Is this positions drawn ??

[Terry McCracken]

Anyone have a 7 man tablebase generator?

It's being worked on...maybe soon?

This is a very different position than the first and Whites g-pawn is a problem. This position coming in cold, would have to be analized to death.

[smirobth]

You make some good points. However there is also a certain irony in your claim about the h-pawn removal "proof" being unconvincing (even though I sympathize with that point of view). In the original position under discussion removal of Black's h-pawn did not affect the drawn outcome (it is not hard to see that the rook has plenty of squares anyway), while in this new position removal of Black's h-pawn results in a position that is no longer a draw. Part of Black's problem is that in the new position without the h-pawn Black actually has fewer tenable squares for his rook, not more, since the rook now needs to try to stay on the 6th rank, to prevent White's pawn push.

The next position illustrates my point better, and it is the one I should have posted earlier. It is Topalov - Mamedyarov shifted one rank to the left.

V.Khenkin, 1966
[d]1Q6/5pk1/4r1p1/3K2P1/8/8/8/8 w - - 0 1

Dvoretsky: As we know, a similar position, without the g6-pawn, is drawn. But here, when this pawn deprives the rook of the second protected square, White wins: he gradually approaches the black pawn using the zugzwang technique.

The main line he gives is 1. Qc7! (zugzwang) Re3 2. Qc2! Kg8 3. Kd6 Re6+ 4. Kd7 Re3 5. Qc4! (the Rook must leave the e-file) Ra3 6. Qe4 Kg7 7. Qe5+ Kg8 8. Qb8+ Kh7 9. Ke7 +-

Indeed this position is lost for Black. However I partially disagree with Dvoretsky's reason why. I believe the primary reason is that in this position Black's rook is no longer safe on e6! For example in the position given, if White's queen is on c8 instead of b8, then playing 1.Qxe6! immediately wins! After 1...fxe6 2.Kxe6 we have this tablebase win:
[d] 8/6k1/4K1p1/6P1/8/8/8/8 b - - 0 2
Tablebase win in 17

If Black had not had the g-pawn, then 1.Qxe6? fxe6 2.Kxe6 leads to an equivalent position where Black has one pawn less ... but now can draw the game!

[d] 8/6k1/4K3/6P1/8/8/8/8 b - - 0 2
Tablebase draw

I should add that although in the 2 positions above I show it as Black to move, since White just played Kxe6, the results of neither position would change, even if it were White to move (thus it is not only when Black can capture White's pawn that the ending is drawn)

Thus it is largely the king and pawn ending where Black's extra pawn hurts his defense. In the original position where White's pawn is on the h-file, the king and pawn ending is always an easy draw since Black's king is nearby.

But never-the-less the main point of your post is still good.

[Terry McCracken]

Thanks for your insult. Removing the h6 pawn changes nothing, except one more redundant protected square.

I didn't mean to insult you and now that I have reread my post, I don't think I did. I just disagreed with you.

It doesn't matter. I've been reactive due to people who wanted scientific proof whether the initial position was drawn by force.

[Terry McCracken]

You make some good points. However there is also a certain irony in your claim about the h-pawn removal "proof" being unconvincing (even though I sympathize with that point of view). In the original position under discussion removal of Black's h-pawn did not affect the drawn outcome (it is not hard to see that the rook has plenty of squares anyway), while in this new position removal of Black's h-pawn results in a position that is no longer a draw. Part of Black's problem is that in the new position without the h-pawn Black actually has fewer tenable squares for his rook, not more, since the rook now needs to try to stay on the 6th rank, to prevent White's pawn push.

The next position illustrates my point better, and it is the one I should have posted earlier. It is Topalov - Mamedyarov shifted one rank to the left.

V.Khenkin, 1966
[d]1Q6/5pk1/4r1p1/3K2P1/8/8/8/8 w - - 0 1

Dvoretsky: As we know, a similar position, without the g6-pawn, is drawn. But here, when this pawn deprives the rook of the second protected square, White wins: he gradually approaches the black pawn using the zugzwang technique.

The main line he gives is 1. Qc7! (zugzwang) Re3 2. Qc2! Kg8 3. Kd6 Re6+ 4. Kd7 Re3 5. Qc4! (the Rook must leave the e-file) Ra3 6. Qe4 Kg7 7. Qe5+ Kg8 8. Qb8+ Kh7 9. Ke7 +-

Indeed this position is lost for Black. However I partially disagree with Dvoretsky's reason why. I believe the primary reason is that in this position Black's rook is no longer safe on e6! For example in the position given, if White's queen is on c8 instead of b8, then playing 1.Qxe6! immediately wins! After 1...fxe6 2.Kxe6 we have this tablebase win:
[d] 8/6k1/4K1p1/6P1/8/8/8/8 b - - 0 2
Tablebase win in 17

If Black had not had the g-pawn, then 1.Qxe6? fxe6 2.Kxe6 leads to an equivalent position where Black has one pawn less ... but now can draw the game!

[d] 8/6k1/4K3/6P1/8/8/8/8 b - - 0 2
Tablebase draw

I should add that although in the 2 positions above I show it as Black to move, since White just played Kxe6, the results of neither position would change, even if it were White to move (thus it is not only when Black can capture White's pawn that the ending is drawn)

Thus it is largely the king and pawn ending where Black's extra pawn hurts his defense. In the original position where White's pawn is on the h-file, the king and pawn ending is always an easy draw since Black's king is nearby.

But never-the-less the main point of your post is still good.

Well done Robin!

Terry