>game is technically "solved" in the sense that there's almost always a correct way to play, but the skill ceiling is still unfathomably high
Vidya like this?
CRIME Shirt $21.68 |
>game is technically "solved" in the sense that there's almost always a correct way to play, but the skill ceiling is still unfathomably high
Vidya like this?
CRIME Shirt $21.68 |
Brood war
Classic WoW
HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA HA
Bot game for bot people
Chess is not a solved game, technically or otherwise.
You don't know what "solved game" means.
so I shouldn't feel dumb for not always knowing the best move? even grandmasters and super computers don't?
Knowing the best move is not what solved game means.
even computers can only estimate what the best move is, only positions with 8 or less pieces on the board have been solved by supercomputers
That being said, Stockfish running on contemporary high-end hardware and long time controls could probably play against God (hypothetical 32-man tablebase) from the regular starting position and probably manage to draw most games. That you can't prove through exhaustive search or as a theorem 1. d4 is a drawing move (and that there are no winning moves), doesn't mean it isn't one among the equally good moves.
To be fair, most of those victories are from busted meme openings they are forced to play, such that either both engines win on their turn to play B/W, or that starting evaluation is at the precipice of being objectively convertable and one engine manages to win and then barely hold when defending, while engines playing the regular starting position would almost certainly just result in a draw. That being said, the top engines do from time to time (although very rarely) take games off each other from even positions as well. Clearly, they are by no means perfect, even if they're overwhelmingly superhuman.
>meme openings
a strat that works well is the opposite of a meme
In this case, it's a meme opening because both players have their opening dictated by the tournament organizers in order to make the games more interesting. If either player were unshackled, the meme opening would never occur.
Engines left to their own devices would play standard openings. What engine tournaments do to make for better games and more decisive results is make them play a new opening in each match (each player taking turns playing as White and as Black to make it fair), and some of them are quite standard, others not so much. For example, at the moment of writing this, the ongoing Computer Chess Championship game started from move 9 after Caro-Kann Defense: Advance, Tal Variation. The engines didn't pick the first 8 moves. Ideally, the tournament-organizers tend to aim to pick poor openings that are difficult but still possible to hold (one side making self-handicapping moves they'd never make themselves), so as to get decisive 1-½ or ½-1 game pairs (sometimes much stronger engine manages to win even from the handicapped position, but this very very rarely happens with evenly matched engines).
A game isn't solved until you can force someone to play a certain way and cutting off every single option they have consistently for every match you play.
does that mean chess is unsolvable? if an answer to that doesn't exist currently is it probable to assume that chess will.be solved at some point?
Chess cannot be solved.
no but there's so many possible variations that it might as well be at this point in time
"It is estimated there are between 10111 and 10123 positions (including illegal moves) in Chess. (If you rule out illegal moves that number drops dramatically to 1040 moves. Which is still a lot!). "There are even more possible variations of chess games than there are atoms in the observable universe"
copy paste didn't get the exponents right
It is estimated there are between 10^111 and 10^123 positions (including illegal moves) in Chess. (If you rule out illegal moves that number drops dramatically to 10^40 moves. Which is still a lot!). "There are even more possible variations of chess games than there are atoms in the observable universe
lol that makes more sense. I'm like wtf this homie thinks there's only 1000 chess positions
It can't be solved because there are more board states than atoms in the universe.
we just don't know, why is it so hard to believe?
it's an extremely complicated calculation, we' haven't found an answer yet
op here. It's not solved. That's just a meme to trigger the elo-gays. Tic-tac-toe is solved, for example, because there's always a way to play to prevent a loss. Chess isn't because the middle game is too vast to be feasibly solvable. Endgame configurations are often considered solvable, but that doesn't count because you can only reach this point after someone makes mistakes that they didn't have to make.
unsolvable? no
is it current unsolved? yes
FIDE's time controls and repetition = draw rules make it essentially close to being solved. Let's say you're at move 40 and there's a +0.2 advantage, the machine may see a path to winning but obviously the human would rather settle for a draw in an obviously (human perspective) drawish position.
close to being solved is not solved
even from many moves made that make humans want/forced to draw, there are different variations and timing at achieving such similar board states
This is not relevant from a human perspective. If you have 5 minutes left on the clock and the position appears drawn, you settle for a draw. There's no point in mulling over the possibility of Stockfish or whatever engine finding a mate in 37.
Solving Chess is clearly a computable problem. The way was described in Claude Shannon's 1949 paper: you just run min/max search to the end.
Whether it's solvable IN PRACTICE is a more difficult question. Arguments like
don't matter because you can recompute part-solutions and reuse memory, you can solve Chess with very little memory in fact. So it comes down to speed, and it's in the territory where it can come out both ways depending on your exact assumptions: if you imagine a star system sized computer (Matrioshka brain) running as efficient computations as laws of physics are thought to allow (Landauer's limit etc), until the end of natural star formation, you might or might not have enough time. Quantum computing almost certainly doesn't offer exponential speedup (8-Chess is a PSPACE problem, and BQP ⊂ PSPACE, where BQP is the name of complexity class for problems that can be "efficiently" solved with a quantum computer), but it might provide some polynomial speedup similar to Grover's algorithm. Exact details don't matter, it reduces to the same kind of order-of-magnitude questions that make up the other assumptions. If Chess is solvable in the real universe, it's definitely very difficult indeed.
