How many home runs would Pujols hit if he was pitched to like Juan Pierre?

Juan Pierre has 13 home runs in over 6000 plate appearances in his career so far. That’s not very good. In fact, Albert Pujols has hit more home runs this year in the month of June alone. Based off of that large discrepancy in power between the two, we can be pretty confident that pitchers are going to be much more aggressive when facing Pierre than when facing Pujols.

But what if they weren’t? What if the two magically switched bodies, and Pujols suddenly saw life through the eyes of Pierre? For one, you would probably start to see some pitchers getting demoted for allowing 450 foot bombs to “Juan Pierre”, along with some steroid allegations which would subsequently result in a “accountability” sermon from Ken Rosenthal. Secondly, before the pitchers caught on to the guise, you would most likely see a large increase in the amount of home runs hit by Pujols. How much? Well let’s find out.

Using the wonderful Pitch f/x technology, which, believe or not is, probably not ruining baseball, we can see the location of each pitch thrown to Pujols this year. So, I split up the strike zone into 9 different zones, and figured out the percentage of (total) pitches thrown in each zone:

image

As you can see, he doesn’t see a lot of pitches in the strike zone, with a Zone% of just 45.1% by my calculations (FanGraphs has him a 44.9%). League average is a little over 49%, so he really has been getting strike starved this year. You’ll also notice that he’s seen very few pitches on the inner third of the strike zone, with just over 10%. That seems rather smart for the pitcher when you consider Pujols’ prolific bat speed.

The next step is to take a look at the percentage of pitches in each zone that Pujols has hit a home run:

image

Predictably, he’s most efficient on pitches right down the middle, hitting over 8% of them out of the park. Other than that, he likes the high outside corner, the middle inside part of the plate.

You can reverse engineer Pujols’ home run total this year using those two graphs. Multiply the percentage of pitches hit for home runs in each zone by the percentage of pitches seen in each zone by his total pitches seen that year. That’s how you’ll transpose his HR/Zone on Juan Pierre’s pitch distribution.

First, however, we have to look at the distribution of pitches to Juan Pierre. I’ll use the same method as with Pujols obviously, and I reversed the two sides to adjust for the fact that Pierre is a lefty:

image

As you can see, Pierre gets pitched to a lot more, as over 55% of pitches seen are in the strike zone. He also sees a lot more pitches down the middle, which would be nice for Albert.

Anyway, doing the process I described above, we can estimate that Pujols would have hit roughly 62 Home Runs so far this year if he saw the same distribution of pitches that Pierre did. Ridiculous.


Print This Post
Sort by:   newest | oldest | most voted
The Real Neal
Guest
The Real Neal

Follow up question, of course is “What would Juan Pierre’s OBP be if he was pitched to like Pujols?”

RZ
Guest
RZ

very cool.

Nick Steiner
Guest
Nick Steiner

Thanks a lot, RZ and studes. 

Neal, that would probably take a lot more work.  Home runs are a singular and generally isolated events.  I’m already taking somewhat of a reach by not including variables like count.  OBP made up of many difference variables (strikes, balls, fouls, contact ability, O-Swing, etc.) that are much harder to control for. It’d be much harder to OBP than home runs, although it’s certainly doable.

Maybe I’ll take a look at that next time.

Quasibozo
Guest
Quasibozo
A Boston based radio sports show once had a discussion something to the effect of rookies who start off blazing and then taper off. They used Jacoby Ellsbury as an example, who’s numbers dramatically decended for a while. They noted why he struck out so often his first year. He rarely took a base on balls then. Opposing teams were aware of his speed threat when on base and pitched to him to avoid hits. When he was asked why he never took pitches his answer was “..because they were strikes..” The point was made, until a batter establishes a… Read more »
RedRobot8
Guest
RedRobot8

Nick, does this mean you’ve gotten your Pitch F/X database up and running?

Dan Novick
Guest
Dan Novick

As ridiculous as 62 home runs sounds, it’s actually too low. I don’t know what it should be, but Pierre has seen 69.7% fastballs this season, while Pujols has seen just 52.8%. That’s a difference of over 400 fastballs extra that Pujols would see over a full season.

Jeremy Greenhouse
Guest
Jeremy Greenhouse

Nick, good idea for a post.

I tried applying the local regression technique I’ve gotten familiar with. Given the same spread of pitch location, not adjusting for handedness or height, I got a difference of .6 homers per 100 pitches, and given Pujols 2500 pitches this year, that comes out to a difference of 15 homers, which added to the 47 he already has, gives me exactly 62 as well.

Nick Steiner
Guest
Nick Steiner

Quasibozo – I think it’s a little late for that wink

Dan – I didn’t adjust for pitch type yet, obviously that would make a difference. 

RedRobot – No, I used Jeff Zimmermans’ SQL dump one – it doesn’t have Pitch type, which is partially the reason that I didn’t adjust for it.  Have you got your’s up?

Jeremy – Thanks for testing it out, it’s pretty cool that we got similar results despite a different process.

Dave Studeman
Guest
Dave Studeman

Very neat idea, Nick.  Thanks!

harmon
Guest
harmon
What you are saying is that if pitchers didn’t try to pitch to a particular hitter’s weaknesses, but to some other hitter’s weaknesses, the first particular hitter would hit more home runs. Well, yeah. Stands to reason. Probably even true. So what’s the point? No new information that I see. What would be interesting is to know if any other hitter sees THE SAME pitch distribution that Pujols sees. And also, to see where the pitchers are placing the balls outside the strike zone, and what Pujols is doing with those. Looks like if the pitch goes up, the ball… Read more »
RedRobot8
Guest
RedRobot8

I got the 2007 data parsed, but stopped there for now.  I am reworking the parser script a bit to make sure I get all the information I want in a format that is easy to play with before I bother adding 2008 and 2009.  Primarily I am adding G/L/F classifications as well as making it easier to access the Event data.

edward brooks
Guest
edward brooks

, lets leave the what if questions for the middle school kids. Can’t we make better use of our time and technology.  Ok, here is a question how many home runs would Juan pierre have if he was as big and strong and great as Albert Puhols, or how many stolen bases would Juan Pierre have if when people looked at Juan pierre they thought he was Don Zimmer but it really was Juan Pierre.

Nick Steiner
Guest
Nick Steiner

RR – do you think you can let me “borrow” your parser once your finished wink.  I’d really be interested in parsing batted ball classifications, and I wouldn’t be able to figure out how to do so by myself.

RedRobot8
Guest
RedRobot8

Sure thing.  I’ll let you know when it is bug-free.

Devon
Guest
Devon
This was a cool idea! You just showed that about 15 HR’s have been lost to history based on pitch strategy. We don’t have Pitch/fx data for 1998-2001 do we? Has me thinkin’ now (but I’m sure someone’s already done this & I just haven’t stumbled onto it yet)… across the leagues, what % of HR’s are hit from each part of the K zone? And does it differ significantly between the 2 leagues? This also makes me wonder… can this kind of thing be used to determine possible steroids users? I mean, if a player is suddenly hitting more… Read more »
awayish
Guest
awayish

as great as pujols is, he’s no barry bonds. would have been fascinating to see similar pitch/fx analysis with bonds.

wpDiscuz