Solutions for 8/1p2k3/8/8/5Q2/8/ppp1p3/qrrbK3 w - - 0 1

[fkarger]

Hello Frank:

Maybe there is an error in that game.
Because black was able to play Kg2.
If I understand the position you cant force the black King to leave that corner
once it stands near to or on h1.

We discussed a problem called 'The Untouchable King' a month ago, that dealed with the mating king side not allowed to move (like here) and the queen doing all the work, mostly staying at knight jumps (like here again):

Re: Minichess puzzle.

There were ways to force the mated king side to leave corners and drove it anywhere. I do not know if that can be applied here since the chessboard is more populated than in that problem.

Regards from Spain.

Ajedrecista.

Hi Ajedrecista,

thank you for the hint!
I think in this case it is not possible to force the black King to leave the corner
because the usual pattern to do that with the Queen alone does not work:
you would need Qf1 once Black plays Kh2, but there is a Pawn on e2...

Best regards

Frank

[peter]

Maybe there is an error in that game.
Because black was able to play Kg2.
If I understand the position you cant force the black King to leave that corner
once it stands near to or on h1.

Give your line longer than 50 moves to mate from latest pawn- move.
I'm successful with Chest so far backwarding the CorChess' line till move 12 of Black's. There:

[d]8/8/8/1p4Q1/8/5k2/ppp1p3/qrrbK3 w - -

Engine: ChestUCI Ver.5.2 (0 MB)
von Franz Huber

40.00 0:00 +M40 13.Dg7 Kf4 14.Dg2 Kf5 15.Df3+ Kg5
16.De4 Kh5 17.Df4 Kg6 18.De5 Kh6
19.Df5 Kg7 20.De6 b4 21.Dd6 Kh7
22.Df6 Kg8 23.Dg6+ Kh8 24.Df7 b3
25.Df8+ Kh7 26.Df6

Bester Zug: Dg5-g7 Zeit: 0:00.000 min


That would be only 52 all in all from starting with 1.Qc7+?!, so there will be Black blunder(s) in first 12 moves probably, yet to prove them such (giving away the 50 moves invetably) I still see as rather difficult

[fkarger]

Probably black should avoid to move the b Pawn like 6...b6.

I see two defences for Black: run to h1 or run to a8.
In both cases I dont see how you can stalemate the black King.

So 6...b6 could be a blunder.
Better move the King closer to h1.

[peter]

Probably black should avoid to move the b Pawn like 6...b6.

I see two defences for Black: run to h1 or run to a8.
In both cases I dont see how you can stalemate the black King.

So 6...b6 could be a blunder.
Better move the King closer to h1.

It won't help to just give single alternative moves for one side without the line following with best moves from both sides till mate. Or do you really doubt the mate to be possible at all after other one first moves but 1.Qd2, don't you?
If it's only about the 50 moves (which aren't of special interest for mate in x- puzzles normally anyhow), you have to give best DTM of your personal analysis from the move you change in given line onwards, regards

[lech]

And a little longer DTM after 1.Qc7 Ke6:

[d]8/1pQ5/4k3/8/8/8/ppp1p3/qrrbK3 w - -

Engine: Sting10x (8192 MB)
von Marek Kwiatkowski
...
29.00 0:00 +M203++ 2.Dd8 Ke5 3.Dd7 Ke4 4.Dd6 Kf5 5.Dc5+ Ke6
6.Dd4 Kf5 7.De3 Kg4 8.Df2 b6 9.Dg2+ Kf4
10.Dg1 Ke5 11.De3+ Kd5 12.Df4 (3.299.150) 15785

Then I tried to let Sting play out on its own after 1.Qc7+ Ke6, but that failed, because with 10'+2" Sting comes to 0.00- eval soon, obviously seeing 50 moves- rule getting into horizon, not knowing, it doesn't count here, because we judge with "puzzle- rules", regards

BTW. Why you stop the search at depth 29 after time 0.00 ?
You don't like me or Sting ?

