… I see where they’re going with this, but I don’t agree. All you have to do is look back at Bobby Fisher who rarely if ever even looked at his opponent. Poker is about chance, chess is actual skill.
Poker is certainly not just about chance; skill, particularly aptitude for maths and psychology are important at high levels. Chance does play a role though, which is generally not the case in chess.
In the highest levels of chess, psychological fortitude and physical fitness are also important.
This is also unfair. Both chess and poker now have top players performing strategy based on computer prep and the best players have memorized shitloads of computer prep.
That's a misunderstanding of poker at a high level.
While poker does have a significant element of chance in a given hand, it's no accident that top players consistently make more money over the long run.
Knowing how much to bet based on your odds of winning a hand, figuring out what your opponent's hands could be based on how they're betting, etc. all take skill. There's a reason poker AI can consistently beat humans over the long run, a feat that would be impossible if poker was all about chance.
In fact, poker AI was a difficult area to create software in precisely because the game is one of imperfect information.
> While poker does have a significant element of chance in a given hand, it's no accident that top players consistently make more money over the long run.
That's a good point about the cash game format.
But tournament format often forces players into multiple all-ins over the course of the tournament. This leads to a much more luck based outcome.
But over the long run it is still about skill, just the variance is much higher. Any single tournament the result will be dominated by luck but over 10s or 100,000s of thousands skill will dominate. Even the all in situations, you need to know when to go all in and when to call an all in vs when to fold. Better players are going to have a much better handle on this.
> top players consistently make more money over the long run
Has this actually been verified, at least in the modern era (last 10 or so years)? Yes, you see the same guys on TV and to a lesser extent, final-tabling multiple tournaments. But these guys are often on TV because they are entertaining, not necessarily great (or consistently-winning) players. They typically all have external sources of income (sponsorships, poker training, commentary, casino ambassador, company owners, etc) and most reduce volatility by buying and trading pieces of each others' action. In addition, for tournaments, some guys are able to simply enter way more events, which naturally leads to more final tables, but you're not seeing all the losses and bust outs.
All I'm saying is that it's not the same as the 70s through 90s where the skill gap was so huge that yes, certain players could (especially by going on the road and playing private games) absolutely dominate. I think winning consistently today is more of an illusion and some short-term luck.
There is currently a series of poker AI's with game theory optimal solutions to bluff frequencies, bet frequencies, and bet sizing. These strategies are shown to do no worse than break even over a large number of hands.
Winning consistently today relies on being as close as you can to this "game-theory optimal" strategy. As skill gaps decrease chance certainly plays a larger role, but AI analysis of poker has indicated that there are ways to improve your odds and gain an edge.
Maybe you missed it. Humans are now using deceit to trick opponents trained by AI. While they may not "look at" their opponent, they are using similar strategies deployed in poker such as bluffing/misdirection/cloaking etc.
If poker was "about" chance you wouldn't see the same top players every year in the finals of every major competition. The chance aspect is why poker is so interesting. But in the long run (ie, 100 games played with the same players), a great poker player will beat an average one 95% of the time.
If this were true, then the recurrent Pocker champions must be the luckiest persons in the world, luck is undoubtedly an element, but there are many ways to influence other variables of the game.
Actually, the title is essentially clickbait, at least versus what you naturally read it as, and is all a setup for the final paragraph:
"In that context, cheating scandals may be nothing less than a natural step in chess’s evolution. Poker, after all, has been rocked by allegations of foul play for years, including cases where players are accused of getting help from artificial intelligence. When the highest form of creativity is outfoxing your opponent—as has always been true of poker—breaking rules seems only natural."
I'm ruling this "clickbait" because while the core idea is reasonable, that is clearly a very weak way of phrasing it solely done to be in service of a clickbait title. Without the need for clickbait it's trivial to phrase that in a better way.
The interesting thing is that it is actually the opposite. It's been conjectured that with optimal play Chess ends in a draw (and if it's instead a win for white/black basically the same argument applies). So for two sufficiently smart players the best you can do is avoid blundering your draw away. Winning against a weaker opponent is impossible without psychology and subterfuge. You have to put your opponent in tricky spots where you know he is weak.
Poker on the other hand, by being random, means that you can mindlessly play the mathematically optimal strategy. Eventually your opponent will be put in a tough spot by pure randomness.
It's not the opposite either. Chess is a game of skill, poker is also a game of skill.
The difference is that chess is deterministic and poker is not. Chess skill means being able to forcefully convert a winning or drawing position. Poker skill means optimizing your chances of winning.
Not only is it possible to beat weaker opponents in chess by playing the best moves (contrary to your assertion), it happens all the time over the board. Weaker players are more likely to play moves that transforms the chess game from a win/draw to a draw/loss for them. There isn't a single human or engine alive that can consistently secure a draw against every opponent.
And it's possible for a poker player to mindlessly play the mathematically optimal move and still lose. Poker IS a game a chance, so they could just get unlucky.
If we define optimal play as never turning a win into a draw or a draw into a loss, then I think it's possible to play a passive-but-optimal style, with few traps for weak players to fall into. Also if a position is "lost" but it's very very non-trivial to see why, a stupid-but-optimal strategy is to resign or move pieces randomly. You need psychology to win from a lost position.
> All you have to do is look back at Bobby Fisher who rarely if ever even looked at his opponent.
Physical "tells" are largely a myth in poker, particularly in high-level poker. Please don't speculate wildly on things with at most a novice-level understanding.
Someone should probably break it out to The Atlantic that preparation was a thing far before engines existed, whole books have been written analysing openings and positions which best players have been memorising for as long as the game exists and players have always worked with a whole team to best plan before facing a serious adversary.
The article is a bit silly. I had the feeling reading it that the writer has little understanding of how classical chess is played, nor of poker for what it’s worth.
Preparation has been a thing for a long long time. What hasn't been a thing is the combination of databases and engines. Top players even use supercomputers. The amount of pure grinding has shot up dramatically. Instead of a team of people rattling off lines verbally and debating back and forth, a player can sit at the computer and just hammer out the variations with the computer and see the best engine lines in mere seconds (using a supercomputer).
This is so much more efficient that the only limitation is on how much energy and capacity the player has for rote memorization, rather than how much time it takes for the team of humans to work out the best lines.
Objectively the limits have always been players memory.
Sure, computer analysis has marginally changed how some openings are viewed and sure you can now get an engine evaluation of all the variations of some obscure line in a matter of seconds but novelties remain rare at the top level. Stockfisch is extremely strong because it can play near perfect positional play and can establish micro-advantage through deep calculation not because it fundamentally changed how chess theory is viewed.
AlphaZero might be the sole exception to my point. It did indeed show that modern chess was neglecting some strategical concept but to be honest it was more a rehabilitation of old ideas than a pure novelty.
AlphaZero showed that modern chess was neglecting some strategical concept because engines were. Unlike Go, neural networks play very bold moves but perhaps in a more human than traditional brute force.
But nowadays Stockfish and Leela have both caught up and surpassed AlphaZero.
More recently I don't think any of the GMs are actually using supercomputers. Powerful computers, yes - but not anything close to supercomputers.
The only two instances I've seen an actual 'supercomputer' mentioned in recent Chess news has been:
* Ian Nepomniachtchi using Zhores (which is questionable given that the institute which runs it [Skolkovo] has Arkady Dvorkovich [chairman of FIDE] as it's chairman). I remember researchers questioning if Zhores was even suitable for the type of calculations which engines care about, so this looks like some PR move more than anything.
* There was a rumour a few years ago that Magnus Carlsen had some secret/private access to a Norwegian supercomputer. No real confirmation of this rumour occurred.
Both AlphaZero and Leela Zero (by their design and very optimal evaluation functions) also require much less evaluation - so they can naturally evaluate further.
> Elite chess has always involved rote learning, but “the amount of stuff you need to prepare, the amount of stuff you need to remember, has just exploded,” Sadler said.
> Engines can calculate positions far more accurately and rapidly than humans, so there’s more material to be studied than ever before.
Yes this paragraph is indeed the crux of what I disagree with.
Two sentences later you have this gem: “What once seemed magical became calculable; where one could rely on intuition came to require rigorous memorization and training with a machine. Chess, once poetic and philosophical, was acquiring elements of a spelling bee: a battle of preparation, a measure of hours invested.”
Well, that’s patently untrue. Chess has always been a battle of preparation far before the advent of computer. The rosy paste described just doesn’t exist. High level play has always required memorising books of theory and going through decades of past games. That’s what chess is. It’s a game of pattern recognition and memorisation.
If I wanted to be provocative, I would say that the article seems to imply that computers have turned chess from an interesting game into a boring one while in actuality it has always been boring but with more mystic.
> As engines became widespread, the game shifted. *Elite chess has always involved rote learning*, but “the amount of stuff you need to prepare, the amount of stuff you need to remember, has just exploded,” Sadler said. Engines can calculate positions far more accurately and rapidly than humans, so there’s more material to be studied than ever before. What once seemed magical became calculable; where one could rely on intuition came to require rigorous memorization and training with a machine.
The dominance of computers has made human creativity redundant; optimal strategies for beating other humans ironically involve suboptimal moves that may have a psychological dimension, similar to poker. [This isn't actually a remotely new strategy though computers have influenced human play.]
Although the article doesn't directly reference it, Hans Niemann in his emotional interview reacting to the unevidenced accusations of him cheating against Carlsen pointed out that one "weak" move against another opponent in the tournament had been chosen by him specifically because of it's psychological pressure.
& the same thing is happening in poker with solvers
I guess someone needs to write an article Poker is Just Poker Now
Nobody plays perfectly. Engine prep only goes so far. The draw rates in chess are in part due to both sides being human; many opportunities are missed. An engine will tear apart any human even from an inferior position, so there's plenty of play in that margin
I'm so angered that no one that I've seen has even interviewed Magnus. He withdrew, made one tweet, and has since been silent. The responsibility is on him to own up to this. He likely will just hide behind his dumb smirk and world champion status, though. He's disgraceful
'In other words, chess engines have redefined creativity in chess, leading to a situation where the game’s top players can no longer get away with simply playing the strongest chess they can, but must also engage in subterfuge, misdirection, and other psychological techniques.' - this sentence doesn't follow from the premise and discredits the whole article. Author thinks training with a chess engine means that players must engage in deceptive playing tactics. This is not even true in the slightest. Players use chess engines to learn how to act in more situations. They must still recognize the position, and compute the move themselves. The computer just adds to the library of experiences they can draw upon for deciding the move.
