Saturday Summary, Sans Svidler-Shankland, Starting Sunday

I guess the supposed starting date for the Svidler-Shankland match was based on the opening ceremony. They'll start playing tomorrow (Sunday).

There was plenty of action elsewhere though. First, let's get the computer action out of the way: Lc0 is still the chess entity of the future. Stockfish beat it 4.5-2.5 in one final, and in the other semi Houdini beat Komodo 4.5-3.5. Stockfish will beat Houdini tomorrow (Sunday) in the final of the TCEC Cup.

Now for the main event, round 1 of the Isle of Man International. Of the 82 games in round 1, only one saw the lower-rated player beat the favorite, and that was all the way down on board 68. But when we change the subject and talk about draws, it's a very different story. Despite gaps from nearly 300 points to slightly over 400, 20 players managed to draw their heavily favored opponents. Sticking only to 2700+ players, Anish Giri, Vladimir Kramnik, Wesley So, Hikaru Nakamura, Radoslaw Wojtaszek, Richard Rapport, Le Quang Liem, Michael Adams, and Zoltan Almasi were all held to draws. Some of them were losing or much worse (Kramnik, So, Wojtaszek, Rapport) and had to eke out the draw, while others (Kramnik again, Rapport again, Adams) were winning or at least much better before their opponents saved themselves.

Worse still for the top players, this was their supposedly easy round. In round 2 the top players are already facing mid-to-high 2500s, and it's only going to get tougher from there. This will be an entertaining event.

The Other Computer Championship

Meanwhile, Chess.com's version of the Computer Chess Championship, blitz (5'+2") edition is nearing the halfway point, with Stockfish holding a relatively slim lead over Houdini, with Lc0 in a relatively distant third. Stockfish has 55/61, Houdini 53/61, and Lc0 has 47.5/60. Komodo, surprisingly, is only in 6th, with 44/60, trailing Ethereal (47/61) and Fire (45.5/62 - Komodo could pass Fire when it catches up in the number of games played).

The Final Four, TCEC-Style

This is not the main TCEC event, but I like the format of the TCEC Cup, a 30' + 10" knockout event that's down the final four. Very unsurprisingly, the semi-finalists are Stockfish, Lc0 (Leela), Houdini, and Komodo. The first semi is underway, featuring the first pair.

All the matches in the event are best-of-eight, with pairs of tiebreak games played if necessary. Stockfish has demolished its opponents so far while Lc0 has squeaked through its matches (scroll down the page linked above to see the brackets), but sooner or later one assumes Lc0 will devour all the conventional programs. Has that time come yet? Probably not, but we'll see.

Starting Tomorrow: Isle of Man, Svidler-Shankland

I hope you're all enjoying your day off after the European Club Cup, because tomorrow not (just) one but two high-level events get underway: the Isle of Man Open and a six-game match between Peter Svidler and Sam Shankland. (The latter is part of a chess festival in Hoogeveen, in the Netherlands.)

The IoM is an open event, as stated, but at the top the field is strong enough to create a Candidates field: Anish Giri, Vladimir Kramnik, Maxime Vachier-Lagrave, Sergey Karjakin, Wesley So, Hikaru Nakamura, Levon Aronian, and Viswanathan Anand are the top eight.

Good times, fellow chess fans!

European Club Cup, Rounds 6 and 7

The 2018 European Club Cup is now history, and was won by a St. Petersburg team on tiebreaks over one ostensibly from the Czech Republic. All seven players on the St. Petersburg squad were born in that city, including not just their captain, Peter Svidler, but even their board 6, Maxim Rodshtein, who now lives in and represents Israel. The "Czech" club, by contrast, featured two Czechs, two Poles, an Austrian, and three players from India. (They don't call it for the "European Club Cup" for nothing! Or maybe they do.)

Anyway, for most of us the main interest in the event is the elite-level chess being played, so let's follow our usual procedure and recap the action of the 2800s and the hitherto struggling Peter Svidler. In round 6 Magnus Carlsen, the world champion and still world's #1 despite his best efforts to give that title to Fabiano Caruana, had the white pieces against Ding Liren. Whether for positive reasons or just to avoid showing prep for next month's match with Caruana, Carlsen played the Scotch Four Knights. Every so often someone manages to find a new wrinkle that gives the line a spark of life, but overall it has a reputation for being insipid and harmless.

It lived up to (or maybe down to) its reputation. Carlsen got nothing and then blundered with 23.g4? His 25.Re1 made things even worse, and Ding had a winning advantage he almost surely would have converted against any other player. The size of Black's advantage ebbed and flowed, but it was generally decisive for a long time. Ding probably lost the win when he played 44...Bxd3 instead of 44...Rxc2 (45.Nxf4 Rc1+ and only then 46...gxf4, and Black's d-pawn will decide). That was a huge result not just because it left Carlsen at #1, but it also prevented Ding from leapfrogging Shakhriyar Mamedyarov into third place on the rating list. And while I kid about the team competition, it was huge for that match. Carlsen's team was badly outrated by Ding's except on board 1, but they drew five games and won on board 2 to win the match.

