Unpredictable, rare, and occasionally effective…but always entertaining.

Archive for February 2011

Infield Fly Balls and the Coliseum

with one comment

In his most recent entry, the acclaimed sabermatrician Paapfly does a lot of research regarding what makes Matt Cain good.  The whole entry is worth a read, but one part caught my eye:

Season in and season out, Cain gives up a lesser percentage of home runs on fly balls than the average pitcher, both at home and away. Why? I can’t say with confidence, but Dave Pinto at Baseball Musings has come up with a theory that’s as good as any I’ve read: he surmised that Cain’s fastball is dropping less than the hitter expects, resulting in the ball being struck somewhere below the center of the ball, and either going straight into the air (as his 12.9% career infield fly ball percentage (IFFB%) might suggest) or at least somewhere inside the park*.

*Quick aside that I had not considered at any point until NOW: wouldn’t Oakland be ideal for him, given the enormous foul ground territory? *Checks*  In a small sample, granted, he’s pitched to a 1.16 ERA in 3 starts, his best ERA at any park where he’s had two or more starts. This could be coincidence, but I’m intrigued.

Do more infield pop ups matter more for pitchers in Oakland?  I too was intrigued, so I tried to look it up.  Not for Matt Cain.  But for every Oakland pitcher to pitch during the period for which we have the data.

Batted ball data has been recorded for nine seasons (2002-2010), creating a sample of 99 pitchers and 190 pitcher-seasons.  For all these pitchers, I created a database with about 40 columns, with everything from wins and losses to balls and strikes thrown to FIP and tERA.  That might seem like a lot of data, but after it’s parsed, there’s very little left.

Another item to note is that the batted balls are recorded only when they are hits or outs.  Many foul balls that are caught in Oakland go unnoticed in other ballparks and result in something else (or another foul ball).  There should be more fly balls recorded in Oakland because of these additional outs.

Of the 99 pitchers, only 38 of them threw at least 100 innings for the Athletics (I only looked at time with the Athletics for this study), allowing for roughly 50 innings in the Coliseum and 50 on the road.  If I further parse  the data, only 15 pitchers threw both 100 innings at home and on the road while with the A’s in the last nine seasons.

At first, I checked to see if there was any correlation between infield fly ball percentage (IFFB%) and ERA, RA, BABIP, FIP, xFIP, and ERA-FIP using the home and road stats.  More infield pop ups would lead to more outs thanks to the additional foul territory.

I used all three sets of pitchers (all 99, the 38, and the 15); if the theory has any weight, we would expect to see a negative correlation with the first three statistics (they go down as IFFB% goes up) for the home data, and there should be no correlation for the road data.   FIP and xFIP are theoretically supposed to normalize batted ball noise, and I’m really not sure what to expect from those, though I suppose there should be no difference.  If ERAs are lower though, then ERA-FIP should also be lower, and that would have a positive correlation.

Here are a few sets of results:


ALL 99 -0.078 -0.041 -0.127 -0.031 -0.006 -0.064
100+ IP 38 -0.049 -0.094 -0.175 -0.123 -0.196 0.066
200+ IP 15 0.110 0.203 -0.100 -0.035 0.032 0.174


ALL 99 -0.166 -0.143 0.010 0.201 0.056 -0.244
100+ IP 38 -0.031 -0.044 -0.305 0.026 0.162 -0.071
200+ IP 15 -0.279 -0.279 -0.134 -0.500 -0.088 0.094


I could analyze them, but I realize I made a mistake.  IFFB% is the percentage of fly balls that are infield fly balls.  What I really needed was the percentage of batted balls that are infield fly balls.  To get that I multiplied IFFB% by the fly ball percentage (FB%), to get what I called TOTAL_IFFB%.  Here are those results:


ALL 99 -0.127 -0.081 -0.166 -0.021 0.010 -0.130
100+ IP 38 -0.019 -0.086 -0.210 -0.062 -0.051 0.041
200+ IP 15 0.027 0.120 -0.280 -0.035 0.085 0.077


ALL 99 -0.164 -0.132 0.017 0.152 0.104 -0.223
100+ IP 38 -0.096 -0.131 -0.434 0.014 0.281 -0.153
200+ IP 15 -0.257 -0.272 -0.210 -0.423 0.058 0.048


I also did R-squared tests with the same data:


ALL 99 0.016 0.007 0.027 0.000 0.000 0.017
100+ IP 38 0.000 0.007 0.044 0.004 0.003 0.002
200+ IP 15 0.001 0.014 0.078 0.001 0.007 0.006


