FB% And ISO
One of the things we like to do in the statistical community is test common wisdom to see if the empirical evidence lines up with what is believed to be true. One of these truisms is that pitchers are not afraid to high contact slap hitters, and thus will feed them a steady diet of fastballs, knowing the worst case scenario is probably a groundball, even if they turn on it.
Since I’ve been in a graphing mood and we have pitch type stats for hitters on the site, I decided to take a look at this theory, plotting the data from the 121 batters who have racked up enough plate appearances over the last three years to qualify. On the x asis, I put percentages of fastballs seen, and on the y axis, isolated slugging percentage.
Here’s the chart.
As you can see, the data definitively supports the truism. The correlation between FB% and ISO is -.59, suggesting a strong inverse relationship – the higher your ISO, the lower your FB% will be, and vice versa. The average hitter sees fastballs 60% of the time, but that’s the upper bound for low ISO hitters – Melky Cabrera’s .108 ISO and 60.8% FB% are about as far left as you can go on the graph and still be talking about a guy without much power.
However, it’s interesting in that the limit isn’t symmetrical. Notice how there are some high ISO guys on the right hand side of the graph. Matt Holliday, especially, stands out – he’s got a .248 ISO and has still been thrown fastballs 64.8% of the time over the last three years. Coors Field is likely a factor there, but it isn’t with Carlos Lee – .237 ISO, 62.9% fastballs.
Two other interesting players are Johnny Damon and Garret Anderson. They have identical .161 ISOs over the last three years, but Anderson has seen the fewest fastballs of any hitter in the sample (48.9%) while Damon is up near the top (67.5%). Do pitchers perceived Damon as a slap hitter, due to his frame? Or perhaps Anderson just really struggles against breaking balls, and pitchers are exploiting this? Maybe both?
We don’t have all the answers. I’m sure there are scouting reports at play here, indicating some hitters are more vulnerable to bendy pitches than others, but the trend is still clear – pitchers really will challenge no power hitters with fastballs while sticking to their off-speed stuff against the guys who can launch a baseball 500 feet.
Dave is the Managing Editor of FanGraphs.
Fastball numbers for Garret Anderson aren’t surprising; he’s a “dead-red” guy, always looking for fastball. He still gets around very well on any fastball.
Pitchers tend to nibble out the zone with off-speed stuff because, well, he tends to fish for it.
Very interesting, Dave. It looks to me like the relationship changes a bit at very low power levels, with a ceiling of about 70% FBs. It would be interesting to look at pitchers-as-hitters — theoretically, they should get 90% FBs. But I’m sure they don’t. For really bad hitters, the opposite logic can apply: don’t throw a FB unless you get behind in the count, because a FB is the only pitch these guys can hit.
FYI, I think this graph would be more intuitive with ISO on the horizontal axis, so as ISO grows the line slopes down showing declining FB%. But that could just be me….
It seems to me that guys with declining skills should see more fastballs, as their bat speed decreases off-speed pitches might be easier to handle than a good fastball.
Am I the only one who idiotically saw “FB%” and, despite having read the intro, was momentarily confused by the apparent inverse relationship between Isolated Power and Fly Ball Percentage?
Yeah? Only one? Okay.
lol… I totally spent the first three minutes thinking the same thing.
Hahaha it took me so long to realize it was fastball%. I was wondering why in the hell all the power guys had such low fly ball percentage
Who is in the top left hand corner of the graph? The high iso/low FB guy
My money’s on Ryan Howard. He swings at all sorts of trash that’s way low and away, so they keep feeding it to him.
Howard: .318, 51.9%
Good job on Howard. Whos the guy close to Howard up in the top left? Looks like about 54FB% and 290 iso or so.
Ortiz: .298, 53.6%
Drew, my hunch is that it’s Ryan Howard. Over the past two seasons, pitchers have really given him a steady diet of breaking pitches.
I owe Don a Coke.
Matt Holiday is an interesting point on the graph. I wonder if him seeing more fastballs has anything to do with the thin air in Colorado. It makes breaking pitches less effective than in normal air (it also makes fastballs move less, but a curveball that doesn’t break is easier to hit than a fastball that doesn’t move).
I actually built a plot like this for the BtB contest, but I didn’t end up using it. I used the last 3 years of data, but I set the minimum PAs to 400 I think and ended up with 393 players.
With a linear fit, I got a slope of -0.683 with an R squared of 0.321 (0.567 R value). If I switch it to a power fit the R squared improves to 0.371 (0.609 R value) and the equation is ISO = 3.235*e^(-5.09*(FB%)) with the FB% entered in as a decimal. I wonder if a power fit works better for your data set as well.
Is pitcher ISO allowed a stat that you guys have?
Dave, awesome little study. Thanks again!
I wonder how we can determine about how effective the general strategy is? Probably use WPA.
I’m guessing the dot about half way between 60 and 65% FB and at about .090 ISO is Ichiro!.
Ichiro: .083, 62.6%
You’re on a roll, Ryan.
Would it be at all possible to have the excel versions of these graphs/tables for download, so that all of the points can be seen? I suppose I could also just do the work myself…haha.
If you go to the leaderboards, there’s an Export To Excel button that will put the data in excel for you. From there, just insert a scatter plot and tell it where to find the data. Voila.
Awesome, thanks!
How do you plot the line?
I hit the trendline button in excel.
Haha thanks.
Excellent use of graphic, David. Thanks!
This just in: The Sun will rise in the East.
I think these numbers are a bit skewed because a team will play a guy who can’t hit a breaking ball as long as the balls he does hit go very far (Geoff Jenkins, Russell Branyan, etc.). But, they generally wont play a guy who can’t hit a breaking ball that is a singles hitter unless he can pitch. And, no one plays guys who can’t hit fastballs much! In other words, power hitters are less likely to be able to hit a breaking ball on average and therefore should see more of them regardless of where in the lineup they hit!
Also I have read that Matt Holliday is known for killing breaking balls, so people only throw him enough breaking balls to keep him honest. So he is probably not entirely a Coors effect. Although the fact that breaking balls break less there should affect the entire team. If you took that factor out of your data, then you would likely find Holiday still gets more fastballs.