I have to admit I've never really felt comfortable using ops. I may just be nitpicking here, but adding two percentages - one based on at bats, the other on plate appearances - isn't statistically sound.
HOOTIE's right, of course: there's a strong correlation between ops and performance, but in a sense, it's like adding a team's winning percentage in day games to it's winning percentage on the road. You'd also get a very strong correlation to total winning percentage, but you'd be using a very artificial number.
I realize ops has been very much en vogue for a while now, but I still can't bring myself to like it... especially when I can just look at obp and slg separately.