Novice chess players know that a player’s two bishops are always on different-colored squares. More experienced chess players know that that isn’t necessarily true: It’s possible to promote a pawn into a bishop, and so to end up with two (or more) bishops on the same color.
But it seems like this would be an incredibly rare occurrence, in actual play. First, you’ve got to promote a pawn to something other than a queen. That’s already pretty rare, because a queen is, in almost all situations, the strongest piece. And on the rare occasion where someone does underpromote a pawn, it’s usually to a knight, because that’s the one piece that can do something that a queen can’t.
The only reason it’d ever be to a player’s benefit to choose a bishop or rook instead of a queen would be if promoting to a queen would mean immediate stalemate, since that’s the only way that threatening more squares could ever be detrimental. And even if a queen promotion would be stalemate, the advancing player is likely to be able to wait a turn for the board position to change, and so a queen promotion would be safe.
And then we also have to assume that, at the time of the underpromotion, the advancing player also still has a bishop on the board. Which means that the other player had to lose enough pieces that stalemate was possible, while the other player still had at least one significant piece, without the game already being over. And even if we assume all of that, there’s still a 50-50 shot that the bishop still on the board would be the other color, anyway.
And so I wonder: Has it ever happened, in the history of recorded competitive games, that a player has actually had two bishops on the same color of square?
There are examples where it has been done as a joke; here’s one from very high-level players:
And here’s one from one high-level player messing around against a high-level computer:
I’m not aware of any “real” examples. As you say, it’s difficult to even construct situations where it would make sense.
There are some examples in constructed problems, too. The “Babson Task” (create a position where it’s necessary to promote to whatever piece the opponent promotes to) has produced a few.
That’s a real knee-slapper!
Here’s a good article about the Babson Task. Tim Krabbé’s stuff is always good.
So if I’m understanding this right: Both players saw that the position was drawn anyway, no matter what the promotion was (if it were anything but a bishop, the king would have taken it), so there was nothing lost by having a little fun.
This one, though, was genius, even if it’s a trick that would never work against a human opponent. As I’m reading that, the computer was programmed to never accept a draw, if it thought that it was in a superior position. So the human player arranged a sort of pawn-zugzwang, or took advantage of one that arose naturally, where the first player to move a pawn would be put at a disadvantage. He then deliberately sacrificed some material to ensure that the computer would think itself superior, and started wasting moves. At 49 moves after the last capture, the computer felt itself compelled to move a pawn, so as to avoid a 50-move draw, which resulted in the whole board falling apart in such a way that the computer’s material couldn’t save it. At this point, the human decided to insult the computer in a manner which would be a horrid breach of etiquette against a human opponent, by winning with a whole bunch of bishops instead of going in for the clean kill.
I had actually considered the possibility of underpromotion as a deliberate insult, but discarded it on grounds that no serious player would be that unprofessional. I hadn’t considered, though, that etiquette wouldn’t apply to a computer opponent.
And I was also, of course, aware of constructed problems where this either has already happened, or is forced to happen. That’s why I put in the “competitive match” qualifier.
You’re right about the first game. It’s kind of a practical example of the kind of lessons given to explain why opposite-coloured bishop endgames are often drawn. You can give white as many light-squared bishops as you want and it’s still a draw.
Nakamura has a bunch of computer games like that. There’s one where he ends up with 6 knights, etc.
The “insulting” promotions happen sometimes in online games. One player will refuse to resign a hopeless position, so the other player will make the win take forever, sort of trying to give a message about what it feels like to have your time wasted. It turns into a kind of game of chicken about who values their own time less :).
I’ve seen a tournament game where one of my team captured with a pawn and promoted it to a knight (because it gave check.)
It was in the early middle-game, so for one move the player had three knights.
(It was also the best move.)
I did that once, because I was about 11, and didn’t fully understand the rules of chess. I thought you had to promote to a captured piece, so if your queen was still on the board, you had to choose another piece, and I chose a bishop. I ended up with two bishops on the same color, which had not been my plan, but I didn’t realize it in time. I chose a bishop over a rook or knight because the bishops governed different squares, and ended up not getting what I wanted. I still had on rook and one knight on the board.
[hijack](You have to know I’m Jewish to get why this is funny, but most people know that.) My brother and I were with our family at an Episcopal church for someone’s confirmation, and the bishop invited all the little children up to the front of the church for an short talk, and he begins by explaining that he is the bishop: “does anyone know what the bishop does?” My brother was about 5, and had gone up with the other kids. When the bishop asked that question, we heard my brother’s voice loud and clear “He moves diagonally.”[/hijack]
A player from my club used to play the Albin Countergambit, so he was often looking for a third black knight.
Honestly, Rikah, the bishop was probably half-expecting one of the kids to say that.