The Nepo Shuffle -- The 2021 World Chess Championship

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
I’m nowhere near advanced enough to analyze these games as they are happening and when they are heavily focused on knight moves like this one was it is way too complex for me. I have to rely on the commentary. (I switched back and forth from Anish and Judit Polgar to Hikaru. I can only take Hikaru in small doses but he had Wesley So on his stream so that was nice.)

So most of my observations are big picture. What strikes me the most is how razor thin the margins are. Carlsen gave up the pawn for development — actually to slow Nepo’s development as much as to get his own bishops on those scary diagonals.

Nepo’s bishop on C1 kept him from being able to link his rooks early and Carlsen was able to push his pawns on the queen side to get some initiative. It felt as though Nepo was just a tempo behind and if he could have found a tempo with one of those knight moves it would have changed everything. But that seemed to be the difference between white having to defend as opposed to attacking through the end game — one early game tempo.

Maybe that is always the way with these Spanish games. I don’t know opening theory well enough to know.
 

SumnerH

Malt Liquor Picker
Dope
Jul 18, 2005
28,708
Alexandria, VA
Naroditsky was saying on his stream (while playing random games and such) that Magnus seems particularly sharp and on form at the moment in the speed matches they've played.
 

Bread of Yaz

lurker
Mar 12, 2019
120
Carlsen has historically had the reputation of being super solid, grinding small advantages into larger ones. In the first two games (and especially today) he's been playing much more sharply. He gave up a pawn to get his night to d6 then sacked the exchange to keep up the pressure - Nepo has displayed good nerves and held firm. Today bodes well for the excitement level in the rest of the match.
 

SumnerH

Malt Liquor Picker
Dope
Jul 18, 2005
28,708
Alexandria, VA
Carlsen has historically had the reputation of being super solid, grinding small advantages into larger ones. In the first two games (and especially today) he's been playing much more sharply. He gave up a pawn to get his night to d6 then sacked the exchange to keep up the pressure - Nepo has displayed good nerves and held firm. Today bodes well for the excitement level in the rest of the match.
And a game 3 that's a bit of a yawner.

But according to lichess, apparently, game 3 was the most accurate game in world championship history.
 

Bread of Yaz

lurker
Mar 12, 2019
120
And a game 3 that's a bit of a yawner.

But according to lichess, apparently, game 3 was the most accurate game in world championship history.
Nepo got nothing out of the opening, and Magnus again deviated from well-trod lines and equalized easily. Nepo has yet to pressure Carlsen in any serious way, but OTOH has looked relaxed and comfortable during the games and has used his time well. He doesn't seem fazed by the magnitude of the event at all.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
Carlsen looks like he is really comfortable in the Ruy Lopez. If Nepo keeps opening with e4, I don’t think Carlsen is going to respond with the Sicilian for so long as he is ahead or tied in the match.

If it is 2-2 in game 5 and Nepo opens with the king’s pawn again I bet we see something different on move 3. Maybe d4.
 

Jnai

is not worried about sex with goats
SoSH Member
Sep 15, 2007
15,318
<null>
Carlsen looks like he is really comfortable in the Ruy Lopez. If Nepo keeps opening with e4, I don’t think Carlsen is going to respond with the Sicilian for so long as he is ahead or tied in the match.

If it is 2-2 in game 5 and Nepo opens with the king’s pawn again I bet we see something different on move 3. Maybe d4.
2018 saw the Rossolimo like 3x in a row when Caruana had white I think. The danger of shifting into something that's even slightly less well prepared seems very high.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
2018 saw the Rossolimo like 3x in a row when Caruana had white I think. The danger of shifting into something that's even slightly less well prepared seems very high.
Yeah. I guess what we don't know is where Nepo has done the deepest preparation. When he has the white pieces it's not his choice.

I don't really have access to databases or know how to use them but you would imagine that Nepo expected a very high probability that Carlsen would play the Sicilian, because, well, it's the championship and it's the Sicilian. Maybe Nepo only had backup preparation for the Spanish, and even if he thought Carlsen might have played it I bet he was probably more expecting a Berlin.

Game 5 will be interesting. Nepo is probably going to be much better prepared for the possibility of playing the same lines, so Carlsen may pull a switcheroo.

One thing that has been interesting is that Carlsen has seemed pretty comfortable in all three games potentially sacrificing a pawn early for positional advantage. I don't think it has always happened, but I think that's a major purpose of the Catalan and he hasn't been shy about giving a pawn to keep Nepo undeveloped with the white pieces.
 