Well, that's assuming you are looking for a strong solve. There's no principled reason why there couldn't be some neat proof that relies upon arguments about symmetries, copying moves, or such like. For example, in that aforementioned paper, Shannon points out that if the rules were subtly different and you could pass your turn, you could prove as a theorem White can at least draw (if first-mover wins, White can make their move, if second-mover wins, White can pass, else it's a draw).
Also, it is quite conceivable that you could have such good heuristics that there's no position where an engine would draw a winning position or lose a drawing one. It's just that you couldn't PROVE you have such an engine, without solving the game.
Try that again, in English this time
There is a known algorithm that can definitely solve Chess (checking all possible games). The question isn't if Chess is solvable in principle (it is), but if it's solvable in practice. It's definitely not solvable with a desktop PC. It's definitely not solvable with a supercomputer (as of now it has been solved for positions with 7 pieces or fewer, and some select 8-piece positions, and the difficulty keeps skyrocketing as you increase the number of pieces, and solve would require all 32). Or a quantum computer (Quantum computers aren't "faster" than classical computers in the sense of having higher clock speeds or whatever, but there are specific types of problems they can solve in fewer steps and Chess isn't one of those problems).
But you could try imagining a star system or galaxy sized computers running until the end of the universe. Depending on what kind of computer exactly, and how long exactly, and how difficult the problem is exactly, it can come out both ways. Maybe Chess can be physically solved in this universe, perhaps not.
>The question isn't if Chess is solvable in principle (it is)
No it isn't. There's too many pieces for there to be one set of moves that guarantee a win regardless of what the opponent does.
nta but if we ever come up with a solution to chess, it won't be a simple set of moves. It'll be an incomprehensively large decision tree that a player can traverse and force a win/draw every time regardless of opponent behavior. This is a question of computability, and that's a topic every cs major has to learn to graduate. Chess isn't magic. It can be described algorithmically, and there is a definitively finite number of possible board configurations, each with solution to win/draw. Just because the hardware doesn't exist to compute a solution doesn't mean it isn't computable.
>It'll be an incomprehensively large decision tree that a player can traverse
That's not solved.
That's quite literally what the solution to a game is - an algorithm, a decision tree
*each with solution to win/draw
Let me correct myself here. Not every board configuration is solvable for both sides, but it's solvable for at least one, and starting out, there is necessarily a decision tree that can lead to a win/draw regardless of opponent behavior.
It is. Why is tic-tac-toe solved? Because I have a (small) decision tree regardless of whether I'm x or o. No matter what the opponent does, I have a move that I know will force a win/draw. Chess is no different. It's just ridiculously resource intensive to compute a decision tree that large. I'd be surprised if it ever happens, but that doesn't mean it's not mathematically possible. Solved != solvable.
>Because I have a (small) decision tree regardless of whether I'm x or o
Wrong.
Its because as the first person to go, you have very specific moves that force a draw.
That's why Tic-Tac-Toe is solved.
yes there is no one set of moves that guarantee a win regardless of what the opponent does, but there is a different set of moves that guarantee a win for every possible move the opponent can make.
or there is no such set of moves and instead there is a different set of moves that guarantees a draw for every possible move the opponent can make
Solvable just means you can necessarily prevent a loss from the starting configuration. Chess is solvable. It just hasn't been and likely never will be solved because the search space for the middle game is too large for even all the computers buildable with all the silicon on the planet to compute before the heat death of the universe. If it's ever solved one day, it'll be because of advancements in mathematics and hardware.
You have a decision tree, which is a mapping from player (either you or the opponent) behavior to a new decision tree, each eventually resulting in a win/draw. If the opponent fricks up, you can force a win. If not, you can at least force a draw. I'm not going to argue computability with you. My senior thesis was on automata theory. You're just some random shitter whose self-image will be diminished for some reason if chess isn't a mythical game of cosmic proportions.
Tic-Tac-Toe does not have a decision tree.
You always make the same moves and you always get a draw.
>tic-tac-toe doesn't have a decision tree
>you just react to your opponent's moves when deciding what to do in order to force a draw
Do you just not know what a decision tree is, or are you merely pretending?
>showing board positions means there is a decision tree
You are still making the same moves regardless of what the opponent is doing.
Look back at the pic, anon. See those arrows? What do you think they represent? What about all the different game boards. What do those represent? Come on, man. You can't be this moronic.
>It just hasn't been and likely never will be solved because the search space for the middle game is too large for even all the computers buildable with all the silicon on the planet to compute before the heat death of the universe.
You sound insane. Almost like one of these fanatical religious people who claim the most ridiculous things. You do realize that right?
Atheists are more fanatical than any religious organization on earth.
