What initially attracted you to making a computer play a game?
That actually started back in 1966, when I was a high school sophomore, and a friend of
mine named David Zeuch introduced me to the Avalon Hill board wargames. We played
those, and I thought they were a lot of fun. I played them into college, though I didn’t
have a lot of free time during my college years. When I was in graduate school, I ran into
a fellow who worked at the computer center, and he was trying to getBlitzkrieg,an
Avalon Hill game, running on the computer. I told him he was crazy. I said, “That can’t
be done, forget it.” But that conversation planted a seed. I thought about it, and about a
year later I decided I was going to attempt it. So I went to work and it turned out to be
nowhere near as difficult as I had feared. So I ended up putting together a little program
on an IBM 1130 in FORTRAN. It actually ran a computer game, a little tactical armored
simulation. The debut of that game came early in 1976 when I showed it off at a little
wargame convention that we held. Everybody played it and thought it was a great deal
of fun. So then I bought myself a KIM-1 and redid the whole thing around that system.
That design was unmatched for many years, because you had genuine hidden move-
ment. I had built little tiny terminals, as I called them, and each player had his own little
map and little pieces, and a screen to divide the two players. Two guys played this
wargame, each one unaware of the position of the other. It was a lot of fun, and that was
1977 or ’78.
What made you at first think it would be impossible?
The difficulties of organizing the artificial intelligence for it. I thought, “That’s just
going to be impossible.” And the hex-grid motion, I figured that was probably comput-
able, and in fact it turns out it’s not that difficult. But I figured that doing armored
tactical planning on the computer, at the time, seemed ridiculous. Now, you have to
remember that was twenty-five years ago, and given the state of AI back then, I was
really on rather solid ground thinking it impossible. But as it happens I solved that prob-
lem, marginally, within a year.
What made you think it would be worthwhile to put games on the computer?
I was driven by one thing and that was “blind” play. I was very concerned that, no mat-
ter how you looked at it, with board games you could always see what the other guy was
up to. And that always really bothered me, because it was horribly unrealistic. It just
didn’t seem right, and I thought the games would be much more interesting blind. And,
in fact, when we did them, they were immensely powerful games, far more interesting
than the conventional games. And as soon as I saw that, I knew that this wastheway to
go. And board-play technology has never been able to match that simple aspect of it. It
was so much fun sneaking up behind your opponent and, as they say, sending 20 kilo-
grams up his tailpipe. It was really impressive stuff, very heady times.
So from that early work, how did you come to work at Atari?
Well, actually a bit more transpired first. I got a Commodore Pet and programmed that
in BASIC with some assembly language routines to handle the hex-grid stuff. I had
shown my tactical armored game at some wargame conventions and everyone had been
very impressed. So then I actually madeTankticsinto a commercial product and sold it
258 Chapter 14: Interview: Chris Crawford