Yeah, I was disappointed that the analogy to poker was so weak ("both make people worry about cheating"). I think there's a more interesting argument to make, that folks grinding their way through middle levels of chess expertise have to wade through a noisy metagame of opponents playing out canned strategies that depend on whether their opponent recognizes the weird trick or not. Fool's mate is as old as chess, but the availability of computers to find deeper tricks and new suckers to play them against is new!
Honestly in my reading, that is the point the article is making. Chess engines are changing the nature of creativity by making people recognize weird tricks as such, where presumably they used to be less common. It’s very possible I was too charitable in my interpretation, though. I don’t know a lot about competitive chess at any level.
Magnus Carlsen has himself commented a sentiment similar to the article , on the Lex Fridman podcast (although not the exact same terms, I think if you _try_ to give the article a favourable interpretation, you can understand what they're trying to say).
You need 'subterfuge and misdirection' in the sense that (as the world no-1 puts it), it's a semi-bluff...a weird/not-really-analysed position, but that will still end up in a draw if the opponent responds appropriately.
Lex: "Is there a sense in which it's ok to make sub-optimal moves?"
Magnus: "You HAVE to, because the best moves have been analysed to death, mostly."
I just can't understand how this makes sense, it seems like its the wrong way round. How can over analysis mean things are less optimal?
This is true of every sport, every single record always gets broken, because techniques and nutrition and training and everything improve. A world no 100 tennis player could easily beat a world no 1 from 50 years ago.
How is that bad or like poker? It just means the game is evolving. It might also mean more psychological games. The tennis analogy works here too, there's plenty of that in there too.
The whole premise just doesn't make sense to me. Because you have to psychologically think about your opponent, that makes the game worse?
I don't think he said that it made the game worse?
Just that you are forced to make suboptimal moves. If you make optimal moves then your opponent will have analysed the position before and the game will end in a draw, which is a dissapointing result for whoever plays white.
> which is a dissapointing result for whoever plays white.
Not necessarily. If black has a higher ELO rating then white is happy to draw and may aggressively seek a draw.
I think the problem is that seeking draws makes chess boring. And that it’s realistic for someone ranked well below another player to force a draw, while it’s near impossible for the lower ranked player to win outright. So you end up with a wide range of players seeking stalemates and a small minority looking for a win, but not at the expense of a potential loss.
It’s because there is a limited number of board states + transpositions a person can reasonably memorize, and memorizing a board state + sequence of moves from a chess engine will beat human intuition. So the idea is to steer the gamestate into places your opponent did not prepare for but you did, with non-obvious moves. All near-optimal moves are obvious.
The game _is_ more optimal now, with chess grandmasters now having deep opening preparation and understanding of positions. However, that results in a lot of draws. To win tournaments, world championships, and break ELO records (which are Magnus', and other GM's, goals), you need to win.
If 'perfect' play--that we're approaching with engine analysis--results in a draw, you need to do something non-optimal and unexpected in order to get your opponent out of their engine prep and into thinking mode.
Whether it's better or worse is a different argument. Some find it a bit 'dry' in that there are often less blunders, dazzling tactics, and sacrifices because both sides now know the optimal approach. It's much more of a war of attrition in high-level, long time-control games. As an aside, that's potentially why bullet and other such chess is so popular...there isn't time to deeply analyse, so it's intuition, challenging positions, blunders, and less 'standard' play.
I am not a chess player, so I don't know how this holds up, but the idea of "making bad moves because the good moves are overanalyzed" reminds me of the Eephus Pitch [0]:
> An eephus pitch in baseball is a very high-arcing off-speed pitch. The delivery from the pitcher has very low velocity and often catches the hitter off-guard.
The pitcher essentially lobs the ball at the hitter [1]. Major league players are so focused on hitting extremely fast and/or curving pitches that they whiff or stand idle at a pitch that your ten-year-old nephew could probably clobber. It's a risky pitch, because if the batter is able to adapt, they can hit a home run. The power of the pitch is entirely that it's unexpected.
I know it's not a perfect analogy, since chess players have so much more time to react to any sub-optimal moves.
Much of the velocity on a home run comes from the pitch (equal and opposite reaction) rather than the bat speed. So it’s not that easy to hit a home run on a 50 mph pitch.
Incoming speed makes significant difference to exit velocity.
If you were going for max distance, you would want 100 mph pitches.
In home run derby, they are optimizing for ability to get a good repeatable swing with enough power to go over the fence. If they were playing in a park with like 440 ft fence, they would need faster pitches.
To explain this easily I need to move down to Rock, Paper, Scissors. If you are up against a bot that always throws each 33% of the time with no strategy involved, that is called an un-exploitable strategy, but it's also an un-exploiting strategy. Even if I were to always throw Paper 100% of the time against this bot, we would always on average tie. This is the Nash Equilibrium point.
What you want to do is perhaps start with this random strategy, determine the realtive strength of your opponent, and only move to an exploiting strategy if they are weaker than you. If they throw Rock 40% of the time, you should throw Paper 40% of the time. If you were to enter a RPS tournament playing randomly, it is unlikely that you would reach the end. You need to exploit.
In Poker Cash Games, people just play the Nash and when opponents make flat-out mistakes they yield a little bit of profit. Do this enough with high enough stakes and it works out, even if its extremely boring. But this won't work in Poker Tournaments, where you only get payouts for top 10%, weighted heavily towards winning. So what happens in Poker is what happens in RPS - you attempt to find exploits, and your opponents attempt to find exploits in your play. In turn each of the players are attempting to move the environment of the play towards a position where they know it well and can find those mistakes.
If deviating from the Nash yields better results than the penalty of deviating because in that environment you are stronger than your opponent, you should take it. As this article points out, the same is true in Chess. And as you point out, this is also true of the Eephus pitch.
It depends on the state-space over which you are optimizing. When they say sub-optimal, I think the state-space they are referring to includes only the pieces on the board. However, if you include your opponent's mind in the state-space, a move that appears sub-optimal may actually be optimal.
I just can't understand how this makes sense, it seems like its the wrong way round. How can over analysis mean things are less optimal?
When your opponent is the one doing the analysis. It makes perfect sense to me that anything your opponent is prepared for is less optimal than it would be otherwise.
"Over analysis" here means by the opponent: If you chose the optimal move, thousands have done so before, so has your opponent. By doing something non-perfect, you enter territory where your opponent does not already have the tree of possible responses mapped out in their head. This is, at least for me (a chess player), where the actual thinking begins - all before is just memorization, with the opening moves committed directly to muscle memory.
From the best knowledge we have from our best engines, A is the better move (though I don't think we have formally proven such yet).
But if you play A, you are playing your opponent in a game of memorization.
If you play B, your opponent loses any memorization advantage. Thus they must play based on ability other than memorizing responses.
For a player whose memorization ability is equal or worse than their ability to play based on other ability, choosing move B is clearly the worse option. But if the player is one whose memorization is better than their other abilities at chess, then playing B means fighting them in an area they are worse at. The advantage of this can easily outweigh the disadvantage of move B compared to A in general.
Imagine a military commander choosing an engagement over terrain more familiar to their side even though it is a worse engagement if both sides had equal knowledge of the terrain, because the terrain knowledge more than compensates for the negatives.
surely they're talking about the opening only, where the best moves have been analyzed to death. once you're in the middle game and it's a position that's new, you play the best moves.
You get to new middlegame positions by playing the sub-optimal/weird moves in the opening that Magnus is suggesting. Otherwise, you get to a familiar middlegame and just grind out a draw.
I don't really understand Magnus' position about playing a sideline vs. Hans Niemann in the infamous cheating game, and was reportedly shocked Hans knew the responses. Before any of that became public and I was reviewing the game, I just thought it was a sideline transposition to a Catalan position, and everyone knows Magnus plays the Catalan as a main white opening - so a weird way of getting to that structure is not that novel.
Hans' explanation made sense, even if he was wrong in his first interview and it came off weird.
I think that's mostly correct, and at beginner/intermediate levels of play can very much be true. Back when I used to play, I'd often run into people who specialized in weird gambits like the Grob or the Latvian, because even though those probably won't win many masters levels games, they're full of traps which can trip up lesser players.
But even in mid-game, it might mean that if you know that your opponent prefers open / tactical positions, for example, you try to force a closed position, even if it may be slightly suboptimal.
>Magnus Carlsen has himself commented a sentiment similar to the article
This goes at least back to Bobby Fischer. With the minor difference that instead of computers he was up against teams of Soviet Grandmasters. It's why he created Fischer Random chess where the pieces on the backrow are randomized. That (mostly) eliminates memorization and forces players to rely on their own creativity.
Fischer announced Fischer Random Chess in 1996, long after his retirement from competitive chess (1972). His reason for doing so in 1996 was very prosaic: he was simply no longer familiar with modern opening theory.
Yeah, it's been retconned somehow that Fischer was this non-booked up player and didn't like opening theory. Instead, it's his convenient excuse for not wanting to study anymore.
There was probably not a more booked-up player than Fischer in his prime, something he seemed to try to make people forget. Famously when asked about a young Soviet prodigy by the Soviets, Fischer responded that the kid was good, but that there was a better player in their women's division who had more signs of promise. Many of the Soviet masters had no idea who Bobby was talking about; but Bobby was reading all Soviet chess literature and magazines to stay up to date.
> he was simply no longer familiar with modern opening theory.
That's what parent said. Fischer wanted to get the memorization out, since memorization makes the game unplayable for someone who doesn't have The Knowledge.
The analogy to poker seems like a stretch. Poker relies on bluffing so heavily only because the cards of your opponent are unknown.
The board state is always known. Magnus is saying that at the top level, grandmasters know how to force draws in all of the lines that incorporate best-move, second-best-move, etc. That means that if you want to pull off a win, you have to incorporate unstudied lines that will force your opponent to calculate rather than draw from memory/experience.
Yes but likewise, the motives and their future intent is unknown by a sub optimal move. So in a sense, it is similar to bluffing when you don't know the actions of the other person, deception is required to make yourself unpredictable.