That's an exaggeration, but ultimately a dumb comparison, since organizations can't be fanatical because they aren't people
Atheists are more fanatical than any religious person on earth.
There, happy?
That's also false, since religious people are by default fanatical and the nature of atheism goes against fanaticism. If you mean fanatical atheists are c**ts, that's a different story.
>religious people are by default fanatical and the nature of atheism goes against fanaticism
You are fricking moronic.
Point out what's insane about a long form version of
>chess is solvable but it's too complex to feasibly solve
Take all the time you need.
What's too complex about 32 pieces on a 64-square board that all the computers in the world can't solve before the "heat death of the universe" you fricking dumbass nerd?
When your computer does something, how do you think it's doing it? Magic?
You have no idea what you're talking about and you aren't owed an explanation from anyone.
I genuinely think Ganker might be the most moronic double digit IQ board on this site
How is that insane? To represent all of the board states you would need a complementary amount of bits.
That's literally a decision tree in the image you're responding to - you do know this right?
seeing Ganker respond to simple acknowledgements of hardware restraints by seething made me remember why i make six figures for plugging APIs together. you morons are reason. thanks, genuinely
but most of the moves are obviously bad moves, so those can be pruned, this reduces the search space dramatically.
an "obviously bad" move might become a brilliant combination a few moves deeper.
however for both human and machine, the key to reduce search space is just finding good heuristic to eliminate all the paths that have low probability of leading to winning
pruning decisions that have a low probability to win won't solve chess. that's the question here, not whether pruning simplifies the problem at the cost of precision, which is trivially true because that's already how AI chess engines work.
Not every problem can be solved with a greedy heuristic. A locally good choice might be ultimately incorrect. This is a basic concept in algorithm design.
But they're "obviously" bad because of heuristics, not PROOF. Right, contemporary engines have pretty good heuristics and then they 100M-check there isn't some unexpected consequence 10++ moves down the line, or they have superhuman heuristics and they 10k-check the same, they will not be fooled by -2682/256 materiel from a Queen sacrifice and will happily sacrifice Queens if it's the winning move. Particularly recently, top engines sacrifice materiel like the old masters. But just like I can't spot all compensating factors for a sacrifice that Stockfish makes, there's no assurance that some move that seems bad at a glance (or after looking for a couple of million lines) doesn't in fact turn great in the long run. Engines can and do misevaluate positions: that's how they can win games off each other! Hell, I've seen games where Stockfish has found forced mate-in-30-something and the opposing engine (Lc0/Komodo usually) might not only win slower, or draw, but lose!
Basically, the only way to DEFINITELY know there isn't any unexpected consequences to the move is to, well, check all moves. Of course, engines USUALLY make the best moves anyway. I'm pretty confident that Stockfish running on contemporary high-end single CPU would mostly draw 32-man tablebase. And I've checked that it can find the longest known mates (mate-in-549) without tablebases. But that it makes the winning moves most of the time isn't proof, or solve, of anything.
Why do you need to check "all possible games"? Why not just solve the most common possible endgames? Like it's king and a pawn versus king, isn't that solvable/solved?
>Like it's king and a pawn versus king, isn't that solvable/solved?
That's not a position. It matters where on the board those pieces are located.
In the case of fewer than some number of total pieces on the board (7? 8?), there are endgame tables where any position of the pieces is totally solved.
They're fricking our (chess) pussies
It is effectively solved. Two engines will always draw. The only way for super GMs to beat each other is either to intentionally get themselves in a weaker position to take their opponent out of theory, and trust themselves to be better in the scramble. Or to keep drawing and drawing until it gets to stricter time controls and trust themselves to be faster. Its a good board game for children but its ludicrous that it is still played professionally.
>It is effectively solved
No it isn't.
>Two engines will always draw
No they won't.
Demonstrate a chess engine's victory over another equally advanced engine.
Have you never watched chess engine tournaments?
Do you know anything about chess?
Personally no but you didn't say much more than that either.
Because I actually assumed you weren't a moron. My bad.
Other people in the thread have explained everything. Why should I repeat it?
see
Idk anon, they actually explained their position plausibly and all you have in response is nuh-uh. It's pretty clear they're right and you don't know what the frick you're talking about
NTA, but they were just lies that offered no argument to be countered. Two engines will not always draw, and it is not effectively solved.
>The only way for super GMs to beat each other is either to intentionally get themselves in a weaker position to take their opponent out of theory
it's like claiming soccer players won't kick when there's a goalkeeper or something and then saying soccer is solved, you just kick when there's no goalkeeper
Because it's full of falsehoods.
>two engines will always draw
There are actual tournament competitions between different extremely powerful engines where one engine will be victorious over the other. https://www.youtube.com/watch?v=9yoXlr8x5xg&list=PLDnx7w_xuguGOTrXZ_I26qCFA-VviWO21
>The only way for super GMs to beat each other is either to intentionally get themselves in a weaker position to take their opponent out of theory, and trust themselves to be better in the scramble
Wrong as well. GMs play all sorts of lines. They will not intentionally get themselves in a weaker position. That is how you lose. What they will play is unpopular lines to throw opponents off but matches quickly get away from book moves.
