Of course I've seen a fortress like this somewhere before -- it's even probably farily well known -- but I have no experiance other than vaguly remembering it's supposed to draw. Any advice on general considerations would be appriciated :)
Fortress or not? R+B vs R
Sort:
tonydal wrote:
The starting position must be winnable.
You're a master, surely you're able to elaborate. If trading pawns doesn't help, and white can't win the f7 pawn, please let me know how you expect to win this.
I'm no master, and my analysis leaves out a lot of technique, but I tried the ideas I thought of to no avail -- I posted it here for help -- someone help me out :)
tonydal wrote:
I expect to win because I'm a piece up. As long as I avoid positions with the wrong rook pawn, I can't imagine how Black should be able to draw. Obviously it's going to be a grind...but White should come out on top (some way or other). And it's much too complicated (for me anyway) to analyze in some general way.
Well you only have one rook pawn and it's the wrong one :) So no passers are likely -- but I understand what you're saying, trading down like I did was too trivial.
Obviously white is up a piece, but the pawnless endgame is drawn. I understand you don't want to throw around general ideas, but it seems like white is running out of them. To say it very generally, white must use king and pawns (without trading down to the bad rook pawn) to limit space and create a zugzwang.
***Edit
This fantasy position seems to work for white. The variation I give of course is bad :). This should work fine as a win for white. Like you said it'd be a grind OTB, but I'm fairly convinced it's not a real fortress.
If you Fritz it you would know for sure.
zxb995511 wrote:
If you Fritz it you would know for sure.
That's the problem with programs though, They're worthless at endgames that don't have a chain of forcing moves (like pawn endgames). I drew the original positoin against Rybka3 on quad cpu about a dozen times over the course of an hour before I decided to make a post.
Of course it thinks it's winning, but mindlessly shuffles pieces back and forth. When the 50 moves rule approaches it moves a pawn, that's about the extent of it's technique.
The same for any theoretical or book draw -- programs immediatly let you have them because they think they're winning. That's why they come with TBs, to help them not be so stupid :).
Never throw an endgame into a program to look at the evaluation. Always analyze it yourself first.
In a different topic rednblack asked about an endgame. I set off to try to draw for black against my prog Rybka (of course you have to help computers find useful moves in endgames -- that's why I need help) and after a few tries I ended up with the following position which I think can be drawn. I'm wonding if someone with more experiance with this type of position could tell me if this is a fotress or not?
So in either final position h6 is too weak for white for him to make use of the passer right? Or was there a better way for white to handle the original position?