3-D Tic-Tac-Toe

[devwebcl]

Hi All,

 

I was researching of the game 3-D Tic-Tac-Toe. It is mentioned in Wikipedia and Atarimania that both versions (800 & 2600) were done by Carol Shaw.

Is this correct ?

 

Also, anyone has beaten this game ?

(in different levels)

 

Cheers,

 

Devwebcl

[Kiwi]

http://www.atariage.com/software_page.html?SoftwareLabelID=521

 

This page have the information you need. It has the programmer name, Carol Shaw, which is correct. I'm not sure about the 800 version.

 

Videogamecritic.net review of this game is full of baloney and they gave it an F. I think this game is great again the CPU and there's strategy involve to defeat your opponent. Also, you can play against a player. I always wondered what's going on during the "thinking" process and why does it take so long.

 

Here's a blog of a good review of this game and I had a small discussion with Mezrabad about this game. http://www.atariage.com/forums/index.php?app=blog&blogid=87&showentry=6772

[Pitfall Harry]

Hi All,

 

I was researching of the game 3-D Tic-Tac-Toe. It is mentioned in Wikipedia and Atarimania that both versions (800 & 2600) were done by Carol Shaw.

Is this correct ?

 

Also, anyone has beaten this game ?

(in different levels)

 

Cheers,

 

Devwebcl

 

I've beaten the game, on all its levels. But then, I've been playing 3-D TTT since before it even made its debut as a 2600 cartridge game. I dedicated the entire summer of 1989 to developing a computer program to play "unbeatable" 3-D TTT. I never quite achieved that goal, but I did come fairly close to it. The program I wrote does an exhaustive game tree search for "winning combinations," which is a succession of 3-in-a-rows that force the opponent's options down to a single move (the block) to avoid losing on the next turn. The computer immediately follows one 3-in-a-row with another, then another, until eventually, the computer finds a move that creates a 3-in-a-row in two different paths simultaneously, which forces a win on the next move. Sometimes the winning sequence of forced 3-in-a-rows is as much as 21 moves per side deep. But no matter how deep the search had to be to force a win, if it existed, then my program would find it.

 

The forced-win search algorithm lies at the heart of my program, although other strategic elements were necessary to program. I had to rely on my own personal experience with the game to program solid opening game play, and I had to develop a game scoring heuristic to determine the computer's best move when no winning sequence was found.

 

If you read the instruction manual for 3-D TTT, you will see that the Atari 2600 is capable of only 9 move lookahead on its most advanced level. With my program's ability to sniff out winning sequences 21 moves deep and beyond, the Atari 2600 version was no match at all for my program.

 

To be fair about it, though, 1991 computer technology (a 10 MHz 80286, equipped with a 30 MB hard disk and a full megabyte of RAM!) is an awful lot of muscle to pit against a 1977 vintage Atari 2600 that wheezes along at 1 MHz on 128 bytes of RAM. Not megabytes. Not kilobytes. I mean, the Atari 2600 has only 128 measly BYTES of RAM! How anybody could have ever gotten this machine to do anything at all with only 128 bytes of RAM to work with never fails to astound me. Considering the limitations of the eras, Carol Shaw's 3-D TTT programming accomplishment far exceeds my own.

 

Also, my program is not very special at all. The game of 3-D TTT is now completely solved and has been since as early as 1976 (Google search on Eugene Mahalko, Oren Patashnik, and Victor Allis, for more details). Then again, my 3-D TTT program will always be special to me, because what I learned about the game as I developed my killer program made me understand the game at a far deeper level than I ever had before. It was only after my programming project was completed that I had developed skill enough at the game to be able to consistently beat the Atari 2600, unaided, at its highest level of play.

 

It's an unending source of personal pride for me to be able to demonstrate mental abilities that surpass that of a 4K ROM from 1979 and 128 bytes of RAM.

Edited December 1, 2010 by Pitfall Harry

[devwebcl]

Probably you already know this interview.

