It seemed that Magnus realised as early as move 19, that even 3 vs 4 pawns with better structure, he could draw the game. Game 2, Magnus traded the entire left side, and didn't change his pawn structure at all, even though commentators thought it was a mistake. So can't you assume those moves are no good, and prune to actually practical positions, that would be nearly as good as solved chess?īut why do you need perfect play? Wouldn't 1000th best route still be good, as long as it leads to a win? I have trouble seeing why forcefully trading would lead to worse position, because if it would lead to losing position, you would surely trade differently? Maybe i'm still underestimating how much calculation is there to force trading. And the vast majority of moves would result in absurdly high advantages or outright mates for the other side. If you followed up only the most promising leads? Take the best engine line, calculate just before it becomes too exponential for current hardware, take the best 1 or 2 lines, do it again, until you reach a solved position? Wouldn't that prune a lot unnessary lines?ĭuring this last match, was looking at sesse, and it could in nearly real time analyze 30-50 ply depth. Solve or analyse extremely deeply only a select few lines? Can you have a solved thread, linking endtable and an opening? Since you would be ignoring vast majority of positions, can it be possible have a few openings solved? I guess a couple variations of this: Btw, does anybody know if anybody is close to compiling 8 man end game tablebase? While I don't see chess being solved any time soon, can't it be approximated? For example:įorcefully trading pieces until you're left with a solved position? Is there any way your opponent can avoid trading pieces if you are determined to do so? And since we have 7 man end tables, 8 man probably with a large percentage solved, and many other select and more likely endgames solved as well, is it feasible to force trade into solved winning/drawn position? That should prune a lot of nodes, and significantly speed up calculations.
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |