Magnus Carlsen and Jan-Krzysztof Duda have the only perfect scores after two rounds of the Meltwater Champions Chess Tour Finals after overcoming Arjun Erigaisi and Shakhriyar Mamedyarov respectively. Anish Giri has also won both matches, in tiebreaks, after staging an incredible recovery from 2:0 down to beat Praggnanandhaa.

Giri and Praggnanandhaa faced off with a view across the Bay of San Francisco

Once again it was only Anish Giri’s match that went to tiebreaks on Day 2 of the Meltwater Champions Chess Tour Finals.

Watch the Meltwater Champions Chess Tour Finals live here on chess24!

Two matches were over in just three games, with the surprise of the day Wesley So getting swept aside 2.5:0.5 by his old colleague at Webster University in St. Louis, Liem Le. Wesley was making no excuses afterwards.

He’s just very strong and plays some very good moves… I probably should have won Game 1, but other than that I think Liem played a near perfect match.

That chance in Game 1 came when Wesley went for the piece sacrifice 25…Bxb2?!

Although 26.Qxb2 c4 won back the piece, Liem was able to make a draw from a position of strength.

Instead it seems Wesley should have gone for e.g. 25…Qa6 and combined pressure down the a-file with pushing his c-pawn, when he'd have a winning advantage.

From there on Liem Le was in top form, dominating Game 2 before 31…Qd7! clinched victory.

Qh3, threatening Qxg2# and Ng3+, is an unstoppable threat.

Things didn’t get better for Wesley in Game 3, where he soon ended up in a technically lost position. Liem at times let the win slip, but Wesley returned the favour, and by the end was clearly upset as he decided to make Liem demonstrate that he knew how to checkmate with a queen against a bare king!

The other quick victory was for Magnus Carlsen, who followed his win over So by taking down one of his big rivals from the new generation of chess talent — Arjun Erigaisi. Once again Magnus did enough, though he also made some glaring blunders. Tania Sachdev asked him if he was happy with his play.

No, but I feel like I generally do okay. My instincts are good, but I’m just missing stuff. I’ll try to do better, but obviously the score is good.

Magnus went for the Modern with an early g6 and d6 in Game 1, later noting that openings are currently a weakness of Arjun’s.

I think his openings aren’t good enough, so I think a lot of people like Pragg, for instance, he often puts a lot of pressure in the openings, and Arjun doesn’t seem to catch me there, so then it’s not so easy.

Magnus gradually equalised and then took over until he got a rook endgame where it felt like his winning was just a matter of time. There were some big twists, however, and none more dramatic than when Carlsen played 54…Kb6?

When Arjun replied 55.Ra6+ Magnus briefly panicked, confessing afterwards:

Kb6 was not perfect technique, let’s say that! For one second I thought I was lost, then I thought, ok, I can at least play Ka7 and make a draw, and then I saw wow, actually I’m very lucky to still have chances to win.

The move Magnus played did in fact let the edge slip, but only because of an incredible stalemate with 55.Ra1!! Kb5 56.Ra5+ Kxb4 57.Rxb5+! Kxb5 — the kind of thing you only find because the computer evaluation suddenly alters. Peter and Rustam couldn’t believe it!

The rook endgame became a queen endgame, before Magnus finally won in 74 moves.

Magnus should have all but wrapped up victory by winning Game 2, when Arjun got into deep trouble in the opening.

15.Ba6! here would have been the just punishment for Arjun delaying castling too long, though the slower 15.Rc2!? by Magnus kept a large part of the advantage.

After many more adventures the World Champion got another winning rook endgame, but he described what followed as “embarrassing technique”. He mistimed a b5-break near the end and had nothing better than stalemate in 86 moves.

It was still Arjun who needed a win, however, and despite a good start to the third game Magnus was able to fight back, with 27…c5! a key move.

Arjun went for 28.exf6!? but after 28…Bxf3 the bishop on f3 was a thorn in White’s side, while the c-pawn would prove a monster. Magnus summed up:

After I blundered, he played very well for a while, so he was much better, but it wasn’t very easy to win, and then when I got this move c5 I thought I was sort of ok. He probably should take the bishop and then the pawn, but then I have very good drawing chances. In the game I think he has nothing, but clearly this game was a product of the score in the match, not much else.

Magnus moved to 6/6, earning a combined $15,000 for winning his first two matches in rapid chess.

The one player to match Magnus was Jan-Krzysztof Duda, who didn’t allow Shakhriyar Mamedyarov any revenge for the Aimchess Rapid final. Duda held comfortably with Black in the first game and then resisted the temptation to simply take a draw when down to around 15 seconds in a complicated position.