[peter]

And a little longer DTM after 1.Qc7 Ke6:

[d]8/1pQ5/4k3/8/8/8/ppp1p3/qrrbK3 w - -

Engine: Sting10x (8192 MB)
von Marek Kwiatkowski
...
29.00 0:00 +M203++ 2.Dd8 Ke5 3.Dd7 Ke4 4.Dd6 Kf5 5.Dc5+ Ke6
6.Dd4 Kf5 7.De3 Kg4 8.Df2 b6 9.Dg2+ Kf4
10.Dg1 Ke5 11.De3+ Kd5 12.Df4 (3.299.150) 15785

Then I tried to let Sting play out on its own after 1.Qc7+ Ke6, but that failed, because with 10'+2" Sting comes to 0.00- eval soon, obviously seeing 50 moves- rule getting into horizon, not knowing, it doesn't count here, because we judge with "puzzle- rules", regards

BTW. Why you stop the search at depth 29 after time 0.00 ?
You don't like me or Sting ?

I like both of you very much, Marek.
It just took quite a while for next output- line of Sting's at that point, so I copied it after having waited some time, really don't remember, how long exactly it took me, till I did that, but for sure not exactly after 0.00", as fast as that I'm not at my advanced age

Problem here is, as you'll probably have seen already too, Sting doubt's the mate within 50 moves- boundary and changes eval between DTM- one and 0.00, as well in outplaying (with not very long TC) as in infinite analysis.

How about your trials to the given position? Do you think there isn't any mate to be reached after other first moves but 1.Qd2 or do you think it's the only solution as for DTM shorter than 50 moves to latest pawn- move?
You're the expert with Sting of course, so please let it answer the question about best line after e.g. 1.Qc7+? in any way you think best usage of the engine here.

Thanks for the latest update of your fine engine btw and for the long lasting great work of yours with it. I'd really regret if you'd stop developing it, as you wrote in the other one thread.

[peter]

[d]8/8/8/1p4Q1/8/5k2/ppp1p3/qrrbK3 w - -

Engine: ChestUCI Ver.5.2 (0 MB)
von Franz Huber

40.00 0:00 +M40 13.Dg7 Kf4 14.Dg2 Kf5 15.Df3+ Kg5
16.De4 Kh5 17.Df4 Kg6 18.De5 Kh6
19.Df5 Kg7 20.De6 b4 21.Dd6 Kh7
22.Df6 Kg8 23.Dg6+ Kh8 24.Df7 b3
25.Df8+ Kh7 26.Df6

Bester Zug: Dg5-g7 Zeit: 0:00.000 min

One more move backwards, neither Chest nor I see a way (broke off in depth 43) to get mate after 12...Kh1! instead of Kf3?, so here CorChess blundered for sure and it would still be necessary to have a better way for White to avoid the 12...Kh1 for proving mate after 1.Qc7+?! (or yet ? if it changed the outcome at least with 50 moves boundary yet counting).
I'm outa here now, having to admit, Sting could well be right with 0.00- eval after 1.Qc7+ yet too, as long as I can't prove the opposite, I appologize to Marek

[peter]

It really fooled me for quite a while way back then and so I had to return to it again. Traps to be seen clearest are such like this one line:

1.Qc7+? Ke6 2.Qc5 and now 2...Kf6? (in main line at 1.move, that's ok, all moves lose anyway) because here White can win with 3.Qd5!, reaching same position as after 1.Qd2 Kf6 2.Qd5.
If Black answers correctly to 1.Qc7+ Ke6 2.Qc5 Kd7! the triangle maneuver of White has failed and it's drawn.

I'd still like to know the author of this very fine puzzle, I know it's in ACT- suite and I guess I saw it in Glarean once in a while too, but neither in HHdbVI nor with YACPDB it's to be found, regards

[fkarger]

Unfortunately I also dont know much more about the puzzle.

I agree it is a very fine and subtle one!
In fact it is the only one I do not fully understand despite deep analysis.

[Ajedrecista]

Hello:

[...]

I'd still like to know the author of this very fine puzzle, I know it's in ACT- suite and I guess I saw it in Glarean once in a while too, but neither in HHdbVI nor with YACPDB it's to be found, regards

Unfortunately I also dont know much more about the puzzle.

[...]

I searched the authorship before Peter's post and I did not find anything, sadly. It is an excellent puzzle, indeed.

I am remembering right now a puzzle at TalkChess with white to play and win where the first move was Ke1 (I would say from d2) with a similar idea, with lots of black pieces on the kingside, all blocked. I think that a white a-pawn could run and promote, then pursue the black king through the board at knight jumps, to finally giving checkmate with something like ..., Kh3; Qh5#. However, I can not find it. I do not remember if it had such subtleties.

Regards from Spain.

Ajedrecista.