Will AlphaZero show that this generation in top level chess was a waste?

I have often wondered where are the immortal games of the previous generation. Where is Magnus Carlsen's immortal game? It simply doesn't exist. I feel that the current generation doesn't play immortal games anymore. They instead stick to the 'computer' mode of playing, just slowly improving your position and eventually making that one extra pawn count and grind out a result.



I will be the first one to admit that I am oversimplifying and the current generation has played some great and exciting games as well. However for the most part I think everyone will agree that the era of romantic chess was done and dusted.



The kind of chess that masters of the past played, just always looking to go to any extent to get an advantage and not being happy with equality was, one felt, done with in the history books.



Also I don't want to sound too delusional in criticising the games of the best chess players in the world. Just as one doesn't need to be a great singer to know a great song from one that isn't, similarly one needn't to be the greatest chess player in the world to tell am immortal game from one that isn't.



I am fully aware of my very limited chess prowess as I write this. This computer generation of chess players doesn't do immortal games anymore. The last true immortal game of chess was played by Vishy vs Levon in 2013, and even Vishy could hardly be said to be a computer generation chess player.



The fact is that since 1997 when Kasparov lost to Deep Blue the new players were taught that a computer could take down the best chess player in history (which is what Kasparov could very reasonably lay claim to being), so lets learn how the computer plays. The computer model is the way to go forward.



From that point on the focus shifted from doing what one could to gain an initiative to playing scared, and always sticking to equality. For if you give the computer an inch, it will crush you, and so will the player who has learned chess from its computer teacher.



Even after the defeat to Deep Blue, Kasparov played a true immortal game against Topalov, but the players were already regarding the computer as the new supreme being in chess.



Gradually everything the human players of the present or past did came to be looked at through the engine's glasses. Suddenly great immortal games of the likes of Adolf Anderssen and Paul Morphy started being 'refuted' by the computer. If their opponent had play 'this' move instead of what they played they would have got equality.



This assessment has since come to be treated as the gospel unshakeable truth spoken by a higher being.



Today after Stockfish's humbling at the hands of Alphazero, one could definitely dare to re-look at all those assessments. After stockfish in its game always felt it was playing moves that gave it equality, but it still lost. So clearly stockfish's vision has a limit.



So what if the limits that the chess engines of this world have placed on this great game of ours are also flawed.



Stockfish refuted the line, is a phrase I have heard countless times in reference to the games of the old masters. However what if stockfish was wrong. What if in those games there is a counter to stockfish's refutation that we so far have been unable to spot.



The point is not that there must be one. However just think of how an entire generation of chess players have been restricted by the limitations of the chess engines which they regard has the all powerful being in chess.



They have all been taught to play chess a particular way - the computer way, and have been taught to be timid in their approach and always to toe the equality line. What if this equality line is flawed and the old masters had it right all along.



Chess is not a game to play to draw, but a game in which to be creative to always push for the win. There is a saying in chess that each generation stands on the shoulder of giants. However we may not be able to say that of this generation if alphazero's games hold up to scrutiny.



This generation of chess players will be considered the aberration, the generation that left the path of the romantic chess of the previous generations and embarked on a journey with myopic vision of the game, limited even further by the fear of losing equality rather than a sense of adventure and willing to push for the win.

So far from providing a shoulder for the generation to come to stand on, the generation to comes will have to unlearn what this genertaion of computer players did and stay away from their approach.


This may come to be the legacy of the current generation and that is a shame.

AlphaZero has shown NOTHING in the way of chess theory. It defeated an older version of SF running on a laptop which had all it's databases disabled.

What is interesting is the Monte Carlo algorithm's used to learn and play the games. Evaluations of each position were absent. There was no human programing to suggest the material value of the pieces.

DeepMind ran on massive hardware. It was all a publicity stunt in which Google stacked the deck. Current SF8 running on multiple processors found mistake after mistake by SF in the match; rated it's performance at 2200/2300.
The OP is making way too much of the match result imo. The only thing that AlphaZero showed was the ability of AI research to self teach given only the rules of a board game. It had almost unlimited hardware resources. A one time publicity event. There are no plans for DeepMind to ever play chess, go or shogi in the future.

current generation of players at the top don't cretae flashing combos against other top players as mistakes are not there.

And computers have nothing to do with it. I means cars go faster than humans run. Calculators compute faster than humans could ever. Just an another - fairly complex- calcutarot emerged. And will not play ever again as there is no science there. Only science was to show it can be done

Ah but you see the whole point of what is or isn't a mistake is under scrutiny. So far what was a mistake or not, was seen through Stockfish and chess engine's eyes.

However stockfish would tell you it made no mistake against Alphazero (I mean if a move was a mistake why would it play it), and still got destroyed.

So this mentality that one could be aggressive only if the opponent makes a mistake is what I am saying is the wrong approach.

One doesn't need a mistake from an opponent to play brilliant moves.

You have to remember that chess players usually wanna win games.

Or do they go in thinking atleast lets not lose first, and then if the opponent presents a window, only then go for the win. Or else just happily get a draw and be happy.

That is the problem. The attitude of the current chess players is that till the opponent will always be equal, till he blunders. So the win first attitude isn't there. The truth is that the platers today just want to not blunder first. This restricts the spirit of enterprising chess which was found in abundance in the romantic era greats of the past.

You can only win if your opponent make a mistake, with no mistakes a game is very probably a draw, just look at endgame tablebases. In fact, it also means that there are not such a thing like "good moves", but only ok moves and bad ones. A player like one of the 1800s would never win nowadays. For instance, it is said that Morphy was ahead of the rest of players of his time not because he was better at combinations (Andersen was as good as him at tactics) but because of his superior understating of positional play.

@mdinnerspace "DeepMind ran on massive hardware. It was all a publicity stunt in which Google stacked the deck. Current SF8 running on multiple processors found mistake after mistake by SF in the match; rated it's performance at 2200/2300."

Do you have any proof of what you are writing ? There was a thread on reddit where the stockfish creator said that at most the SF used vs A0 was 100elo less than the current stronger version...

The top players produce immortal games in large amounts. Hundred years ago fantastic games were pretty rare so their games rich in errors became immortal. A top players today produces a better quality within a year than a the old masters in their life time. The old masters would today win no money in a open tournament.

Do you want back the time of poor resistance and defense? They were simply lucky that they had no engine who showed them what clowns humans are.

SF played the match on a standard lap top with opening and end game table bases disabled. It becomes no more than a tool to check for blunders. It's performance was rated at 2200/2300 with all the mistakes made.
The primary factor of playing strength for SF, Komodo, Houdini is often the hardware used to run their programs. Playing the match on a standard lap top hardly represents a level playing field vs AZ playing on a super-computer.