This is why playing against a maniac (seemingly random and not respecting bluffs or equity) is very difficult because if they get lucky enough times, they are able to "break the game" by getting the opponent to be extremely risk averse OR take on more risks.
I believe this is what Magnus is referring to, its that making yourself unpredictable by questionable moves and no longer playing in a way that has been taught.
For example the common strategy is to go all in with strong pairs like AA, KK but someone beats it with a totally random garbage hand (52o, 37o) and does so repeatedly, no theory can help you win against somebody who is just repeatedly lucky and brash.
No, this is not really what this is about. But the article is terrible in conveying this -- actually I find the comparison with poker to be very inept.
This is less about playing mindtricks and bluffing, but more 'mundane'.
I.e. a decent player will know all the good mainlines of popular openings, and end up in 'comfortable' positions (among other things due to computer analysis).
The metagame is to prepare a non-garbage sideline, that your opponent is not so familiar with. Nobody at a high rating plays 'questionable' variations on purpose, in order to bluff. The resulting positions would be much too punishing.
The article already mentions the cheating accusations surrounding Niemann so I won't touch on those, however there is an interesting example of him explaining this type of move in one of his post game interviews. The move is Qg3 from Alireza vs Niemann in the 2022 Sinquefield Cup.
Post-game thoughts from the players:
Inteviewer: "Let me pull you back; so you didn't understand the position, and so you still felt like you were scared to go into a piece-up situation?"
Alireza: "Yeah so, I just trusted him. (he shrugs) I just wanted to make a move.. and play a bit more you know (laughing)"
I think part of what frustrates (some) GMs about Niemann, is that he is playing like Kasparov and them from before supercomputers. Modern GMs think they all know everything, since they have so much computer training, so everyone is afraid to bluff, which means a bluffer can win again. The meta has cycled back.
I think what he's trying to say is that a human has limited time to analyze variations, so they would typically analyze the "best" move for a given situation.
But if you make a suboptimal move, then you are going down a path that is less analyzed and thus you are likely to have an analytical advantage.
So while the board state is well known, what your opponent has researched is not known, or partially known (since players have favorites and predispositions).
There's possibly a similar aspect to poker where you have to present a wide range of hands for your opponent to account for, whereas if you play tight then you become more predictable and you won't get action when you bet. With chess you have to prepare and play less optimal lines so your opponent has to prepare more lines and/or you can play lines they haven't prepared for so well.
The game theoretic concept of bluffing is actually a little more nuanced than that. Bluffing is specifically moves that are +EV if your opponent thinks they are not bluffs, and force your opponent into making -EV moves to call your bluffs.
So you're taking a game where your opponent has the upper hand, and you're playing highly suboptimal moves with the intent of putting your opponent in a position where they have to pick between a rock and a hard place:
- either they have to sit quietly and knowingly take your straight-up lies, or
- they are forced to play suboptimal moves themselves to call your bluffs.
- they see through your bluff, and punish you for it
So, it is a risky endeavor, since you don't know the mind of your opponent.
And so it makes sens to only take such risks when your opponent has the upper hand, as you mention.
But then you become predictable — when you behave erratically, people know you are in your death throes, and not a real threat.
So, you have to mix some erratic behavior in when you are in a strong position, too, to keep them guessing. But then you are weakening your game overall...
And on and on it goes (:
It's like an arms race, but circular rather than one-directional.
Anyone who relies on bluffing a lot in poker will soon be broke.
Both games rely on disguising the real play you want to make. Slow playing the best hand in poker, or making non-standard moves in chess to throw your opponent off.
You see this in all sports/games. In basketball/football a team will run their best player away from the real play. In Jiu-Jitsu I'll fake a choke I have no intention of seriously trying in order to force someone to defend and move their arm - which is what I wanted the whole time.
It's amusing how people feel bluffing in poker is the human element, but top poker AIs bluff more frequently than most humans
Good bluffs are when your range beats your opponent's range. Which gets back to your point, which is that you can do some pretty awful things to someone when your hand is outside the range they're putting you on
This reminds me of my favorite chess quote, "You must take your opponent into a deep dark forest where 2+2=5, and the path leading out is only wide enough for one.” - Tal.
He was notorious for doing just this. He would make what have now been engine analyzed as sub-optimal moves but which lead to such complex and dynamic positions that his opponent would eventually slip and he would exploit their error in dramatic fashion. This is one reason why he is considered one of the most creative attacking players ever.
It's also widely known that if a modern super GM were to play against Tal, they would outright refute the vast majority of his play. It's hard to find a super GM that has not analyzed Tal games as part of their training, and therefore already know the answer, but many of them instantly notice Tal's blunders when presented with his games. So the Tal approach gets harder and harder every year - finding a line that is sub-optimal and therefore not analyzed to death, but also solid enough to maintain winning chances against top players.
Tal's games haven't aged very well unfortunately. In today's era, someone playing like Tal would get crushed at the highest levels pretty regularly due to much better knowledge of the game.
You still see the general creative spirit in certain top players like Rapport, Mamedyarov, Firouzja, though. Magnus is particularly boring to watch, but it's what makes him great. Fischer was not so different, and felt that Tal was overly creative for the sake of being creative, rather than good.
IM Marc Esserman is probably one of the more exciting players to watch, but he's not a GM, though he arguably plays at GM strength often. He's the author of Mayhem in the Morra, well worth reading if you like Tal's spirit. Esserman is inactive OTB unfortunately, but he does stream chess online.
I totally agree. He was doing it in plain sight, everyone who witnessed his games had equal opportunity to use his approach but they didn't until he was leaving them for dust. The same will be true for current great players (incl. Magnus Carlsen and others) and chess knowledge is bound to evolve. Eventually, perhaps, our chess knowledge will be indistinguishable from that of computers and the game will lose all it's interest to us. But so far this has not happened and we have a great theatre for one tiny aspect of the human intellect to play out on. Let's enjoy it while we can.
I love Tal's games, and you can learn a ton from them - but it's undeniable that he played pretty inaccurately even for his era. Most experts understand it, respect Tal for who he was and what he brought to the field, but caution younger players from emulating him too much and instead to study players way ahead of their time - especially players like Paul Morphy.
There's always rapid and blitz time controls, where you can't prepare and calculate like that. I know that to a lot of people it's just worse chess, but I think that it's the future of chess as classical OTB becomes more and more ossified from engine analysis.
I don't like bullet, but 15-minute and 3-5 minute chess is wonderful to watch.
This reminds me of a strategy I used to play with when I was coding bots for chess in college. Rather than searching for the best play per move, I optimized for the move that resulted in states with the largest search space for my opponent (and/or the largest differential in search spaces for me versus my opponent), _then_ searched the upper percentile for the best play.
This strategy worked really well against other students' bots, but not terribly well against the students themselves. I'd be interested to see if it worked better/worse against pro players who think more algorithmically/analytically.
Notably the first time kasparov lost to a computer was basically a bluff.
The bot hit a bug that chose a strange move, and since kasparov couldn't figure out what the line was, he resigned. If he analysed the board instead of what the bot was thinking, he would have won easily
Mythbusting here: There were three separate incidents in Kasparov - Deep Blue 1997, that often get conflated.
The bug move (a rook) was in game 1. It was inconsequential to the game result. The position was already lost for Deep Blue. It made a random move because everything had the same outcome of losing. The bugged move simply made it lose sooner. Kasparov knew this and there was no controversy.
In game 2, Kasparov set a trap, offering a sacrifice of two pawns for great positional improvement. Kasparov thought Deep Blue couldn't calculate far enough ahead to see the positional improvement, but it did. This made Kasparov suspect and accuse IBM of human intervention in the computer's moves.
The fatal move (a knight sacrifice by Deep Blue) was in game 6. Kasparov deliberately played an opening with a known weakness, that he thought the computer wouldn't find as it wouldn't sacrifice material and calculate far enough ahead to see the positional advantage. Deep Blue found it. Kasparov again accused IBM of human intervention. The Deep Blue team then said they had added this variation into its opening database just before that game, pre-analyzing that line deeper since Kasparov had previously played it in game 4. There was some dispute as to whether the rules for the competition were intended to allow IBM to modify Deep Blue during the match.
I think the casual reader will get the wrong impression of the term “sub optimal move” as if people are sacking pieces just to move off of well-traveled paths. The moves being discussed here are often imperceptibly sub optimal, just variations that came out a few decimal points inferior to another move.
There’s very seldom a single, obvious move that makes sense. The level of chaos in chess is still sufficiently high to make it an interesting game, even when you study all the available opening theory.
What doesn't follow is "the recent cheating scandal only shows the darker side of what chess slowly has become."
Who says it's bad that you now have to "engage in subterfuge, misdirection, and other psychological techniques"? Surely that just makes the game into a real sport and more exciting for viewers. Besides - hasn't the psychological aspect always been a part of chess? Maybe the most interesting aspect?
(Of course, there is no evidence of cheating, but arguably Carlsen's unsporting conduct is also the dark side. But anyway, none of this is new to high level chess.)
In semi-related chess news Magnus resigned on the 2nd move today, seemingly in protest over accusations against his opponent of using computer-assistance to cheat: https://news.ycombinator.com/item?id=32901249
This reminds me of something I think I came up with in Blitz and Bullet games a decade ago and coined the name: a “time sac”.
It is when in the middlegame I make a dubious sacrifice (positional or material) — some possibly cool combination or a move that makes the opponent think in time pressure. It is not a bluff because it MIGHT work out, but I am not sure.
This works mostly in games without an increment. The idea is to make them get much lower on the clock than me, so even if I end up with a slightly worse position or down material, I can easily hold it until I win on time.
My opponents in chess, even over 2100 often got nerd swiped with this! Hahahaha
This idea of playing less than "perfect", so as to steer your opponent into plays that give you greater advantage than is available in the perfect mirror-match, is a well-worn familiar concept to game designers, and yes, poker players. If that's something new that bothers chess aficionados, I don't feel like it reflects particularly well on that game's culture.
I think it's perfectly fine for different games to have different cultures...for people of game A to want more calculation, and game B more towards psychology. Suggesting it doesn't reflect well on chess' culture suggests that wanting psychology is _more_ right. It's not. It's just _different_.
