What other variants should be added to chess?

@nadjarostowa Honestly, people play what is already available. We don't know how the status of S-Chess would have been different if it had been greenlit 7-8 years ago when it was considered and on the agenda for discussion. It's just pure guesswork. Would it take off or would it remain a pure curiosity for most people?

I just looked up the numbers of games last month:

That's 1.5 M games in variants (and this includes 367 k games of chess960, which is not really a variant in that sense), vs 91.5 M in standard chess.

That's 1,5 % (or 1,2 % - depending on how you treat 960) of games. And this is for the currently available variants, which are probably the most popular.

Also, the less popular variants have considerably less games, like Racing Kings with only 58 k games played. That's 0,06 % compared to the amount of standard chess.

Adding more variants will likely split "variant players" into more variants, but not really bring many new players.

Looking at chess.com, which has more variants, give a similar expression: of approx. 186 k players online, 450 are enganged in variants. That's 0,2 %.

Those numbers are just the opposite of "we all want more variants". It rather suggests that dropping variants would be more appropriate than adding new ones (I don't recommend, either).

I am uncertain what would be considered optimal or ideal, given the evident frustration expressed by many. There have been requests for 960 to be made available for existing variants. However, I seldom observe titled players above the FIDE Master level participating on the variant server on Chess.com, with the exception of one active International Master who performs well in several variants. Most GM's seem to ignore the variant server entirely.

"we all want more reasonable variants" – which builds on common chess understanding. This will allow people to use already accumulated knowledge without starting from scratch. Ease the transition.

None.

Chess engines analyze, AI can arbitrate time.

Code has already changed the game of chess, especially when it comes to analyzing positions, but what if it could do more? Imagine AI taking over the clock and scoring system to make games faster, more dynamic, and less prone to annoying draws. Lichess could try to replace the old 19th-century clock approach with something that truly fits the AI era. We don't have to wait for FIDE, do we? Can we not create a variation to the Bronstein increment time?

1. Bonus Time for Captures Only: Ditch the standard increment time that rewards players for shuffling pieces around and go for bonus time that rewards meaningful actions like capturing opponent pieces.

2. Proportional Bonus Time: The amount of bonus time would be proportional to the value of the piece you captured. Take a queen and you would get a bigger bonus time than capturing a pawn. This would keep the clock in sync with the actual flow of the game. Bonuses would be an incentive to seek exchanges and aggressive play. A simple formula could be to use standard piece values (Pawn=1, Knight/Bishop=3, Rook=5, Queen=9) and multiply it by a time bonus value of 10 seconds. Give smaller bonus time for bullet games and larger bonus time for Classical games.

AI can do much more than just analyse a game. It can be active in a game without interfering. Engines seem to love draws, because I assume they are programmed not to lose. So cheaters will have to learn to avoid the best moves and start picking exchanging moves to gain time. I don't think chess engines can do that yet. It definitely does not exchange the appropriate pieces to gain the required time to win. It would require a new training approach to do that.

See we don't have to change chess into a variant, we just need a AI-driven clock. We had sand clocks, mechanical clocks, electronic clocks and now AI-driven clocks is the next best thing for chess.

Look at the graph in the link that Grok found. It sure seems like we are changing the way we play chess !!
Do engines exchange less than human players or does it only depend on skill levels?

https://randalolson.com/2014/05/27/a-data-driven-exploration-of-the-evolution-of-chess-moves-captures-and-checkmates/

@Toscani said in #26:

1. Bonus Time for Captures Only: Ditch the standard increment time that rewards players for shuffling pieces around and go for bonus time that rewards meaningful actions like capturing opponent pieces.

So long fine manouvres are now penalized? Captures are pretty limited. What is best depends on the position.

What an absurd and horrendous idea.

2. Proportional Bonus Time: The amount of bonus time would be proportional to the value of the piece you captured. Take a queen and you would get a bigger bonus time than capturing a pawn. This would keep the clock in sync with the actual flow of the game.

No, it would be random, with no relation to what the position demands or not.

See we don't have to change chess into a variant, we just need a AI-driven clock. We had sand clocks, mechanical clocks, electronic clocks and now AI-driven clocks is the next best thing for chess.

It's not "the next best thing". It's utterly stupid, sorry. It makes absolutely no sense at all, and it ruins any serious game play.

@tcooke said in #7:

but about ten thousand play Chess 960, which is probably the most popular variant).

No actually, according to database antichess is the most popular.

@nadjarostowa
How about inventing your own ideas instead of criticizing and trying to shame others publicly. Don't squish others down to raise your self in public view.

I realized after playing a game with the thought of exchanging a piece to increase my clock, I would have to give more time to the sacrifices made to help compensate for the material loss. The invention of the clock is what caused the issues that chess has today.

The clock race is the wrong approach for chess. It's like using a car to drive over water and a boat to float on pavement.

Now that we're on the subject, what's the reason Lichess doesn't choose to go for other possible time formats?