|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing> |
| Sep-26-22 | | TheaN: Is this technically a queen trade Monday? <9.Qf7+ Qxf7+ 10.exf7#>. Not a sacrifice, but a queen is lost and a mate in two so I'll take it. |
|
| Sep-26-22 | | TheaN: The Alekhine is a bit of a double-edged minefield, though safe to say it's likely not Black's best attempt at creating one: the French and Sicilian are more theoretical but have a slightly more solid base. Said that, here Black does in fact achieve a 'haze move' from White, as 5.Bxf7+? is bollocks (according to opening explorers a novelty in fact) and Black was defending properly up to 8....h6?? This line with Bc4 is already less good as 4....Nb6 can be safely played and it's roughly equal. How Black could think of h6 I don't know. Says the person that left a rook en prise with check in his last classical game, so do as I say not as I do. |
|
| Sep-26-22 | | ClassZPlaya: For me the most interesting part of today's puzzle is that Stockfish gives 1 ... Nf6 a " ? ". Does this mean that modern engines evaluate Alekhine's Defence as unsound or (unlikely) is it just not in Stockfish's opening book? |
|
| Sep-26-22 | | goodevans: <Cheapo by the Dozen: I presume Dake was giving the simul, and didn't realize how strong a player he was facing?> That's something I pondered upon and I think you're probably right. Doesn't excuse falling for a mate-in-2 though! |
|
| Sep-26-22 | | TheaN: <ClassZPlaya: For me the most interesting part of today's puzzle is that Stockfish gives 1 ... Nf6 a " ? ". Does this mean that modern engines evaluate Alekhine's Defence as unsound or (unlikely) is it just not in Stockfish's opening book?> It's SF11. We're up to 15 at this point, and NNUE confirms that 1....Nf6 is likely not Black's best move, it's not worth a ?. In a sense, this is where purely theoretical chess and practical chess clash. The Scandinavian is in a similar type of evaluation. As with most space surrendering moves, objectively White must be quite a bit better, but it's very tough to convert and the slightest mishap and the advanced pawns fall like flies. |
|
Sep-26-22
 | | chrisowen: I glades job it buck with Qf7+ jeep aoe its doh axioms juggle v i dang it's fade Qf7+ aid; |
|
| Sep-26-22 | | saturn2: 8...g6 like others said kept black's advantage. The incorrect sacrifice 5 Bxf7 paid off for white. |
|
Sep-26-22
 | | PawnSac: <Brenin: The wrong pawn: 8 ... g6 would have given Black's K enough breathing space, and Black a winning advantage.> agreed. and after Nf7 Bxe6 Nxh8 Kx black has two minors for the rook. A healthy plus. <White's 5 Bxf7+ was somewhat optimistic.> somewhat overly optimistic, indeed.
<Cheapo by the Dozen: I presume Dake was giving the simul, and didn't realize how strong a player he was facing?> Even if.. It's hard to fathom how a player like Dake can beat Alekhine, put up best results in Warsaw teams, and win a strong blitz event 12/12, then fall for this simple M in 2?? |
|
| Sep-26-22 | | Hercdon: The names must be inverted |
|
| Sep-26-22 | | DrGridlock: <newzild: Interesting that Stockfish 11 considers Alekhine's Defence bad enough to warrant a question mark.> Fritz 17 - to depth 32 - gives 1 e4 f6 an eval of .4. By contrast, Fritz 17 - to depth nn - gives 1 e4 e5 an eval of .25. So modern engine evals rate the Alekhine worse, by about .15 points, for black - but certainly not enough to gain a "?" annotation. |
|
Sep-26-22
 | | keypusher: Just for the hell of it, I looked at various responses to 1.e4 with SF 15 running on my Mac desktop out to about 40 ply. As interesting to me as the evaluations were the opening lines it preferred. In order of White advantage: 1.e4 e5 -- +.18 to about .3. It played the Ruy Lopez/Berlin throughout, but just before I stopped running it SF15 switched to the Giuoco Piano (reached via 3.Bc4 Nf6 4.d3 Bc5). 1.e4 e6 -- +.42. To my surprise, SF played 3.e5. 1.e4 c5 -- +.52 SF preferred the Sveshnikov (with 7.Nd5 Nxd5 8.ed Nb8). 1.e4 c6 -- +.57. SF played the old classical line 2.d4 d5 3.Nc3 de 4.Nxe4 Bf5. I left it running and at around 45 ply it switched to 2.Nc3 d5 3.Nf3 Bg4. 1....d6 -- +.67. SF went to a Philidor after 2.d4 Nf6 3.Nc3 e5. 1....g6 -- +.69 SF comes up with humdinger for your next blitz game: 2.d4 Bg7 3.Nc3 d5!? 4.ed Nf6. Fairly telling that the engine would rather sacrifice a pawn than play a normal Pirc/Modern. 1.e4 Nf6 -- +.7 to .9. The evaluation was still moving around a fair amount past 40 ply. SF played the Four Pawns Attack, which is not too surprising. If, like me, you make -3 or worse moves with some frequency, you shouldn't worry too much about these results. But I think they go some way toward explaining the prevalence of 1....e5 at the very top levels, and the rarity of hypermodern openings. |
|
| Sep-26-22 | | Pawn Slayer: I think the names may have been reversed.