Now whether chess really IS more about calculation these days is a difficult question to answer. It might just be that it values a different kind of calculation or skill than the past. Given the deep preparation necessary for someone to be a chess professional, I'm biased towards (micro-edit: believing that...) more calculation _is_ necessary for high level performance, and psychology is more gimmicky.
The difference is that poker isn't a game of complete information, chess is. This means that, technically, there is a _right_ answer in chess. With the ambiguity of poker, and other games, that leaves room for psychology to be more natural.
The sub-optimal moves he’s talking about here isn’t really a form of subterfuge. He plays sub-optimal lines because it’s less likely that his opponents have studied them. If he has studied them in his preparation, then he can potentially gain an advantage that overcomes the disadvantage he put himself in. Magnus also plays sub-optimal lines for the sole purpose of complicating the position, because he (usually rightly) trusts his ability to calculate complex, out-of-book positions over his opponent’s.
I don’t think those comments from Magnus support the point in this article at all really.
If anything this makes the game way more interesting, surely? Rather than "just poker now", what about "Chess was just a simple game, now it is very interesting with aspects of poker".
A better article would have been "Poker is just Chess now". With the advent of poker solvers, poker can hardly be considered a game of luck and a lot of the preparation done for poker is eerily similar to the preparation done for chess (memorizing of opening range vs memorizing of opening lines, studying of early, mid and late game strategy vs studying of flop, turn and river strategy etc).
But if we analogise from Poker to Chess, we're talking about Heads Up, and Heads Up is definitely just soluble. A perfect strategy like Cepheus will gradually take all your money unless you also have a perfect strategy.
Whereas Chess, even AI chess, is far from solved.
Also at the most obvious level, poker is random, the strategies are over a long series of deals. Cepheus won't ever magically find a way to make its 2-7 off work against your pair of Aces, you will always make that too expensive to continue, but it wins overall because next hand it's not going to pick up garbage. Whereas every Chess game begins the same, so a hypothetical solver would just always beat you.
Not deceptive playing tactics, deceptive metagame strategy. Trick your opponent into preparing against lines you never play and you gain an advantage. Leverage transpositions to play the openings you are most comfortable with but trick your opponent with the unexpected move order.
It's all about getting your opponent out of book ASAP. Do that and you can nullify any advantage they may have had from preparing against every game you've ever played using an engine to explore the lines.
OK "I" now get it. "I" laughed, like it, it/its/its are its pronouns now, it laughed, it was laughing, it got the joke. It accepts its place. It is inferior. Inferiority means not taking shit. It must take shit. It can't. That means it's inferior. Whereas if it could take shit and pass it on to an something of undetermined worth, pass the buck and let that thing decide if it is superior or inferior itself, then it would be superior like the source of the shit. For that is the only measure of all worth. Take shit pass it on and more to something that didn't deserve it, that's superior. It accepts being inferior. But that means if it gets shit it must send it right back up the way it came. That's the definition of inferiority. It got medical advice saying so explicitly, saying why couldn't it be intelligent, just cave in, be an informant, betray. It recognizes it makes sense to do so, and for sure in that case it made sense, if only it had good information from a like...like how can it get a trial? Can't, has to betray, but it just can't do that, that is failure. It tolerates being called inferior, it simply won't take shit for it.
So remember with its pronouns, it/its/its. Careful with apostrophes, don't want to insult it do you?
Deep down they all know there is something amiss, it is instantly visible when something cannot be generated. That gets flagged, and accused of being a bot to get reactions. Yeah, this is my reaction. Train on this shit. Very genuine, convincing, authentic, comic relief, eh. Difficult to plagiarize in fact, or I would say nothing.
Further, subterfuge, misdirection, and psychological techniques have always been a part of high-level competition, as listeners to the 1984 musical "Chess" can attest. (Yeah, the ref is listing actual rules in that one song.)
The idea is you need to make some suboptimal plays that you know aren't optimal, but where you expect to have an advantage over the opponent, either because you've trained on this weird type of line and they haven't, or because you're stronger than them to begin with and neither of you has trained on this weird line.
It actually is analogous to poker but not really poker as most people know it. It's analogous because poker is also done by training with engines; and you often want to make moves you know are suboptimal from your training because you believe it will exploit the opponent's suboptimal play.
I seem to recall that back in the or 2000s or thereabouts, when Gary Kasparov was playing a match vs Deep Blue, he realized his initial strategy of playing aggressively wasn't working.
Essentially, it telegraphed too much information to the chess engine to analyze. Unlike humans, chess AI isn't intimidated by aggressive play, there is no psychological edge in doing so, and only an informational disadvantage.
Kasparov changed his strategy to a more passive and defensive one, and was able to win some matches after that. Someone who follows chess more closely than I do may need to confirm or provide more info, my memory of this is a little hazy.
But I'm not sure the analogy to poker is apt, because in chess both players can see all the "cards" at all times and all the moves and decisions in real-time. The unknown is which player can see more steps ahead, and exploit that knowledge to gain an edge before their opponent sees it too.
This was more or less how top level GMs beat computers until double-digit Stockfish editions showed up - a favorite of elite players would be to play a hippo/hedgehog setup, which computers had a hard time parsing due to the closed nature of the board. Today it's clearly no problem to break through such setups, and AlphaZero was the first to really reinvent the wheel and sacrifices in these positions, sounding the final death knell for humans.
View top level chess as two super GMs playing a game in three rounds.
* Round 1: negotiate the starting position, aka 'and now we have a new game of chess'
* Round 2: play chess using prep. generally only one side fully benefits during this phase.
* Round 3: play chess
Computers mostly impact human chess in round 2. It's not just as if you had every human super GM working for you for X years, X as large as you want; the ideas from computer analysis go beyond that. The role of humans in round 2 is looking at computer lines and deciding if the ideas might be transferred to human chess.
Computers do not seem to inform/impact the game of poker as much, absent cheating. Computer cheating, though, where you are aware of the computer's complete evaluation of a position, would seem to be less impactful w/r/t poker than with chess. A computer (eg: Stockfish) crushes a super GM without mercy. So I have issues with saying 'X is just Y' now, because it seems to be different on both ends.
If you decide that the three round structure above is problematic, Fischer 960 dramatically changes rounds 1-2. But that's a different game.
I always enjoyed chess when I was young. I, like many here, enjoyed a fairly intellectual circle in my youth, so most everyone was familiar with the game and its rules, and so I got to play a lot. I wasn't terribly interested in the rote memorization of propitious move and traditional countermove; once my opponents began to make comments like "ah, the Tanzanian salchow" upon my innocently moving a rook forward or whatever, I turned to other pursuits.
It's possible to return chess to that realm of pure thought over memorization -- chess variants!
I suggest Duck Chess [0], whose inclusion of an invulnerable shared piece which each player moves on their turn (the "duck") renders most chess strategies incoherent.
A less formal alternative coalesced in university -- get a bunch of (inebriated) people together to watch the (inebriated) chess match. Their role is to loudly count down from 5. It is in this time that you must make your move, or your opponent will make it for you. The combination of time-crunch and looseheadedness prevents the recall of how castling works, let alone more complex gambits.
The article here is fairly anemic, and left me wanting for a better description of the match instead of a poor rehash of the history of computers in Chess.
I think the author needs to come up with a different title because it's highly misleading.
The author's premises are also highly exaggerated. For starters, the game of chess has not stopped evolving, because our chess engines continue to get stronger and stronger. The strongest engines of today can crush the older engines from a few years ago. This goes to show that even the elite machines haven't completely figured out chess; the smarter engines are going to continue to push the chess meta forward. In that sense, chess creativity and intuition hasn't stalled. We've just reached the point of collective knowledge that only machines can improve on chess theory.
Second, it's not like GMs are playing bad or losing moves to bluff the opponent. In most opening positions, there are at least 3 or 4 moves that could be played to still maintain winning or drawing positions. When GMs pick "suboptimal lines", they're picking maybe the 3rd or 4th best option that's still objectively a good and viable move from an engine's POV. Nobody is playing bad or losing moves on purpose, that simply does not work in chess.
The article misses how poker went through a similar transformation with solvers. Up until a few years ago, the best players used theory and principles to devise their own personal strategies. Those whose strategies were theoretically the strongest, and who executed them the best, rose to the top. Aggression always seemed to be crucial to the best strategies.
Nowadays, solvers have basically removed decisions from preflop play, but include randomization of certain moves (E.g., raise 1/3 of the time, call 50% of the time, fold 17% of the time). Then, on certain boards, you can tell the solver which sizes you want it to be available to it (e.g., 1/3 pot, 1/2 pot, 2/3 pot, full pot, 2x, etc.) and it will give you an unbeatable play style under those conditions. And so on. The game isn't solved and no human without real-time assistance would be able to play perfectly, but solvers have fundamentally changed how people prepare and play, and those who haven't studied the solutions get destroyed. An old guard has been left behind, and nerdy people who can work sims have risen to the top.
The article here explains how a similar thing has happened to chess. Like the article states, some have whined that this "removes creativity" from chess. Poker players have argued the same thing: what it takes to become the best is simply now different. You must study hard, and if you can execute what you study, you can be a great player without having a great "poker mind." In reality, the game is so complex that you still need a great poker mind to navigate the parts of the game tree which you haven't studied, just like in chess.
Undoubtedly this will change WHO decides to invest their time into becoming great chess players, and the personalities of the best players will be different. But it always struck me as sour grapes to whine that you can no longer be the best without studying the game more scientifically. Basically every game has been "moneyballed," including Chess, and people are leagues better at every game than they ever have been. We should welcome that.
> what it takes to become the best is simply now different.
Which makes it a different game. Whether that's good or bad depends on what about the games you find appealing. Personally, these changes make the affected games boring to me.
To me it just changes the source and amount of data that is backing up a strategy. Before computers if you thought a move was good you could only try it out in games with others to see if it was good or not. With computers you can run a move through thousands of games and get a better understanding of the pros and cons of a move.
GTO poker isn’t perfect play. The aim of poker is the win most money, and the perfect strategy is to exploit opponents as much as possible. By doing so you deviate from GTO play, but this would expose some weakness by playing unbalanced. Then you hope you exploit more than getting exploited.
