From the 3 ..Re8 position, the tablebases say it is a win in 31 moves.
http://www.shredderchess.com/online-chess/online-databases/endgame-database.html
From the 3 ..Re8 position, the tablebases say it is a win in 31 moves.
http://www.shredderchess.com/online-chess/online-databases/endgame-database.html
EDIT: Apparently a tablebase has been developed somewhere for KRB v KBB. That solves that question!
All 6-piece endings are publicly available in the tablebase linked by notmtwain - with the exception of 5 pieces of one color against 1 on the other side which almost always wins unless stalemate.
The solutions to all 7-piece endings have been generated as well but are only available commercially. Some users on chess.com have access and occasionaly help us with them, like n9531l.
It is always better to consult a tablebase for endgames within its range than a chess playing engine since the tablebases are 'perfect' and the engines are not.
Many pre-2000 problems and studies have been cooked (proofed wrong) by engines and tablebases. Recent ones are usually tested before publication with the help of the same scoundrels that destroyed our older creations; they have become our 'friends'.
I'm letting SF think about this for awhile. After about 15 minutes it's getting somewhere.
Eval started at -5 then went to -7, then -10 and now at a depth of 53 it's at -15.71 starting with g3+
It will eventually find a mate. I'll let you know.
To be at such a big minus scores, SF must already have won a bishop. The rest is not very interesting. The question remains "did it generate optimum counterplay?".
At a depth of 56, the eval was -54.58 .....and then my laptop battery drained when I was asleep. Oops. I woke up this morning eager to see the result, looked at my computer and rolled my eyes.
I'll set it up again today and let it run while watching the Super Bowl.
Wish I had a better computer. I'm just using an old dual-core processor and 4gb.
I made all the moves in the top line and it found a mate in 33 or there abouts. Turns out 3.Ba6 is no good. It now likes 3.Bh1. If instead 3.Be4 in an attempt to guard b1, it's mate in 46 or less with 3...Re8
Recently I came across an endgame study while reading an old endgame book (Jeno Ban's The Tactics of Endgames). The book claims the following position is a draw however Stockfish disagrees. The computer is often wrong in these kinds of endgame positions though so I wanted to see if anyone else could determine if this position is definitively lost or won. I'm not aware of any tablebases that cover this 6 piece endgame. The study is by A. and K. Sarytchev 1929. Is White drawing or does Black win?
EDIT: Apparently a tablebase has been developed somewhere for KRB v KBB. That solves that question!