Bread of Yaz

lurker
Mar 12, 2019
120
Nice performance from Nepo today. Carlsen blitzed out the first 20 moves or so in a Petroff without blinking an eye in a deep prep line. Nepo kept calmly finding accurate moves in the kind of position that Carlsen often prevails in.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
I'm looking forward to the analysis on it, but it looked to me as though it was a line that you had to be super prepared for to defend. Carlsen had a mating net set up with a rook and knight, but Nepo was able to counter it by pushing his a pawn and making it a risk to promote, which occupied Carlsen's second rook. Carlsen spent a lot of time looking for a possible mate before settling for the draw but when you looked at the most likely lines, it seems to me that Nepo was always going to be just in time with the promotion with his a pawn to avoid getting mated. The key was that when the pawn promoted, it would have a clear line to Carlsen's knight to just be in the nick of time. I don't think that's something you see over the board. That's the result of very deep preparation.

From the article on chess.com it sounds as though Nepo said after the game that he was prepared for every line and remembered his moves, right until the end. Carlsen played a couple of novelties, including one that surprised Caruana during the commentary so you know that Nepo is very very prepared.

If you're rooting for Nepo then this is definitely encouraging.

I don't really like the Petroff. I expect though that Nepo has it so well covered that we may see it again if Carlsen starts with the kings pawn.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
I've dug more deeply into the game today.

If you're not a fan of classical time format head to head chess there's really nothing about today that would change your mind.

It appears that Nepo did not really play a single minute of chess today. Once e4 was on the board, he played the Petroff for a draw and then the entire rest of the game was him playing a game of memory recall to make the moves that his preparation with the engines told him to make.

It was very impressive that he was prepared for some novelties by Carlsen but this is basically what this event is going to be from now on.

There is some super exciting chess going on among super GMs. There have been many exciting and decisive games in the qualification tournaments for the candidates and elsewhere. And that's all very positive. But the glamor of this particular event may have seen better days with what it has become and likely will continue to be.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
A couple hours in and Carlsen still playing the Ian-bot.

Hopefully, we actually get some chess soon today. I am starting to feel foolish for waking up early to watch this stuff each day.
 

Bread of Yaz

lurker
Mar 12, 2019
120
A couple hours in and Carlsen still playing the Ian-bot.

Hopefully, we actually get some chess soon today. I am starting to feel foolish for waking up early to watch this stuff each day.
Chess has been around a long time. Its not surprising that the world's greatest players avoid dubious lines and can get 15-20 moves down the best lines once centuries of play have shown that many moves lead to clearly worse positions. The ability to put the right pieces on the right squares at precisely the right times is phenomenal.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
Chess has been around a long time. Its not surprising that the world's greatest players avoid dubious lines and can get 15-20 moves down the best lines once centuries of play have shown that many moves lead to clearly worse positions. The ability to put the right pieces on the right squares at precisely the right times is phenomenal.
Well, I think the last 10 years has been more significant in terms of learning how not to put a piece on the wrong square than the previous 2,000 years by a wide margins. I think overall that computers will add to enjoyment of the game as new lines and techniques are discovered, but this event in particular suffers for it. I don’t mean to diminish at all what Nepo is doing. It is an incredible display. But as soon as he got out of preparation the queens came off and here we are again. Maybe we will get an exciting end game.

I’m sorry if this is coming off as dumping on an event that others are enjoying and I understand that these head to head tournaments almost need to viewed as one game of 14 rounds. But I was looking forward to this and am disappointed so far.
 

Bread of Yaz

lurker
Mar 12, 2019
120
Well, I think the last 10 years has been more significant in terms of learning how not to put a piece on the wrong square than the previous 2,000 years by a wide margins. I think overall that computers will add to enjoyment of the game as new lines and techniques are discovered, but this event in particular suffers for it. I don’t mean to diminish at all what Nepo is doing. It is an incredible display. But as soon as he got out of preparation the queens came off and here we are again. Maybe we will get an exciting end game.

I’m sorry if this is coming off as dumping on an event that others are enjoying and I understand that these head to head tournaments almost need to viewed as one game of 14 rounds. But I was looking forward to this and am disappointed so far.
I get it. There have not only been no decisive games, but no games with tactical sparring, castling and attacks on opposite sides etc. Its been slow strategic maneuvering. I do view this as a small victory for Nepo, who is known more as a tactical, thrown-all-caution-to-the-wind player. Maybe he wanted to get his nerves under control before launching into a Sicilian or a Benoni. Fingers crossed we get something like that soon.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
I get it. There have not only been no decisive games, but no games with tactical sparring, castling and attacks on opposite sides etc. Its been slow strategic maneuvering. I do view this as a small victory for Nepo, who is known more as a tactical, thrown-all-caution-to-the-wind player. Maybe he wanted to get his nerves under control before launching into a Sicilian or a Benoni. Fingers crossed we get something like that soon.
Good points. There is lots left in this tournament. It will be interesting what happens if we get closer to the end with draws and if Calsen can hold off Nepo with the black pieces. Nepo getting to start each round with white starts off as an advantage but may turn against him if it gets closer to the end. Sort of like penalty kicks.

Yesterday's game was the one that bummed me out. When you dig deeply into it you see that one of the main reasons that it was comfortable for black was that he knew he could win the pawn race on A and make it just in time to put white's knight in jeopardy and avoid a rook/knight mate.

That's just not something you see over the board. Maybe that's not something that having 10 Russian GMs as your seconds helps you see even if they're running Petroff lines for 24 hours a day in preparation. That's something that a computer sees and Nepo basically had it memorized and started pushing the A pawn confidently with like a minute between moves.

I'm definitely not saying that Carlsen would have won, but in prior years that game would have been a lot more fun if Nepo had played maybe a more human move after Carlsen's novelty. Maybe I'm not giving anyone enough credit and that was a fairly easy to spot line but I think maybe not. I also understand that I'm a bit like an old man yelling at clouds here or talking about when movies used to cost a nickel.
 

blueguitar322

Member
SoSH Member
Sep 20, 2005
995
The challenge is that these championship games are so easy to draw with the obscene preparation these guys come in with, that losing a single game can change things massively. That's why there's way more Berlins than Najdorfs and more QGDs than King's Indians. The risk is just too great. Lose one game, and your opponent has a decent chance of drawing every other game. Both sides essentially need to agree to go towards an unbalanced-but-even game. If you want unbalanced, but they want a draw, usually they can force you into making concessions in order to get that imbalance.

So far in 2021: 5 games, 5 draws.
In the 2018 championship: 12 games, 12 draws.
In the 2016 championship: 12 games, 10 draws.
In the 2014 championship: 11 games, 7 draws.
In the 2013 championship: 10 games, 7 draws.
In the 2012 championship: 12 games, 10 draws.

That's 62 games and 51 draws, or about an 84% draw rate. It's 92% if you only look at the last two championships + the first five games of this one.

This isn't purely a 21st century problem, either. Under a different format, with draws not counting, the 1984 Karpov-Kasparov match went 48 games with 40 draws (including 27 consecutive) over 5 months before being controversially terminated. (I am cherrypicking a little bit; their later matches still featured a high percentage of draws, but not this bad).

All this to say: the main strategy in these matches is mostly to out-prepare the opponent by finding positions that the computer thinks are even (thus perhaps outside the scope of your opponent's prep) but where your familiarity yields a decent chance of applying pressure. That's very different than in rapid where intuition and rapid calculation are more important than prep. It's also very different than classical tournaments, where broad prep is more important than deep prep, and the mismatches between deep and broad prep lead to more decisive games.

This is why Nepo was so discouraged after today's game. He managed to get a position that was objectively drawn, but he had a slight positional edge and greater familiarity. But then he couldn't do anything with it.
 

YankeesIsrael

Member
SoSH Member
Sep 21, 2009
108
The challenge is that these championship games are so easy to draw with the obscene preparation these guys come in with, that losing a single game can change things massively. That's why there's way more Berlins than Najdorfs and more QGDs than King's Indians. The risk is just too great. Lose one game, and your opponent has a decent chance of drawing every other game. Both sides essentially need to agree to go towards an unbalanced-but-even game. If you want unbalanced, but they want a draw, usually they can force you into making concessions in order to get that imbalance.

So far in 2021: 5 games, 5 draws.
In the 2018 championship: 12 games, 12 draws.
In the 2016 championship: 12 games, 10 draws.
In the 2014 championship: 11 games, 7 draws.
In the 2013 championship: 10 games, 7 draws.
In the 2012 championship: 12 games, 10 draws.

That's 62 games and 51 draws, or about an 84% draw rate. It's 92% if you only look at the last two championships + the first five games of this one.

This isn't purely a 21st century problem, either. Under a different format, with draws not counting, the 1984 Karpov-Kasparov match went 48 games with 40 draws (including 27 consecutive) over 5 months before being controversially terminated. (I am cherrypicking a little bit; their later matches still featured a high percentage of draws, but not this bad).

