|Sep-10-03|| ||courts: on .....29 why not Rh8? Black had chances, but played poorly. |
|Sep-10-03|| ||Honza Cervenka: 32.e6 looks very good. If 32...f6, then 33.Rxg6+ Kxg6 34.Qh5+ Kg7 35.Qf7+ with mate. If 32...Rf8, then 33.Nxd5 cxd5 34.Qf6+ |
|Sep-10-03|| ||Brian Watson: If 29...Rh8 30.Qg5, doesn't black still have to contend with the threat of 31.fe? 29...exf4 was a logical move, because it allows black to escape the pawn fork by Qd2+ |
|Jun-14-04|| ||Gypsy: <Honza Cervenka: 32.e6 looks very good...> Although White should still prevail, lots of work remains after Black counterstrikes with 32...Nc3 33.bxc3 Qd1+ 34.Rxd1 Rxd1+ 35.Kb2 gxf5 and if 36.Qe7 then probably 36. Rg1. |
|Jun-14-04|| ||Gypsy: This game won the 1-st Brilliance of the Prague 1908 tournament. The brilinces were judged by K. Traxler of the Traxler Counter-Attack fame.|
Prague 1908 results were: 1-2.Duras & Schlechter(13.5/19), 3.Vidmar(13), 4.Rubinstein(12.5), 5.Teichmann(12), 6.Maroczy(11.5), 6-9.Leonhardt & Marshall & Salve(11), 10.Janowski(10.5), 11. Duz-Chotimirski, 12.Alapin, 13-14.Mieses & Suechting, 15.Spielmann, 16.Prokes, 17.Bardeleben, 18.Kvicala, 19.A.Rabinovic, 20.F.Treybal.
|Jun-15-04|| ||Calli: Yes, and I think Rubinstein lost his chances when Janowski won a Brilliancy prize for Janowski vs Rubinstein, 1908 |
|Jun-15-04|| ||Gypsy: Looks like a great game <Calli>. Cool finnish, but, at first glance, also strategically interesting stuff before that. Thanks. |
|Jul-27-04|| ||patzer2: Interestingly, the best move in this "brilliancy prize" game is the one that was not played in lieu of 32. Rxf7+. |
Fritz 8 analyzes 32. Rxf7+ as near equal and leading to only a small White advantage after 32...Kxh7 33. Qh7+ Kf8 34. Qxg6 Nxc3 35. bxc3 Qd1+ 36. Rxd1 Rxd1+ 37. Kb2 Rd7 38. e6 Rc7 39. Qf6+ Kg8 (+1.12 @ 15 ply and 1417 kN/s).
However the best move and the one not played here is 32. e6!! which is given as the solution to number 1745 in Chess Informant's Encyclopedia of Chess Middlegames/Combinations. The combination is classified under the theme "Demolition of Pawn Structure" via Sacrifice at g6 (g3)."
The reason the 32. e6! solution is classified as a sacrifice on g6 is that after 32.e6! fxe6 [32...f6 33. Nd5 ; 32...Rf8 33. Nxd5 cxd5 34. Qf6+ Kg8 35. exf7+ Kh7 36. Rh1+ ], the thematic followup move 33. Rxg6+! quickly wins after 33...Kxg6 34. Qh5+ Kg7 35. Rf7+ Kg8 36. Qh7#
|Jul-27-04|| ||patzer2: <Honza Cervenka> I just noticed you had previously recommended 32. e6!! With Informant, Fritz 8 and you recommending the move, it definitely looks like a winner. |
|Jul-27-04|| ||Gypsy: Ok, maybe you can help me <patzer2>. What is wrong with Duras finnish in the game? The game went 32.Rxf7 Kxf7 33.Qh7+ Kf8 <34.e4> and Duras blew Suechting out of the watter. Your Fritz 8 argues for <34.Qxg6+> instead, with about +1.12 advantage to White. Therefore we need to conclude that either Duras <34.e4> was an error and there is a hidden resource for Black or that Fritz 8 did not see the game conclusion. I can not find that hidden resource. |
|Jul-28-04|| ||patzer2: <Gypsy> I wouldn't argue that 32.Rxf7 Kxf7 33.Qh7+ Kf8 <34.e4> is an error as compared to <34.Qxg6+>. Although I think the latter Fritz 8 move offers more resistance and better chances for a draw, I'm not so certain that White couldn't find a win with best play in even the Fritz 8 line. |
My point is that 32. e6!! gives White a much bigger advantage and an easier win than with the 32. Rxf7!? move played. With best play against 32. Rxf7!? White's task is certainly not so easy as to make it a blowout win. Long tortuous wins or draws are possible in such positions and are exhausting and difficult to analyze. So I prefer clear strong advantages, where they can be found, such as 32. e6!! offers here.
|Jul-28-04|| ||Gypsy: Ok, this puzzled me <patzer2>, so I cranked out some variation to get to the bottom of this. First let me present the variations. Since they are mostly human generated (on top of that, some by me), take their veracity with caution. |
(A) Variations ofter <32.Rxf7>
A1: Game <33...Ke6> 34.Rxg6 Kxe5 35.Qg7 Kf5 (35...Nf6 36.Qxf6#) 36.Rg5+ Ke6 37.Re5+ Kd6 38.Ne4# -- a checkmate as clean and pretty as they come.
A2: Game <35...Rd7>, objectively the best, 36.exd7+ Qxd7 37.Qg8 Ke7 38.Qxa8 (or 38.Re1+ 39.Qxa8) with R vs 2P and easy win.