ALL 99 0.027 0.017 0.000 0.023 0.011 0.050
100+ IP 38 0.009 0.017 0.188 0.000 0.079 0.023
200+ IP 15 0.066 0.074 0.044 0.179 0.003 0.002


I would call it “mixed results,” and there aren’t any real patterns, at least not that I can see.  More importantly, this still isn’t what I really want.  What I want to know if having a higher TOTAL_IFFB% means something in the Coliseum.  For this I divided the pitchers into three groups: high TOTAL_IFFB% (above 6%), medium TOTAL_IFFB% (between 3 and 6%), and low TOTAL_IFFB% (below 3%).  For reference, Cain’s career TOTAL_IFFB% is 5.9%.  Here are some results:


>6% 28 3.87 4.25 6.91 3.32 1.04 44.4% 0.273
3-6% 35 3.56 3.90 6.90 3.04 0.83 37.2% 0.277
<3% 36 3.56 3.90 5.54 2.99 0.72 28.9% 0.281
ALL 99 3.62 3.97 6.48 3.08 0.84 36.1% 0.277


As a control, here is the same test for the groups on the road (grouped again by Home TOTAL_IFFB%):


>6% 28 4.17 4.44 7.19 3.79 1.12 42.9% 0.281
3-6% 35 4.16 4.53 6.89 3.35 0.96 37.0% 0.295
<3% 36 4.66 5.15 5.56 3.22 1.02 28.8% 0.305
ALL 99 4.31 4.69 6.57 3.41 1.01 35.9% 0.295


I included K/9, BB/9, and HR/9 for each group to determine if maybe one set of pitchers was simply inferior to the others.  All of these statistics are weighted by the innings pitched by each pitcher, so there’s a lot more Barry Zito in the >6% than Seth Etherton.  There are approximately as many innings in the 3-6% group as in the >6% and <3% groups combined.

Obviously there is less scoring in the Coliseum because it is harder to hit home runs, but we also see a decrease in walks (perhaps because pitchers are less worried about mistakes).

Quick aside and important point: holy crap does the Coliseum depress BABIP.  And it has to be the Coliseum, because the defenses, pitchers, and hitters should be relatively the same in both sets.  I checked and checked and checked this, and I’m pretty sure it’s right.  I’m not sure where I would have messed that up.  Plus everything else makes sense.

The pitchers combined to have a 4.03% TOTAL_IFFB% at home and a 3.95% TOTAL_IFFB% on the road.  The 0.08% difference is for roughly 20,100 batted balls at home and 19,500 batted balls on the road.  The tenth of a percentile difference, when applied to the 20,000 batted balls, is roughly 20.  2 per season.  Seems so wrong.  Then I found this:

One of the more interesting effects I found is that parks have a strong affect [sic] on the proportion of batted balls that are infield flies.

Team            FlysIF
Brewers         1.15
Mariners        1.12
Marlins         1.12
Reds            1.08
Devil Rays      1.07
Phillies        0.93
Royals          0.93
Indians         0.92
Diamondbacks    0.92
Giants          0.90

A player is 28 percent more likely to hit an infield fly in Milwaukee than he is in San Francisco. Why is that? My guess is that it has to do with foul territory. Since infield flies are only recorded when the ball is put into play, parks with a lot of foul territory are more likely to see foul pop-ups stay in and get caught, whereas ballparks with little foul territory will see a lot of pop-ups land in the stands and go unrecorded.

The problem with that theory is that while the parks at the bottom of the list do tend to be a little smaller in terms of foul territory, those at the top seem to be pretty average on the whole. Perhaps my data source is off, but there may be some other variable I’m not thinking of.

The A’s aren’t on either side of the list.  While a couple of years old, it hasn’t changed.  The park factor for infield flies in the Coliseum is close to neutral.  I have to agree with David’s last sentence: “Perhaps my data source is off, but there may be some other variable I’m not thinking of.”

There’s nothing there to suggest that getting more infield fly balls makes you a better pitcher in the Coliseum.  Matt Cain’s ability to limit home runs is what would continue to make him a very good pitcher in Oakland.  However, there’s so much bias in the manner that I carved up the data that I can’t be sure it’s correct.  Would love some reviewers to take a look…


ALL 99 0.010 -0.244 0.201 0.056 -0.166 -0.143
100+ IP 38 -0.305 -0.071 0.026 0.162 -0.031 -0.044
200+ IP 15 -0.134 0.094 -0.500 -0.088 -0.279 -0.279

Written by Dan Hennessey

February 2, 2011 at 9:54 PM

Posted in Uncategorized