The Pirates aren’t known as one of the more Sabermetrically-savvy teams like the Rays and Athletics, but they do have Dan Fox (formerly of Baseball Prospectus) as the director of baseball systems development. Thanks to Fox and others, the Pirates are able to use stat reports to prepare lineups and pitching match-ups with greater specificity. SB Nation Pirates blog Bucs Dugout talked to manager Clint Hurdle, who provided more details about the way he utilizes the tools at his disposal despite not being a number-cruncher himself.
Many statistically-oriented writers have criticized managers’ tendency to give -- to use an example -- a bench player a start because he is 4-for-9 against the opposing starter. The reasoning behind that is nine at-bats is a terribly small sample size, making any information gathered from those nine at-bats largely meaningless. However, the Pirates’ method of grouping similar pitches together is quite interesting and makes sense. Grabbing, say, Andrew McCutchen’s performance against Cole Hamels won’t tell you much since it’s only 17 plate appearances. But grouping together left-handed starters with a 91-93 MPH fastball and an 82-84 MPH change-up would expand your sample size to a level where you can start to make conclusions about a player’s performance level.
We have a system analysis that is so unique that what we do is, we have player batting averages, swing and miss rates, on base percentage, OPS, it is all laid out for that pitcher and 15 comparables. So truth be told, at times you can get a player that is 10-for-20 off a guy in real time and he doesn’t rank maybe in the top of your batting order if you were constructing one sabermetrically over the long haul. But also you can get a pretty good feel on what that kind of guy can do against those kinds of pitchers. It’s tool. It’s a useful tool. I’ll say that.
I’ve grown in the time I’ve been here by being open-minded, knowing I have some people upstairs that are really, really smart. But you have to mesh the two.
Many statistically-oriented writers have criticized managers’ tendency to give -- to use an example -- a bench player a start because he is 4-for-9 against the opposing starter. The reasoning behind that is nine at-bats is a terribly small sample size, making any information gathered from those nine at-bats largely meaningless. However, the Pirates’ method of grouping similar pitches together is quite interesting and makes sense. Grabbing, say, Andrew McCutchen’s performance against Cole Hamels won’t tell you much since it’s only 17 plate appearances. But grouping together left-handed starters with a 91-93 MPH fastball and an 82-84 MPH change-up would expand your sample size to a level where you can start to make conclusions about a player’s performance level.