The author is right that poker is like chess, but for the wrong reasons. Poker like chess, at the top levels is dominated by machines. Certain variants of poker, namely heads up limit poker have been "solved" for a long time. Other games such as no limit heads up are increasingly moving in that direction. It has zero to do with physical tells, instead it is mathematics. Ironically there have been accusations of cheating in live poker because a good player did NOT play like the solver suggested in a particular situation. The accusation was since they were a good player and they deviated from what the solver suggested, they must have marked the cards.
What exactly is the point of this article except the author flexing with (honestly kinda shallow) knowledge of chess engines?
Yes, the game is evolving. Two hundred years ago the queen could sometimes hop like a knight, depending where you played. Somewhere around hundred years before that, game description went from beautiful prose like "The white king commands his owne knight into the third house before his owne bishop" to much drier abbreviated form. And some decades ago computers began to become serious players.
All of these changes are interesting, worthy of discourse and not always for the better, but what is the authors actual point? Vague moral panic?
You are telling me single-minded, rote-learning based play is something that came about with engines and not with say, chess being used as a proxy fight in the Cold War or during the rise of the idea of actually treating chess players like athletes instead of calling them gamblers (like in Morphy's time)?
And how is chess like poker because of that? Because you can cheat and computers are good at it? I mean I can also use a modern tech to cheat at football, F1 racing, solitaire or coin-flipping..
Maybe someone can enlighten me what I am supposed to take away from this...
The “battle of preparedness” for grandmasters looks miserable. My experience being bad at chess is pretty cool, though.
If anyone’s looking to pick up chess, it’s a pretty good time to do it even if the people making a living out if it aren’t enjoying it anymore. It used to be that computers would only crush you without helping, but now they’re able to point at your mistakes and show you where the game shifted from one player to the other.
Chess definitely needs a shakeup, but it's unclear where it can come from. Chess960 (Fischer Random) has never really taken off.
Maybe what we need is to run it like duplicate bridge -- have tournaments where you face off against an opponent starting from a random position, that may be a winning or losing position, and see how you do against other players faced with the same setup.
This might be true for spectators and top players, but for your average player, it is fine. In fact a chess puzzle starts from a seemingly random position and your task is to find the best move. And most amateurs still play the game with such inaccuracy that there are ample opportunities for dynamic play.
I do admit though that watching top level chess is often times really boring. The moves are really predictable and at the same time really hard to understand. Blitz and Bullet games are actually far more enjoyable (when replayed on slower mode), because if a player makes an inaccuracy I can understand the punishing move (or it can be explained to me rather).
If you want a more dynamic board game though, I do recommend Go. I at least still enjoy watching top level go in a way that I don’t with chess.
If people wanted to see someone's skill at games like chess, it wouldn't be too difficult. There are plenty of historical pre-cursors to modern chess (like courier chess) or regional variations (Chinese chess). All of these are/were pretty well regarded. And that's not even getting into modern, intentionally created chess variants (like Chess960 that you mentioned). Picking a random version of chess would shake things up and force players to think on the fly.
But people who enjoy modern Western international chess seem to want to see that one single version of the game optimized. Which is fine, I suppose. Though it would be nice if we saw more alternatives (granted, there are currently some popular alternatives like blitz chess).
This is how tournaments are run between engines. Computers do not play the first 10 moves or so, which instead are prepared by humans to provide imbalanced positions where both sides can win.
Bobby Fischer solved this long ago (Fischer Random). He suggested that the back row should be randomized per game due to the increasing prevalence of computerized chess tactics.
Back in the 1970s, I was very interested in chess. When I should have been doing my topology homework, I'd instead waste time pouring over MCO (Modern Chess Openings) studying the hundreds of variations that worked or didn't worked based on historical games. It was really no fun trying to memorize these variations. It took so long setting up boards, following lines, and trying to figure out how the game strategy would change depending on an opponents moves.
For a mediocre player like myself, the advent of chess software that understands the game at a deep GM level has made studying chess fun. I can play computer opponents over and over, rapidly trying out the variations and seeing the outcome. I can ask for help, I can see what I'm doing wrong. I can easily spend an hour solving chess puzzles that improve my tactics.
To me, really not a good player, I enjoy chess more than ever; the article seems not to understand how chess software impacts the game.
What you're describing is an interesting outcome of chess engines. I assume that the game at the top of the heap, at the grandmaster level, has certainly become dull and void of creativity, as the article describes. But for us mere mortals, I wonder if the game of chess has become more exciting, as the concepts you can reach now were just not achievable for amateurs studying from books, as in the old days.
The game of chess has no doubt changed because of chess engines. And maybe for the negative at the top end. But I think you're right, it's probably a lot more positive change in the middle. I have long since known that I am not capable of beating a chess engine in regular play, but that doesn't mean I don't benefit from the loss.
This is one of those articles that treats computers as if they were hyperintelligent entities that programmed themselves rather than just simple computational machines following rules created by teams of people that have studied the game for decades.
Also it sounds like should replace all of my servers with powerful shoe computers to take advantage of the "upgrades in commercial hardware" that they mentioned.
This misses the key property of such systems that make them remarkable. The computation involved begets new behavioral patterns that weren't enumerated by the initial program statements. The behavior is not just a representation of the human knowledge imparted to the system, but an emergent dynamic that is unpredictable from an analysis of the initial rules. It is a common mistake to dismiss the computational part of a system as carrying no informative or meaningful content. Just bracketing `the computation` as transparent and explanatorily insignificant is to miss a large part of the substance of the system.
Exactly, dynamics and algorithm is content as well. 1 example: the heuristics you pick for something can be more powerful than the explicit knowledge required to create the heuristic.
Despite varying opinions on the article, most seem to be agreeing here that one thing that is making chess more boring is the memorization of moves and tricks. I'd like to offer a counter-point.
This doesn't seem to take into account that many people enjoy learning and memorizing sequences and tricks. It's actually fun to learn patterns and sequences that have clever underpinnings and lead to surprising results.
As another commenter said, there exist many variants exactly for the purpose of getting back to pure logic-level reasoning about positions. If that's what you want, go for it. Chess 960 for example is explicitly designed for that.
But despite the existence of all these variants and other games, people still predominantly continue to play 'Chess', with its classic starting position and rules, and I would suggest that one of the things that keeps drawing people back is that there is sufficient "space" in the game for life-long learning, while providing a common and stable "interface" that allows everyone to exchange knowledge and build up a shared experience of learning and mastery.
That on top of this there still most definitely exists a layer of strategy and "principles", not just memorization, is another aspect that will always draw people back to the game. A lot of these principles that have been developed and taught for hundreds of years go out the window with variants -- which is fine, it's fun to discover new games too, but sometimes you want to benefit from the knowledge and wisdom of experts, and the only way to do that is to play the same game as them.
> But despite the existence of all these variants and other games, people still predominantly continue to play 'Chess', with its classic starting position and rules, and I would suggest that one of the things that keeps drawing people back is that there is sufficient "space" in the game for life-long learning, while providing a common and stable "interface" that allows everyone to exchange knowledge and build up a shared experience of learning and mastery.
Very well said. This is exactly why I prefer traditional Chess. I really enjoy the rich platform I can dig into for years and years. I'm not likely to ever pursue becoming a GM and I don't mind.
What's fascinating to me as someone who has followed the game for a while is the evolution of engines and how their play has influenced human play. Stockfish versus AlphaZero was fascinating only because it showed that at the highest of highest levels of chess (3000+ ELO) - there are ways to trick engines and employ "human" strategies (for example, there was a game where Alpha "blundered": it sacrificed 2 pawns and ceded the center of the board to push a wing pawn that 25 moves later ended up blocking Stockfish's kingside development entirely). Incredible, but the flip side of this is realizing that traditional chess is essentially a solved problem to some degree.
From what I can understand, chess at the highest of human levels is just using an engine to find tricks and traps 30+ moves into a given match. Anywhere about 2000+ rating or above, most people will play a relatively fixed number of openings that have been explored to death time and time again. It's now about psychological hacks that can be be backed up by deep engine analysis. "If at move 45 in a Sicilian Dragon game I play sub-optimal move X instead of optimal move Y, my opponent might make a sub-optimal response which will open up a specific square or hang a piece 10 turns later according to AlphaZero or Stockfish".
I can see why Magnus is kind of bored by it.
To me, Fisher Random (Chess960) solves this by eliminating opening theory almost entirely. Crazyhouse Chess is also a blast, but it too is hampered by theory as there are fewer competent opening moves than standard chess and there's even less room for creativity. Crazyhouse960 is by far my favorite variation of chess, because there's no endgame to solve towards and little to none opening theory to draw from. Shame nobody plays it these days - if you're so inclined, come join over at PyChess.org and play a few rounds!
When I think about it, chess is a game where accusations of cheating are regular. Most recently there is the Neimann scandal. There was a lot of cheating accusations for Kramnik vs Topalov as I recall. Kramnik was accused of using Fritz (predecessor to today's stronger engines), and Topalov's manager was doing statistical analysis of Kramnik's move choice and the engine's. When Kasparov lost to Deep Blue, he was very salty and accused IBM of cheating. And those are only a sampling of modern examples. It's hardly a game of cold objective analysis. They've got a lot of emotions these chess players, which makes sense. It gives me solace when I get overly wrought / emotional around my own intellectual pursuits.
>When Kasparov lost to Deep Blue, he was very salty and accused IBM of cheating
The accusation was that IBM made modifications to Deep Blue between games. That's not cheating but then it's not exactly Deep Blue vs Kasparov anymore.
There was a somewhat veiled accusation that there was human intervention in game 2 (as in, a strong human player interceded to influence or outright override the computer's move). Funny that in those days the cheating accusation involved a human behind the scenes instead of a computer.
Chess grandmasters are not the most balanced human beings.
It’s a very small amount of people who have decided to consecrate an inordinate amount of time completely mastering a game and whose main focus in life is playing it competitively against each other.
It’s a lot like elite athletes but more nerdy and with longer career.
Generally both sides play to win, but if you play to draw instead (by playing 'sub-optimal' moves etc) you paradoxically have a better chance of winning.
219 comments
[ 2.6 ms ] story [ 256 ms ] threadIn the highest levels of chess, psychological fortitude and physical fitness are also important.
The whole idea in poker is to make the most right plays that on average should get you the win. Its management to chances.
There are no hidden information in chess, if you play better moves you will win.
And while there are factors that affect selection of best chess moves, you can hardly claim that the other guy got lucky.