Speaking of Mamedyarov, he faced less stellar opposition, defeating IM Tamas Petenyi (2444) with the black pieces in just 34 moves. White's problems began as early as move 8, when 8.Qe2? f4 already gave Mamedyarov a clear advantage.

How about Svidler? He had lost four games in a row entering the round, but at last he had a stroke of good luck rather than his usual kind. He managed to get an extra pawn in a rook ending, but his opponent, GM Erik Blomqvist, could have held it with best play. He defended accurately for a while when it got down to rook and two pawns vs. rook and one pawn, but he eventually committed the one error Svidler needed to squeeze out the point. White needed to play 51.Rd3+ before scurrying to g3 with the rook. Interestingly, he can (and must!) even meet 51...Ke5 with another zwischenschach, 52.Re3+, as 52...Re4 53.Kf3 (or 53.Kd3) 53...Rxe3+ 56.Kxe3 is a trivially drawn ending. Svidler played the remainder perfectly and won the game.

Just in time, too, as his last round opponent was Carlsen. The game was a draw, with Svidler enjoying some pressure most of the way but never the sort of advantage that Ding enjoyed. Svidler's team won the match comfortably, and that was good enough for team victory, as noted already. (Or at least it turned out that way. If the match on board 2 hadn't finished in a 3-3 tie, I suspect whichever team won would have had the better tiebreaks and won the event.)

As for the other 2800s: Mamedyarov drew a pretty tame game with Radoslaw Wojtaszek, while Ding looked shaky against Zahar Efimenko until the latter played 25...Ng7? (25...gxh5! gives Black a large, nearly winning advantage) 26.Rh3 Be7??, transforming a better position into a lost one. Perhaps Efimenko missed that after 27.f3 the natural retreat 27...Ng5 gets rolled by 28.Bxg5 Bxg5 29.f4 Be7 30.hxg6 fxg6 31.Bxg6. So he tried sacrificing a piece with 27...Nxh5, but it wasn't enough after 28.fxe4 dxe4 29.Bxa4, and Black resigned on move 40.

The games mentioned above, with the analytical comments incorporated therein, can be replayed here.

Aleksandr Dronov the 29th World Correspondence Chess Champion

Or if you prefer, the winner of the 29th World Correspondence Chess Championship. Either way, hats off to Aleksandr Surenovich Dronov, whose score of 9.5/16 (three wins, thirteen draws) gave him the title, putting him half a point ahead of Jacek Oskulski, Leonardo Ljubicic, and Horacio Neto. Neto had the last game to finish, against Thomas Mahling, and that game dragged on at least six months after all the others were finished. (In fact, I'd had a tab open on my browser for six months or maybe even a year, waiting for it to end, and finally gave up on it in the last month or two. On a whim I checked tonight, and voila! - it finished September 25.) Neto was pressing, but after 80 moves had to acknowledge the draw.

I don't plan on following the 30th Championship as closely, but if you want to the crosstable is here. 103 games are finished, 33 remain, and there have been a less-than-whopping four decisive games so far. (There were 13 out of 136 in the 29th Championship, in case you were curious.) While it is easy to make fun of or be turned off by the overwhelming percentage of draws in correspondence chess, the games are generally very interesting and feature lively openings, so it is worth keeping an eye open for correspondence games. Do note that you can find the PGNs of finished games just below the crosstable on both pages linked above.

A Brief Recap of the Top Players in Round 5 at the European Club Cup

Peter Svidler is having the time of his life - the bad time of his life. He lost his fourth straight game, three of which have been with White, and once again it was with a blunder. He was winning against Christian Bauer, three pawns up. Bauer did have some pressure against Svidler's kingside; enough to give him practical chances though not enough to make up for the missing material. Svidler's 40th move was fine, though it encouraged further trouble. A move like 40.h4, returning one of the pawns to close up the avenues to his king would have been a good idea. A move later, however, his 41.Bc3 didn't just make things more complicated; it lost. (41.Qb1 maintained a winning advantage.) Or rather, it gave him a losing position after Bauer's 41...Nd3! There was still some fight in the position in case of 42.Qc4, but Svidler's 42.Kh1 led to a forced mate, and he resigned after 42...Nf3 43.h4 Nxh4. The poor guy must feel bad for his team, and the loss of 25 rating points must hurt as well.

As for the 2800 club, two of them faced off. Shakhriyar Mamedyarov had White against Magnus Carlsen. The game was equal throughout and finished in a quick draw. Mamedyarov's position looked more pleasant at the end, but there was no way to make progress. Ding Liren also drew quickly - in fact, both his game and Mamedyarov-Carlsen were drawn in 30 moves. He had Black against his countryman Wang Hao, another rented European for the week. (At least China is part of the same land mass containing Europe. Another team is borrowing a couple of low-rated players from the U.S. and Canada.)

