Creating chess variants on the spot

[gbtami]

Do you have any info of the status of chess variants in lichess? I mean, I know they are playing Shield Tournaments (Shield Crazyhouse starts in 1 hour) and money tournaments for chess variants, but do they have plans to implement another, and if so, which one has better priority? Are they thinking about implementing double boards for Bughouse or boards for different layouts? Or will all this be normal looking variants for a while?

https://github.com/ornicar/lila/issues/441

[Nordlandia]

I proposed wild9 for lichess. For unknown reasons it got downvoted and closed within hours.

https://github.com/ornicar/lila/issues/3442

[Ovyron]

Do you have any info of the status of chess variants in lichess? I mean, I know they are playing Shield Tournaments (Shield Crazyhouse starts in 1 hour) and money tournaments for chess variants, but do they have plans to implement another, and if so, which one has better priority? Are they thinking about implementing double boards for Bughouse or boards for different layouts? Or will all this be normal looking variants for a while?

https://github.com/ornicar/lila/issues/441

Okay, thanks! Surprising to see there was recent activity there.

Lichess looks like the perfect solution, I just played a King Of The Hill tournament there and it's full of interest and strong players. It required 20 rated games to enter, but I had no problem finding people to play with on the hour prior to entering.

So what got people in the fist place to implement the changes necessary for crazy house? It's not trivial, and seems like a lot of effort because you have to change the interface to allow players to see the bench of the pieces and to be able to drop them and all that, and it happened. Was this popularity alone so we will not see something major implemented until it's as popular?

[Ovyron]

I proposed wild9 for lichess. For unknown reasons it got downvoted and closed within hours.

https://github.com/ornicar/lila/issues/3442

My problem with Wild9 is that it gets rid of the Bishop pair for no reason and the whole mechanics of it are a part of what makes chess great, having to decide if you capture the opponent knight or run away to the h file when attacked by the h pawn and stuff like that.

Having the capturable king shouldn't mess with this, plus, two knights are weaker than two bishops because they get in the way of each other and are redundant, so the better choice would have been to replace the b knight with the king.

[hgm]

Actually this redundancy of Knights seems to be a red herring. Knights cooperate extremely well, and 7 Knights crush 3 Queens, where 7 Bishops have a very hard time to even achieve a draw. Still, I agree that this TwoKings is a silly idea ('non-designed variant').

I wonder what would happen to the value of a Bishop pair on a 7x8 or 9x8 cylinder board. It would be intersting to test that.

[Nordlandia]

Actually this redundancy of Knights seems to be a red herring. Knights cooperate extremely well, and 7 Knights crush 3 Queens, where 7 Bishops have a very hard time to even achieve a draw. Still, I agree that this TwoKings is a silly idea ('non-designed variant').

I wonder what would happen to the value of a Bishop pair on a 7x8 or 9x8 cylinder board. It would be intersting to test that.

Right. The redundancy of two knights is relatively small. And TwoKings is probably poorly designed.

Alternative setup:
Same as Standard Chess setup, with Queens replaced by Kings.

http://www.chessvariants.com/invention/kings

[Ovyron]

Actually this redundancy of Knights seems to be a red herring. Knights cooperate extremely well, and 7 Knights crush 3 Queens, where 7 Bishops have a very hard time to even achieve a draw. Still, I agree that this TwoKings is a silly idea ('non-designed variant').

Yes, I read all the threads on the subject (though, they were mostly Tsvetkov arguing the opposite of what was true, and people arguing with him.)

However, in practice, in my experience, it's always better to have different pieces rather and redundant ones.

My order would be the following:

Two Bishops (Opposite colors)
Bishop and Knight
Two Knights

Things may go haywaire if you add a third piece of any kind, and multiple Knights may work well together, but the fortresses Knights can build to defend each other doesn't seem as good as being able to mate the opponent's king.

Note on Nordlandia's addendum: I don't like the idea of replacing the queen with a crippled queen that may only move one square at a time.

What about this?:

A Commoner appears once per game whenever the player develops a Knight or Bishop, on the vacant square.

This would make it a 34 pieces on the board variant, but players would be able to have both Knights, both Bishops, and decide where they want the Commoner to start.

[Nordlandia]

Only issue about that is it increase length of castling.

[Ovyron]

Well then, maybe allow the player to decide whether they summon the commoner when developing a piece or not. Perhaps we can allow them to make it appear in any vacant square of the first rank, left by a piece that hasn't moved yet (say, you could summon it by castling, in either the king's square or the rook's square.)

[hgm]

You could also gate the co-King on the board on squares passed over in a double push of a Pawn. Then it starts better developed, which seems an advantage for such a slow piece.

It only takes 10 sec to invent a new Chess variant...

That reminds me: someone started a new forum about Chess variants:
https://chessvariantforum.createaforum.com/index.php .

They currently run a design contest for a variant on a 10x10 board. You are all encouraged to participate!