>Or to keep drawing and drawing until it gets to stricter time controls and trust themselves to be faster.
I don't even know what they mean here. There are many different time controls but in any formal tournament they are using classic time controls which mean games can easily run into multiple hours.
You are stupid. You are a black gorilla moron redditor.
>What they will play is unpopular lines to throw opponents off
If they were the strongest lines they would be popular, ergo weaker lines are indeed being played and that anon was right
>If they were the strongest lines they would be popular
You have no idea what you're talking about
If you actually had a refutation you would have provided it. All you have is "nuh uh".
The fact is that in any competitive endeavor, the strongest tactics and strategies are the most popular, because they are the strongest. If less popular lines were as strong as the most popular ones, they wouldn't be less popular.
his explanation hinges on the acceptance of specific facts which are flat out false, "nuh uh" is perfectly a apposite response
yes it is solved, an upper bound in possible board states is 13^64 or approximately 10^71, if a program evaluated each 1 million states per second it'd only take 10^57 years to solve chess
That's not what solved means.
yes it is because chess has no time limit so I can just wait that long for computer to finish running before I make that first move
>chess has no time limit
Wrong.
I think the problem here is people conflate
>this problem is solvable if given to a sufficiently powerful computer that has sufficient time to perform some super-universal task
with
>this problem is solvable within seconds from an app on my smart phone that I can pull out while I'm taking a piss during a break at a tournament
If Chess engines solve everything why don't you go and become world champ?
If AI-assisted weapons can hit a target with a nearly 100% accuracy rating, why don't you go and become the ISSF world shooting champ?
Thanks for proving my point with a shitty analogy.
What's your point? That you don't understand how a computer can be capable of solving problems ridiculously fast in the same world where competitions are usually limited to humans in a proctored environment in which it's considered cheating to use AI to win?
ok pajeet, what is your definition of the game getting solved? Any position can be solved by any computer these days even phones in a matter of seconds. Chess can be solved , if a domestic computer has no trouble going depth 40.
Magnus computer goes to 70+ in seconds, if this doesn't mean it's solved then you are correct, chess will never be solved cause pajeet is mad that the game is in fact solved.
https://analysis.sesse.net/
Engines investigate moves at such depth because they're only looking at duodecillionth of possible lines.
You do realize that in order to solve any puzzle, you don't choose the incorrect solution to work the puzzle out right?
Yes, but that comes down to the issue discussed some messages above: how the engines (well, αβ-engines, anyhow) prune the search tree. Not because there's mathematical certainty there's no winning moves in the discarded set, but because of heuristics, rules-of-thumb that can be wrong. In some cases the developers have DELIBERATELY picked evaluation functions that are empirically known to evaluate certain types of positions incorrectly (one example would be how Stockfish deals with fortresses) because it's faster and the engine is stronger on average without such consideration. In this case there's a Stockfish variant Crystal that specializes in correctly evaluating positions like fortresses, but it loses to regular Stockfish. And sometimes it just happens unintentionally.
A solved game is one where every single configuration has been mapped. Tic Tac Toe is a solved game as is Connect Four. Chess is a game in which individual positions can be solved but to my knowledge it has not been solved yet despite computers being way better than any human player.
>A solved game is one where every single configuration has been mapped
No it isn't.
A solved game is where there are certain moves you can play that win 100% of the time, regardless of what the opponent does.
>moves you can play that win 100% of the time, regardless of what the opponent does
Yes, and doing so would require you to find a game tree for which under each possible enemy move on each turn there is a subtree that leads to you winning. That hasn't been done
...Which is why the whole thread is people saying chess isn't solved.
Can you read the fricking thread?
you are acting stupid again. understanding each possible move should be a draw, not a win
how the frick do you draw in chess, either you capture their king or they capture yours. Sorry you learned some snow falke everyone gets a participation trophy version of the game
>how the frick do you draw in chess
If you dont know anything about chess, why are you here in a chess thread?
Stalemating is one of the first topics discussed in beginners books, and it's the goal of every player who's down by multiple pieces in the endgame since it's considered a draw rather than a loss. It's part of what makes the game fun.
Reeks of soccer
You'd also be getting called a moron if you joined a soccer thread and complained about the rules of the game there despite demonstrating a genuine lack of understanding about how the game plays at the most basic level.
If someone intentionally stalemates they're a homosexual and not worthy of playing with
Metagays kill everything, go back to taking BBC in your ass
is there a chess version of scrubquotes?
You've never played chess have you?
no, that's why my opinion is more correct, I am unbiased
>I came into a thread about a game i've never played. Lemme tell the GMs how to play.
GMs intentionally stalemate. Most of them were better players at 5 than you'll ever be in your entire life.