 

But here it is http://www.vintagecomputing.com/index.php/archives/800#more-800

Carol Shaw talks about her games, including 3D Tic Tac Toe.

 

Actually, answers my questions

[+SpiceWare]

That was interesting, thanks for posting the link!

[sega_SHARK]

Thanks for the link, I'm drooling over the Colecovision prototype!

[devwebcl]

In this iterview is mentioned that Atari Checkers is better than Activision Checkers... does anyone know more details on this ?

[devwebcl]

I've beaten the game, on all its levels. But then, I've been playing 3-D TTT since before it even made its debut as a 2600 cartridge game. I dedicated the entire summer of 1989 to developing a computer program to play "unbeatable" 3-D TTT. I never quite achieved that goal, but I did come fairly close to it. The program I wrote does an exhaustive game tree search for "winning combinations," which is a succession of 3-in-a-rows that force the opponent's options down to a single move (the block) to avoid losing on the next turn. The computer immediately follows one 3-in-a-row with another, then another, until eventually, the computer finds a move that creates a 3-in-a-row in two different paths simultaneously, which forces a win on the next move. Sometimes the winning sequence of forced 3-in-a-rows is as much as 21 moves per side deep. But no matter how deep the search had to be to force a win, if it existed, then my program would find it.

 

The forced-win search algorithm lies at the heart of my program, although other strategic elements were necessary to program. I had to rely on my own personal experience with the game to program solid opening game play, and I had to develop a game scoring heuristic to determine the computer's best move when no winning sequence was found.

 

If you read the instruction manual for 3-D TTT, you will see that the Atari 2600 is capable of only 9 move lookahead on its most advanced level. With my program's ability to sniff out winning sequences 21 moves deep and beyond, the Atari 2600 version was no match at all for my program.

 

To be fair about it, though, 1991 computer technology (a 10 MHz 80286, equipped with a 30 MB hard disk and a full megabyte of RAM!) is an awful lot of muscle to pit against a 1977 vintage Atari 2600 that wheezes along at 1 MHz on 128 bytes of RAM. Not megabytes. Not kilobytes. I mean, the Atari 2600 has only 128 measly BYTES of RAM! How anybody could have ever gotten this machine to do anything at all with only 128 bytes of RAM to work with never fails to astound me. Considering the limitations of the eras, Carol Shaw's 3-D TTT programming accomplishment far exceeds my own.

 

Also, my program is not very special at all. The game of 3-D TTT is now completely solved and has been since as early as 1976 (Google search on Eugene Mahalko, Oren Patashnik, and Victor Allis, for more details). Then again, my 3-D TTT program will always be special to me, because what I learned about the game as I developed my killer program made me understand the game at a far deeper level than I ever had before. It was only after my programming project was completed that I had developed skill enough at the game to be able to consistently beat the Atari 2600, unaided, at its highest level of play.

 

It's an unending source of personal pride for me to be able to demonstrate mental abilities that surpass that of a 4K ROM from 1979 and 128 bytes of RAM.

 

Returning to his thread... I tested 3d Tic Tac Toe in 800xl and looks to have implemented a perfect game.

 

The important point is to let to the game go first, as we know whoever go first then must win if plays a perfect game.

 

So, returning to the 9 depth lookahead search, it could be a good constraint, probably it is not necessary to search deeper.

[Tempest]

Did anyone ever dump her preliminary version of River Raid with the changing colors?

[davyK]

I never understood the hate 3D TTT gets. I found it a solid enough game. The display is well thought out and it gives you a good game. It is what it is.

[Xot]

A couple of weekends ago my 14 year old son found this cart in a stack and asked to play with me. Once he got the hang of making lines across the different planes, he was hooked and we ended up splitting the next eight or so rounds. Fun game, no doubt about it.

[7800]

Really, people hate this game? WTF? It's one of the best for Atari 2600 In my opinion.