The Pitcher Type You Can Exploit
One of the things that I try to do as a fantasy owner—and what’s happening here—is sort players into buckets based on what “type” of player they are. We’re using batted ball profiles in this course to define player types, but we could really include anything we want as long as we weigh it according to how much it aids in projecting players.
Side note: I think this is actually the future of BvP, and really overall player analysis. I don’t really use Batter vs Pitcher data myself except to help predict ownership levels in tournaments, but I don’t think it’s useless in theory. It is natural to assume that certain batters will hit certain pitchers better than others due to a whole bunch of factors. The problem is that it’s just tough to identify who is “good” and “bad” based on past numbers because we normally don’t have a sufficient sample size.
If a particular batter can be expected to hit, say, .350 over the long run against one pitcher and .250 versus another pitcher—which could certainly be a realistic scenario based on all the different factors that affect batter versus pitcher matchups—we’d still have a very difficult time determining that batter’s “true” long-term average after even dozens of at-bats. Even if we knew the long-term averages, we still might be able to say with only 60% or 70% confidence if he’s “good” or “bad” versus that pitcher, which isn’t going to get it done.
The point is that our problem with BvP shouldn’t be that it doesn’t “exist,” but rather that we aren’t seeing enough one-on-one matchups to be confident that the conclusions we can draw and the data we get from those matchups are accurate.
In effect, we’re creating a revised version of BvP here by sorting players into these buckets. We’re saying, “Ideally I want to know exactly how this specific batter matches up with this particular pitcher. Their limited history won’t tell me much, so I’m willing to trade in a small amount of relevance (by considering similar players and situations) to expand the sample size in a major way.”
I think this type of “comps analysis” is the way of the future in DFS, particularly because we’re generally working with very specific situations and need ways to overcome the small sample sizes of relevant past data we have.
Now back to the topic at hand…looking at batters based on which types of pitches they crush.
Batted Ball Profiles and Pitch Types
One of the most consistent stats in baseball that few daily fantasy players consider is how well a batter hits certain types of pitches. We spend all our time breaking down a player’s ISO and wOBA—and rightfully so, that’s awesome data—but don’t go any further from the level of handedness to the level of pitch type.
Below, I charted the correlation between ground ball and fly ball percentages with how well batters hit four pitch types: fastballs, sliders, curveballs, and changeups. For more information on how those numbers are calculated, visit FanGraphs.
For off-speed pitches, there isn’t much of a correlation. Fly ball hitters hit curveballs and changeups better than ground ball hitters, but the former group is just better in general. These differences are negligible.
But take a look at the fastball correlations. What this information is telling us is that ground ball hitters do not hit fastballs well relative to other pitches, whereas fly ball hitters crush fastballs relative to other pitches. That makes intuitive sense, too; it’s more difficult to put a slider into the air than it is a fastball.
With this data in mind, I think there are some basic conclusions we can draw:
1) Try to avoid ground ball batters. They still suck.
2) Target fly ball hitters versus pitchers who have weak fastballs. You can find which pitchers struggle with types of pitches at FanGraphs.
(By the way, we’re concerned with how a pitcher gets hit when he throws a fastball—not the velocity of it, although the two are often linked.)
3) Target fly ball hitters against pitchers who throw fastballs in abundance.
4) Avoid fly ball hitters when they face pitchers who throw a ton of junk.
The End.