While poker does have a significant element of chance in a given hand, it's no accident that top players consistently make more money over the long run.
Knowing how much to bet based on your odds of winning a hand, figuring out what your opponent's hands could be based on how they're betting, etc. all take skill. There's a reason poker AI can consistently beat humans over the long run, a feat that would be impossible if poker was all about chance.
In fact, poker AI was a difficult area to create software in precisely because the game is one of imperfect information.
That's a good point about the cash game format.
But tournament format often forces players into multiple all-ins over the course of the tournament. This leads to a much more luck based outcome.
Has this actually been verified, at least in the modern era (last 10 or so years)? Yes, you see the same guys on TV and to a lesser extent, final-tabling multiple tournaments. But these guys are often on TV because they are entertaining, not necessarily great (or consistently-winning) players. They typically all have external sources of income (sponsorships, poker training, commentary, casino ambassador, company owners, etc) and most reduce volatility by buying and trading pieces of each others' action. In addition, for tournaments, some guys are able to simply enter way more events, which naturally leads to more final tables, but you're not seeing all the losses and bust outs.
All I'm saying is that it's not the same as the 70s through 90s where the skill gap was so huge that yes, certain players could (especially by going on the road and playing private games) absolutely dominate. I think winning consistently today is more of an illusion and some short-term luck.
Winning consistently today relies on being as close as you can to this "game-theory optimal" strategy. As skill gaps decrease chance certainly plays a larger role, but AI analysis of poker has indicated that there are ways to improve your odds and gain an edge.
Maybe you missed it. Humans are now using deceit to trick opponents trained by AI. While they may not "look at" their opponent, they are using similar strategies deployed in poker such as bluffing/misdirection/cloaking etc.
If this were true, then the recurrent Pocker champions must be the luckiest persons in the world, luck is undoubtedly an element, but there are many ways to influence other variables of the game.
"In that context, cheating scandals may be nothing less than a natural step in chess’s evolution. Poker, after all, has been rocked by allegations of foul play for years, including cases where players are accused of getting help from artificial intelligence. When the highest form of creativity is outfoxing your opponent—as has always been true of poker—breaking rules seems only natural."
I'm ruling this "clickbait" because while the core idea is reasonable, that is clearly a very weak way of phrasing it solely done to be in service of a clickbait title. Without the need for clickbait it's trivial to phrase that in a better way.
Poker on the other hand, by being random, means that you can mindlessly play the mathematically optimal strategy. Eventually your opponent will be put in a tough spot by pure randomness.
The difference is that chess is deterministic and poker is not. Chess skill means being able to forcefully convert a winning or drawing position. Poker skill means optimizing your chances of winning.
Not only is it possible to beat weaker opponents in chess by playing the best moves (contrary to your assertion), it happens all the time over the board. Weaker players are more likely to play moves that transforms the chess game from a win/draw to a draw/loss for them. There isn't a single human or engine alive that can consistently secure a draw against every opponent.
And it's possible for a poker player to mindlessly play the mathematically optimal move and still lose. Poker IS a game a chance, so they could just get unlucky.
It is decidedly not skill.
Physical "tells" are largely a myth in poker, particularly in high-level poker. Please don't speculate wildly on things with at most a novice-level understanding.
fyi poker gets mentioned in last paragraph and is barely even connected to chess.
ie, you can use AI to cheat in chess and poker, and there have been cheating scandals in both... therefore chess is poker now <question mark>
https://incoherency.co.uk/blog/stories/sockfish.html
The article is a bit silly. I had the feeling reading it that the writer has little understanding of how classical chess is played, nor of poker for what it’s worth.
This is so much more efficient that the only limitation is on how much energy and capacity the player has for rote memorization, rather than how much time it takes for the team of humans to work out the best lines.
Sure, computer analysis has marginally changed how some openings are viewed and sure you can now get an engine evaluation of all the variations of some obscure line in a matter of seconds but novelties remain rare at the top level. Stockfisch is extremely strong because it can play near perfect positional play and can establish micro-advantage through deep calculation not because it fundamentally changed how chess theory is viewed.
AlphaZero might be the sole exception to my point. It did indeed show that modern chess was neglecting some strategical concept but to be honest it was more a rehabilitation of old ideas than a pure novelty.
But nowadays Stockfish and Leela have both caught up and surpassed AlphaZero.
The only two instances I've seen an actual 'supercomputer' mentioned in recent Chess news has been: * Ian Nepomniachtchi using Zhores (which is questionable given that the institute which runs it [Skolkovo] has Arkady Dvorkovich [chairman of FIDE] as it's chairman). I remember researchers questioning if Zhores was even suitable for the type of calculations which engines care about, so this looks like some PR move more than anything. * There was a rumour a few years ago that Magnus Carlsen had some secret/private access to a Norwegian supercomputer. No real confirmation of this rumour occurred.
Both AlphaZero and Leela Zero (by their design and very optimal evaluation functions) also require much less evaluation - so they can naturally evaluate further.
> Elite chess has always involved rote learning, but “the amount of stuff you need to prepare, the amount of stuff you need to remember, has just exploded,” Sadler said.
> Engines can calculate positions far more accurately and rapidly than humans, so there’s more material to be studied than ever before.
Two sentences later you have this gem: “What once seemed magical became calculable; where one could rely on intuition came to require rigorous memorization and training with a machine. Chess, once poetic and philosophical, was acquiring elements of a spelling bee: a battle of preparation, a measure of hours invested.”
Well, that’s patently untrue. Chess has always been a battle of preparation far before the advent of computer. The rosy paste described just doesn’t exist. High level play has always required memorising books of theory and going through decades of past games. That’s what chess is. It’s a game of pattern recognition and memorisation.
If I wanted to be provocative, I would say that the article seems to imply that computers have turned chess from an interesting game into a boring one while in actuality it has always been boring but with more mystic.
The dominance of computers has made human creativity redundant; optimal strategies for beating other humans ironically involve suboptimal moves that may have a psychological dimension, similar to poker. [This isn't actually a remotely new strategy though computers have influenced human play.]
Although the article doesn't directly reference it, Hans Niemann in his emotional interview reacting to the unevidenced accusations of him cheating against Carlsen pointed out that one "weak" move against another opponent in the tournament had been chosen by him specifically because of it's psychological pressure.
Update: Niemann interview link, as requested:
https://youtu.be/8NQF60RT0b4&t=3m48s
https://youtu.be/8NQF60RT0b4&t=3m48s
Is this the level of analysis expected from a grandmaster?
I guess someone needs to write an article Poker is Just Poker Now
Nobody plays perfectly. Engine prep only goes so far. The draw rates in chess are in part due to both sides being human; many opportunities are missed. An engine will tear apart any human even from an inferior position, so there's plenty of play in that margin
You need 'subterfuge and misdirection' in the sense that (as the world no-1 puts it), it's a semi-bluff...a weird/not-really-analysed position, but that will still end up in a draw if the opponent responds appropriately.
Lex: "Is there a sense in which it's ok to make sub-optimal moves?"
Magnus: "You HAVE to, because the best moves have been analysed to death, mostly."
https://youtu.be/0ZO28NtkwwQ?t=1450
This is true of every sport, every single record always gets broken, because techniques and nutrition and training and everything improve. A world no 100 tennis player could easily beat a world no 1 from 50 years ago.
How is that bad or like poker? It just means the game is evolving. It might also mean more psychological games. The tennis analogy works here too, there's plenty of that in there too.
The whole premise just doesn't make sense to me. Because you have to psychologically think about your opponent, that makes the game worse?
Just that you are forced to make suboptimal moves. If you make optimal moves then your opponent will have analysed the position before and the game will end in a draw, which is a dissapointing result for whoever plays white.
Not necessarily. If black has a higher ELO rating then white is happy to draw and may aggressively seek a draw.
I think the problem is that seeking draws makes chess boring. And that it’s realistic for someone ranked well below another player to force a draw, while it’s near impossible for the lower ranked player to win outright. So you end up with a wide range of players seeking stalemates and a small minority looking for a win, but not at the expense of a potential loss.
If 'perfect' play--that we're approaching with engine analysis--results in a draw, you need to do something non-optimal and unexpected in order to get your opponent out of their engine prep and into thinking mode.
Whether it's better or worse is a different argument. Some find it a bit 'dry' in that there are often less blunders, dazzling tactics, and sacrifices because both sides now know the optimal approach. It's much more of a war of attrition in high-level, long time-control games. As an aside, that's potentially why bullet and other such chess is so popular...there isn't time to deeply analyse, so it's intuition, challenging positions, blunders, and less 'standard' play.
Not quite true; you can break ELO records by performing equally well in a larger pool of people-with-ELO-ratings.
> An eephus pitch in baseball is a very high-arcing off-speed pitch. The delivery from the pitcher has very low velocity and often catches the hitter off-guard.
The pitcher essentially lobs the ball at the hitter [1]. Major league players are so focused on hitting extremely fast and/or curving pitches that they whiff or stand idle at a pitch that your ten-year-old nephew could probably clobber. It's a risky pitch, because if the batter is able to adapt, they can hit a home run. The power of the pitch is entirely that it's unexpected.
I know it's not a perfect analogy, since chess players have so much more time to react to any sub-optimal moves.
[0] https://en.wikipedia.org/wiki/Eephus_pitch [1] https://www.youtube.com/watch?v=ikLlRT2j7EQ
Incoming speed makes significant difference to exit velocity.
If you were going for max distance, you would want 100 mph pitches.
In home run derby, they are optimizing for ability to get a good repeatable swing with enough power to go over the fence. If they were playing in a park with like 440 ft fence, they would need faster pitches.
The distance the ball flies has a lot more to do with bat speed than ball speed (something like 5mph for 1mph, if I recall correctly).
What you want to do is perhaps start with this random strategy, determine the realtive strength of your opponent, and only move to an exploiting strategy if they are weaker than you. If they throw Rock 40% of the time, you should throw Paper 40% of the time. If you were to enter a RPS tournament playing randomly, it is unlikely that you would reach the end. You need to exploit.
In Poker Cash Games, people just play the Nash and when opponents make flat-out mistakes they yield a little bit of profit. Do this enough with high enough stakes and it works out, even if its extremely boring. But this won't work in Poker Tournaments, where you only get payouts for top 10%, weighted heavily towards winning. So what happens in Poker is what happens in RPS - you attempt to find exploits, and your opponents attempt to find exploits in your play. In turn each of the players are attempting to move the environment of the play towards a position where they know it well and can find those mistakes.
