There's a few in the iPhone app store but they all seem to have pretty dismal ratings. We have any developers here that could/would take this on? A solid 42 app would be spectacular.

+1

I would pay money for this.

+2

And if somebody here does take it on, you know you've got well trained beta testers.

+1
For droid

someone provide the algorithms\AI and it could get done.

+1

I wrote a 42 windows app in college for an independent study class (485) with TCP/IP network capabilities... unfortunately, my computer crashed and I lost it. This was a number of years ago...

I actually coded and wrote custom ActiveX controls for the original version for the domino elements. I am so far out of coding that it would take me too long to get up to speed on the graphics engines, display elements, and what not.

However, the algorithms were the easy part and my computer 42 AI was pretty good.

If someone gave me a good looking shell game on an Android platform (with domino controls to work with), then I would consider getting involved in this project.

Actually working on an iPhone 42 app as we speak. I'm to the point where you could play a full game as long as you control all 4 players. At this point, I just need to figure out how to handle the AI...

[This message has been edited by Philip j fry (edited 9/19/2011 9:53p).]

Sweet! Keep us informed.

quote:

---

However, the algorithms were the easy part and my computer 42 AI was pretty good.

---

playing who, you?


to make any 42 AI worth their weight in code, you have to introduce a controlled randomness in their playing.

The first 42 app for the iPhone was complete crap... the second is a little better, but the AI is lacking.

No, the AI algorithms are the impossible part of coding 42 for a single player. Getting the play functions down is quite simple, you just have to make your own algorithm for the rules of bidding and play.

I attempted to take this on about two years ago, except I found that implementing an app is neither easy, nor cheap... so I stuck with c++, and stalled where I started to need to create graphics for the dominoes, but the game works in text.


The biggest problem with coding the AI is that there is no set algorithm for play, it depends on who bids what, when, and what the score is... then what trumps are, what is lead, and what you are looking at in your hand.

If you code all of the AI to think alike, the game becomes predictable... which is boring. If you code the AI to make absolutely random plays, the game is just stupid and no fun to play.

For a good starting point, first, you have to look at the bid, the trumps, and what has been played, then identify, while looking at least three tricks ahead, what the best play is... If you think making a chess engine is difficult, keep in mind, you get to see both sides of the game... there is always a "perfect" play on a chess board. While in 42, there is always a "best" play, but determining that for a fact involves making decisions about dominoes you can't see and have to make a guess on... at a table it's not so difficult... programming it, impossible.

Rebel, that's exactly what I'm running into. I've been taking notes on my own thought process while I play and found that it difficult to express in Boolean form. Pretty amazing what the human brain can do eith just a quick glance.

So far, coming from a limited C programming background and zero experience in creating sprite artwork, just getting to a workable game was a challenge. I'm excited to start working on the guts of the matter now though.

I've been working on an iPad version of 42... I don't have any of the graphics, but have been focusing on the algorithm.

I've kind of got a decent idea on HOW to do it down, but this is my first iOS app and I've been converting my Python code to CPP.

Fry, email me if you want at my username at gmail

quote:

---

Leela: We can't use our ship; we have life support but the engines are wrecked.

Takei: Ironic, because our engines work but our life support systems don't.

Leela: Hey, if your engines work--

Takei: And your life support systems work--

Fry: Stop! You're just goin' around in circles! (to himself) Think, Fry, think! Everyone's depending on you.

---

Ya, something like that....

quote:

---

playing who, you?

---

Nice presumption of my coding skill versus yours...


I graduated with a B.S. in Computer Science and Eletrical Engineering. The algorithm direction was assisted by Dr. Jan Wolter, the at-the-time professor for Artificial Intelligence. We tested it out considerably and then during a second 485 class I added the TCP/IP stuff. I think I knew what I was doing... and so did pretty much everyone who played it. I wish I would have tried to market it commercially, as I invested quite a bit of time in it.

But, glad you feel the need to be superior and what-not on an internet forum...

The AI is not impossible because this guy did it:

http://www.ccdominoes.com/download.html

I have had that game for a few years, and it is pretty fun. Some other folks on TexAgs have it, too, because this is where I learned about it.

I used to pick out the bugs in curtis's code...

for instance... Did you know that you used to be able to prevent the AI from playing a domino by clicking and dragging it around?

Mr. Cameron's AI was very easy to beat, had no hint of indication, hated if you lead a trump that wasn't high, wouldn't listen to it's partner, always held for the 10 on an 84 hand, and just flat out did not play the game as a team player...

to program a decent algorithm for 42, you have to create AI that can learn how to assess a situation it isn't programed for, and play accordingly.

42, when played correctly with skilled opponents is a game of 'why'... The one question a computer cannot answer without being spoon fed the answer beforehand.

...and no hard and fast rules of strategy will win 42.