Many times it makes sense to “underpromote” a pawn to a Knight - rare, but not excessively so. And in the “exceedingly rare” category, I’d place underpromotion to a Rook. I have never seen it in an actual game, but in a few chess puzzles, and a discussion of a game that I believe was authentic. The only imaginable reason to do so is to avoid an immediate stalemate.
My question - is there ever a “real” case for underpromotion to Bishop? I have never seen it - not even in the contrived cases that make their way into the puzzle books. Some of us were discussing the “Rook” rarity, and I was just wondering if the “Bishop” underpromotion case has ever come up.
I suspect - not. (But by stating that, I’m more apt to get someone to correct me.)
I recall seeing a player promote to a bishop so that he could practice mating with either two bishops or a bishop and knight. He later resigned, casting serious doubt about whether he was really trying to win.
Excellent! Thanks (all) for the interesting examples. I’m particularly intrigued with the first one (i.e., from tanstaafl’s reference) as it is elegantly simple, yet also looks like it might have really happened! Some of the others as well.
Although I must confess, I would NOT want to have to win a BBK v. NK endgame! Above my paygrade, I think. Thanks again.
How about 1. d7, pawn moves; 2. d8=B, pawn moves; 3. Bh4, pawn moves; 4. Be1, N moves; 5. Be1xN mate
A problem is that if Black’s first two pawn moves are with the b and e pawns, he’s stalemated after 3. Bxh4. (But maybe it just means I’ve found a wrong try and the real solution is something else.) Changing the position by moving White’s h pawn from h2 to h5, and Black’s h pawn from h4 to h7, would fix that problem.
If Black pushes the h-pawn, promote to Bishop (releasing the Knight from e1).
If Black pushes the b-pawn, promote to Knight (releasing the Rook from d4).
If Black pushes the e-pawn, promote to Rook (releasing the Rook from c3).
Oops, corrected those algebraic mistakes (I still use English and reversed the board!).
The pawns only come into play for the rook or knight underpromotion. After Rc3 or Nd4, R moves, the WR (or WN) could block the BR from getting back to b2, but it could interpose the Bh8 check to lengthen the solution.
Well, yes. But they aren’t used in the part of the solution I saw, underpromotion to a bishop. And then I was trying to make the idea work of Bh8-c3-e1, Nd2; Bxd2+, Rb1; Bc3 mate. But that’s a mate in 6 because Black has 3 pawn tempos and I couldn’t get it to work in 5, and it didn’t use those pawns on the light squares either.
[size=150]EXCELLENT![/size] Great job working out the solution.
I have to say, in my relatively limited exposure - this is perhaps the most fascinating chess problem I’ve ever seen.
I thought I was on track, but didn’t get it. But I enjoyed it immensely, and am happy I started this thread.
I first saw this problem at the old No Exit Cafe in Chicago, which (many years ago) ran blitz tournaments every Tuesday night. 3 or 4 masters were intently studying the position, which they told me was supposed to be a mate in 5. I took a brief look but then departed the scene to play my tournament games. After about 30 minutes, while I was playing blitz, they found the N promotion, and after another hour, they found the R promotion. At that point, I returned to the scene, and they showed me what they’d discovered so far. I immediately blurted out, in jest, "well, if case 1 is a N, and case 2 is a R, then surely case 3 must be a B. HA HA HA HA! " -------- “Hey, wait a minute --” and, of course, the masters then found the rest of the solution within seconds.
I don’t know where the problem came from, but the position appeared monthly, without comment or instructions, on the envelope Canadian Chess Chat arrived in. Eventually, one issue of that mag had a feature article on the position, but it was coy about the source, and only barely hinted that it “might” be a mate in 5 or something.
Does anyone know who composed this problem, and how old it is?
Thanks again.
HA HA HA HA!