Is Karmageddon the future of chess?

[lkaufman]

Karmageddon is a variant of chess based on Armaggedon where black can't castle, but draws give black the win.

After several correspondence time control games against witchesbutt it was clear that I wasn't strong enough to beat him in a chess game. We ceased playing because it would just be a draw after another.

So we tried Karmageddon and it was an entirely different picture:

In this game I was black
[pgn]1. d4 d5 2. c4 c6 3. Nc3 Nf6 4. e3 Bf5 5. cxd5 cxd5 6. Qb3 Bc8 7. Nf3 e6 8. Bd3
h5 9. a3 Be7 10. O-O Kf8 11. Qc2 Bd7 12. b4 a6 13. Qb3 Nc6 14. Na4 Bd6 15. Nc5
b6 16. Nxd7+ Qxd7 17. Re1 Ne7 18. Bb2 Qb7 19. Ng5 Ra7 20. h4 {0-1/2}[/pgn]

Despite being unable to castle I managed to equalize and win the game.

In this game I was white
[pgn]1. c4 c5 2. e3 g6 3. d4 cxd4 4. exd4 Bg7 5. Nc3 d6 6. h3 Kf8 7. Nf3 Nc6 8. Be2 h5 9. d5 Ne5 10. Nd4 Nh6 11. O-O Nf5
12. Ncb5 Bf6 13. Nxf5 Bxf5 14. Be3 g5 15. f4 gxf4 16. Rxf4 Bg6 17. Qd2 a5 18. Raf1 Kg8 19. Rxf6 exf6 20. Bf2 Qf8 21. b3 h4 22. Bd4 Qd8 23. a4 Kg7 24. Bc3 Qb6+ 25. Kh2 Qd8 26. Kh1 b6 27. Qf4 {1/2-0}[/pgn]

Here I could take advantage of black's inability to castle and defeat them.

The only way to make this fair is to play a match of two games, but while on chess we were equal, here we were not, so Karmageddon has a much higher skill ceiling.

This variant could put an end to chess draw death, if it was unusable I'd have expected for black to hold the draw in both games, or for white to win both games, but it seems to have a healthy complexity.

Anybody that thinks black should always hold or white should always win is welcome to take on me, specially, I don't have the Komodo version with Armageddon scoring, it'd be interesting to see if it makes a difference.

I have done a lot of testing of this along with Kai Laskos and S.Pohl over the past year, and our conclusion was that this precise variant is quite playable but somewhat favorable to White. With one rule change, namely that Black can still castle long but not short, the results became so close to 50-50 that it was impossible to say which side had the better chances, and this seems to be the best version. We call it NBSC Armageddon (No Black Short Castling). Check out https://www.sp-cc.de/ for more on this.

Have you spoken with Kai Laskos over the past year?

No, I had some private conversations with him when I was writing my book, but that was over two years ago, before the Pandemic.

[towforce]

Apologies for being negative, and I hope I'm not discouraging anyone, but it seems to me that alterations to the starting position are only delaying the inevitable. IMO, it won't take long for the engines and their nets to adapt to the changes, and my guess is that with each new variation, the time from its inception to a new death-by-draw will only get shorter.

You could make the board (and number of pieces) larger, but IMO this would result in games that are too long to be of interest to humans: correspondence games would go from "months" to "lifetime", and then to "multi-generational battles".

"Come on grandchild - buy the extra computing power we need - the family's pride is at stake!"

At least it would incentivise future generations to strongly solve the game...

[Ozymandias]

Apologies for being negative, and I hope I'm not discouraging anyone, but it seems to me that alterations to the starting position are only delaying the inevitable. IMO, it won't take long for the engines and their nets to adapt to the changes, and my guess is that with each new variation, the time from its inception to a new death-by-draw will only get shorter.

Totally agreed, but what are you to do?

[MonteCarlo]

Delaying the inevitable can have some value.

I'm quite certain I will die eventually, but I'm not in any rush to get there

Cheers!

[towforce]

Apologies for being negative, and I hope I'm not discouraging anyone, but it seems to me that alterations to the starting position are only delaying the inevitable. IMO, it won't take long for the engines and their nets to adapt to the changes, and my guess is that with each new variation, the time from its inception to a new death-by-draw will only get shorter.