>pro players know best
Clearly you've never played any "balanced for esports" video game before. Why do you think nobody plays fighting games, they're for gays
>noobs stalemate
>intermediate players stalemate
>IMs stalemate
>GMs stalemate
>"muh esports muh fighting games muh meta"
God damn. I didn't think it was possible to be stupid because you don't play chess, but you're starting to change my mind. Playing a few games would do you some good. You think a game is bad because there's a rule that discourages sloppy mating patterns. It forces players, even beginners, to think about endgame strategies. When you see somebody looking at a chessboard for an hour, they're probably doing something like that. They're learning the game. You'd do it too if your IQ weren't in the basement covered in mildew.
imagine you bought a ticket to the superbowl and they just decided the game was a draw halfway through because they didn't think they could get the ball futher. You would be pissed, because it's objectively bad game design
Arguing in analogies shows how stupid you are.
imagine if you were in rhetoric class and the teacher told you analogies were a completely legitimate form of argument. You would be pissed because you are a pseudintellectual moron
>apples and oranges
Black folk and israelites
Hamburgers and Fries
guys and gals
>comparing a game where you can't draw to one where you can
False equivalence. Nobody going to a chess tournament will be surprised when some games end in stalemates. That's how the game is played. If a football game ended that way, people would be mad because that's not the expectation people have when they take a week off work and pay $1000 for tickets for the super bowl. Are you just arguing to argue? I don't even understand what your point is anymore.
so football fans have higher standards, understood, thanks for the concession
Football fans expect playoffs not to end in a tie because it's literally impossible for that to happen. That's not how chess works, moron.
yes because chess is worse, when john madden made football he was smart enough to make it that the game couldn't draw. Chess fans aren't intellectual enough to do that
You jumped the shark.
If you didn't realize 15 minutes ago it was already over
>merely pretending for attention
I hope you suicide isn't too hard on your parents, anon.
>when john madden made football
His point was because he doesn't play chess, he's more qualified to nitpick its rules than people who have been playing for their entire lives.
He's an idiot.
Anon, chess is a drawish game after the PCs solved the game
>he thinks no one ever had draws in the centuries before computers
>strawmaning already
i accept the concession zoom
>no argument
I accept your concession zoom.
I'm sure the only reason why that hasn't happened is because shit openings are not even considered by the computer, your tiny mind thinks that every garbage moves needs to be accounted for the game to be solved but that's not true, you don't solve puzzles by acting moronic and to increase the time it takes to solve it.
There is actually a general theorem in Game Theory that when applied to Chess, states that there exists a winning strategy (sequence of moves) that guarantees a victory for EITHER white, black or a draw. But the game space is too large to compute the solution.
Computers being able to sometimes make an educated guess on what they think the "best" move is, based on terabytes worth of training data, does not make a game solved. Chess will only be solved when a strategy or series of strategies is found that guarantees a win regardless of what your opponent does
>Chess will only be solved when a strategy or series of strategies is found that guarantees a win regardless of what your opponent does
What happens if two unbeatable chess programs play against themselves?
Then presumably White always wins because it goes first or every single match draws (ie tik-tak-toe)
This happens all the time.
Its called a chess engine tournament.
melee
There are only two possibilities:
The game has a solution and one player is guaranteed to win.
The game has no solution and the other player can always force a draw.
Both of these situations make playing such a game a meaningless endeavour.
Chess is the TF2 of boardgames
As in, nobody plays it except bots?
Counter-Strike is pretty much a simplified version of chess where everyone is a pawn that can randomly go Arnold Schwarzenegger mode
chess is not a solved game, otherwise it wouldn't exist anymore, delete this thread
Basically any games? Is there anything thats the opposite where there isnt a correct to play
Actually I have a perfect example: trying to get the lowest time on a reaction speed test
The fastest time requires randomly guessing when to click a button, not reacting at all.
Not every vidya is as theoretically rich as chess. In fact, I'd argue basically none are except certain simulation games and maybe competitive shit like cs. Saying pokemon yellow has an unfathomably high skill ceiling is disingenuous because you really only need to be as competent as a child who can understand rock-paper-scissors to be good enough to win. To win at competitive chess, you basically need to have been playing consistently for years, studying openings, strategies, combinations, and endgame configurations. The skill ceiling isn't even comparable. Again, very few games are that complex.
Of course chess is a solved game. It's pure math.
chess with only 7 pieces left is solved.
>Competitive and fun
>Rewards pattern recognition
>0 trannies
Take the chess pill
Chess is "solved" in the sense that engines have ruled with almost 100% certainty certain openings are straight up losing so everyone starts with a limited set of tools which takes out much of the mysticism the game used to have. It's "solved" as far as humans are concerned for the first 10 moves or so. And as for the rest of the game, it doesn't matter if a machine can find a winning solution 50 moves from now on because humans will never see these exact combinations and settle for 3-fold repetitions aka draws.
>game has a meta
>means its solved
It's an ever shrinking meta, which is the issue. For chess to be an interesting game to watch, the quicker players are thrown into uncertainty, the better. The more they converge towards a draw, the closer the game is to dying, from a human perspective.