If deviating from the Nash yields better results than the penalty of deviating because in that environment you are stronger than your opponent, you should take it. As this article points out, the same is true in Chess. And as you point out, this is also true of the Eephus pitch.
When your opponent is the one doing the analysis. It makes perfect sense to me that anything your opponent is prepared for is less optimal than it would be otherwise.
Move A.
Move B.
From the best knowledge we have from our best engines, A is the better move (though I don't think we have formally proven such yet).
But if you play A, you are playing your opponent in a game of memorization.
If you play B, your opponent loses any memorization advantage. Thus they must play based on ability other than memorizing responses.
For a player whose memorization ability is equal or worse than their ability to play based on other ability, choosing move B is clearly the worse option. But if the player is one whose memorization is better than their other abilities at chess, then playing B means fighting them in an area they are worse at. The advantage of this can easily outweigh the disadvantage of move B compared to A in general.
Imagine a military commander choosing an engagement over terrain more familiar to their side even though it is a worse engagement if both sides had equal knowledge of the terrain, because the terrain knowledge more than compensates for the negatives.
Hans' explanation made sense, even if he was wrong in his first interview and it came off weird.
I think that’s what Magnus is saying is that if you play by the book you go from opening to end game without much creativity or challenge.
So, the top “creative” players will try to go off script to make their top “book smart” opponents have to think on the fly.
But even in mid-game, it might mean that if you know that your opponent prefers open / tactical positions, for example, you try to force a closed position, even if it may be slightly suboptimal.
This goes at least back to Bobby Fischer. With the minor difference that instead of computers he was up against teams of Soviet Grandmasters. It's why he created Fischer Random chess where the pieces on the backrow are randomized. That (mostly) eliminates memorization and forces players to rely on their own creativity.
There was probably not a more booked-up player than Fischer in his prime, something he seemed to try to make people forget. Famously when asked about a young Soviet prodigy by the Soviets, Fischer responded that the kid was good, but that there was a better player in their women's division who had more signs of promise. Many of the Soviet masters had no idea who Bobby was talking about; but Bobby was reading all Soviet chess literature and magazines to stay up to date.
That said, FRC is pretty interesting.
That's what parent said. Fischer wanted to get the memorization out, since memorization makes the game unplayable for someone who doesn't have The Knowledge.
The board state is always known. Magnus is saying that at the top level, grandmasters know how to force draws in all of the lines that incorporate best-move, second-best-move, etc. That means that if you want to pull off a win, you have to incorporate unstudied lines that will force your opponent to calculate rather than draw from memory/experience.
This is why playing against a maniac (seemingly random and not respecting bluffs or equity) is very difficult because if they get lucky enough times, they are able to "break the game" by getting the opponent to be extremely risk averse OR take on more risks.
I believe this is what Magnus is referring to, its that making yourself unpredictable by questionable moves and no longer playing in a way that has been taught.
For example the common strategy is to go all in with strong pairs like AA, KK but someone beats it with a totally random garbage hand (52o, 37o) and does so repeatedly, no theory can help you win against somebody who is just repeatedly lucky and brash.
This is less about playing mindtricks and bluffing, but more 'mundane'.
I.e. a decent player will know all the good mainlines of popular openings, and end up in 'comfortable' positions (among other things due to computer analysis).
The metagame is to prepare a non-garbage sideline, that your opponent is not so familiar with. Nobody at a high rating plays 'questionable' variations on purpose, in order to bluff. The resulting positions would be much too punishing.
Post-game thoughts from the players:
Inteviewer: "Let me pull you back; so you didn't understand the position, and so you still felt like you were scared to go into a piece-up situation?"
Alireza: "Yeah so, I just trusted him. (he shrugs) I just wanted to make a move.. and play a bit more you know (laughing)"
https://www.youtube.com/watch?v=xT9orNSgk5w&t=55s
Niemann: "You have to understand, when I play the move Qg3, this is a purely pyschological move."
https://www.youtube.com/watch?v=CJZuT-_kij0&t=780s
Whether or not Niemann did cheat over the board, this shows that even top players can be susceptible to this kind of bluff.
But if you make a suboptimal move, then you are going down a path that is less analyzed and thus you are likely to have an analytical advantage.
So while the board state is well known, what your opponent has researched is not known, or partially known (since players have favorites and predispositions).
So you're taking a game where your opponent has the upper hand, and you're playing highly suboptimal moves with the intent of putting your opponent in a position where they have to pick between a rock and a hard place:
- either they have to sit quietly and knowingly take your straight-up lies, or
- they are forced to play suboptimal moves themselves to call your bluffs.
- they see through your bluff, and punish you for it
So, it is a risky endeavor, since you don't know the mind of your opponent.
And so it makes sens to only take such risks when your opponent has the upper hand, as you mention.
But then you become predictable — when you behave erratically, people know you are in your death throes, and not a real threat.
So, you have to mix some erratic behavior in when you are in a strong position, too, to keep them guessing. But then you are weakening your game overall...
And on and on it goes (:
It's like an arms race, but circular rather than one-directional.
Both games rely on disguising the real play you want to make. Slow playing the best hand in poker, or making non-standard moves in chess to throw your opponent off.
You see this in all sports/games. In basketball/football a team will run their best player away from the real play. In Jiu-Jitsu I'll fake a choke I have no intention of seriously trying in order to force someone to defend and move their arm - which is what I wanted the whole time.
It's amusing how people feel bluffing in poker is the human element, but top poker AIs bluff more frequently than most humans
Good bluffs are when your range beats your opponent's range. Which gets back to your point, which is that you can do some pretty awful things to someone when your hand is outside the range they're putting you on
He was notorious for doing just this. He would make what have now been engine analyzed as sub-optimal moves but which lead to such complex and dynamic positions that his opponent would eventually slip and he would exploit their error in dramatic fashion. This is one reason why he is considered one of the most creative attacking players ever.
You still see the general creative spirit in certain top players like Rapport, Mamedyarov, Firouzja, though. Magnus is particularly boring to watch, but it's what makes him great. Fischer was not so different, and felt that Tal was overly creative for the sake of being creative, rather than good.
IM Marc Esserman is probably one of the more exciting players to watch, but he's not a GM, though he arguably plays at GM strength often. He's the author of Mayhem in the Morra, well worth reading if you like Tal's spirit. Esserman is inactive OTB unfortunately, but he does stream chess online.
I don't like bullet, but 15-minute and 3-5 minute chess is wonderful to watch.
This strategy worked really well against other students' bots, but not terribly well against the students themselves. I'd be interested to see if it worked better/worse against pro players who think more algorithmically/analytically.
The bot hit a bug that chose a strange move, and since kasparov couldn't figure out what the line was, he resigned. If he analysed the board instead of what the bot was thinking, he would have won easily
The bug move (a rook) was in game 1. It was inconsequential to the game result. The position was already lost for Deep Blue. It made a random move because everything had the same outcome of losing. The bugged move simply made it lose sooner. Kasparov knew this and there was no controversy.
In game 2, Kasparov set a trap, offering a sacrifice of two pawns for great positional improvement. Kasparov thought Deep Blue couldn't calculate far enough ahead to see the positional improvement, but it did. This made Kasparov suspect and accuse IBM of human intervention in the computer's moves.
The fatal move (a knight sacrifice by Deep Blue) was in game 6. Kasparov deliberately played an opening with a known weakness, that he thought the computer wouldn't find as it wouldn't sacrifice material and calculate far enough ahead to see the positional advantage. Deep Blue found it. Kasparov again accused IBM of human intervention. The Deep Blue team then said they had added this variation into its opening database just before that game, pre-analyzing that line deeper since Kasparov had previously played it in game 4. There was some dispute as to whether the rules for the competition were intended to allow IBM to modify Deep Blue during the match.
There’s very seldom a single, obvious move that makes sense. The level of chaos in chess is still sufficiently high to make it an interesting game, even when you study all the available opening theory.
Who says it's bad that you now have to "engage in subterfuge, misdirection, and other psychological techniques"? Surely that just makes the game into a real sport and more exciting for viewers. Besides - hasn't the psychological aspect always been a part of chess? Maybe the most interesting aspect?
(Of course, there is no evidence of cheating, but arguably Carlsen's unsporting conduct is also the dark side. But anyway, none of this is new to high level chess.)
This reminds me of something I think I came up with in Blitz and Bullet games a decade ago and coined the name: a “time sac”.
It is when in the middlegame I make a dubious sacrifice (positional or material) — some possibly cool combination or a move that makes the opponent think in time pressure. It is not a bluff because it MIGHT work out, but I am not sure.
This works mostly in games without an increment. The idea is to make them get much lower on the clock than me, so even if I end up with a slightly worse position or down material, I can easily hold it until I win on time.
My opponents in chess, even over 2100 often got nerd swiped with this! Hahahaha
https://gamedesignadvance.com/?p=2346 https://www.sirlin.net/articles/solvability
Now whether chess really IS more about calculation these days is a difficult question to answer. It might just be that it values a different kind of calculation or skill than the past. Given the deep preparation necessary for someone to be a chess professional, I'm biased towards (micro-edit: believing that...) more calculation _is_ necessary for high level performance, and psychology is more gimmicky.
The difference is that poker isn't a game of complete information, chess is. This means that, technically, there is a _right_ answer in chess. With the ambiguity of poker, and other games, that leaves room for psychology to be more natural.
I don’t think those comments from Magnus support the point in this article at all really.
The best games have bluffing, IMO.
Whereas Chess, even AI chess, is far from solved.
Also at the most obvious level, poker is random, the strategies are over a long series of deals. Cepheus won't ever magically find a way to make its 2-7 off work against your pair of Aces, you will always make that too expensive to continue, but it wins overall because next hand it's not going to pick up garbage. Whereas every Chess game begins the same, so a hypothetical solver would just always beat you.
Not deceptive playing tactics, deceptive metagame strategy. Trick your opponent into preparing against lines you never play and you gain an advantage. Leverage transpositions to play the openings you are most comfortable with but trick your opponent with the unexpected move order.
