Okay, maybe somebody with more years of experience can tell me how I hallucinated this one.
I really, really, really thought that I had read in the rulebook somewhere that a B & N vs. K endgame was 75 moves to win or else it’s a draw. But then when I went to look it up, I couldn’t find any such reference. Maybe that was the case in one of the older versions of the rulebook and it’s changed over the years. My TD in training thought that might be a FIDE rule. Any thoughts?
Which brings me to part two…let’s say that’s the above endgame. If there is no rule posted to the 75 move rule effect, then is it 50 moves and it’s a draw?
There was some tinkering with the 50-move around 1990, but this ending was never included. The extensions were intended for endings that really could take more than 50 moves to force mate from a bad initial position, like two Knights vs. pawn. B+N vs. K takes something like 35 moves to mate from the worst possible start. Perhaps you were thinking of B+B vs N, which I believe was extended to 75 moves for a while.
At present there are no extensions to the 50-move rule, though there is still a provision allowing the organizer to announce some at the start of the tournament. I don’t know of anyone who does that.
The answer to your second question is yes. If you can’t mate in 50 moves, the game is drawn, no matter what the material situation.
I once had to observe a game in a section I was directing that was B+N+K vs K (and less than 5 minutes on the clock, just to make it interesting).
He made (I think) one slipup in the sequence and it took him 47 moves, his opponent and I were both counting them.
Thankfully I’ve never had to actually DO this checkmate in a game, because I doubt I could, and almost certainly not under time pressure.
Back in my youth, the ‘other’ Larry Evans (Larry D) used to hold court at the Chicago Chess Club (on Wabash) and play B+N+K vs K in the skittles room, with him having just 1 minute on a BHB clock (you could have as much time as you wanted), for a quarter a game. You could set up any position that wasn’t an immediate draw. I never saw him lose on time or fail to make the checkmate within 50 moves.
There are no specific provisions in the current rulebook for any extension of the 50-move rule. I don’t think there are in FIDE anymore, either.
A couple of decades ago there used to be a few extensions for a few specific positions, but B+N vs K was never one of them. It’s known that B+N vs K should never take more than 38 moves (or something like that).
I think one of the exceptions might have been R+B vs R. Most of these are theoretically drawn anyway, though with great difficulty. Of the small percentage that are theroetically won, an even smaller percentage require more than 50 moves. (Everything in this paragraph is subject to override by anybody that may know better than I.)
The same thing used to happen with A-player Anthony Novakovic, who, less than a year after his first rated tournament, bet master Craig Chellstorp that he could mate with B+N vs K in less than a minute using a BHB clock. He did it so fast that Craig, even while losing the bet, was rolling in the aisles with laughter.
This is even more amazing because some BHB clocks seem to have an “anti-delay” built in. If you press the button hard enough, the clock will run faster than normal for a few seconds, before settling down. It sounds sort of like TICKTICK TICK-TICK TICK … Tick … Tick … tick … tick … tick … etc.
[quote="Smythe Dakota
This is even more amazing because some BHB clocks seem to have an “anti-delay” built in. If you press the button hard enough, the clock will run faster than normal for a few seconds, before settling down. It sounds sort of like TICKTICK TICK-TICK TICK … Tick … Tick … tick … tick … tick … etc.
Bill Smythe[/quote]
As a time pressure junkie who has wiggled out of many a fine mess thanks to time delay, now I know why I was quick to abandon my old BHB.
Then again now a days if I get stuck with an analog clock it needs to have a big face so I can read it. Another good reason to be punctual for the round!
Okay, thank you all very much…then I’m glad I only drew. Yes, this was a rated event and it took me 78 moves to make the mate. First time I’ve ever had to do it in my life and even though I knew that I wanted the K in the corner of the color of the bishop, it was another ordeal to figure out how to get it there. I know I could do it much faster now that I know the technique, but it was another thing to do it in a rated game thinking out the moves. I knew I wished he had swapped the Bishop for the Knight instead of the pawn.
Until the late 1980s, the rules specified three positions for which the 50-move rule was extended to 100 (though there was some ambiguity as to whether you had to announce that this was in effect before the start of the tournament). The positions were: King, Rook & Bishop vs King & Rook; King and two Knights vs. King and pawn, provided the pawn was blockaded behind a specified line; and one special position of King, Rook and a (or h) pawn pawn vs. King, Bishop and a (or h) pawn. Rook and Bishop vs. Rook does come up occasionally, but I don’t think I’ve ever seen the position that takes more than 50 moves arise. Two Knights vs. pawn is the only one that might actually affect the outcome of a game. (Lilienthal reached it three times and didn’t win any of them, which ought to give some indication of the likelihood you or I would.)
