Why do sinking fastballs cause groundballs?by Chris Constancio
August 03, 2007
There are at least two basic hypotheses to explain why sinking pitches lead to groundballs. If low pitches generally lead to more groundballs than other pitches, it's possible that sinking fastballs are associated with groundballs because the pitches simply end up in the lower half of the strike zone more often than the more traditional four-seam fastball. The second hypothesis proposes that location doesn't matter because the sinking movement causes batters to misjudge and swing over the top of the baseball regardless of where the pitch is located.
As John Walsh has noted, there is a relationship between vertical movement and batted ball outcomes—pitches with downward movement result in more groundballs than other pitches. I do not think we can pinpoint the cause of this relationship without considering pitch location, however.
What follows is a preliminary examination of the relationship of pitch location, pitch type and batted ball types using enhanced gameday data of 12 pitchers. Six of the pitchers are known for their above-average sinking fastballs, while the other six are known for above-average four-seam, or rising, fastballs. Note that rising fastballs do not actually rise but can appear to when the batter expects more of a downward break. With regard to sinking fastballs, I make no effort to differentiate a sinker and two-seam fastball in this study.
This study uses a convenience sample because enhanced gameday data is not in all major league parks. There were not enough data to include some very relevant pitchers, such as Chien Ming Wang and Brandon Webb, in this study. The pitchers included in this study also were selected because identifying the fastball of interest was relatively straightforward given the other pitches in their repertoire.
The sinking fastball group includes data from Roy Halladay, Fausto Carmona, Tim Hudson, Derek Lowe, Roy Oswalt and Andrew Miller.
The four-seam fastball group includes Kelvim Escobar, Daniel Cabrera, Matt Cain, A.J. Burnett, Daisuke Matsuzaka and Tim Lincecum.
For each pitcher, I sampled 40 fastballs that opposing hitters swung at (excluding bunts and foul balls). Each swing was categorized as a swing-and-miss strike, a groundball or an airborne ball (line drives, pop-ups, and fly balls). We can visually represent the distribution of outcomes for each pitcher by sorting the pitches into nine zones.
Here are the results for the four-seam fastballs:
This series of graphs is mapped to portray the pitcher's point of view. In other words, the graph in the lower right corner represents pitches close to right-handed hitters' knees. Each graph includes tick marks at 25% intervals.
Here are the results for the sinking fastballs:
Now, the data are a bit noisy because left- and right-handed batters were mixed together in this dataset and we are dealing with a relatively small sample in each group, but these images should make a few trends apparent. First, pitches over the heart of the strike zone lead to a lot of airborne balls for both types of pitchers. This should come as no surprise. Second, pitchers with good four-seam fastballs get more swings and misses than pitchers with sinking fastballs.
It's also clear that location does matter when pitchers are throwing four-seam fastballs. Throwing four-seam fastballs at or above the letters is a high-risk/high-reward activity that leads to plenty of swings and misses and also a lot of fly balls. Keeping the ball low will lead to more groundballs.
The story is a bit different for pitchers with sinking fastballs. Pitches result in groundballs at least 20% of the time in every one of the zones. If we aggregate the data bit more and use an imaginary horizontal line through the middle of the strike zone to divide all pitches into two groups, we get the following results:
Sinking Fastballs Four-Seam Fastballs Low High Low High Airborne 32.6% 30.6% 33.3% 29.5% Groundballs 31.2% 33.6% 26.2% 13.5% Strikes 36.3% 35.7% 40.4% 57.0%
I think this more clearly answers the question of why sinking fastballs lead to groundballs. The location of sinking fastballs does not appear to be related to the batted ball outcome. In fact, the high pitches were hit on the ground a bit more frequently than the low pitches from this group, though the difference is not statistically significant.
It's too early to make any bold conclusions. Nuanced interpretations will be possible when more data are available to support more rigorous methods. The preliminary results do not support location-dependent theory for why sinking fastballs result in groundballs, however. Instead, this evidence suggests that sinking fastballs lead to groundballs because hitters are swinging on top of pitches in all areas of the strike zone.
Chris Constancio analyzes prospects and the minor leagues at FirstInning.com. He welcomes comments, questions, and suggestions via e-mail.
<< Return to Article