I had this for the Apple ][e and would play it, obsessively. At one point I had gotten good enough against the computer that the average game took ninety secon

I had this for the Apple ][e and would play it, obsessively. At one point I had gotten good enough against the computer that the average game took ninety seconds, with the computer taking about as much time as I did. I eventually learned how to completely beat the computer such that it had no cells of its color. I took photos of a couple of instances just as proof. After a while, that got dull so I wrote a little program that would randomly POKE into the game's memory space. The usual result was a crash, or a program that didn't know how to play, or would only put pieces on the lines instead of the empty cells, that kind of thing. Every so often, though ... it would just play a little differently against me. No crashes, no weirdness. I always wondered what exactly had gone on with my random mutation there: was there some table of position evaluations whose weightings I had tweaked? A rule jumped over in the code?