Talk:Solved game

Definition of weak solution

Some time ago, an inconsistent, unreferenced, second definition of weak solution was added - "That is, produce at least one complete ideal game (all moves start to end) with proof that each move is optimal for the player making it." Contrary to this claim, it is in general possible to have an optimally played example of a game, an (indirect, perhaps) proof that the example game is optimal, but no strategy for playing a position that does not occur in the game, so this claimed alternative definition is not appropriate. Elroch (talk) 21:42, 1 March 2022 (UTC)

@Elroch: I have moved this new section to the bottom of the page, as is traditional here. About the removal, I don't understand your objection (to me, the two phrases appear to say the same thing) but I also don't see the removed text as particularly beneficial so I have no problem with your removal. --JBL (talk) 22:26, 1 March 2022 (UTC)

My objection to the unsourced claim is stated in the sentence starting "Contrary to this claim" above. The deleted criterion seems obviously much weaker than the exhibition of practical strategies (as achieved for checkers, for example). Note that the difference is a matter of practicality. I can describe an algorithm to play chess optimally as white or black - construct a 32-piece tablebase and play according to it - but this is not practical to execute. If the only information in a solution is the value of the game, there can be no distinction between the definitions. Thus the difference can be practical rather than mathematical. Elroch (talk) 15:04, 2 March 2022 (UTC)

Changed the word "moves" to "play" in the definition. There is nothing in the prequel to suggest the game is discrete. E.g. many people will have weakly solved the cat and mouse game where the initial position has the mouse in the centre of a circular pool and a cat at the edge who can run four times as fast as the mouse can swim but not as fast as the mouse can run. It has no consecutive positions so no defined moves. Changed "moves" to "details" in the ultra-weak definition correspondingly. Martin Rattigan (talk) 22:16, 25 May 2024 (UTC)

"ultra-weak" proofs are not the deepest and most interesting

The paragraph starting with "Despite their name, many game theorists believe that "ultra-weak" proofs are the deepest, most interesting and valuable" contains wrong information. According to their definitions, a "strong" proof will always count as an "ultra-weak" proof, so that sentence as written is trivially false. I believe that paragraph is trying to say that a non-constructive proof is deeper and more insightful than a proof constructed by brute-force search, even if the first is ultra-weak and the second is strong. If that's the case, it should be changed to say that. But since that paragraph is already marked with a "citation needed" tag, I believe it should be simply removed unless a source for that argument is found. RRM (talk) 01:41, 5 November 2023 (UTC)

Kalah

Just wanted to provide some context to my previous edit (first day contributing to a wiki articles).

In alignment with revision 1057486730 of the main Kalah page, I have removed reference to original research and a computer program by Mark Rawlings. These contributions did not comply with Wikipedia's No original research policy. :(

I'd like to encourage Rawlings to publish his research in solving Kalah variations. I believe his work is interesting and could be very valuable. The crate catsby (talk) 04:14, 28 April 2024 (UTC)