it says that stockfish8 is 2500rating but i beat him easily,but i cant even beat an 2000 rated human...(sorry for bad english)
i think the rating is for normal chess. im not sure though.
lol
Why the lol? I think @legend is right.
I don't know anything about atomic chess or Stockfish's skill in this variant but you played Stockfish seven times and you lost 6 times. I think this A.I. is a good opponent for you.
And remember: Toadofsky and ubdip are improving the engine almost on a daily basis
And remember: Toadofsky and ubdip are improving the engine almost on a daily basis
Not sure if trolling or ....
Hoping he is a just a troll though.
Hoping he is a just a troll though.
A few remarks from my side:
- Stockfish definitely is weaker in atomic chess compared to standard chess and maybe also compared to crazyhouse or antichess.
- There are a few major weaknesses that are easy to identify, e.g., evaluation of endgame positions, and finding threats by indirect king attacks, but suprisingly tests reveal that patches that try to fix these issues do not improve or even regress playing strength (in selfplay), even though I had expected huge Elo gains. This is very confusing to me, and has not happened to the same extent for other variants, but anyway we still keep improving Stockfish for atomic, e.g. about 150 Elo in the last two months. If you have observed specific weaknesses where Stockfish has room for improvements, please let me know.
- You are rated ~1900 and scored ~14%, so the estimated rating of Stockfish would be around 2200. Furthermore, keep in mind that playing strength of humans as well as engines depend on the time control.
- White has a big advantage in atomic chess, so players are sometimes able to score wins as white against stronger opponents.
- Stockfish definitely is weaker in atomic chess compared to standard chess and maybe also compared to crazyhouse or antichess.
- There are a few major weaknesses that are easy to identify, e.g., evaluation of endgame positions, and finding threats by indirect king attacks, but suprisingly tests reveal that patches that try to fix these issues do not improve or even regress playing strength (in selfplay), even though I had expected huge Elo gains. This is very confusing to me, and has not happened to the same extent for other variants, but anyway we still keep improving Stockfish for atomic, e.g. about 150 Elo in the last two months. If you have observed specific weaknesses where Stockfish has room for improvements, please let me know.
- You are rated ~1900 and scored ~14%, so the estimated rating of Stockfish would be around 2200. Furthermore, keep in mind that playing strength of humans as well as engines depend on the time control.
- White has a big advantage in atomic chess, so players are sometimes able to score wins as white against stronger opponents.
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
Takeback proposition accepted
#8 Also Stockfish has no opening book (and doesn't use the Opening Explorer).
#1 Interesting observation. I'm unsure how the ratings are selected. Take a gander through Advanced Search and advise what you think an appropriate rating would be...
#1 Interesting observation. I'm unsure how the ratings are selected. Take a gander through Advanced Search and advise what you think an appropriate rating would be...
This topic has been archived and can no longer be replied to.