Most of this is self-explanatory. OPS is OBP plus SLG and (forgive me if I'm stating the obvious) is a good, simple indicator of run production. Anything with an "i" in front of it is an index that tells how the player is performing compared to league average. For example, if Joe Blow has an "iOPS" of 34.77, then his OPS is 34.77% better than league average.
I don't have access to all of the stats necessary to compute OBP, so I've used a simplified (H+BB)/(AB+BB) formula, which, unless you're a mutant like Craig Biggio or Jason Kendall who gets hit by tons of pitches, gives a reasonably good approximation. I also don't have pitchers' hitting stats, so the league averages aren't "true" league averages, but since everyone is affected by this in exactly the same manner, I'm not too concerned.
Not much of a glossary, eh?
Addendum (June 2, 1998). Okay, I've found one other minor problem with the numbers. When a player is promoted, demoted, or otherwise moved from one league to another, USA Today drops their record from the previous league and includes only that from the new league. Because I determine league statistics by summing all the individual players' statistics, this leads to some inaccuracies. But again, this affects everyone the same way, so while it's certainly not the optimum, it's good enough for our purposes.