But yeah, I'd never seen that specific trick before, and at the speed I was playing I didn't see it coming until a move or so before it landed on the board. Stockfish is just ridiculously persistant and tricky in these kinds of endings.
That was a wonderful exposition of this method of endgames training: it made me want to do the same! Also, I'm impressed by your perseverance in trying to crack this endgame, day after day. Also, thanks for the recommendation of "Move List" resource, I'd come across it previously, but forgot by now, it's a very interesting one: I'm currently looking at his approach to repertoire building.
Thanks! I haven't done much training against the computer from practice positions in the past, but it's clearly a really good method - it forces you to think actively rather than consume knowledge passively. When I was a more active chess player I did a bunch of the studies in the Dvoretsky Manual, but the other benefit of the MoveLibrary positions is that you get repetitions in against a strong/invincible opponent. It's nice to get the mistakes out of your system when rating points and tournament standing isn't on the line.
Usually when I do this exercise, I alternate playing the white and black pieces, slowly learning how to play the ending correctly for whichever side I was originally trying to win or hold with.
Occasionally I've tried doing this in an ending where the computer just beats me repeatedly from both sides, which is always a maddening and humbling experience. But at least once (a few years ago, maybe) I found an endgame where *I* could actually beat Stockfish (or whatever lichess uses) from both sides, which was delightful. I think it was some bishop endgame with a really long horizon.
Also, I love Dvoretsky's book. In my last tournament I was able to save a lost position by going for a two-pawn-down rook endgame where I was employing the "the frontal attack" against a g-pawn while my king blockaded a d-pawn. It was still winning for my opponent because my king couldn't shuffle between the 6th and 5th rank, but he couldn't figure out how to win it and eventually gave up his d-pawn, which made it a draw. Felt so good.
Yeah, my method was pretty ridiculous: I was basically blitzing my way through the black side of the endgame over and over again without taking much time to reflect or improve my play. That said, getting in a lot of reps led to a fair amount of comfort with the position overall - I stopped thinking of it as unpleasant by the end.
There's nothing quite like holding a bad or lost endgame - reminds me of surviving against Onsichuk at the Amateur Team East back in the day.
Interesting post Andy, you've got more patience than I do. That Rc7! "trapping" the Black rook on c5 is a fun new pattern to keep in mind though.
I think you're mispronouncing stubbornness :)
But yeah, I'd never seen that specific trick before, and at the speed I was playing I didn't see it coming until a move or so before it landed on the board. Stockfish is just ridiculously persistant and tricky in these kinds of endings.
That was a wonderful exposition of this method of endgames training: it made me want to do the same! Also, I'm impressed by your perseverance in trying to crack this endgame, day after day. Also, thanks for the recommendation of "Move List" resource, I'd come across it previously, but forgot by now, it's a very interesting one: I'm currently looking at his approach to repertoire building.
Thanks! I haven't done much training against the computer from practice positions in the past, but it's clearly a really good method - it forces you to think actively rather than consume knowledge passively. When I was a more active chess player I did a bunch of the studies in the Dvoretsky Manual, but the other benefit of the MoveLibrary positions is that you get repetitions in against a strong/invincible opponent. It's nice to get the mistakes out of your system when rating points and tournament standing isn't on the line.
Usually when I do this exercise, I alternate playing the white and black pieces, slowly learning how to play the ending correctly for whichever side I was originally trying to win or hold with.
Occasionally I've tried doing this in an ending where the computer just beats me repeatedly from both sides, which is always a maddening and humbling experience. But at least once (a few years ago, maybe) I found an endgame where *I* could actually beat Stockfish (or whatever lichess uses) from both sides, which was delightful. I think it was some bishop endgame with a really long horizon.
Also, I love Dvoretsky's book. In my last tournament I was able to save a lost position by going for a two-pawn-down rook endgame where I was employing the "the frontal attack" against a g-pawn while my king blockaded a d-pawn. It was still winning for my opponent because my king couldn't shuffle between the 6th and 5th rank, but he couldn't figure out how to win it and eventually gave up his d-pawn, which made it a draw. Felt so good.
Yeah, my method was pretty ridiculous: I was basically blitzing my way through the black side of the endgame over and over again without taking much time to reflect or improve my play. That said, getting in a lot of reps led to a fair amount of comfort with the position overall - I stopped thinking of it as unpleasant by the end.
There's nothing quite like holding a bad or lost endgame - reminds me of surviving against Onsichuk at the Amateur Team East back in the day.