There was some tinkering with the rule around 1990, when computer analysis became popular, but eventually the specific positions were deleted from the rules and the option of extending the fifty moves left to individual organizers. I haven’t heard of any grass-roots demand to bring them back.
It may be that there is no such position, but that the rulemakers (at the time) decided it was so easy for the defender to err, that he should be required to hang on for 100 moves instead of 50 just to prove he knew what he was doing.
It wasn’t exactly new, but the idea (of holding the King back with the Rook on the third rank) had always been dismissed in a footnote (“if the King and Rook are separated, a series of checks …”). It turns out that if the defender picks the one right square for the Rook each time, the stronger side has to play carefully to create a zugzwang and force the defender back one rank/file.
At the 1977 U.S. Open in Columbus, Belle (well, it’s operators) challenged all comers to play Queen vs. Rook. Half a dozen boards were played, mostly with good players. Four of them were drawn. Eugene Meyer offered a draw around move 45 (he’d finally found the winning method but didn’t have enough moves left). The only team that won legitimately was Jean Milton and Jack Peters. One board, manned by a group of coffeehouse players from New York, managed to get to the Philidor position by asking the computer each move how many moves to mate, and adjusting their play accordingly. But then they couldn’t find the idea of losing a move with the Queen (“six moves to mate … six moves to mate … six moves to mate”) until I pointed it out.
B + N vs. K is a tricky one! A couple of years ago I saw a friend reach this position and he was not able to get the job done in 50 moves. That kind of motivated me to practice the ending until I could do it myself. I can usually do it in 50 moves when I try it (but not always!)
I read that all the extensions to the 50-move rule have been revoked.
…The only time I’ve used the 50-move rule in tournament play was when defending pawn and rook against queen. I had two squares for the rook to move back and forth (protected by the pawn) to keep his king from approaching, even so I almost mishandled it (got my king stuck on the wrong side and got my rook stuck right in front of the pawn) but I was able to rebuild the fortress and get 50. Anyone else played this position?
Was that the occasion when the computer, defending K+Q vs K+R, every so often moved the R to a square that seemed to lose quickly (“surely there must be a fork here”) but nobody could ever find a refutation? I heard that, the first few times this happened, all the spectators said “oh, come now” and looked for a forced win, but after the first few times, they finally gave up and said “hmm, I guess the computer found another defense” and didn’t even bother to look for the refutation.
Whether the “third rank defense” is actually “better” is a matter of definition. With perfect play. it doesn’t any longer to mate (might even take a few moves less). But in the “standard” defense (keeping the King and Rook together), all of the stronger side’s moves are pretty easy to find. In the other, he has do some work. In Columbus, the good players realized this after a while, but mostly after they had used up too many moves to mate within 50.
Although I wasn’t aware of that 1977 US Open challenge (K+Q vs computer with K+R), I have in my possession a story about a similar challenge that happened around the same time, in California.
Some early chess programmers had fully analyzed that ending, creating the world’s very first tablebase. Their results showed that it’s harder to win than the books were saying at the time – primarily because a single small inaccuracy can extend the time needed to mate by something like 8 or 10 moves; make just two second-best moves and you’re already in danger of running out of moves.
So the programmers made a money bet with the dominant player then active in their area – GM Walter Browne. A match was set up and stakes agreed to. I don’t recall how many times the ending was played or the result; I do recall that in at least one of those games, the GM failed to finish off the computer within 50 moves.
I have a copy somewhere of the Northern California state chess magazine that narrates this story; I think it is from 1978.
I also recall reading a news article from a mainstream (non-chess) publication back in the 1990s when programmers solved K+R+B vs K+N+N. It’s a forced win for the stronger side – in a maximum of 233 moves. (Don’t ask me how I remembered that number, but I’m pretty sure it’s right…although maybe a still-stronger engine has found a shorter solution, in the years since.)
Also I think K+Q vs K+B+B – an ending no one ever expected would be a forced win – was eventually solved with engines and turned into a tablebase. I think both that one and K+Q vs K+B+N also require more than 50 moves.
It’s also interesting that the man I regularly refer to as the stupidest man on earth (I used to call him the stupidest man on the Internet; but actually the Internet isn’t even big enough to contain his stupidity) devotes a page of his own Web site to a particular K+Q vs K+R position whose solution he claims credit for.
This monumental fool then goes on to boast, apparently with a straight face, “Even Kasparov doesn’t have an endgame position named after him. But I do!”
Another good reason to be punctual for the round!