my board vision getting blurry, I couldn't be sure this was foolproof, but I couldn't see any good way out for black.

If 16...Bf5, then 17.Nxf5 and now I threaten Nh4#.

If 16...e4 (to prevent Bd3), then 17.Bd5! threatens Bxe4# (note: 17...Qxd5 18.R1f6#).

I realized afterwards that I overlooked 16...h4, which creates an escape square for the king. Even if I had considered it, I probably wouldn't have forseen the correct follow-through, which my computer tells me is 17.Bd3+ Kh5 18.Ne4! (threat: Nf6+).

Much easier than yesterday because there is only one line of play: 14 Qh6+ Kxh6 15 Rh4++ Kg7 (forced, because of the power of the double check) 16 Bh6 mate.

POSTSCRIPT: Qh6+ leaps out as candidate move to any chess player brought up to look out for/ beware of double checks as a winning tactic in a tactical arsenal. Georges Renaud and Victor Kahn devote a chapter to the tactic in their book "L'art de faire mat" (English translation has benn published - The Art of Checkmate).

14. Nf5+ Bxf5 15. Qh6+
(15. Rxf5? h6! 16. Qg4! Nxg4? (16...Nbd7 17. Ne4 Kh7 18. Nxf6 Nxf6 19. Rxf6 Qxf6 20. Rf1 Qc6 21. Qh4 g5 22. Bxg5; White is more than fine but Black is still not completely lost) 17. Rf7+ Kg8 Rd7#)

15...Kxh6 16. Rh4++ Kg7 17. Bh6#

You are quite right! 11...Bd7

click for larger view

12 Ne6! works ONLY if the Queen moves. Unfortunately Black simply hands back the material with 12 exf4 13 Nxd8 Kxd8 to leave

click for larger view

and White will lose another piece.

Lol! You are showing your age. Alot of our younger Kibitzers may not get that one.

Lol! You are showing your age. Alot of our younger Kibitzers may not get that one. >

Here you are, kids - educate yourselves!