Fortress?

Apart from the blunder, what do you think?

The evaluation is high but if the position cannot be improved it's no ounce better than 0.00.

Tournament game of mine last night, classical time control, somewhat lucky win after all.

I think even after 59 Kxd6 black should win. Black should play with king and queen on black squares. As white has the bad white bishop, it is almost KQ vs. KR. Ultimately the advance g6-g5-g4 and then h5-h4 should win.

Yes, the given analysis line by sf looks good. Starting with ...Qb8! Actually in mutual time-trouble even three results are in the air... (amongst humans)

Why the name "Fortress"??

For example, if White plays the expected 46.Kb4 - can Black win? Ist it a fortress for white? There are probably some other good opportunities as well.

lichess.org/editor/8/8/3pB1pp/3P4/1K3kP1/1R5P/2q5/8 b

It has some fortress potential - in my opinion.

Just for the record, 25 years ago (1993) I played the KID Saemisch with ...e5 against the same opponent.

Those were the days!

White is lost in this endgame, Queen is very hard to beat with rock, so, black must win

Yeah, but imagine White‘s keeping the rook on the third rank covering h3, Be6 protecting d5. Probably Zugzwang, possibly draw, but winning is not that easy.

If the position is +2 till the cows go home it is: draw

Recently I played a game which looked like a fortress in the end. We drew, as both were of the same opinion, but actually it is not. Interesting is the way to win: black should stalemate white king with queen alone which will force rook to move.

@Sarg0n Do you prefer the c5 or Nc6+e5 Samisch approach? And in either case, where did you learn the theory? As a KID player myself, the Samisch is a variation I really struggle with.