As for the other super-GMs, none but Svidler were harmed in the making of today's film, so I'll leave you with a link to the three aforementioned games.

Note: An Important Update to a Recent Post

In this recent post I reviewed a couple of endgame study books by Sergei Tkachenko. In a not particularly clever bit of non-editing, I neglected to include one of the book titles. That has been fixed, with apologies to my readers, the author, and the publisher.

The Big Guns at the European Club Cup: Successes and Failures

The European Club Cup has been a mixed bag at best for the top players thus far, four rounds into the seven-round tournament. Let's start at the top, with Magnus Carlsen. After taking the first round off, he won in trademark style against Vladimir Potkin, creating enough little problems to give his strong opponent the chance to hang himself, which he duly did. In rounds 3 and 4 he drew. That wasn't a bad result in round 3, as Black against Radoslaw Wojtaszek, but the round 4 draw with White against Alexander Donchenko (2610) was another story. Carlsen played some very weird-looking chess - maybe deliberately, to have some fun - and after achieving nothing with it for a while suddenly got an opportunity after Donchenko's 26...h5(?). Had Carlsen played 28.Re3 or 29.Ree4 he would have had good winning chances; missing (or rejecting) both chances, the game finished in a perpetual.

The second highest-rated player, Shakhriyar Mamedyarov, has only played two games: a win in round 1 against an IM, and draw with Black in round 4 against Zahar Efimenko. If anyone missed any chances in the latter game, it was Efimenko, who appeared too happy to make a draw with his elite opponent.

The third highest-rated player is on the verge of becoming #2 in the tournament, and #3 in the world rankings. Ding Liren has gone 2.5/3, grinding out a win with Black against an IM in round 2, drawing with Jan Werle in round 2 (Ding went a bit too all-out for the attack - it was better to maintain the center rather than throwing all the eggs into the kingside basket with 14.e5 - but fortunately for him Werle was happy to make a draw and didn't press for more), and then he was just the latest guy to capitalize on Peter Svidler's blunderful form.

Svidler is having a nightmarish event, losing all three games he has played (and 17.8 rating points) on major errors (as opposed to being outplayed a bit at a time). He was winning a complicated fight against Aleksandar Indjic in round 2 until he played 38.Qxa6, allowing Black to save the game, and then got a second chance to regain a decisive advantage on move 40 with c4. Instead, he had to find the right move to save the game on move 42, but didn't manage. Only 42. g3! would save the game, ensuring that the Black king could not escape from checks by navigating his way through White's kingside pawns. The key variation is 42.g3 Nb2 43.Nb6! Nd3! (everything else loses for Black) 44.Qd6! R7xb6 (the only winning move in the 42.g4 version of the line) 45.Qe7+ and White has a perpetual. With the pawn on g4 instead of g3, Black's king can escape to f4 or h4.

In round 3 the position was complicated but level against Romain Edouard until Svidler played 32.Qf3?? instead of 32.Kg1, hanging the h-pawn. Perhaps Svidler thought he'd be okay after 32...Qxh2+ 33.Kf1, but the nice 33...f4! spoiled the illusion. Granted, 32.Kg1 Re2 33.h4 seems scary, but Black has no way to break through to White's king.

Finally, Svidler was suffering almost the entire game against Ding Liren after surrendering the center early on, but thanks to a combination of resilient defense on his part and looseness on Ding's side of the board Svidler had a chance to make a fight of the ending with 38...Kf7. Instead, he played 38...Nxa2??, and after 39.Bd2 found his knight dominated. This couldn't have come as a surprise to Svidler, as it's a basic pattern, so he either banked on 39...a5 or 39...b5 (the move chosen), only to realize that neither move worked. The problem with the former is that while the knight gets a momentary reprieve after 40.Bxa5 Nc1, it's "recaught" after 41.Bb4!, e.g. 41...Kf7 42.Rd1 Nxb3 43.Rd3 Nc1 44.Rf3+ Ke6 45.Bd2 and the knight will be collected in at most three moves. He tried 39...b5 instead, but 40.b4 (preventing ...b4 followed by ...Nc3) caged the knight, and Svidler resigned after 40...Kf7 41.Rd3, not needing to see Ra3xa2.

Other members of the 2700 club have given up plenty of draws, but I think only three others have lost to non-2700s: Pentala Harikrishna lost to David Howell (who came into the event 2689 and is now over 2700 on the live list, so that's not much of an upset), David Navara lost to Nils Grandelius (2655), and in what I think was the most notable upset so far, Wang Hao (2722) lost to Nemeth (2484).

All these games can be replayed here (without notes, sorry).

Notre Dame 19, Pitt 14

That was horrible, but despite some painfully poor play against a much weaker opponent, they fought and scrapped and scraped out a narrow victory. Still, winning ugly is still winning, and with one of the teams ahead of them getting crushed today there's a possibility they'll break into the top 4 for next week.

Record to date: 7-0.

Next victim: Navy (in two weeks)

The music of the day might have been a dirge, but with a sigh of relief we present the usual tune: