Engine playing musketeer chess variant

[musketeerchess]

Hi Ferdinand
I missed your post ! Shame on me.

I'll take a serious look at your engine my friend. Hope you found the Musketeer Variant entertaining. All the best and thanks again for your valuable contribution.

Zied

[Angrim]

You will still run into a lack of interest from skilled programmers because of flaws in the design of this variant. The largest of which is that you have 10 new pieces, but only 2 that will ever be played, since you allow the opponent to veto your choice of pieces if you choose something other than those 2. But also the piece placement rules are excessively complex with no clear advantage provided by that complexity.
Most of us(skilled variant programmers) program for fun, and if a variant would not be fun then we ignore it. The offer of money makes it somewhat more interesting, but then the engine code becomes yours which means that we have to write an engine from scratch instead of modifying one of our existing engines. And you sound more like you want to pay for a finished product rather than pay wages, which makes it a very risky investment on our part.

[musketeerchess]

Hi
I'd like to donate and i'm not opposed to the engine remaining in the GNU license if it's based on Stockfish or other well known free engines.

Offering money is another strategy. In fact, my pieces are commercial.

The variation for me is very funny. You really have to try it and you'll see the interest and subtle design. Concerning the initial placement of the pieces, and how to introduce the new pieces, we can change the rules sothat the introduction of the piece in an empty square (1st or 8th rank) when there is an empty square.

The subtle rule behind the way of introduction of the new pieces i currently use is: both parts define their strategy from the beginning, which imho is similar to classic chess where the initial placement of the pieces is known.

[tttony]

wow, first time I heard of this chess variant and it looks dificult, many pieces with many moves, this engine will eat nps

I would not recommend to start with stockfish code, it's big and specific

Good luck with your engine!!

[musketeerchess]

Hi
The chess variant adds only 2 pieces.
To begin a programming we shall only add leopard and cannon.

The code of stockfish is not considered to be very complicated according to good programmers. I'm not a top programmer. I begin to interest in stockfish code as i haven't yet found programmers interested enough in this project.

[musketeerchess]

Hi
I updated the rules for Musketeer Chess.

Take a look at www.musketeerchess.com

3 kits of different pieces are available. They are triple weighted plastic chess pieces that could be used either to customize your classic chess set, to play musketeer chess or to create your own chess variant.

I'm going to begin another post that asks your help concerning the rules of another game. We'd like to create another piece that will replace part or all the pawns. This piece should still be close enough to the pawn but should add or change other features to add more excitment and thrill to the game of chess

[stegemma]

I'd like to finance an engine developer to program an engine playing musketeer chess.

Finance will be proportional to the classic rating of the engine.

musketeerchess(at)gmail.com

www.musketeerchess.org

So I would earn almost nothing, because my softwares are very weak

Out of joking, I have a little experience in changing my engine to play makruk. It was very simple to do but almost impossible, for me, to debug the result. I simply don't know if and why the program plays good or bad moves, because I don't understand the game at all. There were not as many perft information to compare with... definitively it was a pain to program a chess variant, even the easiest one (for me, of course!!!). The musketeer seems more complex than makruk and i think that it would not so attractive, to the mass of chess players. Looking at some piece movement, it seems to me that it was not always coherent; the cannon, for sample, would be more coherent if it could move to all the white squares at distance 2, in the position shown in your site. Pieces that can move as the sum of standard one are easiest to learn (B+K, Q+N and so on) but "fortress" seems... a patched software

A very-strong chess player I think that won't play a chess variant, because doing that he/she could change its "chess sight" (but I don't know so many strong chess players to be sure for that).

[musketeerchess]

Hi
Fortress is in deed a Patched software!
It's name is Fortress, which means it's designed to protect Royal pieces, which explain it's capabilities (it protects the nearby squares better than anyone, it can also launch further attacks (acting like a limited Bishop).

See the rules in the following link

http://www.musketeerchess.com/resource ... Rules.jpg

I also updated my homepage. Better design, better explainations and more pictures of the various pieces.

[stegemma]

Hi
Fortress is in deed a Patched software![...]

Sorry, I was referring to the "fortress" piece in musketeer, not to your software. The kind of movement of that piece looked to me as it looks a software after a set of patches done just to fast solve an emergency, I was not saying anything about the Fortress software itself.

[Evert]

I'd like to finance an engine developer to program an engine playing musketeer chess.

Finance will be proportional to the classic rating of the engine.

This seems like a mistake, since just hacking stockfish to play your variant will maximise what you have to pay without necessarily resulting in the best product for your variant. Being strong in one variant doesn't necessarily imply being strong in another variant. You can tune an engine so that it is strong in multiple variants, but it will never be entirely optimal for any particular variant, and sometimes a term that is a clear improvement for one variant ends up severely weakening another.
Make sure you know what exactly you want to optimise; in this case it isn't (classical) playing strength.

Having said that, I think Stockfish would make a poor starting point anyway. The code is very clean and optimised for playing chess, but it is not so easy to expand it to include more piece types (or at least it wasn't the last time I looked at it, which is a while ago). From what I remember, Senpai would be better suited - but it is of course considerably weaker.