I'll be the first to tell you that I'm not a saber-researcher. Every now and then I'll try to dabble my toes into that pool, but for the most part it's way beyond me and I'm content with the writer/educator role. So...I've got an idea but I don't have the ability to carry it through, or even to know how realistic it is. I'd love some input from whoever feels like offering.
Basically, my idea boils down to this: much like we can calculate a Swinging Strike Percentage, would it be possible to create a Fat Pitch Percentage and/or a Nasty Pitch Percentage? When commenting on a recent Mike Fast article, he mentioned that pitch location and sequencing can have significant impacts on how well a pitcher performs on a day-to-day basis, and so I thought it would be interesting if we could quantify this to one extent or another.
The trick is, I imagine there'd be a heck of a lot of legwork to get something like this implemented, and even then it wouldn't be perfect. What's the league average run value for 90-92 MPH fastballs with x and y movement in this section of the plate? Okay, now what about the run value for that same pitch, but in a different location? Establish run value cutoffs for "nasty" pitches and "fat" pitches, and you'd be able to tell if that last fastball was in a good spot with good motion, or if it was a fat pitch that should have been hit. How many of his pitches classified as "nasty" and how many as "fat"? I'd find it interesting, for sure.
Of course, this strategy would completely ignore pitch sequencing, which is another important factor and might be just as important as location. I know this isn't a perfect idea, well, that's why I'm tossing it out there as merely a concept. I do think that if we could implement some system like this, it'd allow for even better analysis of pitchers on a micro scale.
Anyway, you're all much smarter about this stuff than I am. What do you think?