Skip to content

Rich Harden’s HR%: What Is To Blame?

October 22, 2011

A few days ago I looked at Rich Harden‘s 2011 campaign and came to the conclusion that it in fact wasn’t as good as many thought it was and that simply in this instance some of the advanced stats failed to tell the full story, and actually instead told a contradictory one. I wrote,

“I said hey Harden got penalized by poor luck, but when looking at the data it seems he has lost something with his changeup, it has indeed become more hittable and for Harden that means he is surrendering more home runs. With that it seems his FIP is an accurate gauge of him, that he is merely average and not as good as his xFIP or SIERA make him out to be.”

The reason I came to the conclusion that the changeup was the problem was because ultimately his HR rate on changeups had jumped while on his fastball it had remained fairly static. But reader Ken Arneson – who is great to have comment because he brings a good perspective to debates – argued that I saw it the wrong way writing,

“I’d disagree that he’s lost something on the changeup. I’ve been arguing on Twitter that he’s lost something on his fastball, which makes the speed separation between the two pitches too small. He’s thrown changeups as hard as 86mph, and fastballs as slow as 88mph. That makes the changeup, if he leaves it up in the zone, little more than a batting practice fastball. He can get away with an 85mph changeup in the middle of the zone if his fastball is 95mph. He can’t if it’s 90mph.”

That makes sense because a changeup only “changes up” a hitter when the speed it goes at varies significantly from the fastball. As Ken notes, if the speed separation doesn’t vary much, it makes a changeup incredibly hittable. Now Harden has had a diminishing speed on all his pitches over the past few years as you can see in this chart.

Year Fastball Speed Change Speed Change Diff
2006 93.5 84.8 -8.7
2007 93.8 85.8 -8.0
2008 92.0 84.8 -7.2
2009 92.1 84.0 -8.1
2010 90.5 82.2 -8.3
2011 91.7 83.8 -7.9

As one can see while the pitch speed has dropped, it has dropped across the board resulting in a similar difference still of about eight miles per hour. Ken however noted that over the course of the season that provides a lot of white noise and it got me thinking about looking at each of the 17 home runs Harden surrendered in 2011 and comparing them based upon the speed variation. I figured that I could look at the difference between games where he allowed a home run and games where he didn’t. The logic being that in games where he failed to surrender a long ball, the difference between fastball and changeup speed would be greater, but then I ran into a problem. Harden had only two outings all year in which he failed to give up a home run – his eighth and ninth starts of the year.

Game Fastball Speed Change Speed Change Diff
8/14/2011 91.3 84.0 -7.3
8/19/2011 91.9 84.2 -7.7

These games couldn’t have been more different for Harden. The first, a four inning affair where he got knocked out care of a terribly high pitch count (98 pitches in those four innings) against Texas and the second, another game at the Coliseum when he blanked Toronto striking out 11 in seven innings – yet in both cases the difference in speed was below his average difference for the year. Given the tiny sample size there, it made no sense to compare games with home runs to games without, so instead I decided to look at the very hitters who hit the home run and all the pitches they saw prior to and including the home run pitch, which gave us the chart below:

Date HR Hitter Fastballs Changes CH Diff


Wily Mo Pena





Michael Young





Mark Trumbo





Vernon Wells





Nick Swisher





Desmond Jennings





Casper Wells





Edwin Encarnacion





Russell Martin



Robinson Cano





Jack Hannahan



Jack Hannahan





Billy Butler





Mark Trumbo





Adrian Beltre





Bobby Abreu



Vernon Wells








It turns out that the average home run hitter saw a 7.2 mile per hour differential between fastball and changeup – which is less of a differential that Harden’s season average of -7.9. Interestingly it was a combination of factors reducing the difference, a) his fastball clocked in slightly slower versus hitters that he’d surrender a dinger to at only 91.5 versus a 91.7 season average, but b) his changeup clocked in faster versus the HR hitters at an average 84.3 compared to 83.8 on the season as a whole. These numbers don’t seem particularly stark but ultimately could that be the difference? Interestingly it is the change up that seems to be problematic however as it is clocking in faster for those swatting home runs than the fastball is slower.

Ultimately, I cannot deduce if it is the fastball or the changeup to blame. I open up the comments to anyone who wants to run with my data or who has any conclusions as to whether or not this is even conclusive. Should the A’s choose to re-sign Harden, the A’s new pitching coach (signed today) may be able to solve this one for us. Curt Young is back in the green and gold following a one year assignment as the Boston Red Sox pitching coach. May his September in 2012 be a brighter one than his September in 2011 for his young arms.

6 Comments leave one →
  1. Ken Arneson permalink
    October 22, 2011 2:47 am

    Wow. Atsalotta work.

    So Harden’s basically lost 2mph off his fastball and 1mph off his changeup since 2006.

    I guess from this if I wanted to argue that his main problem is the loss of velocity, I could argue that, and if I wanted to argue that the main problem is a loss in difference in velocity, I could argue that, too. These numbers don’t definitively tell us one or the other.

    I guess from this I’ll mentally blame 2/3 of it on loss of velocity an 1/3 on loss of difference of velocity, and leave it at that.

    • October 22, 2011 2:59 am

      In reality it didn’t take much time.

      I was hoping to find some sort of big difference between his on days and off days and in many ways it is funny and illuminating that of the two starts in which he didn’t surrender a home run one was an on day whereas one was very much so an off day. I think your point is a salient one, regarding the differential however I am just not sure how it seems to manifest itself. I suppose one could argue HRs like the Swisher one were a result of a subpar fastball coupled with a too-fast change, whereas the HR Vernon Wells hit to close out the season was simply Wells hitting well (for a change). I suppose the big lesson is that what I was hoping to see is perhaps impossible to tease out.

  2. Michael F permalink
    October 25, 2011 11:38 pm

    Sometimes the numbers don’t tell you anything. This is one of them. The sample size is too small. Clearly he is losing something off of something. Most of the time what they lose is location and that would be impossible to check.

    • October 26, 2011 12:35 am

      Yeah this is a tough one to peg. I threw it info out there to see if anything stood out but nothing did. I am with you that it doesn’t seem like we learn much from these stats.

  3. Mike permalink
    November 1, 2011 1:55 pm

    Maybe you should do some analysis on the std deviation of the speeds instead of just the average. If he is throwing fastballs or curveballls with a greater variance of speeds, there is a greater likelihood that a hitter sees only a 2-3mph difference between two intended pitches.

    • November 1, 2011 9:53 pm

      Mike –

      Thanks for reading. I think you are right. That will take some more time to figure out but I am curious as to the results.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s

%d bloggers like this: