I say recalled - I only have to see Marco's name and this game pops up.

I'm afraid in my mind he is in the Dufresne, Kieseritzky, Levitsky club (there are other members...)

If they want to select their Sunday Problems from famous games or traps I've had on my demo-board in front of a class then no complaints from me.

Black comes out of it a piece down if either the a8-Rook or the f8-Rook takes on d8.

In the Rf8xD8 line here.

click for larger view

White plays Kf1, The e5 Knight moves and the c5 Bishop falls. (but see below)

This Kf1 idea is not available in the Ra1xd8 line.

click for larger view

Here White takes on c5 and wins as in the game.

Should White try 15.Kf1 in the Ra8xD8 line.then Black drops the Bishop back to b6. If White takes the e5 Knight then fxe5 is a CHECK.

click for larger view

Black is actually good here.

It is worth noting (you might have this OTB as White one day) that even here (in the Rf8xd8 line) with Black to play.

click for larger view

Black can play 15...Rf8 (Reinfeld mentions this try.)

click for larger view

and again White cannot (yet) grab the e5 Knight, He has to prepare it with, according to Reinfeld 16.Ke2 (it looks OK it's one of them better to know than having to dig it out OTB type moves. May even be worth falling for this trick in blitz just to play 15...Rf8.)

I will point out that if White tries 16.Re2 intending Nd3-f2 after taking on e5 there is a neat perpetual lurking.

click for larger view

16.Re2 Rad8 (this stops the Nd3-f2 idea) 17. fxe5 fxe5+ 18.Nf4 Rd1+ 19.Re1

click for larger view

19...Rxc1! 20.Raxc1 Rxf4+ 21.Ke2 Rf2+ 22.Kd1 e3.

click for larger view

Actually threatening a cheeky pawn and Rook Epaulette mate.

And White's best appears to be to settle for the perpetual with 23.Re2 Rf1+ etc...etc...

The final position:

And finally, we must wrap it up correctly and squeeze the last instructive point out of the game, writers (and kibitzers) never seem to go deep enough when discussing this game. Tarrasch had it worked out right down the nth degree. No surprises.

Here Marco resigned.

click for larger view

Try 18... Nd3 19. cxd5 Nxe1 20. Rxe1

click for larger view

And no win of the pinned Bishop with 20..Kf7 as 21.Bb4 or Bh4.

8.Bxc6 Bxc6
9.dxe5 dxe5
10.Qxd8 Raxd8
White exchange Queens so that he would have two protectors of e4 pawn. had he not done so, Blak could have made the exchange of Queens and either Rook or the Knight would be sitting on d1 and e4 pawn would hav had one protector

11.Nxe5
White is now a pawn up. Any attempt to capture e4, will cause further loss to Black:

11...........Bxe4
12.Nxe4 Nxe4
13.Nd3!
e file belongs to White now, there are two unsupported Black pieces there, one has to go:

14...♗c5+?! Better is 14...♗h4 15.g3 ♘xg3 16.hxg3 ♗xg3 where Black get two pawns for the knight.

If 10...♖fxd8 follows 11.♘xe5 ♗xe4 12.♘xe4 ♘xe4 13.♘d3 f5 14.f3 ♗c5+ 15.♔f1! ♖f8 16.♔e2!

a)16...♗b6 17.fxe4 fxe4 18.♘f4 (to prevent 18...♖f2+) g5 19.♘h3 (to prevent 19...♖f2+ again) g4 20.♘f4

l was barely awake when I made that dumb comment. Of course <10...Bxd8> is useless because only a R-capture prevents <13.Rxe4> with the back row mate threat.

On silicon, White's advantage at the end is actually a lot more mild than in many games where the loser trudges on for sometimes quite a long time before resigning. It's not even +3! A computer would probably, in fact, hang around quite obstinately as Black here (hope you're good at endgame judgment if it's blitz, because it might very well drag on long enough to make you lose on time, since it often takes a while to convert an exchange up endgame).

Well, after <14.c4>, my Stockfish NNUE gives it +6 at depth 20, with a slow but constant increase with depth. (I feel that NNUE really makes a difference, creating much more accurate evaluations as a result of just pragmatic unbiased learning, not some puny human trying to tell the machine what and how to evaluate :)

As to the trap, it sure is a nasty one and deserves its (in)fame. <7...0-0?> loses a pawn with no compensation whatsoever, but after <14.Bxc5+??>, Black loses at least the exchange, and while its position doesn't look *that* bad to me, to Stockfish it's as if you're down a rook. And both moves look total natural and sensible, which is a key property of any trap deserving its name.