Why Would A Pitcher Pitch Against The Shift?

On April 10 in an outing against the Toronto Blue Jays, Dallas Keuchel was pitching Jose Bautista outside in the fourth inning.

That much isn’t really surprising, because since September of 2009, Bautista has done the bulk of his damage by pulling the ball. What was curious to me, however, was that Keuchel was pitching Bautista well outside despite the fact that the Astros defense was employing a heavy shift. It seemed counterintuitive, since I had assumed – though I hadn’t previously given it much thought – that a pitcher would be best served by pitching into a shift and giving the batter something he’s more likely to pull.

As it turns out, Keuchel was probably using the proper approach.

For some background, here is Bautista’s spray chart since September of 2009, courtesy of Bill Petti’s awesome spray chart tool.

Two things are clear here: avoiding Bautista pulling the ball is ideal, and he hits a great number of groundballs to the left side of the infield. Considered separately, pitching outside and shifting both make sense, though, again, my assumption was that a team would employ one or the other.

The Astros employed both.


In this particular example, the result didn’t really matter – Keuchel missed outside three times and then didn’t come anywhere close to the strike zone with a 3-0 pitch, walking Bautista on four straight. But the process here is the curiosity, not the result.

As mentioned, the Astros were probably smart to employ a shift and then pitch “out” of the shift by dealing away. Last year, Bautista took 27 outside pitches to the opposite field while pulling 64 outside pitches. More importantly, he hit groundballs to the right side on just six occasions compared to 49 pulled groundballs.

In other words, even with an outside pitch that should theoretically be easier to take the other way, Bautista is pretty unlikely to hit a groundball to the right side, meaning the cost of shifting isn’t very high at all. Sometimes pull-push percentages can confuse the fact that teams don’t care about pull rates on the whole when shifting, they only care about where groundballs are hit (since, obviously, flyballs and even linedrives are less likely to land where shifted players are or are not).

This got me curious as to whether this was Bautista-specific or a league-wide phenomenon. Well, thank godby for Daren W and Baseball Savant, as the PITCHf/x search there allowed me to dive into league-wide rates from the 2013 season.

To establish a baseline of how often pitches get pulled, shot up the gut or taken the other way with pitches in certain locations. That is, I first wanted to see if my initial logic that an outside pitch is easier to take the other way was true or simply something that made sense in my head but didn’t happen in practice. As it turns out, outside pitches are taken the other way far more often than pitches down the middle or pitches on the inner part of the plate, which a batter would have to inside-out to hit to the opposite field. This also presents itself in each hand-versus-hand iteration.

INSIDE	Pull	Pull %	Middle	Middle %	Oppo	Oppo %	Total
RHH v LHP	5013	59.66%	2113	25.15%	1276	15.19%	8402
RHH v RHP	9261	58.72%	4012	25.44%	2498	15.84%	15771
LHH v LHP	1666	56.02%	754	25.35%	554	18.63%	2974
LHH v RHP	5885	60.66%	2290	23.60%	1527	15.74%	9702
TOTAL	21825	59.23%	9169	24.88%	5855	15.89%	36849

OUTSIDE	Pull	Pull %	Middle	Middle %	Oppo	Oppo %	Total
RHH v LHP	4035	38.87%	3441	33.14%	2906	27.99%	10382
RHH v RHP	6396	39.53%	5307	32.80%	4479	27.68%	16182
LHH v LHP	1770	39.01%	1457	32.11%	1310	28.87%	4537
LHH v RHP	9218	39.44%	7435	31.81%	6719	28.75%	23372
TOTAL	21419	39.32%	17640	32.38%	15414	28.30%	54473

MIDDLE	Pull	Pull %	Middle	Middle %	Oppo	Oppo %	Total
RHH v LHP	3291	49.95%	1906	28.93%	1391	21.11%	6588
RHH v RHP	8973	57.54%	3800	24.37%	2821	18.09%	15594
LHH v LHP	1305	46.23%	885	31.35%	633	22.42%	2823
LHH v RHP	5599	50.35%	3171	28.52%	2350	21.13%	11120
TOTAL	19168	53.06%	9762	27.02%	7195	19.92%	36125

To tidy it up, here are the aggregate results by pitch location, regardless of hitter-pitcher match-up.

	Pull %	Middle %	Oppo %
INSIDE	59.23%	24.88%	15.89%
OUTSIDE	39.32%	32.38%	28.30%
MIDDLE	53.06%	28.70%	22.33%

What this tells us is just what I assumed earlier – that it would make sense to pitch a batter inside if you’re going to shift, given the overall pull rates league-wide. But again, it’s only really groundballs we’re concerned with when it comes to shifting the infield.

So here are the numbers again, focusing on just groundballs.

INSIDE	Pull %	GB PULL %	GB Oppo %
RHH v LHP	59.66%	31.12%	1.81%
RHH v RHP	58.72%	34.38%	2.57%
LHH v LHP	56.02%	35.27%	4.44%
LHH v RHP	60.66%	27.91%	2.70%
All Inside	59.23%	32.01%	2.58%

OUTSIDE	Pull %	GB PULL %	GB Oppo %
RHH v LHP	38.87%	28.35%	5.74%
RHH v RHP	39.53%	29.76%	6.01%
LHH v LHP	39.01%	28.01%	6.81%
LHH v RHP	39.44%	27.64%	6.70%
All Outside	39.32%	28.45%	6.45%

MIDDLE	Pull %	GB PULL %	GB Oppo %
RHH v LHP	49.95%	28.22%	2.57%
RHH v RHP	57.54%	23.87%	2.63%
LHH v LHP	46.23%	27.67%	4.64%
LHH v RHP	50.35%	26.76%	2.65%
All Middle	53.06%	26.95%	3.06%

Once again, those are mostly to show that the trends here hold for all hitter-pitcher possibilities. To put it in a more digestible form, here are the aggregates for each pitch location, comparing groundball pull and opposite rates (as a percentage of all balls in play, plus home runs) compared to the overall rates.

Location	Pull %	GB PULL %	Oppo %	GB Oppo %
INSIDE	59.23%	32.01%	15.89%	2.58%
OUTSIDE	39.32%	28.45%	16.10%	6.45%
MIDDLE	53.06%	26.95%	16.42%	3.06%

The important cell is the middle row, in the right-most column. Even given outside pitches, batters only hit groundballs to the opposite side 6.45 percent of the time last season.

It seems that on average, and not just with Bautista, pitching outside with a pull-shift on isn’t all that risky. That won’t hold for all players, of course, and 6.45 percent isn’t zero, but if you see a pitcher who appears to be pitching “out” of the shift, that’s why.