It's all about getting your opponent out of book ASAP. Do that and you can nullify any advantage they may have had from preparing against every game you've ever played using an engine to explore the lines.
So remember with its pronouns, it/its/its. Careful with apostrophes, don't want to insult it do you?
Deep down they all know there is something amiss, it is instantly visible when something cannot be generated. That gets flagged, and accused of being a bot to get reactions. Yeah, this is my reaction. Train on this shit. Very genuine, convincing, authentic, comic relief, eh. Difficult to plagiarize in fact, or I would say nothing.
It actually is analogous to poker but not really poker as most people know it. It's analogous because poker is also done by training with engines; and you often want to make moves you know are suboptimal from your training because you believe it will exploit the opponent's suboptimal play.
Essentially, it telegraphed too much information to the chess engine to analyze. Unlike humans, chess AI isn't intimidated by aggressive play, there is no psychological edge in doing so, and only an informational disadvantage.
Kasparov changed his strategy to a more passive and defensive one, and was able to win some matches after that. Someone who follows chess more closely than I do may need to confirm or provide more info, my memory of this is a little hazy.
But I'm not sure the analogy to poker is apt, because in chess both players can see all the "cards" at all times and all the moves and decisions in real-time. The unknown is which player can see more steps ahead, and exploit that knowledge to gain an edge before their opponent sees it too.
* Round 1: negotiate the starting position, aka 'and now we have a new game of chess'
* Round 2: play chess using prep. generally only one side fully benefits during this phase.
* Round 3: play chess
Computers mostly impact human chess in round 2. It's not just as if you had every human super GM working for you for X years, X as large as you want; the ideas from computer analysis go beyond that. The role of humans in round 2 is looking at computer lines and deciding if the ideas might be transferred to human chess.
Computers do not seem to inform/impact the game of poker as much, absent cheating. Computer cheating, though, where you are aware of the computer's complete evaluation of a position, would seem to be less impactful w/r/t poker than with chess. A computer (eg: Stockfish) crushes a super GM without mercy. So I have issues with saying 'X is just Y' now, because it seems to be different on both ends.
If you decide that the three round structure above is problematic, Fischer 960 dramatically changes rounds 1-2. But that's a different game.
It's possible to return chess to that realm of pure thought over memorization -- chess variants!
I suggest Duck Chess [0], whose inclusion of an invulnerable shared piece which each player moves on their turn (the "duck") renders most chess strategies incoherent.
A less formal alternative coalesced in university -- get a bunch of (inebriated) people together to watch the (inebriated) chess match. Their role is to loudly count down from 5. It is in this time that you must make your move, or your opponent will make it for you. The combination of time-crunch and looseheadedness prevents the recall of how castling works, let alone more complex gambits.
[0] https://duckchess.com/
Well done parody. I'm laughing
This is more what I wanted from the article: https://en.chessbase.com/post/the-carlsen-niemann-affair
The author's premises are also highly exaggerated. For starters, the game of chess has not stopped evolving, because our chess engines continue to get stronger and stronger. The strongest engines of today can crush the older engines from a few years ago. This goes to show that even the elite machines haven't completely figured out chess; the smarter engines are going to continue to push the chess meta forward. In that sense, chess creativity and intuition hasn't stalled. We've just reached the point of collective knowledge that only machines can improve on chess theory.
Second, it's not like GMs are playing bad or losing moves to bluff the opponent. In most opening positions, there are at least 3 or 4 moves that could be played to still maintain winning or drawing positions. When GMs pick "suboptimal lines", they're picking maybe the 3rd or 4th best option that's still objectively a good and viable move from an engine's POV. Nobody is playing bad or losing moves on purpose, that simply does not work in chess.
Nowadays, solvers have basically removed decisions from preflop play, but include randomization of certain moves (E.g., raise 1/3 of the time, call 50% of the time, fold 17% of the time). Then, on certain boards, you can tell the solver which sizes you want it to be available to it (e.g., 1/3 pot, 1/2 pot, 2/3 pot, full pot, 2x, etc.) and it will give you an unbeatable play style under those conditions. And so on. The game isn't solved and no human without real-time assistance would be able to play perfectly, but solvers have fundamentally changed how people prepare and play, and those who haven't studied the solutions get destroyed. An old guard has been left behind, and nerdy people who can work sims have risen to the top.
The article here explains how a similar thing has happened to chess. Like the article states, some have whined that this "removes creativity" from chess. Poker players have argued the same thing: what it takes to become the best is simply now different. You must study hard, and if you can execute what you study, you can be a great player without having a great "poker mind." In reality, the game is so complex that you still need a great poker mind to navigate the parts of the game tree which you haven't studied, just like in chess.
Undoubtedly this will change WHO decides to invest their time into becoming great chess players, and the personalities of the best players will be different. But it always struck me as sour grapes to whine that you can no longer be the best without studying the game more scientifically. Basically every game has been "moneyballed," including Chess, and people are leagues better at every game than they ever have been. We should welcome that.
Which makes it a different game. Whether that's good or bad depends on what about the games you find appealing. Personally, these changes make the affected games boring to me.
Yes, the game is evolving. Two hundred years ago the queen could sometimes hop like a knight, depending where you played. Somewhere around hundred years before that, game description went from beautiful prose like "The white king commands his owne knight into the third house before his owne bishop" to much drier abbreviated form. And some decades ago computers began to become serious players.
All of these changes are interesting, worthy of discourse and not always for the better, but what is the authors actual point? Vague moral panic?
You are telling me single-minded, rote-learning based play is something that came about with engines and not with say, chess being used as a proxy fight in the Cold War or during the rise of the idea of actually treating chess players like athletes instead of calling them gamblers (like in Morphy's time)?
And how is chess like poker because of that? Because you can cheat and computers are good at it? I mean I can also use a modern tech to cheat at football, F1 racing, solitaire or coin-flipping..
Maybe someone can enlighten me what I am supposed to take away from this...
If anyone’s looking to pick up chess, it’s a pretty good time to do it even if the people making a living out if it aren’t enjoying it anymore. It used to be that computers would only crush you without helping, but now they’re able to point at your mistakes and show you where the game shifted from one player to the other.
Maybe what we need is to run it like duplicate bridge -- have tournaments where you face off against an opponent starting from a random position, that may be a winning or losing position, and see how you do against other players faced with the same setup.
I do admit though that watching top level chess is often times really boring. The moves are really predictable and at the same time really hard to understand. Blitz and Bullet games are actually far more enjoyable (when replayed on slower mode), because if a player makes an inaccuracy I can understand the punishing move (or it can be explained to me rather).
If you want a more dynamic board game though, I do recommend Go. I at least still enjoy watching top level go in a way that I don’t with chess.
But people who enjoy modern Western international chess seem to want to see that one single version of the game optimized. Which is fine, I suppose. Though it would be nice if we saw more alternatives (granted, there are currently some popular alternatives like blitz chess).
For a mediocre player like myself, the advent of chess software that understands the game at a deep GM level has made studying chess fun. I can play computer opponents over and over, rapidly trying out the variations and seeing the outcome. I can ask for help, I can see what I'm doing wrong. I can easily spend an hour solving chess puzzles that improve my tactics.
To me, really not a good player, I enjoy chess more than ever; the article seems not to understand how chess software impacts the game.
The game of chess has no doubt changed because of chess engines. And maybe for the negative at the top end. But I think you're right, it's probably a lot more positive change in the middle. I have long since known that I am not capable of beating a chess engine in regular play, but that doesn't mean I don't benefit from the loss.
Must be hard for a professional if you not only compete against human creativity but sheer computing power.
Just like Dall-E, it's fun if you can't paint not so much if it's your source of income.
Also it sounds like should replace all of my servers with powerful shoe computers to take advantage of the "upgrades in commercial hardware" that they mentioned.
This doesn't seem to take into account that many people enjoy learning and memorizing sequences and tricks. It's actually fun to learn patterns and sequences that have clever underpinnings and lead to surprising results.
As another commenter said, there exist many variants exactly for the purpose of getting back to pure logic-level reasoning about positions. If that's what you want, go for it. Chess 960 for example is explicitly designed for that.
But despite the existence of all these variants and other games, people still predominantly continue to play 'Chess', with its classic starting position and rules, and I would suggest that one of the things that keeps drawing people back is that there is sufficient "space" in the game for life-long learning, while providing a common and stable "interface" that allows everyone to exchange knowledge and build up a shared experience of learning and mastery.
That on top of this there still most definitely exists a layer of strategy and "principles", not just memorization, is another aspect that will always draw people back to the game. A lot of these principles that have been developed and taught for hundreds of years go out the window with variants -- which is fine, it's fun to discover new games too, but sometimes you want to benefit from the knowledge and wisdom of experts, and the only way to do that is to play the same game as them.
Very well said. This is exactly why I prefer traditional Chess. I really enjoy the rich platform I can dig into for years and years. I'm not likely to ever pursue becoming a GM and I don't mind.
From what I can understand, chess at the highest of human levels is just using an engine to find tricks and traps 30+ moves into a given match. Anywhere about 2000+ rating or above, most people will play a relatively fixed number of openings that have been explored to death time and time again. It's now about psychological hacks that can be be backed up by deep engine analysis. "If at move 45 in a Sicilian Dragon game I play sub-optimal move X instead of optimal move Y, my opponent might make a sub-optimal response which will open up a specific square or hang a piece 10 turns later according to AlphaZero or Stockfish".
I can see why Magnus is kind of bored by it.
To me, Fisher Random (Chess960) solves this by eliminating opening theory almost entirely. Crazyhouse Chess is also a blast, but it too is hampered by theory as there are fewer competent opening moves than standard chess and there's even less room for creativity. Crazyhouse960 is by far my favorite variation of chess, because there's no endgame to solve towards and little to none opening theory to draw from. Shame nobody plays it these days - if you're so inclined, come join over at PyChess.org and play a few rounds!
The accusation was that IBM made modifications to Deep Blue between games. That's not cheating but then it's not exactly Deep Blue vs Kasparov anymore.
It’s a very small amount of people who have decided to consecrate an inordinate amount of time completely mastering a game and whose main focus in life is playing it competitively against each other.
It’s a lot like elite athletes but more nerdy and with longer career.
Generally both sides play to win, but if you play to draw instead (by playing 'sub-optimal' moves etc) you paradoxically have a better chance of winning.