All this to say: the main strategy in these matches is mostly to out-prepare the opponent by finding positions that the computer thinks are even (thus perhaps outside the scope of your opponent's prep) but where your familiarity yields a decent chance of applying pressure. That's very different than in rapid where intuition and rapid calculation are more important than prep. It's also very different than classical tournaments, where broad prep is more important than deep prep, and the mismatches between deep and broad prep lead to more decisive games.

This is why Nepo was so discouraged after today's game. He managed to get a position that was objectively drawn, but he had a slight positional edge and greater familiarity. But then he couldn't do anything with it.
The 1984/5 match seems like the exception, not the rule, as it was played under the 6 win rule, which was abolished after it (and to think Fischer got a majority in FIDE congress for 10 wins in 1975). I don’t think their subsequent matches had a high draw rate (2/3) even by 20th century match standards, especially when you consider that these were two of the toughest players in history to beat. Even with the wild Tal in 1960 and the inferior Short in 1993 over 60% of the games were drawn.

In general, I see several reasons for the huge amount of draws, as mentioned above:
  1. Better opening knowledge/preparation. You have almost no chance to win a game nowadays in preparation like Kasparov did in game 10 vs. Anand.
  2. Having a solid world champion who seems happy to go to rapid games (see game 12 vs. Caruana).
  3. The short match format, which discourages players from taking risks.
The only way I see to change this is to have at least one player be fully motivated to avoid the rapid games. One option is to return the rule that the champion retains his title in case of a draw, but that seems unfair, and the rapid games are interesting and attract fans. What I would do is have the rapid games played before the match (same in soccer and penalty kicks, BTW). This way at any given moment one of the players is behind and must play for a win. In 2010 there were 5 decisive games out of 12, probably because Topalov didn’t want to go to rapid games and took risks, especially in the last game.
 

coremiller

Member
SoSH Member
Jul 14, 2005
5,330
There's also a reinforcing cycle where, as decisive results become rarer, the risk of losing a game increases because it's so hard to subsequently win a game and catch up, so the players play more conservatively, leading to even more draws.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
The 1984/5 match seems like the exception, not the rule, as it was played under the 6 win rule, which was abolished after it (and to think Fischer got a majority in FIDE congress for 10 wins in 1975). I don’t think their subsequent matches had a high draw rate (2/3) even by 20th century match standards, especially when you consider that these were two of the toughest players in history to beat. Even with the wild Tal in 1960 and the inferior Short in 1993 over 60% of the games were drawn.

In general, I see several reasons for the huge amount of draws, as mentioned above:
  1. Better opening knowledge/preparation. You have almost no chance to win a game nowadays in preparation like Kasparov did in game 10 vs. Anand.
  2. Having a solid world champion who seems happy to go to rapid games (see game 12 vs. Caruana).
  3. The short match format, which discourages players from taking risks.
The only way I see to change this is to have at least one player be fully motivated to avoid the rapid games. One option is to return the rule that the champion retains his title in case of a draw, but that seems unfair, and the rapid games are interesting and attract fans. What I would do is have the rapid games played before the match (same in soccer and penalty kicks, BTW). This way at any given moment one of the players is behind and must play for a win. In 2010 there were 5 decisive games out of 12, probably because Topalov didn’t want to go to rapid games and took risks, especially in the last game.
I think this is a fantastic idea. This needs to happen! Just make it so that the winner of the tie break wins in the event that the score is even after 12 or 14 games. Give the loser of the tie break the choice whether to start or end the classical games with the white pieces. This also reduces the impact of having a world championship possibly decided by Armaggedon. Nobody wants to see the championship decided because a guy gets flagged or accidentally knocks over a piece and loses 3 seconds in the mad scramble. Worst case is there are still another 12 games to go. I also think it would be good to make Armaggedon games playable on computer. It's just not suited for over the board. They could sit in the same room or whatever. I don't think that could ever happen if the Armaggedon game might be the final thing that happens in the tournament. People are going to want to see the players face to face. But if it were just part of a preliminary tie break, it could work.

I should clarify some of my posts above. It's not the lack of decisive games that is getting me down. Players finding draws from losing positions can be incredibly exciting -- there were a number of games like that in the World Cup for example and it was all very exciting. My problem is the games where one player is essentially still playing against Stockfish on move 25 depending on how good his opponent's memory is. Human moves are more likely to lead to exciting games, even if they are ultimately drawn.
 

YankeesIsrael

Member
SoSH Member
Sep 21, 2009
108
According to ChessBase application, 8.c4 was only played twice in the past (8.Bb2 44 times), 8…dxc4 once in 1995, and 9.Qc2 is a novelty.
 

Bread of Yaz

lurker
Mar 12, 2019
120
Today exciting enough for everyone? Queen versus two rooks, time pressure, multiple red moves before the time control!
 

