By far, this is the best chess game I have found. Great job!

This sounds wonderful.

(Actually, since this was a project update, you could have posted another post.  But, if you didn't want to double post, that's fine.  To clarify, you only have to wait for one hour before posting a project update.)

Great job, ThePenguin!

Great!  The greyscale is wonderful!

I see that you changed the front screen. 

thepenguin77, I really have to congratulate you on this one. The AI is actually making decent moves. It's amazing! Whatever your AI is, it's great. I'd like it to be a bit faster, but if that's not possible, that's alright. However, a bullet chess mode where the AI is simplified and you're on a timer would be absolutely amazing! Since it's a calculator game, I'd imagine a fast mode would help its appeal, as not everyone wants to play a full-length game of chess. =)
Also, since the AI is turn-based (I think?...), I'd imagine Chess960 wouldn't be too tough to do either. I'd love it if you included stuff like that!
Madskillz, those sprites are amazing. It'd be pretty awesome if thepenguin77 could make the chess piece "sets" selectable, so you could choose between the 3D and 2D pieces.

Points made:
Bullet chess
Chess960
Swappable chess piece "sets"

The AI has improved a ton. It now can be much more fun to play.

There was another chess game with an ai... and programmable difficulty. I'm not criticising you, I'm saying this because it was open source and you could look at the code. I could find it if you want

I believe the program yoman is looking for is "Chessmaster 83", at http://www.ticalc.org/archives/files/fileinfo/236/23676.html. It doesn't work in Mirage, which prevents me from testing it (I'm lazy.. =D), but I remember looking at it a while back.

Anyways, it DOES have AI, but it's (likely) the same type of AI that thepenguin77 is going for. I posted a comment a couple posts ago asking whether prioritizing center control was possible— I think that would really help the AI, as it would solve a lot of its problems, such as falling for scholars mate, putting knights on the rim, never doing e4/d4 as white, etc...

I don't know if your AI already does this, but I think you should specifically program the AI to prioritize controlling the center, especially early in the game. Right now the AI does strange moves half the time. It only really shines when more pieces come into play, limiting its options. I think making the AI prioritize controlling the center could really give the AI a significant boost in competency, especially when it's playing white. The AI isn't good at playing white because it doesn't really give much thought to the center, allowing you to take control of the game easily.

Of course, whether this would cause significantly more slowdown than there is already, I don't know... Maybe it could be an "hardcore" mode?

I agree with that, I was playing today and I realized that the main reason that it fails so bad is that it has a poor opening. The hard part is telling the AI how to move towards the center because moving towards the center should never put the piece itself in danger, and this is a possibility if I give center moves a value. Also, whatever system is in place should probably only last about 10 moves, because especially in a 3 piece checkmate situation, it is pointless.

One thing I'm going to do to fix this is make the AI consider more move choices, right now it only looks at the last best 4, which is why black always moves the rightmost two pawns. I might also give the move a bonus if it moves a piece away from a wall, but it would have to be small so that it wouldn't put the piece in danger.

There was another chess game with an ai... and programmable difficulty. I'm not criticising you, I'm saying this because it was open source and you could look at the code. I could find it if you want

Sure, I'd love to see it if you can find it, although I believe the one you are talking about is TI-Chess. It does have source but for one it is in C and I'm programming asm. I remember critor talking about this, and he said that he didn't even look at the AI code, he just got it from somewhere else. And lastly, the AI takes like 5 min to make a move and the 68k calculators are about twice as fast as the z80s. This means on the hardest difficulty it would take 10 minutes between moves and even then the AI isn't really all that great.

Edit:
   Sorry, missed a few posts.

Bullet chess
Chess960
Swappable chess piece "sets"

Bullet chess at the moment is not possible because to be fast, the AI would suck really bad.

Chess960 is a maybe, if I could work out the castling situation it would be pretty easy though.

Swappable piece sets will definitely be done. I'll probably also have just a plain letter set too for maximum readability.


@madskillz
    I'll try to use that sprite set when I add in sets. The only thing I worry though is that even zoomed in on that picture, I really couldn't tell the pieces apart.