The middlegame is too complex. Even analysis engines don't typically traverse a search space deep enough to possibly be able to predict the best move. Even when they can and doing so is computationally trivial, the only thing that will come out of it is that high elo chess theory will devolve into running AI engines and literally just memorizing board-configuration-to-moveset mappings. For the average chess player <2000 elo, this will never really matter.
>meta
Meta = What Magnus plays
Magnus plays the Kalashnikov , suddnedly it's the best defense in the world
Magnus plays the Catalan, the best opening in the world is the Catalan.
e4 or d4?
Not vidya but anyone ever wonder about bowling? like how the frick has nobody figured out how to throw a strike every time.
I know all that shit about the oil patterns or whatever but surely someone should know how to knock over the 10 pins every time
That is actually a good point.
I wish someone could answer that.
Bowling is a physical action. Even if you know you need to hit X board at Y speed with Z revolutions, that doesn't mean your body is physically capable of doing it every single time. Hitting a free throw is comparatively simpler and nobody has a 100% FT percentage.
It's all about micro adjustments.
You would think that you can just roll it the same every time, but air pressure matters and holes in the ball create enough of a disbalance that it's not achievable unless you have truly perfect posture, speed, and control. Even the players weight changing because they drank some water or started sweating can have impact on the balls performance. Additionally to this each lane while built to extremely strict controls are still not perfect down to the micrometer with a lane somewhere else. All of these differences are not visible or able to be felt by humans. There is just a bit a leeway when hitting a strike so players have to guess how much to adjust their next shot to keep their ball a strike based on how their ball previously went down the lane.
>I HATE THIS. I HATE THIS. PLAYING WHITES IS SO EASY. IT LITERALLY REQUIRES NO BRAINS OR SKILL TO PLAY WHITE. ALL THEY DO IS CHEESE ALL DAY.
>LITERALLY 0 CHANCES TO COUNTERPLAY DURING ALL GAME. WHEN COULD I HAVE DONE ANYTHING? WHAT CAN I DO WHEN ALL THEY DO IS GAMBLE AND TAKE DUMB RISKS?
>NO. I CAN'T TAKE HIS PAWN OR HE WAS GOING TO DEVELOP KNIGHT AND THEN FORK MY BISHOP- NO I CAN'T MOVE MY QUEEN SO EARLY TO TAKE A PAWN. YOU LITERALLY HAVE NO IDEA WHAT THE FRICK YOU ARE TALKING ABOUT. YOU DUMB 900 ELO APE. YOU ARE BANNED. GOODBYE.
>"""HAHAHA YOU ARE JUST NOT TALENTED ENOUGH, ARTOSIS. MAGNUS WOULD HAVE OF WON, ARTOSIS.""" WELL, MAGNUS DOESN'T PLAY AGAINST STREAMSNIPPING APES. DO YOU THINK THIS IS GOOD PRACTICE? COIN FLIP AFTER COINFLIP? THIS IS SUCH A WASTE OF MY TIME. ARGGGHHHHH
>""""HAHAHAH I GOT YOUR ARTOSIS. YOU WEREN'T LOOKING AT MY KNIGHT, YOU FOOL. LOOK AT ME. WOOOWWW. I CAN MOVE MY PIECE THROUGH YOURS. I AM THE KNIGHTMAN. LOOK AT ME HOP AROUND. NOW YOUR QUEEN IS GONE. GRANDMASTER RANK."""""
>NO CASTLING. NO DEVELOPMENT. JUST FOOL CHECKMATE ALL GAMES. WHAT IF I MOVED PAWN TO F7 OR F6? YOUR QUEEN AND BISHOP WILL BE OUT OF POSITION. EZ FORK. EZ GAME. GG.
>GREAT AND NOW ANOTHER RANDOM LAGGY BARCODE WHITE. FUUUUUUUUUUUUUUUUUUUUUUUUUUCK.
>two computers can play the game perfectly so that means its solved!!
You aren't a computer moron
The higher powers of the reflective intellect are more decidedly and usefully tasked by the unostentatious game of draughts than by all the elaborate frivolity of chess.
Shut the frick up moron
Fischer was right, the game died when computers solved it. Many morons ITT.
>Fischer was right
Careful now.
In chess where the pieces have different and bizarre motions, with various and variable values, what is only complex is mistaken (a not unusual error) for what is profound. The attention is here called powerfully into play. If it flag for an instant, an oversight is committed, resulting in injury or defeat. The possible moves being not only manifold but involute, the chances of such oversights are multiplied; and in nine cases out of ten it is the more concentrative rather than the more acute player who conquers
Enter
*ahem*
What happens when two supercomputers play chess against each other?
stockfish wins
>game is technically "solved" in the sense that there's almost always a correct way to play,
thats not what "solved' means, mathlet. Are you a subcon, out of curiosity?
Do you just not know what bait is, or do you actually feel special for having seen this conversation played out in earnest before?
Chess is only "solved" in the opening stages. After that it's all mindgames.
GMs study each other for years and know all their moves in the opening.