Totally agreed, but what are you to do?

My preference: keep driving standard chess for now. Although correspondence chess has hit death by draw, taking the evidence as a whole, I actually think that we haven't reached the upper limit of the ELO scale yet. We need something different from hand-coded evaluation and NNs to significantly improve the game today IMO.

Another way to stick to standard chess but reopen the game is to change the computer competition from "win against a computer opponent" to "guess more accurately than your opponent how many moves from this position to checkmate".

Or introduce a rule for computer chess banning things like recursion, iteration or tablebase lookups in the evaluation. That would drop computers back to about 2500 elo IMO, and put domesticated apes back on top where we belong. It's OUR planet!

[Uri Blass]

Apologies for being negative, and I hope I'm not discouraging anyone, but it seems to me that alterations to the starting position are only delaying the inevitable. IMO, it won't take long for the engines and their nets to adapt to the changes, and my guess is that with each new variation, the time from its inception to a new death-by-draw will only get shorter.

You could make the board (and number of pieces) larger, but IMO this would result in games that are too long to be of interest to humans: correspondence games would go from "months" to "lifetime", and then to "multi-generational battles".

"Come on grandchild - buy the extra computing power we need - the family's pride is at stake!"

At least it would incentivise future generations to strongly solve the game...

Making the board (and number of pieces) larger does not mean making the games longer.
You can have a rule that a game is finished after 60 moves with rules who say who is winning if there is no checkmate but no need to make the board bigger for correspondence chess and I believe the following version of correspondence chess with the normal chess rules can survive for many years
if people are interested.

1)Order the legal moves of the side to move by accepted order (you can use for example lexicographic order).
2)Both players express their opinion of the expected result for every legal move.
3)If they agree about every legal move then the side to move make a best move based on his(her) choice and we go back for step 1
otherwise find the first move that they disagree and make a chess game when the winner is the player who prove the claim.

for example if after 1.f3 f6 white claim 2.g4 is losing when black claim it is a draw then
white needs to play with the black pieces after 1.f3 f6 2.g4 with the target to win with black when the target of white is to make at least a draw.

There is going to be a winner who achieve the target in case that after enough moves we do not go back to step 1.
The only way there is going to be no winner is if we always go back to step 1 until the game is finished but I believe that it is not going to happen at least in the next 20 years(Edit:it can happen even today in case both players use exactly the same engine for exactly the same time with exactly the same hardware to produce their evaluations but I do not expect it to happen because even if they have the same hardware and the same time and they use some smp engine then by some luck the engine may give different evaluation for both sides).

[lkaufman]

Apologies for being negative, and I hope I'm not discouraging anyone, but it seems to me that alterations to the starting position are only delaying the inevitable. IMO, it won't take long for the engines and their nets to adapt to the changes, and my guess is that with each new variation, the time from its inception to a new death-by-draw will only get shorter.

You could make the board (and number of pieces) larger, but IMO this would result in games that are too long to be of interest to humans: correspondence games would go from "months" to "lifetime", and then to "multi-generational battles".

"Come on grandchild - buy the extra computing power we need - the family's pride is at stake!"

At least it would incentivise future generations to strongly solve the game...

Making the board (and number of pieces) larger does not mean making the games longer.
You can have a rule that a game is finished after 60 moves with rules who say who is winning if there is no checkmate but no need to make the board bigger for correspondence chess and I believe the following version of correspondence chess with the normal chess rules can survive for many years
if people are interested.

1)Order the legal moves of the side to move by accepted order (you can use for example lexicographic order).
2)Both players express their opinion of the expected result for every legal move.
3)If they agree about every legal move then the side to move make a best move based on his(her) choice and we go back for step 1
otherwise find the first move that they disagree and make a chess game when the winner is the player who prove the claim.

for example if after 1.f3 f6 white claim 2.g4 is losing when black claim it is a draw then
white needs to play with the black pieces after 1.f3 f6 2.g4 with the target to win with black when the target of white is to make at least a draw.

There is going to be a winner who achieve the target in case that after enough moves we do not go back to step 1.
The only way there is going to be no winner is if we always go back to step 1 until the game is finished but I believe that it is not going to happen at least in the next 20 years(Edit:it can happen even today in case both players use exactly the same engine for exactly the same time with exactly the same hardware to produce their evaluations but I do not expect it to happen because even if they have the same hardware and the same time and they use some smp engine then by some luck the engine may give different evaluation for both sides).

It seems to me that if we want to drastically reduce or eliminate draws, and/or to reduce or eliminate the value of memorizing openings, we should try to do so in a way that leaves the new version as close as possible to traditional chess. Most of the proposals I have seen both in this thread and elsewhere really feel like someone is just inventing a whole new game. The solution to the memorizing problem is fairly clear, either play Chess960 (or some variant of it, such as Chess324 which chess.com will use in their next Computer Chess Championship event at my suggestion), or else play normal chess with openings chosen by lot from some large list, preferably always playing two game matches with colors reversed. Draws are a bit harder. Either we change the starting position, or else change the draw rules. There are arguments for both approaches. Magnus Carlsen just suggested that simply barring Black from copying White's first move would increase White's edge and reduce draws, and I am sure that he is correct. This is clearly not enough for engines, might be enough for top humans. More severe restrictions on Black's first move or two would do even more in this direction, though at an obvious price in terms of discarding historically best openings. The other way, changing the draw rules, would have much less effect on openings, much greater effect on endings. My latest idea on this is simply to add a rule that says that any non-capturing move by a piece to a square it just vacated loses the game, unless the opponent's last move was a check, a capture, a pawn move, or a move that changed castling rights. This would replace the insufficient material draw rule, since with my rule even king vs king would be a win for the side who first took opposition (the opponent would eventually have to move to the corner if we keep opposition, when his only legal move on the next turn would be a losing move to the square just vacated). This would eliminate most perpetual checks, all insufficient material draws, and the majority of repetitions. There would still be plenty of draws by 50 move rule (and some repetitions and very rare stalemates), so most likely the game would still be drawn with perfect play, but the draw margin would be much reduced and White's advantage would at least be much closer to winning, perhaps close enough to make Armageddon practical with modest time odds even in engine vs engine play. I can't imagine any other simple rule-change (without changing the initial setup) that would reduce draws so drastically without fundamentally changing the nature of the game.

[rjgibert]

Using a similar idea, you can eschew the Black not castling proviso and still balance things by awarding a win to Black for short draws and a win for White for long draws. What constitutes a short draw being decided by an I_cut_you_choose procedure.

Unfortunately, this eliminates incentive to win by checkmate--a weakness of Karmaggedon as well. You can restore incentive to win by checkmate by awarding draws 2 points and 1 points according to whether the draw is short or long. Of course checkmate is rewarded by getting all 3 points.

This would balance classical chess and all postions in chess960 as well. In classical chess, a consensus would develop and the I_cut_you_choose ritual would be skipped with fixed number just as Komi is fixed in the game of Go.

[lkaufman]

Using a similar idea, you can eschew the Black not castling proviso and still balance things by awarding a win to Black for short draws and a win for White for long draws. What constitutes a short draw being decided by an I_cut_you_choose procedure.

Unfortunately, this eliminates incentive to win by checkmate--a weakness of Karmaggedon as well. You can restore incentive to win by checkmate by awarding draws 2 points and 1 points according to whether the draw is short or long. Of course checkmate is rewarded by getting all 3 points.

This would balance classical chess and all postions in chess960 as well. In classical chess, a consensus would develop and the I_cut_you_choose ritual would be skipped with fixed number just as Komi is fixed in the game of Go.

That would change chess more than any other proposal under serious discussion. It is true that there is some number of moves (probably a very large number) that would make the game fair; so White's goal would then be to reach positions where there is no plan, but every 49 moves he moves another pawn, until stalling for the necessary large number of moves. Who would want to play that game? I do like the idea of counting something to break ties, with extra points for Black to make it fair (just like Komi in Go), but that "something" certainly should not be the number of moves played. It has to be something relating to who is doing better in the game, such as counting material plus pawn advancement, or mobility, or some combination thereof. The player who came closer to winning by normal rules should get the win (or the 2/3 or 3/4 point if you like).