Nepo Adds Supercomputer to his WCC Team

[Pargat Perrer]

As I have reached the age where my memory for specific opening lines is failing me I am willing to give Fischer Random a try. On the other hand having bishops on the wrong squares and knights starting on A1 or H1 makes me cringe. I guess even though Fischer Random maybe the way to go there are still alot of adjustments to make.

Why should Knights starting on A1 or H1 make you cringe? They are only starting there, they aren't sentenced to remain there. It what you do with them that counts. It's like having a chance to buy a decrepit house at a good price; think about what can be done with that house rather than what it is when you buy it. This is all part of mind expansion.

[Brad Thomson]

Thanks Pargat. I would add that at the top level, the boys do not mind so much losing at the faster time controls, or at least they do not mind taking the risks, because they have the pretense/excuse of the faster time controls as having been the cause, and the need to excite the fans as having been the reason for the taking of the risk. Perhaps there will always be a place for traditional chess at faster time controls, but the last so-called match for the World Championship of Chess proved that the time has come for Fischerandom when it comes to slower time controls.

[Mario Moran-Venegas]

I would like to see a compilation or list of teams of seconds along with their silicon friends used at world championships.
I do not think I can reach a conclusion as to whether a supercomputer helped or did not without analyzing move by move. I do not think a supercomputer would be the only factor. How they are used is key. From Anand's quote by Francis I read improper use. (“It must be good for the player to have good hardware. The problem is that it often doesn’t tell you
what you want. You want it to find an improvement for white, and it may choose to find it for black instead.)

[Mario Moran-Venegas]

I definitely do not see a future need to modify chess. The memory portion will be larger but human analysis will continue from there on a position that is more drawish than the starting position.
Maybe the change needed is that a draw is a win for the challenger!!

[Mario Moran-Venegas]

Regarding the first post on this thread announcing Zhores supercomputer as the choice by Nepo. This computer is not strong enough to rank in the TOP500. Many of the top500 are dedicated to a task (weather ,biochemistry etc) but many can be rented.

[Wayne Komer]

Nepo Adds Supercomputer to his WCC Team

May 19, 2021

Seconds

Nepo and Carlsen have enjoyed a good relationship for years. The relationship is so close that Nepomniachtchi has on occasion worked as a second for Carlsen. Before the London Chess Classic in 2012, the Norwegian was disappointed when Nepo was denied a visa to the UK and had to do the job from home.

That was not an issue during the 2014 World Championship match in Sochi, when Nepomniachtchi and Vladimir Potkin were already present for a training camp in the same hotel. Preparing for the sixth game, the duo accepted an invitation to visit Carlsen's €1,600-a-night suite and were key in preparing a Siclian variation in the sixth game that was won by the Norwegian.

“It's super-exciting to face Nepo. I couldn't have asked for a better opponent. He showed his level in the Candidates. That's why it will be very interesting from a sporting point of view. In addition, he has an exciting style, plus the fact that we have known each other for so long.”

While it's clear that Peter Heine Nielsen will remain his lead second, the reigning World Champion reveals he is expecting to have a new team of seconds this time.

This time there will be some new people around him.

In the past Carlsen has worked with Jon Ludvig Hammer, Laurent Fressinet, Maxime Vachier-Lagrave, Michael Adams, Pavel Eljanov, Pentala Harikrishna, Daniil Dubov, Sam Shankland and Jan Gustafsson. Even Ding Liren has worked with Carlsen, although not before a World Championship match.

Nepomniachtchi, on the other hand, sent shock waves around the chess world when it was revealed that Meltwater Champions Chess Tour commentator Peter Leko had been working for him during the Candidates. Whether the former challenger will remain in the team remains unclear.

Carlsen also told VG that he thinks the extension from 12 to 14-game matches will lead to more risk-taking and sharp games, and that it's 'wonderful to avoid two rest days, because it's almost more nerve-wracking to walk around and take trips on the days off'.

He also ensured his fans:

I am 100 percent motivated! I have already thought a lot about the match.

https://chess24.com/en/read/news/100...aration-begins

[Ian Findlay]

Not sure these supercomputer clusters are an advantage - Topalov used one - state-sponsored "Blue Gene" (? or something
like that - Rybka based) against Anand in 2010, although it helped him win the 1st game (as I recall), it was no match for Anand!

Karjakin in 2015 prepared similarly for Carlsen with a super-expensive state-sponsored super-computer, not much success!
(OK, that one went down to the wire, but still!)

In 2010, Anand's 'super clusters' were KKC (Kasparov, Kramnik, Carlsen), who helped him formidably.
These are Anand's thoughts on the super-computer advantage:

“It must be good for the player to have good hardware. The problem is that it often doesn’t tell you
what you want. You want it to find an improvement for white, and it may choose to find it for black instead.

“It is also insidious — it can stop you playing your favourite lines because of some obscure problems
somewhere. I would say computers are very useful — especially very powerful hardware — but if working
with the computer means you stop taking risks it’s of course going to kill you,” Anand said.

