Internet Chess club on Facebook Internet Chess club on Twitter Internet Chess club on Google+ Internet Chess club on YouTube Internet Chess club on LinkedIn Subscribe to Internet Chess Club RSS feed

GM Joel’s World Championship recap - Games 1-3

GM Joel Benjamin

Welcome to GMJoel’s World Championship recaps.  I’ll be presenting these reports on the off days, summarizing the developments of the last two or three games.  I’ll look for key moments in the games and trends I think the games are showing us.

No one can be sure what to expect from the outset at a World Championship match.  The fans all hoped, at the very least, that we would not see the type of draw-fest that has been common in recent years.  While the most cynical will say of the first three games, “same old same old,” the draws were not all the usual feeling out the opponent type of affairs, as both players showed (to varying degrees) a willingness to accept risks.

Game One


We saw that right off the bat with Carlsen producing a marvelous opening surprise with the black pieces.  Nepomniachtchi passed on the often-sharp Scotch Opening, which he has frequently essayed in the past, and the nearly ubiquitous Giuocco Piano in favor of the Ruy Lopez.  It seemed reasonable to investigate what variation Magnus would look to defend.  Nepo steered clear of the Marshall Attack, as most grandmasters do these days, not because Black’s attack is so fearsome, but because so many variations lead to drawish positions.  Magnus unveiled his take on the anti-Marshall.

With 8…Na5!? Magnus continued in the 21st-century approach of not caring about the e5-pawn in the Ruy Lopez.  I was a bit surprised I didn’t find any memorable games with this move, though a little research shows that the position reached in the game has been more commonly reached by (from the diagram) 8…Bb7 9.d3 Na5!? (9…d5 being the usual method of sacrificing the e5-pawn).

Carlsen’s pawn sacrifice has appeared in a few grandmaster games, but with a stunning lack of success—White has won almost all of the games.  Coupled with the engine favoring White, we can understand why Carlsen’s gamble hasn’t become mainstream.  We have all heard that Magnus has mined the magical powers of AlphaZero, so we should not be surprised to see him take a deeper look at the position.

We have also heard that Nepo has access to the most powerful supercomputers in Russia, and his 14.Kf1 is a quirky computer-approved move, only seen in a couple of correspondence games.  Carlsen answered with the not-very-obvious 14…Rfb8.  The explanation is that the more natural 14…Rfe8 is strongly met by 15.Nc3, a move not possible one move earlier because of Qxe5! winning material.  In the game, 15.Nc3? Qxe5 again wins for Black, so Nepo was compelled to take the queen and bring Carlsen’s knight into a better position.

I awoke Friday morning to a position that looked extremely ugly for White, and I only understood it after noticing Nepo had one pawn extra.  Still, Black’s position seemed quite a bit easier to play and the momentum went Carlsen’s way for several moves.  22.Bf4?! seemed plausible enough but gave Carlsen a little something to work with.  Nepo dug in by sacrificing his extra pawn to avoid invasions, transferring his king to the queenside, and keeping points guarded in that sector.  Still, it felt like the kind of position Magnus usually wins.  His play was a bit too complacent, and he missed the opportunity to keep the pressure on.

Here Carlsen could have tried 38…Rcb6, with the idea that 39.Ra2 is powerfully met by 39…a5! With the idea of using the a-pawn as a decoy to penetrate with the rooks.  Once White was able to activate his knight Black’s initiative was gone, and Carlsen acquiesced to a repetition.

So, whose moral victory is it anyway?  Nepo’s first white blew up in his face, yet Magnus must be terribly disappointed to waste the surprise value of this gambit, failing to sustain pressure in his usual manner.

Game Two


Nepo was the first one to raise the stakes, but Carlsen seemed only too happy to see the game take an early sharp turn.

