Gamma World, which some would no doubt call a science fantasy game (and, to be fair, that's how its first edition bills itself), but I don't think that alters my essential point, namely that, when I wasn't playing D&D, my first inclination was to pull out a science fiction-y game like Gamma World or Traveller or the FASA version of Star Trek.
Consequently, I loved "The Ares Section" of Dragon, whose articles, even when they weren't of immediate use to me (like the articles on, say, Universe). Among my favorites, though, were the Gamma World articles by John M. Maxstadt, which I often did use in my games. A good example is "Of Grizzly Bears and Chimpanzees," which appeared in issue #89 (September 1984). As its name suggests, the article is devoted to detailing the unique abilities of animals, in this case as stock for mutated animal PCs. Maxstadt provides some basic statistics for a dozen different animal types -- bears, big cats, herbivorous animals, primates, snakes, and birds. These statistics include things like general size, their ability to vocalize and grasp/carry items, in addition to more obvious game stats like armor class and movement rates. The idea behind the article is to rationalize the abilities of mutated animals both from a game mechanical and a logical perspective, thereby making them more attractive to play and easier for the referee to accommodate.
Looking back on the article now, what's fascinating is how simple it really is in the end. There are a couple of pages of game stats, presented as Monster Manual-like entries, followed by a couple of pages of explanation of what the stats mean and how they interact with other aspects of the Gamma World rules. That's probably why I found them so easy to use. At the same time, they carry with them an implicit vision of Gamma World, one that's a bit more limited than the wide open "wahoo!" style usually associated with the game. Maxstadt, for example, doesn't provide stats for insects or amphibians, so the referee is either left to his own devices in coming up with his own or else disallowing such mutated animal types, as Maxstadt apparently did. Now, there's nothing wrong with such a limitation and indeed there's definitely a case to be made for it, but, somehow, the idea of playing Gamma World with any limitations seems to go against its fundamental grain and, were I ever to run a campaign again, I'd probably not use this article's system or else come up with additional stats for other types of animals.