Qf6+ and Qg5+ and the game would end peacefully, but Duda realised he could ask questions at no risk with 36.exf5!, and a few moves later Shakhriyar finally cracked with just a second left on his clock.

It was still very tense afterwards, with Duda admitting he was close to going for a “winning combination” that would have lost him the game on the spot!

Instead he successfully steered the game to victory.

That proved enough to clinch victory in the match, after the remaining two games were drawn. If things keep going this way the Carlsen-Duda clash in Sunday’s final round might be a title decider.

For a second day in a row the only match to go to tiebreaks involved Anish Giri, but this time it was a complete reversal of the day before. While Giri had made four solid draws in rapid chess against Liem Le, he came out all guns blazing against Praggnanandhaa, starting by playing the hyper-aggressive 4.f3 Nimzo that his wife Sopiko Guramishvili made a video series about for chess24.

He went for a pet line of the equally aggressive Salem Saleh, but it would end in disaster when he opted for 22.Nxg7? in a situation when that knight sacrifice simply didn’t work.

Praggnanaandhaa comfortably proved that, and then went on to take a 2:0 lead with some virtuoso play in a knight and rook endgame. Anish described Pragg putting his knight on h8 in what seemed a drawish position as “dirty”, and ultimately Anish lost his way in the complications.

Anish said here he was choosing between 65…Ne6+! and 65…Nh3+? and picked the latter, overlooking the “very resourceful” 66.Kh6! That wasn’t the last twist, but Pragg went on to win.

Anish explained that his opening strategy had backfired.

I knew, to be honest, up front that today I’m going to play risky stuff and it could end badly, but I thought the circumstances were good. I played well yesterday, he played badly yesterday, so I thought if I’m ever going to go and try do something different it’s going to be today.

Anish now had to win two games on demand just to reach tiebreaks, but he got off to a great start as this time the 4.f3 Nimzo brought success, with the double attack 25.Bd2! the moment at which the outcome was all but sealed.

Praggnanandhaa still only needed a draw in Game 4, however, and everything seemed to have gone perfectly for him when he got the following endgame with the white pieces.

Pragg would later lament:

I think everything went wrong. The 4th game was crucial, and I thought I had a risk-free position. I really thought what could possibly go wrong there? I mean, I had this outside passed pawn...

Pragg added, “there are probably 100 ways to make a draw”, but gradually he made concessions, most notably when he played 33.g4?!, and Anish was able to take over. There were later tricky ways for White to escape, but in the end a totally shocked Praggnanandhaa was facing the prospect of tiebreaks.

Anish put the turnaround down to some chess psychology.

People play quite poorly in must-not-lose situations, and my last game I played very poorly. It’s really hard to get that out of your system. Even Carlsen is sometimes falling at this hurdle. It’s just really difficult to play with that mindset.

The first blitz game saw Anish once again play as aggressively as possible despite having the black pieces. The attack was crowned with success, but there was at least one bump along the way.

We reached this position after Giri went for an unnecessary piece sacrifice in a great position. It turns out that 17.Qe3! would have left Pragg on top, but after 17.Qe2 Anish had a winning attack.

He found it, unleashing 17…a5!, with the a8-rook later getting to a2 to hit the b2-pawn. It looked as though White would get mated, but although that didn’t happen Anish traded down into a completely winning position.

For the final “must-not-lose” game, Anish switched away from the 4.f3 Nimzo, but he almost got into trouble.

14…Qxe2! here by Pragg, and, at least temporarily two pawns up, he’d have had every chance to force Armageddon. Instead he believed his opponent with 14…Qc5?! and Anish held on to pull off a remarkable victory.

A match that goes to tiebreaks is split 2:1 between the winners and losers, so that Anish is on 4 points for his two tiebreak victories. That’s enough to put him in 3rd place, while Magnus Carlsen and Jan-Krzysztof Duda lead with a perfect 6/6 for two wins in rapid chess.

In Wednesday’s Round 3 Giri faces Duda, while it’s Carlsen-Mamedyarov in the other match involving a leader. So-Arjun is a bottom-of-the-table clash few would have predicted before the event began, while Praggnanandhaa faces a tough battle against Liem Le to get back on track.

The games begin at 12 noon in San Francisco (15:00 ET, 21:00 CET, 01:30 IST).

Watch the Meltwater Champions Chess Tour Finals live here on chess24!

See also: 

• Meltwater Champions Chess Tour homepage
• Meltwater Champions Chess Tour Finals games
• Meltwater Champions Chess Tour Finals line-up announced
• Magnus Carlsen vs. Wesley So in San Francisco
• MCCT Finals 1: Carlsen beats So in thriller