In every simul I've ever taken part in or given, the player giving the simul has always had white. |
|
| Sep-26-22 | | unferth: I believe the names & game score are correct and that Dake was playing black even though giving a simul. From the chess column of the August 8, 1937 Hartford Courant: "Our last one was played in a simultaneous exhibition in Milwaukee in April of last year, and the usually reliable Arthur Dake was on the losing end. His opponent, however, was not the well-known Hartford player." The game score follows, with white listed as A. Powers & black as A.W. Dake. https://www.newspapers.com/image/le... |
|
Sep-26-22
 | | FSR: <keypusher> Fascinating, as Mr. Spock would say. In practice, ChessBase shows the following percentage scores for White after Black's most popular responses to 1.e4 g6 2.d4 Bg7 3.Nc3: 3...d6 55.8%
3...c6 53.5%
3...c5 (Sniper) 58.6%
3...a6 45.7%(!)
3...d5!? 55.7%
After 1.e4, the percentage scores (again, for White) after every remotely reasonable Black response: 1...c5! 50.6%
1...e5 56.0%
1...e6 53.6%
1...c6 53.0%
1...d6 54.5%
1...d5 54.3%
1...g6 51.2%
1...Nf6 52.3%
1...Nc6 53.5%
1...b6 57.0%
1...a6 54.8%
1.e4 Nc6 is the Rodney Dangerfield of chess openings. It "don't get no respect," but more players play 2.Nf3 (allowing 2...e5 if Black wants) than attempt to refute it with 2.d4. And indeed 2.Nf3 scores 55.6%, while 2.d4 scores only 52.9% (51.6% against 2...d5, and 53.3% against 2...e5). After 1.e4 Nc6 2.Nf3, 2...d6 gives White 53.8%, 2...e5 55.6%, 2...d5 (not good if White knows what he's doing) 52.7%, 2...Nf6 58.6%, 2...f5!? (Colorado Gambit) 55.5%, and 2...e6 (58.5%). |
|
Sep-26-22
 | | MissScarlett: <Our last one was played in a simultaneous exhibition in Milwaukee in April of last [sic] year, and the usually reliable Arthur Dake was on the losing end.> |
|
| Sep-26-22 | | unferth: whoops, my transcription error. "April of this year." thanks. |
|
Sep-26-22
| | keypusher: <FSR> So, in Chessbase, the Sicilian is the clear champ, and just about everything outscores 1....e5. Taking the second item first, I'd guess there aren't enough 19th century games in Chessbase to have much impact on scoring. So what hurts 1....e5, I surmise, are non-Berlin defenses to the Ruy Lopez. For example, SF is not a fan of the old Chigorin (3....a6 4.Ba4 Nf6 5.0-0 Be7 6.Re1 b5 7.Bb3 d6 8.c3 0-0 9.Na5 10.Bc2 c5 11.d4 Qc7) -- a little over +.6 at 40 ply. But that and similar defenses held the field for decades. I was surprised by the relatively low SF score for the Sicilian in my not-very-rigorous experiment. I think the Sicilian has always been favored by confident, aggressive, and strong players (no accident that Fischer and Kasparov both relied on it). So that helps explain its score in Chessbase. |
|
| Sep-26-22 | | Granny O Doul: A "revised" edition of "Common Sense in Chess" declared the Berlin defense "insufficient" or some such. But maybe Lasker was right and Reinfeld was wrong?! |
|
Sep-26-22
| | FSR: <keypusher> I should have made clear that I was talking about ChessBase Online, which I think is ChessBase's most comprehensive database. Actually, your surmise about non-Berlin defenses bringing down the average is mistaken. The Ruy Lopez scores 57.1%, slightly better than the Scotch (56.7%) and significantly better than the Italian (54.3%). After 3.Bb5, the Berlin gives White 58.1%. Black scores better with 3...a6 (56.1%), and even better with 3...f5 (Schliemann/Jaenisch; 53.1%), and 3...g6 (just 52.2% - pretty good for a line that doesn't even have a generally accepted name). It's a mistake to try to draw broad conclusions from these statistics. The Berlin number might look bad because a lot of the Black players are lower-rated players trying to draw higher-rated players. What one really wants is players' performance ratings relative to their ratings. I think Black on average performs 35 points less than his rating just by virtue of being Black. (See my award-winning article https://en.wikipedia.org/wiki/First....) So, for example, an opening for Black that gives Black a -20 performance rating is a little better than average, and a -50 is a little worse. I don't know how to get performance rating statistics, which would be very useful. |
|
Sep-26-22
 | | moronovich: When the topplayers play the Berlin defence, you can be shure it works. |
|
Sep-26-22
| | keypusher: <FSR>
Thanks and oh well. Another theory shot down in flames. |
|
Sep-26-22
 | | al wazir: In 1937 Arthur William Dake played at the level of a GM or a strong master. He <was awarded the IM title in 1954 and the Emeritus GM title in 1986.> Averill Powers <was champion of Wisconsin in 1943, 1945, 1950 and 1954.> So which of the two players was giving the simultaneous exhibition, and which was merely one of the multiple opponents he faced? |
|
Sep-27-22
| | FSR: <al wazir> Surely it was Arthur William Dake, who as you indicated was one of the strongest players in the country at the time. To <Pawn Slayer>'s point, it's unusual but not unheard of for the simul-giver to play Black in some games. Emanuel Lasker (these are Black games only) often do so. |
|
| Jan-28-23 | | Cibator: Bent Larsen gave a simul at our club in north London in 1973, and let his 30-odd opponents have the choice of colour!! The number who did the "honourable" thing and took Black was surprisingly high. |
|
Jan-28-23
 | | perfidious: <Cibator>, a member of CG (<jerseybob>) beat Larsen in a 1970 display playing the White side of a Spanish, though I have no idea whether the Dane allowed others to make the first move. In 1972, I lost to Browne, playing White in a 6.Bg5 Najdorf. |
|
 |
|
|
< Earlier Kibitzing · PAGE 3 OF 3 ·
Later Kibitzing> |
Kibitzing
sac: 5.Bxf7+