The 2004 Phillies tried to hit more fly balls

Back in August, I wondered if the homer-happy tendencies of the new Yankee Stadium had an effect on how the Yankee hitters put balls in play. The theory was simple: even if stadiums do not actually play to their commonplace perception (i.e. maybe Yankee Stadium was actually a pitcher’s park), players will respond based on what they think is true. Looking at the results, Tom Tango commented:

That’s an average of around +4% more FB per BIP at a team level, which must be enormous.

I would guess, total guess, that 99% of all players with at least say 400 BIP would be +/-5% year-to-year. I’ll further guess that all teams would be +/-2%. So, when I see a team level change of 4%, that definitely seems to imply something.

If it’s a question of players responding to the perceived expectation of the park, it would be interesting to see if this change happened at a particular moment. Remember, FB/BIP is a very telling stat that quickly shows real changes in talent.

Based on that response, I decided to look at another park thought of as being particularly hitter-friendly: Citizens Bank. CBP opened up in 2004 and quickly got the reputation of being a nightmare for pitchers and a haven for hitters. But did the players actually respond to that possible incentive? I looked at all of the Phillies hitters with at least one-hundred plate appearances from 2004, with the only exception being Ricky Ledee, who split time with the Giants, making his results tainted. Here’s what I found:

Player          2003 FB%  2004 FB%  Difference
Jimmy Rollins      34.4      35.8       1.4
Bobby Abreu        26.7      36.8      10.1
David Bell         37.4      36.6      -0.8
Jim Thome          38.8        45       6.2
Placido Polan      28.6      29.1       0.5
Mike Lieberth      36.7      45.5       8.8
Pat Burrell        43.8      45.4       1.6
Marlon Byrd        25.4      27.5       2.1
Jason Michael      34.5      39.5         5
Chase Utley        36.3      34.4      -1.9
Tomas Perez        32.6      48.4      15.8
Doug Glanvill      36.9      37.4       0.5
Todd Pratt         33.7      34.1       0.4

Yowzers. Add up all those percentages on the far right and you get a net increase of 49.7. Of the thirteen players with the most plate appearances for the 2004 Phillies, eleven of them increased their amount of fly balls per ball in play, with the only two exceptions being Bell and Utley.

But if we look at the guys who would want to get the most out of their fly balls, then we should see how the sluggers changed their game. In 2004, Thome hit forty-two homers, Abreu hit thirty, and Burrell hit twenty-four. Each saw an increase in their fly ball rates, capped off by Abreu’s dramatic 10.1 increase from the year before. It also would’ve seemed impossible for Burrell to hit any more fly balls than he had the past year, but he was able to muster a modest 1.8 increase. Mike Lieberthal also went from a pedestrian 36.7 FB% in 2003 to being in the top twenty-five highest in baseball in 2004.

We can’t exactly get into the minds of the Phillies hitters, but I think this comes pretty darn close, and that says a lot.


Pat Andriola is an Analyst at Bloomberg Sports who formerly worked in Major League Baseball's Labor Relations Department. You can contact him at Patrick.Andriola@tufts.edu or follow him on Twitter @tuftspat
9 Comments
Oldest
Newest Most Voted
Inline Feedbacks
View all comments
Dan Novick
14 years ago

Pat,

How about looking at pitchers through the same lens? Sure, they have less control over this kind of thing because pitchers are limited to their current repertoire, but maybe they had a change in approach as well.

Craig Glaser
14 years ago

I wonder if this might start some kind of feedback loop.  Guys hit more fly balls because the park is supposed to be HR friendly so more HRs are hit there.  Then the park is seen as even more HR friendly and they try to hit even more fly balls.

Not sure how to look at it but it would be fun.

Daniel Adler
14 years ago

Pat,

Very interesting study.  11/13 seems like a pretty good hint that something may be happening.  However, a big chunk of the overall effect is coming from Tomas Perez, who had less than 200 plate appearances.

Are flyball rates steady through a player’s career or do they vary widely?  Do they increase as players age/develop power?  If that’s the case, we would expect the returning players to naturally increase their flyball rate.  I’m no expert on the subject, so I’ll defer to you/Tango.

Nice work.

-Daniel

Nick Steiner
14 years ago

Pat, can you do a weighted average of the delta for all players?  Or just send me over your data?

Jon Wilt
14 years ago

What are the relative heights of the press boxes of the Vet and CBP?  There’s been some data showing that there’s a bias in how batted balls get coded based on press box positioning.  It could be that the batters didn’t change a thing, but the observers changed how they did their work.

John Walsh
14 years ago

Nice study.

However, this:

Add up all those percentages on the far right and you get a net increase of 49.7.

is a very misleading statement.  What you should show is the overall increase as a team, which you decidedly do not calculate as the sum of the individual player increases. 

BTW, given 100 BIP, 1 standard deviation on FB% is about 5%.  And the error on the delta, assuming both BIP totals are 100, would be 7%.  Even for the full time players in your list, those with 400 BIP, the 1 SD error on the delta will be around 3-4%.

So it’s not clear how many of those delta’s you should really have statistical significance.  I grant that since most go in the same direction, there probably is a real effect, but you haven’t shown it statistically.

Nick Steiner
14 years ago

Using Pat’s data I get a weighted by plate appearances delta of 3.9% in 5802 plate appearances.  You figure that translates to ~4,500 BIP, what’s the SD on that?  Or should you calculate it individually by batter?

John Walsh
14 years ago

Nick,

The easiest and most straightforward way to calculate this is to simply add up team BIP and FB in each of the two years and compare them.  I wouldn’t even fool around with individual values, although obviously those are interesting to look at.

I calculate the SD like this: SD = sqrt( fb*(1-fb)/Nbip ), i.e. a straight binominal error.  (fb is fb pct and Nbip is number of BIP.  4500 is the number of BIP for one team in a season, so the SD for team fb% for a given year is roughly sqrt(.4*.6/4500) = 0.7%.  The SD on the difference in fb% in two different years would be sqrt(2) times that or 1%.

So, if the difference between 2003 and 2004 really is 4%, then that’s a significant difference.

schmenkman
14 years ago

CBP’s HomeRun Park factors for the past 2 years have been only +2.2% and +0.5%, ranking it 11th and 16th. 

It would be interesting to know whether fly ball rates have been lower for some reason—have the Phils gotten more GB pitchers? do opposing teams adjust their rotation when visiting CBP to avoid using their FB pitchers, etc.