Joe Sixpack

Well-Known Member
Lifetime Member
SoSH Member
Sep 9, 2002
5,769
Canton, MA
Definitely a better game but still seems destined for a draw. I've been following the analysis from http://analysis.sesse.net (posted by someone else above) and it's useful for idiots like me to understand the positions better.

they have been dead even on the computer analysis for quite awhile now.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
Nepo kept the queens on the board and it cost him.

What a marathon that one was. Credit to Nepo. I hope in the future this does not turn out to be a blue print that encourages more closed play but for today it was a fantastic game.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
Magnus is so freaking good. The part of the game where he basically had 30 seconds per move and also had to guard against threefold repetition while finding a way to push those pawns was just incredible. The way these guys' minds work is just amazing.
 

Jnai

is not worried about sex with goats
SoSH Member
Sep 15, 2007
15,318
<null>
I haven't checked this closely and it's a weird thought but was that game longer than the entirety of the Queen's Gambit Netflix series?
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
They mentioned to Nepo after the game that he could have possibly kept a draw if he had made a different move there, but he had no idea.

They also talked to Carlsen about some lines where he possibly could have gotten an advantage more quickly and it clearly was not the line he was looking at. This thing that the journos do after where they talk about computer lines that showed a different possible result and then ask "did you see" whatever is so unfair and stupid.

The players were down to 30 seconds a move. Carlsen could have gone another sixty moves and there would have been another sixty moves for Nepo to make an inaccuracy to let Carlsen get one of those pawns forward.

Over the board the players don't see the status bar. I guess they only have a general sense of whether they are winning or whether they are losing. Nepo said he knew that queen against rook and knight should be a draw. But asking him to play perfectly against the world champion with 30 second increments and after 8 hours is probably too much.

tl; dr -- knights are fucking tricky. If it hadn't been move 130 it would have been move 145 or move 160 or move 175. Carlsen could have gone all day.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
This is 35 minutes of sheer delight:

View: https://youtu.be/PmQs1KhB948

I could, and quite possibly will, watch this multiple times.
Gotham has been really good. I wish that Daniel Naroditsky was also doing recaps but I guess he has been actually commentating on the games live with Sasha Grischuk (in Russian!). He reposted his thoughts on the game yesterday that he made during his live stream on twitch but I doubt after 8 hours he's doing that today.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
Just watched some recaps.

This is the fifth or maybe even sixth game where Magnus has offered a pawn to slow Nepo’s development. Clearly not a coincidence. I wonder how having a win on the board may change things or whether Magnus may continue to play these gambit style moves.
 

Jnai

is not worried about sex with goats
SoSH Member
Sep 15, 2007
15,318
<null>
Just watched some recaps.

This is the fifth or maybe even sixth game where Magnus has offered a pawn to slow Nepo’s development. Clearly not a coincidence. I wonder how having a win on the board may change things or whether Magnus may continue to play these gambit style moves.
Was thinking exactly this. Clearly he has prepared this, but wonder if he scales it back. I wonder if he's also trying to intentionally provoke Nepo into playing for more, which did seem to work to his advantage at points yesterday.

I also wonder if these sacrifices and gambits are just convenient exit points for Carlsen to get away from planned openings and into playing Chess.
 

YankeesIsrael

Member
SoSH Member
Sep 21, 2009
108
A very interesting start to today’s game. Carlsen’s 7.Nd2 is rare (played 48 times in the past vs. over 4,500 times 0-0 was played). Instead of playing it safe, Nepo responded with the risky 8…Bd6 (Qe7+ is safer) and 9…h5 (a novelty).
 

Bread of Yaz

lurker
Mar 12, 2019
120
A very interesting start to today’s game. Carlsen’s 7.Nd2 is rare (played 48 times in the past vs. over 4,500 times 0-0 was played). Instead of playing it safe, Nepo responded with the risky 8…Bd6 (Qe7+ is safer) and 9…h5 (a novelty).
I do not understand why Nepo is playing 1.e5 rather than the Sicilian. He is allowing Carlsen to get the kind of slow, positional grinding positions that he excels in. Nepo is superb tactically and needs to get into those positions if at all possible.
 

DennyDoyle'sBoil

Found no thrill on Blueberry Hill
SoSH Member
Sep 9, 2008
32,822
AZ
Ian needs tomorrow to regroup in a big way. Game 6 clearly crushed him and he is playing like a man who has already lost. I think stamina and ability to grind has been huge so far in this tournament and Ian has shown no inclination to do the hard work of grinding the shit out of the position today. Maybe he can find a way but he looks defeated.