clock menu more-arrow no yes mobile

Filed under:

How Base-States Affect The Batter-Pitcher Match-Up

Last week's look at ground ball rates in double play situations got me thinking about other ways the batter-pitcher match-up might be affected by how many runners are on base. For example, when do certain holes open up in the infield? Or, when is a strikeout or a walk most likely?

I've decided, then, to post a few tables here this morning focusing on how various stats change according to the various base-states. I'll offer a few casual observations along the way, but I believe this will serve best as a reference. Maybe it will help to settle a bet you have, or maybe it will encourage you to make a bet with some foolish non-reader of Beyond the Boxscore. Because I'm cool like that, I will even let you keep 100% of the proceeds.

I've broken up the tables into two halves, 'Pitchers' and 'Batters', but obviously the differences between them are meaningless. And since these tables are meant to visualize changes in the batter vs pitcher showdown only, I've also removed from the query all plate appearances where the pitcher is batting.

PITCHERS

BASE STATE PA FIP BABIP K% NIBB% HR% HR/FB GB% FB% Pit/PA
Bases Empty 912667 4.46 .296 17.4 7.7 2.9 10.6 44.1 28.9 3.85
"1--" 297717 4.41 .314 15.7 7.0 2.9 10.4 43 29.4 3.58
"-2-" 146324 4.65 .295 16.5 10.4 2.2 9.5 45.6 27.8 3.95
"12-" 117350 4.32 .294 17.2 7.9 2.7 10.3 43.9 29.0 3.70
"--3" 48197 4.81 .304 16.5 10.9 2.2 9.1 44.5 28.7 3.95
"1-3" 50889 4.47 .308 15.4 7.9 2.6 9.3 44 29.1 3.58
"-23" 37140 4.55 .300 16.1 9.6 2.0 8.8 45.2 28.3 3.96
"123" 42190 4.23 .299 17.0 7.0 2.7 9.5 43.2 29.3 3.66

What struck me first here is that BABIP is clearly at it's highest with a runner on 1st. This is most likely due to the compromised positioning of the first baseman, who is then obligated to hold the runner. Also, the second baseman has shifted toward the middle anticipating the double-play ball. This opens up a hole in the right side of the infield which, of course, allows more batted balls to eke through into the outfield. This is also evidenced by the amount of singles fielded by the right fielder in bases-empty situations, versus those with a runner on first base:

Singles / Ball in Play

Fielded by Bases Empty Runner on 1st
LF 36.4 39.9
CF 38.3 35.7
RF 34.4 42.9

That the hole in the defense contributes to more base hits makes sense intuitively, but I'll admit to being surprised that the effect has been so large (a difference of 18 BABIP points compared to bases-empty at-bats) since 2003. The idea that the baserunner may be "distracting" the pitcher, and therefore contributing to more hitter-friendly pitch-locations, is a tempting one, but that theorized effect has been studied a number of times and proven false or at least insignificant.

Some of the other spikes in the percentages here occur for fairly obvious reasons. For instance, a pitcher is much more likely to walk a batter with a runner on 2nd rather than 1st -- with the difference well over 3%. The walk in this instance enables the force-out at 3rd for the next batter and creates an opportunity for the double play at multiple points around the diamond.

That should come as no surprise to any fan of the game, clearly. But that this walk-rate of 7% with a runner on 1st is the lowest of all the base-states, in a statistical tie with Bases Loaded situations, may be somewhat less obvious. It seems that when the batter and pitcher are battling for a particular outcome, the batter tends to prevail more often. We saw this with the outcome of batted-ball types last week, and today with strikeout and walk rates.

Pitchers are most willing to walk a batter with a lone runner on 3rd, naturally, trying to set up a double-play situation and avoiding contact at all cost. However, strikeout rates tend to be low in these "make-contact" situations at just 16.5% with a runner on 3rd, and 16.1% with runners on 2nd and 3rd.

In general, a pitcher's FIP is at it's worst with a lone runner at 3rd, at 4.81 -- a difference of .35 FIP from the bases-empty state. The pitcher is particularly vulnerable in these cases, as he is obviously most willing to surrender a walk. The batter also has more motivation than ever to simply make contact with the baseball in order to get the runner in, which in turn suppresses the K%.

But, we can also see that while a pitcher's walk and strikeout rates suffer while he is 'living outside the zone' under these "--3" and "-23" conditions -- he is also less likely to give up the long ball in those scenarios. As a consequence, both HR% and HR/FB are at their lowest when there are likely to be fewer pitches in the zone. For batters, this means lower SLG and ISO, but surging OBP's. The overall result of this is a significant advantage for the hitter, but perhaps not as much as one may think:

BATTERS

BASE STATE wOBA BA OBP SLG ISO OPS IBB% BNT%
Bases Empty .327 .262 .325 .421 .159 .746 0 0.8
"1--" .336 .281 .334 .441 .160 .775 0 2.5
"-2-" .329 .255 .365 .400 .145 .766 3.8 2.1
"12-" .321 .259 .321 .414 .155 .735 0 2.6
"--3" .341 .282 .379 .437 .155 .817 3.7 0.6
"1-3" .338 .300 .345 .467 .166 .812 0.6 0.7
"-23" .333 .278 .419 .432 .154 .851 11.1 0.2
"123" .327 .288 .324 .458 .169 .782 0 0.1

Intentional walks occur in a remarkable 11% of all runners on 2nd and 3rd situations for obvious reasons. This enormous jump in IBB-rate, along with the 2% spike we saw in non-intentional walks and a modest boost in BABIP in these situations, means batters reach base most often under these conditions (with a .419 OBP). But, again, ISO and SLG tend to be near their lowest.

Of course we should keep in mind that base-states are only half of the story, as a few of these stats see tremendous swings depending on the out-state as well. I've put together a table, therefore, detailing the changes each of these stats experience across the 24 separate base-out states here. The table can be a bit overwhelming, so I encourage you to take your time when drawing conclusions, but I hope you have as much fun with it as I have.

*wOBA and FIP are calculated using their 'generic' non-season-specific coefficients. All data comes from retrosheet 2003-2011 and includes post-season PA.

@JDGentile says stuff on twitter, but it's what he doesn't say that will really make you think.