In that respect, human analysis has an edge, he said.

“With humans it is fantastic, because some of the things they tell you, you can almost relate to. You can
almost see how it is something you can apply at the board. In the last few days some of the suggestions
came from Vlady (Kramnik) — very intelligent suggestions, with his explanation of why you should do this or that.

“Or sometimes Garry (Kasparov) will tell you a story about the player, and it is some insight, which at a critical
moment it may help you to choose between two moves and that can be decisive. So it is a very, very different kind of help.

Of course these are some of the greatest players in chess, so you can’t really ask for more,” said Anand."

Great insights. Ganguly, on Anand's team, told me that they had access to a super computer which gave him a nice edge vs Kasparov. But, times are a changing.

[Pargat Perrer]

I definitely do not see a future need to modify chess. The memory portion will be larger but human analysis will continue from there on a position that is more drawish than the starting position.
Maybe the change needed is that a draw is a win for the challenger!!

Armageddon chess does this, it gives White more time 5 minutes to 3 I think, but makes a draw a win for Black. I think the results so far have been a big advantage to play Black.

I've long thought that chess should do away with half-point scoring. So perhaps there could be a new scoring that gives Black a slight point edge for drawing and for winning.

So for example, a win with White is 9 points, a win with Black is 10 points. A draw with White is 4 points, a draw with Black is 5 points. However the points for the draw would be reduced by 1 for each player if the game doesn't go at least 30 moves.

Perhaps even losing with Black could be a slight edge: a loss with White is 1 point, a loss with Black is 2 points. This would give the Black player quite an incentive to play more risky.

Giving points for losses also distinguishes games played and lost versus games forfeited, which would award 0 points. Voluntary byes would award 4 points to both players, making byes slightly less than a played draw for Black, so players would be less likely to take a voluntary bye with Black. Perhaps forced byes would be where a half-point comes in, awarding 4.5 points to each side.

These numbers are just off the top of my head, someone else might think of a better system.

[Ian Findlay]

Armageddon chess does this, it gives White more time 5 minutes to 3 I think, but makes a draw a win for Black. I think the results so far have been a big advantage to play Black.

I've long thought that chess should do away with half-point scoring. So perhaps there could be a new scoring that gives Black a slight point edge for drawing and for winning.

So for example, a win with White is 9 points, a win with Black is 10 points. A draw with White is 4 points, a draw with Black is 5 points. However the points for the draw would be reduced by 1 for each player if the game doesn't go at least 30 moves.

Perhaps even losing with Black could be a slight edge: a loss with White is 1 point, a loss with Black is 2 points. This would give the Black player quite an incentive to play more risky.

Giving points for losses also distinguishes games played and lost versus games forfeited, which would award 0 points. Voluntary byes would award 4 points to both players, making byes slightly less than a played draw for Black, so players would be less likely to take a voluntary bye with Black. Perhaps forced byes would be where a half-point comes in, awarding 4.5 points to each side.

These numbers are just off the top of my head, someone else might think of a better system.

Most Armageddon games, White gets 5 minutes to Black's 4 with draw odds. Which probably favours black slightly.

[Pargat Perrer]

Most Armageddon games, White gets 5 minutes to Black's 4 with draw odds. Which probably favours black slightly.

Thanks for that correction, so Black has 4 minutes not 3. I did read on some chess site (can't remember) that whenever the top players are playing Armageddon and given the choice of playing White or Black, they are choosing Black and doing very well with it. I am surprised the time difference is only 1 minute, but I guess at that level a minute means a lot.

I wonder if someone has Chessbase and can do a database query to find out exactly how often Black wins Armageddon games in FIDE tournaments or matches when both players are rated over 2600 or 2500. Probably not a huge sample size, but it still might indicate something.

[Ian Findlay]

Thanks for that correction, so Black has 4 minutes not 3. I did read on some chess site (can't remember) that whenever the top players are playing Armageddon and given the choice of playing White or Black, they are choosing Black and doing very well with it. I am surprised the time difference is only 1 minute, but I guess at that level a minute means a lot.

I wonder if someone has Chessbase and can do a database query to find out exactly how often Black wins Armageddon games in FIDE tournaments or matches when both players are rated over 2600 or 2500. Probably not a huge sample size, but it still might indicate something.

I do not think that these games make it to the databases.

At the Banff Open, we have an Armageddon playoff (thanks to an anonymous donor who donates a special Armageddon prize each year). In 2018, I beat Omid Malek as black (my choice), to lose to Mark Ginsburg (who chose black), in a 3-way where Mark got a bye to the final. In 2019 Belsar Valencia chose White?! to lose to Agnieszka Matras-Clement. I personally do not think 1 minute offsets draw odds, but I have not seen any statistics. Maybe 5 minutes to 3:45. There is also no increment which adds to the excitement. It is very popular with the spectators.