Last season, Grant Balfour threw 1,027 pitches. 938 of those pitches were classified as fastballs. This year, 996 pitches and 820 are fastballs. That's 82% versus 91%, the difference is in Balfour's slider usage, which is at 12.5%. Most impressively is the results the pitch is producing.
It goes without saying that Balfour's best (and only, for the most part) pitch was his fastball last season. FanGraphs pitch type run values have his fastball worth 23.4 runs overall and 2.53 per 100. Every other pitch was in the negative, which is a bad thing. His slider was worth -1.45 runs per 100. That's bad. This year, his fastball has been worth 6.6 runs overall and 0.86 per 100, not as good as last year, but still pretty good. His slider though, has been worth 1.8 runs and 1.57 per 100.
These numbers are defensive dependent, so maybe his slider is getting hammered and Carl Crawford/Evan Longoria are making phenomenal plays every time he throws it, but I really doubt that's the case. It's too small of a sample to state that Balfour has developed a dependable second pitch moving forward. It isn't too small of a sample to say he's done a far better job of throwing another quality pitch this year, even if the sustainability of the results are up in the air for now.
I wanted to touch on the fastball run value a little more too. In his career he's posted a per-100 fastball run value under 1.00 only once, and that was -1.68 in 2007. Every other season has been well above a run gained per 100 fastballs. His velocity is down a bit as is his horizontal movement (about a seventh of an inch) so maybe that's a reason for decline. If we simply work off the prior three seasons as a future prediction, Balfour's fastball should bounce back next year.
Frankly, I'd love to see Balfour's 2009 slider with a bit better fastball.