Alphazero vs. Stockfish2024 resulted in 844 draws out if 1000 games! What if Alphazero won every game? That would mean chess is broken, or even better what if white won every game? I'm so glad chess is a draw!
Interested where can I find the games between the two best ever @swarminglocusts
Just YouTube it. They even played the f4 bird for a draw!
yeah!
As far as I know Alpha Zero wasn't continued for or made available for years. This youtubevideos look like cheap scam and clickbait
Engines cannot complain and are told what to do and what opening to play. I don't consider that fair play. To solve chess, engines will need to learn what tempo is really all about. Teach the engine with a tempo handicaps, not with a biased database. Avoid all human biased databases affected by the clock and the chances to draw because their rating was lower than their opponents.
Engines need to learn to outplay their opponents, by using tempo advantages, not clock advantages. The real weakness in a chess position is running out of tempos. It's like missed the bus to get to the destination. The engine must run out of tempos, not run out of time.
The clock was invented for humans limits, not engine limits. The real race is finding a solution and it does not have to be the shortest path. At the moment the CPU cores and Ram of my home computers affects the results of the games. If an engine can stalemate the opponent, I would consider that a win. But the engines don't try to stalemate. So the real limit is the rules that are causing the draws. If a piece lost it's mobility, it got over powered. So chess engines need to be made to solve for stalemates. Then you will discover a new game of chess. One that is less limited by human biases. Include the stalemate as a win and see the draws diminish. There is not many stalemate games by engines, because engines are not tuned to accept stalemates as wins. Tune the engines with that approach and then see the standings of the engines change. Modify the draw rules and chess engines might solve chess. Many factors can cause discrepancies in engine evaluations.
www.ozproblems.com/problem-world/paradox
www.chess.com/blog/Rocky64/kling-combination-a-self-stalemate-theme-that-perplexes-stockfish
woochess.com/en/lesson/chess-clocks-history
citeseerx.ist.psu.edu/document?doi=2dd285dc736fb8fa54ff5916c4797ee4736175fa&repid=rep1&type=pdf
Engines need to learn to outplay their opponents, by using tempo advantages, not clock advantages. The real weakness in a chess position is running out of tempos. It's like missed the bus to get to the destination. The engine must run out of tempos, not run out of time.
The clock was invented for humans limits, not engine limits. The real race is finding a solution and it does not have to be the shortest path. At the moment the CPU cores and Ram of my home computers affects the results of the games. If an engine can stalemate the opponent, I would consider that a win. But the engines don't try to stalemate. So the real limit is the rules that are causing the draws. If a piece lost it's mobility, it got over powered. So chess engines need to be made to solve for stalemates. Then you will discover a new game of chess. One that is less limited by human biases. Include the stalemate as a win and see the draws diminish. There is not many stalemate games by engines, because engines are not tuned to accept stalemates as wins. Tune the engines with that approach and then see the standings of the engines change. Modify the draw rules and chess engines might solve chess. Many factors can cause discrepancies in engine evaluations.
www.ozproblems.com/problem-world/paradox
www.chess.com/blog/Rocky64/kling-combination-a-self-stalemate-theme-that-perplexes-stockfish
woochess.com/en/lesson/chess-clocks-history
citeseerx.ist.psu.edu/document?doi=2dd285dc736fb8fa54ff5916c4797ee4736175fa&repid=rep1&type=pdf
What is opening theory when the dust settles? (after the "all the rages", or on average integrating over all the past rages?).
This topic has been archived and can no longer be replied to.