A3: Game with <35...Rd5!?> with the hope of <36.cxd5?> Qb4+! and the draw by perpetual.
A5: Game with Fritz' <34.Qxg6> Nxc3 35.bxc3 Qd1+ 36.Rxd1 Rxd1+ 37.Kb2 Rd7 38.e6 Rc7 39.Qf6+ Kg8 (+1.12 @ 15 ply and 1417 kN/s). But I realy do not see the point of diverging like this from Duras.
(B) Variations after <32.e6>:
B1: <32...fxe6?> 33.Rxg6 Kxg6 34.Qg4+(!) Kh6(Kh7) 35.Rh5# which is just a wee faster than the continuation 33...Kxg6 34.Qh5+ Kg7 35.Rf7+ Kg8 36.Qh7#
B2: <32...f6?> 33.Rxg6 Kf8 (or 33...Kxg6 34.Qg4+ Kh6 35.Rh5# again) 34.Qh8+ Ke7 35.Qg7+ and either 35...Ke8 36.Qf7#, or 35...Ke6 36.Nd4#.
B3: <32...Rf8?> 33.Nxd5 cxd5 34.Qf6+ Kg8 35.Rxg6+ fxg6 36.Qxg6+ Kh8 37.Rh5#.
B4: <32...Nxc3+!?> 33.bxc3 Qd1+! <34.Rxd1> Rxd1+ 35.Kb2 gxf5 36.Qe7 Rg1, where White still has to work for his win.
B5: <32...Nxc3+!?> 33.bxc3 Qd1+! <34.Kb2!> and Black can not stop the disasters, e.g., 34...Qxg1 35.Rxf7+ Kg8 36.Qh7#.
As far as I have pieced together, Duras rejected 32.e6 in favor of 32.Rxf7 because of the variation B4. He, Traxler, and all commentators after that, till at least 1953, missed the line B5 (I just noticed the B5 possibility today) which makes B (32.e6) the slightly more accurate continuation of the two. I presume that Duras knew that he had an easy win in A (32.Rxf7) and so he just went there without spending much extra time. Apparently Duras had prepared 32.e6! as a repply to passible 31...Ng8.
The only puzzle that remains is why Fritz suggests A5?
|Jul-28-04|| ||Calli: <Gypsy> Fritz (see previous) points out that 33...Ke8? in the game is a mistake. It prevents the defense 35...Rd7 because exd7 is check. If Suchting plays 33...Kf8! instead, White's edge is very small. Therefore 32.e6! is preferred.|
Unfortunately after 32.e6 Nxc3+ 33.bxc3 Qd1+ 34.Kb2!! and white will mate in a few moves. Good idea, though.
|Jul-28-04|| ||patzer2: <Gypsy> Thanks for your analysis of the merits of 32. e6!! versus 32. Rxf7!? <Calli> Appreciate your strong observation on how Fritz 8's analyzes of 32. Rxf7+!? 32...Kxh7 33. Qh7+ Kf8!? (as opposed to Suechting's weak 33. Ke8?) gives Black counter-chances. White might still be able to win, but the job is much harder after 33...Kf8!? I think we're in agreement that 32. e6!! is the best choice here. |
|Jul-29-04|| ||Gypsy: Thank <calli>, you are a great pal! The idea of 33...Kf8! did not even cross my mind. At least I feel good about finding the 34.Kb2!! on my own.|
I am begining to get a notion that some moves by a king-under-attack tend to ride in human-players' blid spot. Here we have two examples that were missed by a whole slough of distinguished tacticians. And I remeber noting other examples (though did not make bookmarks at the time and will have to search for those). Any hypotheses what is at play here? What characterizes moves like 33...Kf8! and why do we subconsciously discard them?
|Jul-29-04|| ||patzer2: <Gypsy> Perhaps it is because we subconsciously accept the premise that all protected Kings are adjacent to pawns or pieces of their own color. I admit I was startled when Fritz 8 first recommended 33...Kf8! putting the naked King all by its lonesome away from any apparent close protection. Yet, in this case, breaking that premise or rule of thumb maximizes resistance. However, it's often hard to break those rules or to find the exceptions, whether in the heat of battle and under time pressure or analyzing and trying to apply human reasoning to chess positions. |
|Jul-29-04|| ||Calli: I agree. There is also the "carrot" of a win. The player is the rabbit instinctively seeing the escape to the queenside and a win. It is very difficult to dimiss that possibility and calculate cooly that its better to give back the rook with Rd7 exd7 Qxc3 and go for a draw. |
|Jul-29-04|| ||Gypsy: (Funny, at one point in my Jul-27 note I actually typed Kf8, but thought only in terms of Ke8. Sorry for the miscom <p2>; that Kf8! did not register untill two days later!)|
Lessons from Kf8!
Natural tendencies are:
(1) King under an attack wants to flee.
(2) King under an attack wants to hide.
(3) It is generally good to be greedy in chess.
Thus we blind-spot the moves that:
(1') Move king towards danger.
(2') Move king into more open environs.
(3') Return material.
Lessons from Kb2!!
Natural tendencies are:
(4) Blast away the attacking piece, especially queen. (5) Rook is certainly less dangerous than queen.
(6) RxQ must be a great trade of material.
(7) Wantonly droping a rook must be bad.
(4'-7') that soo-natural reduction, 34.RxQ RxR+, blocks all other
moves from consideration.
|Jan-10-09|| ||WhiteRook48: the weirdest discovered check I have ever seen.
It would be even weirder if white played e8=R+!