Nepo’s 7…b5 is fairly well-traveled, but not nearly as much as the standard 7…a6.  There 8.Qxc4 has fallen out of favor, with 8.a4 Bd7 9.Qxc4 Bc6 gracing a lot of chessboards.  Black is extremely solid, while White hopes for his space advantage to help his chances.

White has tried 8.a4 in almost all the games, with the exchange sacrifice 8…b4 (8…Bb7 9.axb5 a6 is a recent development) 9.Ne5 Qxd4!? getting a workout.  Magnus opted for the immediate 8.Ne5, when Black can still sacrifice the exchange, but with subtle differences. 8…c6 is an almost unknown response, though 8…Nd5 9.a4 could well transpose into the game continuation.

By 11…Qd7, we were surprisingly in a completely new position.  Magnus had promising attacking chances but spoiled his position with a major oversight. 

White could continue with something straightforward like 17.Nxf6+ gxf6 (17…Rxf6 18.Ne5 would execute Magnus’ idea successfully) 18.Bh6 Rf7 19.Ne1.  Black is probably still okay, but I think the onus is on him to play accurately, as White has only invested a pawn.  But after 17.Ne5? Bxe5 18.dxe5 Nc5! the knight came flying in and compelled Carlsen to sacrifice more material than he wanted to.

The engines told us that Magnus was in trouble (Stockfish more so than my other engines), but like the first game, suffering did not materialize.  The extra exchange may have conferred an advantage to Black, but it didn’t make Nepo any more comfortable.

Black can play for a win with some cold-blooded play.  24…g6 is one option, as both discovery captures backfire—25.Nxb5 and 25.Nxc4 both allow 24…Qg7! and Black breathes a lot easier after snagging the e5-pawn.  24…Qe7 25.Qh5 g5 apparently does not lead to mate, though I admit I wouldn’t be sure over the board.

24..c3 was kind of a bail-out option.  Carlsen was distracted from an attack but steadily scooped up pawns  and the danger of losing passed.  Nepo was content to sack back the exchange to reach a dead drawn endgame, despite a pawn minus.

Boys and girls, this is why you study endgame fundamentals.  I once reached R &2 v. R&3 against Kasparov, and he didn’t even play it out, assuming no self-respecting grandmaster would fail to draw it.  Carlsen had no illusions of winning here, playing it out to make the fans happy.

A second draw, but one with plenty of action for the fans.  Carlsen showed himself to be human with a tactical oversight, but Nepo did not come close to making him pay for it.

Game three


Judit Polgar suggested that after such a tense start it made sense for game three to be a quiet affair.  We saw that typical WC match game we have gotten used to.  Nepo switched to a different anti-Marshall, replacing 8.h3 with 8.a4.  While White has trouble squeezing out winning chances, he reaches a fairly safe position, which might have suited Nepo after two rough rides.

I have noticed in the past that Carlsen would often choose a move a few rows down on the database.  Historically, 10…Na5 has been played most often, leading into a setup similar to the Chigorin Variation.  In 2021 10…Nd7 has gotten a workout from top grandmasters.  But Magnus opted for 10…Re8…certainly a good move, but one that his opponent had perhaps not given quite as much thought to.

Nepo seemed to do everything right but found his position swiftly neutralized.

Black has all but one piece on the back rank, though that changed after 20…Be6.  White has more space of course, but the Black pieces bear down nicely on the central squares.  What to do?  21.Nxa5?! c5 backfires on White.  21.Qd3 looks to suppress Black’s central intentions, though even here 21…d5!? 22.Bxf6 Qxf6 23.exd5 Bf5 looks like a position Magnus would almost never lose.

21.h3 looked a bit slow, and it was.  Carlsen broke in the center and completely equalized.  The ending would only be dangerous to him if he traded all the rooks AND the bishops.

The dreaded draw death has started, though the first two games at least looked like they might produce a winner.  Carlsen looks fearless in the early going.  Nepomniachtchi has a reputation for playing that way, too, but I don’t think we’ve seen it thus far.