Chess is a game that has been around for over 1,500 years. The version of the game we know showed up roughly 600 years ago. Yet, despite its age, chess remains unsolved. The term "solved" when we're talking about a game or gaming in general, refers to the state a game reaches when all the possibilities have been explored and we have a concrete idea of what works and what doesn't. All games can experience some level of "solvability" including popular video games and board games. Some are even already solved completely.
For example, games like tic tac toe are what we would call "solved". If you know what you're doing in tic tac toe, it is impossible to lose. If two individuals know to avoid the obvious losing moves and common tactics then the game would simply end in a forced draw every time and there is no way to force a win. With perfect play, both players can force a draw. However, there are only 255,168 iterations of possible games in tic tac toe, and while for many people that number seems quite large, it's not impossible to figure it out after discounting the obvious losing strategies and simply "brute forcing" each game with calculation to avoid losing. "brute force" refers to a type of calculation in games in which players consider every possibility that exists to find the correct solution. Once you learn the basic rules and strategies of a game like tic-tac-toe, it becomes much easier to draw constantly.
What about a game with more possibilities though? Let's take checkers, which has around 500 billion billion (or 5 quintillion) possible positions. Yet, in 2007, the game was solved using AI computer models. In fact, according to the project's lead, Jonathan Schaeffer, the starting position of checkers is a draw is played perfectly. However, this computation took 16 years (1989-2007) and the model didn't begin clearly outperforming human players until 1996. The way this was done was by beginning at the "endgame". The endgame refers to a period of play in which only a few pieces are on the board. By essentially reverse engineering checkers, the model was then able to reverse the search once it reached the starting position. By searching all the possibilities again, this time from front-back, the model was able to "prove" that its solution to checkers was true. You can read more about the model that accomplished this "solve" here: https://www.researchgate.net/publication/231216842_Checkers_Is_Solved
Chess is undergoing a similar pursuit, this is called Tablebase. Tablebase is an effort to brute force calculate chess from 3 pieces on the board to 32 pieces. Currently, positions of 8 or more pieces on the board (including pawns), there is no concrete formula or solution to approaching the game at this stage. There's still an element of uncertainty to these positions. But Tablebase has effectively explored all positions with 7 or less pieces on the board. This means that with 7 or less pieces on the board, chess is solved. The tablebase database has thoroughly examined every position possible that it can tell whether or not a move is a win/loss/draw for either side with perfect play.
However, this process has a major flaw when it comes to realistically solving chess in our lifetimes... The number of possible iterations and possibilities drastically increases with each piece added. The 7 Piece tablebase finished in 2012, so where's the 8 Piece tablebase? It's STILL being worked on. Such a solution is not easy. In an interview with Chessbase India in 2022, Marc Bourzutschky who is working on contributing to the table base said there are no fewer than 4,795 iterations with 8 pieces on the board. It could take days to solve an iteration. Let's assume 3 days per iteration on average which would result in close to 40 years of constant calculation. This is not the only issue, while multiple computers with ridiculous amounts of computing power could help reduce the amount of time theoretically, there's still no way for these computers to effectively share the information and storage is a big problem too. The 8 piece tablebase is projected to contain 700 Terabytes of information and relies on information from previous tablebases to help computations. For reference, the office in which I work contains 10 Terabytes of data across all computers and servers. The 7 piece tablebase was 18 Terabytes meaning you need a computer even most companies wouldn't need access too to run this information. This project, unfortunately, has neither the funds or infrastructure to make this go faster.
Without rules and regulations chess would have 10^123 power of possible positions (more possible positions than the number of atoms in the universe), with rules and regulations that number drops in shades to 10^40th power (more possible positions than the number of atoms in Saturn). In other words, a computer system would have to be able to contain a database the size of Saturn in order to completely solve chess via the tablebase. This most likely means that chess will not be solved anytime in the near future. It can also be a daunting thought for players of the game.
Without rules and regulations chess and would have 10^123 power of possible positions (more possible positions than the number of atoms in the universe), with rules and regulations that number drops in shades to 10^40th power (more possible positions than the number of atoms in Saturn). In other words, a computer system would have to be able to contain a database the size of Saturn in order to completely solve chess via the tablebase. This most likely means that chess will not be solved anytime in the near future. It can also be a daunting thought for players of the game.
For example, take the position below where it's black to move:
This position is quite tense, each side has two passed pawns on the kingside but the white pawn on e7 is very close to promoting to a queen. Black can quickly realize that his pawns are perfectly positioned on light squares which cannot be reached by the bishop. By evaluating that this pawn on e7 is the most important threat and that moving their own pawns is fruitless, we can almost instantly narrow down this position to only two possible moves. Either Nc7 guarding e8 as the promotion square or Kf7. Kf7 is clearly the best move because it gives us the opportunity to win the pawn whereas Nc7 freezes the black knight to prevent the promotion of the pawn long-term. By using basic chess principles and experience we were able to find the only winning move from back in just a quick moment. The more you do this and study these types of position your understanding will improve.
This is also the same principle with life in general. Life is full of endless possibilities and choices that we have to make. Some we will only see once or twice and others we will have to make daily. We make those decisions by using past experiences and principles we've learned throughout life. Chess is such a unique game because it allows us to practice the same thought process we use in every day life and teaches real world principles.
Comments