Re: Turn Based Games

[EMAIL PROTECTED] (Jason Reposa) wrote in
news:[EMAIL PROTECTED]: 

> Hello all,
> 
> I'm a professor that has been given the honorable task of creating a
> video game programming course. One of my assignments in AI will be to
> have the students implement a function called algorithm() using a code
> base that I would provide.
> 

It turned out that you are interested in computer games (table games
conversions subcategory). 

Please be more clear next time. 

> Using a common protocol (tcp/ip of course) I was hoping to have the
> students pair off, complete algorithm(), and have a tournament to see
> who's algorithm worked the best.
> 
Isn't better to teach them cooperation? 

If you'd like competition do a virus developing. Goal would be to mutate
virus to vaporise FAT. ~_^ 

> I'm having a hard time trying to find a good turn based game to base
> the lesson on. I was thinking tic tac toe, but using a* you can come
> up with the best solution in near nano-seconds. I want something with
> a little more variance than that, but not too complex.

Honestly what are you trying to do? After litening rumblings about how
University students are awsome and easily employable and how I should by
some miracle should get 4 years of secondary Highschool and 5 years of
University in as little time as posible, and blast my programming skills
and experience in game development in the process (offices for employing
in Czech republic are based on trying to employ people by paper and never 
by abilities), I become very uneasy about University students that were 
taught something in theirs AI courses.
What would you try to do to them? Teach them programming? Tech them math? 
Teach them some introduction lessons about AI? Strong or weak? 

If you'd like advice. Things that programmers shouldn't be taught are:
tic tac toe, and buble snorting something like algorithm. 

> 
> I was looking into Othello, but I wanted your input before I made
> anything concrete.
> 

You wouldn't get definite input before you'd state if you'd like to
teach them the strong or the weak AI, and what is your qualification to
do so.

Of course you migth have some problems with teaching becose AI research 
looks somehow like this: Most of work that has been done was done in game 
development and is closed for public. Majority of other AI researchers is 
doing mainly just math like articles that are directly unusable and in my 
opinion majority of that articles are useless. I seen also some rambling 
that strong AI research is unwanted, mostly becose majority of AI task 
could be done by weak AI. So I'm curious what would you like to teach. 

I very don't recomend to create checkers, or othelo becose it's somewhat 
easy to come with best tactic so it'd become boring quickly. Especially for 
AI creation. You wouldn't want to teach just tree searching. 

I would recommend to take one of this three posibilities. It might be more 
difficult way, but more revarding at the end.


You could try to write mob for playing Nethack. With this modification: 
Before start you'd have nice GUI with menu and choice between psycho, stab 
from behind, pacifist with big club. You get the idea. That were 
modifications of mob AI that would hopefully alow reach goal of Nethack by 
different playing style.

Or you could do some work with that soccer that has been offered to you 
somewhere in this discusion. Use the Java version please. And use AI as 
multithreaded. I would recommend to add a training of players to the soccer 
so it wouldn't be just about creation of AI that would control each player, 
but also about creation of AI that would be able to do manager and trainer 
work on team. Add comments too. That would be fun.

And of course you could create program for playing go. I remmember on my 
program for generating introduction moves. That wasn't based on library, 
and was able to generate all strong starting moves that are commonly seen. 
It didn't went for useles starting battles either. 
Actually best strategy for go is Play whatever you want I'd beat you 
anyway. And it's possible to be played by AI.


That three possibilities would work, possibly would teach students some 
basic for real game developing, and wouldn't be just substitution of a math 
teaching class. 
Thought it could be work for more that one year. So you should hope that 
your Uni would alow you to create courses that would leap over two years.