A blockhead blunder!

[Blockhead]

www.stansco.com/cgi-bin/nc_game.cgi?217652

The Ancient Brit (TAB) has bullied me into posting this game here, as follows:

BH: As you advised, I may submit it to the SNC MB.
TAB: I really think that is the best place for an answer. I predict you will get a big response.
TAB: Go on, give the ‘boys’ something to play with, they will love your for it.

Can anyone find a win for black?

This is a lose-lose situation for me. Having a clearly won game I blundered it away, and now, if someone comes up with a winning continuation I'll feel doubly inept and stupid and ever the blockhead!

I think I hate TAB!

[savin]

Blockhead,

That is a fun poistion. Its a draw. There does not seem to be anywhere you can move the king that does not allow white to check in a way that would allow you to either block or take with the rook.

[torphichen]

I agree with Savin and have done a little analysis.

White's initial objective will be to force you to either the h-file or the d-file.
On the h-file you will not be able to avoid g-file checks.
On the d-file he can check from the e-file at will. If you ever play Ka/b/c/d2, Re2+ forces you to take the rook whilst maintaining the stalemate.

If the rook is allowed to start checking the king 'above' (4th/5th rank with K on 3rd or 5th with K on 4th) the repetition rule will eventually force you do to one of these unless you eventually play Ke2 Re4/5+ Kf1, where Re1+ again forces the capture.

From the final position this immediately rules out Kf/g3 due to Rf/g4+. Kf5 also hits the same problem as after Re5+ Kf/g4 Rf/g5+ the rook has achieved its ideal position.

This leaves only Kg5
Rg4+ Kf5 (Kh5 Rg5+)
Rg5+ Ke6 (Kf/e4 Rf/e5+ transposes to Kf5 type lines)
Re5+ Kf7 (Kd6/7 Re6/7+ as detailed above above)
Re7+ Kf/g8
Rf/g7+ and you can never get off the 8th rank

[The Ancient Brit]

My gut feeling was there was a win for Blockhead in the game. I have done a great deal of analysis or rather my copy of Chessmaster 10 has. Taking my computer offline and shutting down all non-essential processes so Chessmaster would run all the faster, I set it to find for mate and then went to bed.

Eight hours later, it had reached the seventeenth move having examined all the possible combinations on the way and had still not found a mate.

I had taken the precaution before doing this that the program is capable of finding mates. I substituted the black pawn at h2 with a white pawn thus removing the obstacle in the game, that of stalemate. Chessmaster found a mate in the seventh move in just over one minute.

It is therefore with reluctance that, drawing myself to the full height of my 1475 rating, I am forced to agree with my betters, Savin, torphichen and Blockhead, that this game is a draw.

Cheers. Denis. The Ancient Brit.

[yassen]

theancientbrit said:

Eight hours later, it had reached the seventeenth move having examined all the possible combinations on the way and had still not found a mate.

I had taken the precaution before doing this that the program is capable of finding mates.

It is therefore with reluctance that, drawing myself to the full height of my 1475 rating, I am forced to agree with my betters, Savin, torphichen and Blockhead, that this game is a draw.

You checked if CM is capable of finding mates It sounds about equal to checking if Michael Shumacher is capable of driving a car

Anyway, it's another good example of the superiority of the man vs machine. It took CM 8 hours, and it probably took Savin no more than 8 minutes to figure out the game is drawn.

[The Ancient Brit]

S’mae (Hi) yassen,

It must seem obvious that Chessmaster can find checkmates but as any chess player will tell you, it is important to look at the obvious, check that it really is obvious, then assume it isn’t and recheck. You must also bear in mind that my experience of computers and chess programs, especially when used in combination, is that they will do exactly what I tell them to do, which may not necessarily be what I mean. Neither of them runs on the same fuzzy logic that I do.

I have no doubt that both Savin and tophichen dealt with the problem swiftly and efficiently and I am grateful to them for their contributions. Lacking the chess brains of either and having prodded Blockhead into starting this thread, I felt obliged to contribute. My contribution, which I am sure you will agree, was different, although coming at the problem from another angle, nevertheless arrived at the same conclusion.

[torphichen]

This is the sort of position where we hold the advantage over machines. Unfortunately they have the advantage in most of the others, at least against me.

Just in case anyone notices, I have just edited a couple of typos on my posting. Rf1+ (para 3) now reads Re1+ and Rf/g8+ (para 5) now reads Rf/g7+.

[perseus]

Draw, it took me about a minute to make sure.

Interesting position. I hope I get the a chance to try it out one day!

[savin]

The Ancient Brit,

Its actually quite a fun position in that the way to solve it is with lateral thinking more than chess. Hee is how I tackled the problem:

1) Black can not take the rook or its stalemate. So the black King must move.

2) Can the black King move anywhere where the white rook can not check the King by being next to the King because anything else is just a repeat of one. Ths is nasty. You just have to take all the pieces off the board except for the Black King and White rook to see the answer. The king can never move to a square that does not allow the White rook to check and be next to the King.

3) The problem is the Black rook and pawn and the position of the white King. The only way to break out of this is to get the Black rook to move. But this could only be done to either take the white rook or block a check of the White rook.


4) There is no way to force a position where the black rook can block a check. But perhaps we can make a position where the white rook can only check the black king from a row where the black rook can take it? This can only happen on f3, f4 or a2, b2, c2, d2 and e2. But it quickly becomes obvious that the rook always has more than one potion of the check and can't relly be forced into anything. Hence its a draw.

Its interesting to note that if the pawn at f6 was at f5 or f7 the game would be a win for black.

[torphichen]

I am probably missing something, as I cannot see the win with the other pawn positions.

Do you have a line or two to show how it is achieved?

[The Ancient Brit]

S’mae Savin,

I doubt if Blockhead thinks of the position as fun in any way. I suspect he is still gnashing his teeth.

Thanks for the explanation of your method. It is certainly more elegant than my brute force number crunching method.

[Blockhead]

Many thanks to everyone who has contributed to this 'fun' example of rank stupidity on my part. Since the game could theoretically continue for another 50 moves I was particularly taken by savin's clear exposition and general ('lateral'?) thinking (zero annotation). I have learned little about chess but a great deal about how some fine SNC players think! Again, many thanks.

I think the general lesson for us all is: Never ever treat any move, especially so called 'obvious moves', in a casual off-hand way. Blunderland awaits those like me who do not heed this simple advice! :{

Next time I'm in the stocks I'll hand out the rotten fruit first! :}

I wondered if any of the contributors (TAB!! ;-) might have something similar for me and others to chew on! Come on TAB!

I'll pop down to the market and buy up all the fruit & veg leftovers! :}