AN Mailbag #4: Verducci Effect on A’s pitchers? (Hint: no)

Do you think the amount of innings that our young hurlers (Parker, Milone, Cook, Doolittle) threw last year will negatively effect them this coming season?

-rhymeswithelephant

Hi,
I haven't looked at the actual innings pitched for our young pitchers last year (and the year before for comparison), but are you at any bit concerned with the Verducci effect on our young pitchers (especially Parker and Milone)? They sure pitched a lot of innings down the stretch last season and if we are going to make a run this upcoming season, these guys need to be healthy.

I'm sure you guys know what the Verducci Effect is, but I've included his 2012 article just in case.
http://sportsillustrated.cnn.com/2012/writers/tom_verducci/01/18/year.after.effect/index.html

Thanks.
-OdyMandrell

First, it's useful to separate out innings thrown by relievers and innings thrown by starters.

Relievers

Our relievers in this group are Ryan Cook and Sean Doolittle. Here is Cook's IP breakdown (all data from Fangraphs, grouped together with league and level):

2008: 33

2009: 142.2

2010: 108.1

2011: 68.2

2012: 73.1

The high totals in 2009 and 2010 represent the fact that he started games through 2010, after which Arizona converted him into a reliever. Cook made 7, 25, and 20 starts from 2008-2010, respectively. His innings total decreased as a result of this switch in roles. Here are the same data for Doolittle:

2011: 1

2012: 47.1

Of course, Doolittle had a well-known conversion from first baseman with knee issues to power lefty reliever between 2011 and 2012. But, he did pitch before at UVa, before he was drafted by the A's in 2007 (data from The Baseball Cube):

2005: 49.1

2006: 90.2

2007: 82.1

He made 1 start in 22 appearances in 2005, 15 in 18 in 2006, and 14 in 16 in 2007. What does all this mean, though? Probably nothing. We know that reliever performance is relatively fungible, except for the most elite guys (Mariano Rivera, et al). What's more, as you can see, it's very difficult to bucket pitchers into discrete groups. If you do, then you start having a sample too small to draw any serious conclusions from.

Starters:

Starters are the actual focus of the Verducci effect, specifically Jarrod Parker and Tom Milone on the A's, if they meet the criteria. First, though, let's define the effect. Here is Verducci's own description of it from the article OdyMandrell linked:

For more than a decade I've been tracking this price, which I call the Year After Effect, and which some places, including internal metrics used by at least one organization, referred to as the Verducci Effect. I began tracking it because Rick Peterson, when entrusted as the Oakland pitching coach with the golden arms of Barry Zito, Tim Hudson and Mark Mulder, believed in managing the innings for a pitcher from one year to the next. Too big a jump for too young a pitcher would put a pitcher at risk the next season for injury or regression.

From his philosophy I used a rule of thumb to track pitchers at risk: Any 25-and-under pitcher who increased his innings by 30 or more I considered to be at risk. (In some cases, to account for those coming off injuries or a change in roles, I used the previous innings high regardless of when it occurred.) I also considered only those pitchers who reached the major leagues. Mariners GM Jack Zduriencik, for instance, agrees that major league innings create more stress than minor league innings, so the effect is more profound.

First, for both Milone and Parker, they do not meet his classical definition. While they increased their workload significantly year-over-year, both threw mostly MiLB innings in 2011, and Milone did not throw more than 30 IP more than his previous high.

Parker (24 y/o):

2008: 117.2

2009: 97.1

2010: (Tommy John Surgery recovery)

2011: 135.2

2012: 202

Milone (25 y/o):

2009: 151.1

2010: 158.0

2011: 174.1

2012: 190.0

So, they don't fit the Verducci effect candidacy. That being said, what of the effect itself? First, the one thing that should be concerning about the description of the process is that appears to use relatively arbitrary cutoffs: reaching MLB, Age <25, 30+ increase in IP, etc. It should indicate to a careful observer that there is significant sample bias. I'm not the only the one who thinks this, too: there are many articles that say that this is a bunk theory. First, Baseball Prospectus' Derek Carty wrote that its biggest problem is that there is no control group and he created one! (To properly test a theory scientifically, one needs two equally matched groups. One that is untouched -- the control -- and another group that receives the "treatment", in this case, the Verducci flag). He concluded basically the opposite of the theory: the Verducci effect guys did BETTER than the control group.

Deadspin doesn't think the theory is worthwhile, either. The Hardball Times, other BP writers, and J.C. Bradbury from Sabernomics (all linked in the Deadspin and BP articles I pasted here) are also all entities that think this isn't a rule that doesn't hold muster.

As fans, specifically A's fans, what are we to think? Personally, I respect the writings of all those guys, as they are a lot smarter than me. I'm inclined to believe them. Is that to say I'm not concerned about pitcher workload for the A's young rotation anchors? