#### Only Chess Forum

1. 26 May '06 16:08
I read today that a 7-piece endgame position has been discovered where a forced win exists, but against the best defence requires 517 moves.

Black to play. White wins.

Anyone want to try figuring out the best line? ;-)
2. 26 May '06 16:31
I don't believe it!

One has to take the 50 move limit for a draw into account.
Since there is 5 picese on board then the 50 move rule can apply for 5 times 50 move, i.e. 250 moves.

If there is no mate within 250 moves there is a draw. Mate in 517 is impossible.
3. 26 May '06 16:54
Originally posted by FabianFnas
I don't believe it!

One has to take the 50 move limit for a draw into account.
Since there is 5 picese on board then the 50 move rule can apply for 5 times 50 move, i.e. 250 moves.

If there is no mate within 250 moves there is a draw. Mate in 517 is impossible.
Your reasoning is flawed. A piece can be forced to move more than once.

Here is an article about it: http://www.xs4all.nl/~timkr/chess2/diary.htm
4. 26 May '06 16:56
These discoveries are more important from a theoretical point of view rather than a practical one. I agree that in a normal game, the 50 move rule could come into effect.

Such discoveries question the validity of the 50 move rule for future games, especially those between computers. This rule has been revised in the past, and may be revised in the future.
5. 26 May '06 19:30
Originally posted by Mephisto2
Your reasoning is flawed. A piece can be forced to move more than once.

Here is an article about it: http://www.xs4all.nl/~timkr/chess2/diary.htm
If (a) a pawn not has moved or (b) a piece is not taken, then it is draw after 50 moves. The 50 move rule is one of the reasons (besides other rules) a game turns into a draw.

In this particular game (a) there is no pawns to be moved and (b) no piece has been taken in the first 50 moves. This gives us that this game is draw after 50 moves and there is no mate to be found.

Why is theese players continuing a game that is an obvious draw?

Why do you think that "a piece can be forced to move more than once" has anything to do weather this game is a draw or not? Why do you think my reasoning is flawed?

And even that - there is a 518'th move in your URL-link and even after that ithe text says "518.Qxg5+ and White wins" So it takes more than 518 moves to win, not precisely 517 moves to mate.
6. 26 May '06 20:18
Originally posted by FabianFnas
I agree with what you state about the 50 move rule.

>> "Why is theese players continuing a game that is an obvious draw?"

This is not an actual game. It's a theoretical chess discovery. And even with the 50 move rule, it's not an obvious draw.

You are correct that my title is inaccurate; I shouldn't have used the word 'checkmate'. But my initial post is correct.
7. 26 May '06 20:57
Originally posted by FabianFnas
If (a) a pawn not has moved or (b) a piece is not taken, then it is draw after 50 moves. The 50 move rule is one of the reasons (besides other rules) a game turns into a draw.

In this particular game (a) there is no pawns to be moved and (b) no piece has been taken in the first 50 moves. This gives us that this game is draw after 50 moves and there is ...[text shortened]... 5+ and White wins" So it takes more than 518 moves to win, not precisely 517 moves to mate.
You are right about the rule. I misinterpreted what you wrote, mainly because I had read the article in the link myself shortly before and was probably too eager to support the first poster than to read more carefully. Sorry for that.

The 518 vs 517 moves is because it is black to move, and white to win in 517 moves.

Theoretical analyses like these are often used as arguments against the 50-move rule, especially in the arena of computer chess.
8.  Wulebgr
Angler
26 May '06 22:06
Originally posted by Varenka
I read today that a 7-piece endgame position has been discovered where a forced win exists, but against the best defence requires 517 moves.
If I'm not mistaken, that shatters the previous record of something close to 250 moves. Where did you read the article?
9. 26 May '06 22:08
Originally posted by Wulebgr
If I'm not mistaken, that shatters the previous record of something close to 250 moves. Where did you read the article?
See the link posted by Mephisto2...

i.e http://www.xs4all.nl/~timkr/chess2/diary.htm

It mentions the previous record.
10.  Wulebgr
Angler
26 May '06 22:10
Originally posted by Varenka
See the link posted by Mephisto2...

i.e http://www.xs4all.nl/~timkr/chess2/diary.htm

It mentions the previous record.
Thanks. A lot has happened since October: 243 moves stood as the record for 15 years, then 290 for less than six months; then 330, and now 517. How long will this record hold? Bets, anyone.
11. 26 May '06 22:21
Figured it out. Yep, 517 moves.
12. 26 May '06 22:32
517 moves to mate.....its certainly interesting, but i don't think it does anything to undermine the 50-move rule....I should imagine calculating 517 moves ahead under time standard time controls is beyond most humans.....

Computer torlaments however...
13. 27 May '06 05:28
mate in 30 moves with black to move
14.  bosintang
perpetualEditMonkey
27 May '06 09:54
Here's the original post by the guys who accomplished this:

fascinating stuff here.

"Generation and verification took
a little over a month on a 3.8 GHZ machine. Yakov and I had thought it so
unlikely that any 7-man ending would require more than 511 moves that we had
hard-coded 511 in a number of places. "

I assume the solution was brute-forced? If so, a month is really not that long of a time to generate such a complex solution. This could potentially lead to some incredible computer tournaments in the future&#12288;if they decide to remove the 50-move rule.
15.  bosintang
perpetualEditMonkey
27 May '06 10:05
"In total, we have
now generated results for over 21 trillion legal positions, covering about 20%
of all pawnless 7-man endings. About 1.5 million positions are mutual
zugzwangs, and 13 are full-point mutual zugzwangs."

Does anyone know the difference between a mutual zugzwang and a full-point mutual zugzwang?