Promoting chess discussion.
|
|
Chess Forum Promoting chess discussion. |
I have tried to download this kind of endgame from the Robert M. Hyatt Site and similar sites with 6 pieces endgames, but i can't find it. With the file Tbgen.exe or tbgenm.exe i can't create endgames more than 5 pieces and i don't understand it, there are some endgames done more complicated to calculated like KQQKQQ for example- Does anyone can help me or explain me why this endgame can't be created or downloaded? Thanks. |
> Hello. > > I have tried to download this kind of endgame from the Robert M. Hyatt > Site and similar sites with 6 pieces endgames, but i can't find it. > With the file Tbgen.exe or tbgenm.exe i can't create endgames more > than 5 pieces and i don't understand it, there are some endgames done > more complicated to calculated like KQQKQQ for example- > > Does anyone can help me or explain me why this endgame can't be > created or downloaded? It _isn't_ generated yet. Generating 6 man takes a lot of time - see tbgen generate a five piece as compared to a four or three pieces and you can guess what time and machine it takes for 6 man TB's. Tbgen wasn't developed for generating 6 men TB's, so you can't use it for more than 5 man. -- CeeBee Archer: "Houdini could get out of this." T'Pol: "Perhaps you should invite him on your next mission." Picard: "Mr. Worf, do you know Gilbert and Sullivan?" Worf: "No sir, I have not had a chance to meet all the new crew members since I have been back." |
>> complicated to calculated like KQQKQQ for example- For KRPKRP you need -all- promotion cases too: KRQKRS for instance. For a human, KQQKQQ may be more difficult, but for a computer pawn endgames are far more difficult, due to the promotion, and due to symmetry. Joost -- Du hast mein Herz zerrissen, meine Seele geraubt Das es so enden würde hätt` ich nie geglaubt [Aus der Ruinen -] Ohne Rücksicht auf Verluste, hast Du meine Welt zerstört [L'Âme Immortelle] Eine Welt, die vor kurzem nur uns beiden hat gehört |
rec.games.chess.misc: > For KRPKRP you need -all- promotion cases too: KRQKRS for instance. > For a human, KQQKQQ may be more difficult, but for a computer pawn > endgames are far more difficult, due to the promotion, and due to > symmetry. IIRC Crafty has some code to circumvent promotions - i.e. the readings from the TB in question doesn't screw up due to missing follow-up TB's. I don't know if there are other programs capable of this as well, but strictly it might not be necessary to have all promotional cases in TBs. -- CeeBee Archer: "Houdini could get out of this." T'Pol: "Perhaps you should invite him on your next mission." Picard: "Mr. Worf, do you know Gilbert and Sullivan?" Worf: "No sir, I have not had a chance to meet all the new crew members since I have been back." |
> the TB in question doesn't screw up due to missing follow-up TB's. I don't > know if there are other programs capable of this as well, but strictly it > might not be necessary to have all promotional cases in TBs. But to calculate the table in the first place, you do need the followup tables. Joost -- Du hast mein Herz zerrissen, meine Seele geraubt Das es so enden würde hätt` ich nie geglaubt [Aus der Ruinen -] Ohne Rücksicht auf Verluste, hast Du meine Welt zerstört [L'Âme Immortelle] Eine Welt, die vor kurzem nur uns beiden hat gehört |
rec.games.chess.misc: > But to calculate the table in the first place, you do need the > followup tables. True. -- CeeBee Archer: "Houdini could get out of this." T'Pol: "Perhaps you should invite him on your next mission." Picard: "Mr. Worf, do you know Gilbert and Sullivan?" Worf: "No sir, I have not had a chance to meet all the new crew members since I have been back." |
> > IIRC Crafty has some code to circumvent promotions - i.e. the readings from > > the TB in question doesn't screw up due to missing follow-up TB's. I don't > > know if there are other programs capable of this as well, but strictly it > > might not be necessary to have all promotional cases in TBs. > > But to calculate the table in the first place, you do need the followup tables. > > Joost Hello Again. Yes, but these tables have been made, in fact, in the site there are all endgames of 3 pieces vs 3 pieces, but only the endgames without pawns. Maybe it's hard to calculate for the computer, i see that the computer spends more time and need more memory to calculate endgames with pawns. Anyway that's a pitty, i have calculated the probability in my games-database, and KRP vs KRP is one of the most played endgame, of course, in the first position is KRPP vs KRP, but it seems me that we won't see this endgame in the nalimov tables for many years. greetings. |
> Yes, but these tables have been made, in fact, in the site there are > all endgames of 3 pieces vs 3 pieces, but only the endgames without > pawns. Maybe it's hard to calculate for the computer, i see that the > computer spends more time and need more memory to calculate endgames > with pawns. Eugene Nalimov once mentioned that generating, compressing and verification took up to five weeks for one tablebase and -again- IIRC needs a 64 bit system/OS. -- CeeBee Archer: "Houdini could get out of this." T'Pol: "Perhaps you should invite him on your next mission." Picard: "Mr. Worf, do you know Gilbert and Sullivan?" Worf: "No sir, I have not had a chance to meet all the new crew members since I have been back." |