When a GM makes an unexpected move, it causes the other GM to sweat profusely as he tries to figure out what the frick is he trying to achieve. This causes stress and makes them prone to mistakes which is what makes chess so exciting to watch.
Computers can solve everything obviously (make an AI that can aimbot and see through walls in CS.GO for example) but the human factor is what makes the sport exciting.
>Chess is only "solved" in the endgame stages
ftfy
Chess openings are no where near solved.
>When a GM makes an unexpected move, it causes the other GM to sweat profusely as he tries to figure out what the frick is he trying to achieve. T
Not when each player has 2 hours plus increments to figure it out, lmao. And if the solution is to gradually decrease the time each player has, then that's just choosing another flavor of death.
Does anyone watch these supercomputer vs supercomputer matches? No? Then why do you gays act like it's relevant to anyone interested in chess rather than quantum mathephysicmagic?
>Does anyone watch these supercomputer vs supercomputer matches?
Yes. Chess fans.
Anyone wants to play a round right now? I'm bored.
Post link
league of legends
Shmups by Cave, Raizing, and Takumi.
Chess is complex.
We didn't solve it.
We don't know if it's solvable.
That's just it, why is there a debate about this?
because it is 100% solvable
Chess is not solvable.
Explain why so many people win xhess games then? You're thinking about tic tac toe
and you're 100% morons
i can put any combination on the board and the chess.com scuffed cloud computer will solve it with relative ease all the way to 40-50 depth, magnus computers goes to 70+depth in question of seconds. There ain't nasa computers and games hardly go beyond 70+ moves, specially once they get out of the opening.
Chess is solved.
You don't know what solved means.
You are simply using your own definition at this point to defend the fact that at least the main openings can be solved entirely to a draw in matter of seconds.
Openings are not solved. Endgame is.
You can use google to actually find out what solved means.
i understand your definition but i choose to not care about it cause i know it's wrong, people that live from chess are scammers and understand the dance.
I can put any combination on the board and it can be solved, chess is solved, people are now playing computer lines from start to finish.
>i choose to not care about it cause i know it's wrong
Stopped reading there.
have a nice day.
Chess is not solved.
It is, you simply choose to ignore it for your own benefit.
Your the one ignoring the definition of "solved" because you think it is wrong.
>i understand your definition but i choose to not care about it cause i know it's wrong
Well, at least opening with this means the post can't get any more moronic. So there's that.
>people are now playing computer lines from start to finish.
No, wait, I was wrong. The reality is that people are in fact not "playing computer lines from start to finish" and with good reasons. Learn a little bit about Chess beyond "this is how horsey moves" and you'll begin to understand why. You don't even have to study for long to get it, so your brain won't melt. Promise.
Chess is the best board game, prove me wrong
csgo
Chess is not solved.
Chess is not close to being solved.
Stop spouting this misinformation.
>People that don't actually play Chess but rather watched that Netflix series thinking Chess isn't solved
To get good at chess you will spend next 10000 hours memorizing the correct moves.
Yes you heard it correclty, you will just be memorizing the correct moves.
The part where you are learning tactics, strategies etc. passes really fast.
>you will just be memorizing the correct moves
Yes, and you won't win every game.
Therefore chess isn't solved.
>Yes, and you won't win every game.
You will literally win every single game opponent deviated from the exact correct answer and fricked himself over by not memorizing as good as you did.
The best of the best have memorized more than doctorate graduates or have a way of remembering the exact moves with pattern recognition.
Basically you would compute the winning strategy is to map out all possible moves as a tree, then from the leaves of the tree (win/loss/draw), pick the optimal move that leads to your preferred outcome. Recurse upwards until you get the winning strategy.
The theorem was every finite game with perfect information in extensive form has a pure nash equibilibrium. Since Chess is a perfect information game and has no random elements, and Chess is a finite game, then the theorem does apply.
The fact that opening theory has existed since medieval times really makes some people screech autistically.
For the game to be solved, white would have to be able to force a positive outcome no matter what black does. We don't know if it's solvable because we haven't run enough simulations yet. We would only know for sure if we wasted an insane amount of computing power trying to ruin a popular game. If there's a sequence of moves that causes victory 100% of the time, then it is. If at any point, the other player could make the right decision(s) and win, then we could at most boil it down to a rock, paper, scissors game.
ITT: guys who can't even know what fizzbuzz is try to explain to software developers how computability works
>who can't even know what
yes sir I great software engineer from mumbai university yes do not redeem the computer sir
Love how fricking CHESSBABIES think their shit isn't solved.
A fricking machine can beat the world's best GO player, and that game is infinitely more complex.
Chess is 0 IQ shit for morons.
>Go
>literally only has one piece compared to chess's 6
>doesn't even have jumping mechanics like checkers
Foolish weaboo.
>go
>complex
You gotta make your bait less obvious, anon
Programming enthusiast but not an expert here. Instead of trying to calculate the possible sequences of moves from start to finish, and then checking which ones end in a winning game, wouldn't it make sense to collect all possible winning configurations of boards and calculate each sequence of moves that can possibly lead to that board state? E.g. if in configuration 1 white wins and white has both rooks, can we rule out all move sequences where White's rook is taken by t3 or t4? Or would this not matter because the amount of calculation would still be incomprehensibly high?
As I understand it, endgame tablebases (you can search for more detail using that term) are created by working going backwards from legal final positions. Doing this in a real-time game isn't useful not least because the engines can already utilize the tablebases that use a clever indexing scheme to read win/lose status of the given position. Currently there are complete tablebases for positions for 7 or fewer pieces and select positions with 8 pieces, although 7-man tablebase is some ten terabytes and due to latency involved reading it from mass storage (since it tends not to fit in RAM), it might not be worth using. Anyway, Stockfish is particular is really good at reducing a position to a solved one.
But crucially, they don't calculate all possible lines but discard lines not worth investigating. The core algorithm for optimizing the exhaustive min/max search is known as αβ-pruning (https://en.wikipedia.org/wiki/Alpha%E2%80%93beta_pruning https://www.chessprogramming.org/Alpha-Beta), but the engines use all other kinds of trickery to make this pruning more efficient, like iterative deepening (the idea is to first look at all moves at depth of n=1, then n=2, etc: this requires you to recalculate lines at each iteration, but it guarantees that all refutations at lower depth are discovered, allowing more efficient pruning at greater depth, when almost all moves being investigated are at maximum depth), or storing already evaluated positions in a hash table where they can be looked up if the same position can be reached in multiple ways, and a plethora of other more advanced techniques.
chess vidya
Skyrim
Melee
>Ganker trying to discuss complex math terms
I would correct a lot of you frickers but I'm fricking exhausted, so call me based or moronic as you see fit
backgammon
>it's solved
>BECAUSE IT ISNT OK
The devinition of solved (simplified version) is that we know what the perfect game looks like. A game wherein each player plays the exact best move where every other possible move is objectively worse.
The fact that chess has multiple openings alone proves chess is not solved. If it was there would be only one opening, which is perfect, and one defence, which is perfect. We don't know which opening and defence are perfect, so it's not solved.
Chess is fun as long as you stay below 1400
NetHack comes to mind. In some sense it's a legendarily difficult game: some people have been playing it on and off for decades and hundreds upon hundreds of hours and aren't even close to ascending. However, NetHack is also beatable by good old-fashioned AI techniques.
Solved game
Play Mahjong instead
Here's your solved game.
>game is bad if you're top 0.001% player
most people who like chess just want to jerk themselves off because they play chess. hardly anyone is any good at it.
i'm not saying fricking anything before you greentext me i'm just saying that. most people just egogay chess
>most people just egogay chess
Isn't this why most people do anything competitive? What's special about chess?
has someone the webm of the two chess pros who look like their soul left their bodies at the end of the game?
Dude was looking directly into the crowd grabbing onto random things playing against the chinese dude who also looked like he had enough of chess for the next 10 years or so
nvm found it lol amazing webm
The amount of autism in this single webm is insane, Ganker could never compete
what's happening here?
>They've just finished playing 14 games of Classical chess where they each had like, 90 minutes in their bank.
>So probably over 30 hours of chess.
>ended in a tie
>this goes to rapid (10 minute games) to decide the winner
>Chinese man starts winning
>Italian man realizes he's not becoming the world champion, and this may or may not be his only shot over
>no way he can come back
>has a bit of a dissociative episode
>Chinese man is obviously happy to won, but just too physically exhausted to do anything but hold his head in his hands
Most physical athletes will never be this tired
Sorry even worse, they had 120 minutes, with 60 being added after 40 moves, and 15 being added if the game goes over 60. Some of the games went up to 80+ moves.
man
Italians aren't smart enough for Chess, guy is Russian
>Italians aren't smart enough for Chess, guy is Russian
You're right, I should know this.
>Italian man
Nepo is russian you moron.
>Chess
>Skill
lol
I’m assuming you’re rated under 2k ELO if you think that.
Everyone knows you only get above 2k ELO by memorizing as many openings and variations as the computer says. It’s not about skill, it’s about memorizing what a computer says are the best moves.
>im assuming you're not in the top 0.01% of ranked players
How'd you guess?
Chess has the potential to being solved. There's no guarantee we can get enough computing power to actually run the algos we need to count for every possibility.
All that menas is that if the computer power is enough then yes, it can be solved.
To put this into an easier to understand scenario
X = unknown value
X = amount of computer power needed
solving chess means solving X
We cannot achieve the value of X because we lack the resources
Mahjong Soul, but it doesn't always goes your way even if you play perfectly, which keep things excitings.
Check with rook check with queen mate with queen?
melee
>high skill ceiling
>smash bros
Nice joke.
Just because there is always a right thing you should do doesn't mean more than 3 people on the planet can actually do it. I would call that a high skill ceiling.
Smash Bros other than Melee is substantially easier though, yes, with the less fricked up movement options.
Why do smash gays think their shitty simplistic game is so deep?
You can hate Melee, but claiming it takes no skill is low effort bait.
It takes no skill.