Fun. Only if two players of unequal strength would play this, then there would be no difference at all with normal chess. The stronger player could stick to normal chess strategy and win each game that way.
Prisoners' Chess
Sort:
Mostly agreed, the whole thing assumes the two players are more or less equal strength. I see this best used in ending stages of tournaments, or against a chess program.
However, a 5 games match is not won until all the games are played, because you never know how many points you'll score (or loose) on the next game. A single blunder in the last game can trash your 4 wins in a row, so the weak player can still hope to overturn the deficit by winning only the last game.
Or the strong player can take risks in the first games, because he knows he will eventually overturn the deficit in the last games, should the risks end up in a lost game.
Likewise, the strong player can take risks in the last game, after he's already won the first 4 games of the match, so that both players have fun till the end of the match.
I've been playing this variant against the computer (Stockfish at skill level 0, which is more or less my strenght) during the last 2 months, and I can confirm it is quite a fun. Please note I'm playing against standard unmodified Stockfish, so the computer does not know about these scoring rules and it plays standard chess.
However, even using the Prisoners scoring on my side only, I realized there are actually much more new strategies caused by the Prisoners scoring than the forced promotion shown above. When the game evaluator shows "# -2" (mate in 2), if one of those two moves includes capturing an opponent piece then you're almost sure those aren't the best two moves and the best mate will take some more moves. Problem is finding it.
Even more fun would be to play against someone/something aware of the Prisoners scoring: when you have a mate in 2 and avoid it, your opponent knows you'r not willing to capture pieces, and he knows you feel strong enough to reach checkmate without capturing.
Or, in a match, maybe you need to score at least 11 points in the last game to win the match, and you are playing white in this position:
Here the obvious mate shown above is not an option for white, because the moment white captures the black queen, he looses any chances to score 11 points and he looses the whole match. White needs to find a different mate that does not involve capturing the black queen and black player knows that.
Hello
I propose this new chess variant named "Prisoners' Chess". The starting position, the chessboard and the playing rules are just the same as FIDE. The difference is the score value of a won game. Where in FIDE Chess a win scores 1 point, in Prisoners' Chess a win scores the same points as the sum of the loosing player's pieces values on the chessboard at the end of the game, king included. Each piece, except the king, has the common approximate integer values:
The king value is also 1 point like a pawn (see below for the motivation).
For example, consider the following checkmate, that could be the first game of a match:
White wins and scores 5 points, because he gained 2 prisoners, that are the black pawn (1 point) and the black knight (3 points) plus 1 point for the king. The score is noted and the match continues. In game 2, black wins:
Here black scored 18 points, because his prisoners are a queen, a rook and three pawns (9 + 5 + 1*3) + 1 for the king. The match score is now 5-18.
Now consider the third game in the match:
White has three mates in one (Rh8#, Qe7# and Re6#), but those are not the best moves. White can force black to promote the pawn, so that he will gain more points on the checkmate in the next move. Black better promotes to a knight or bishop, because that gives away less points to white and a queen in that position couldn't save the game anyways.
Resigning a game in Prisoners' Chess is not allowed and it usually does not make sense, because it gives more points to the opponent. Games always reach checkmate or draw. Even in the case of forced promotion, like in the example above, the loosing part can't resign. If the loosing player decides to resign anyway, he must resign the whole match.
Draw games score 1/2-1/2 as usual, since there are no real prisoners in a draw game.
The reason why the king scores 1 point is because a checkmate without any other prisoners except the king itself, must give at least 1 point to the winner, so that the game score is 1-0, or 0-1, but not 0-0. For example in the following checkmate white scores 1 point and the game result is 1-0:
What do you think of this variant? Implementing this in a chess engine should also be quite straightforward, at least for the minimax algorithm.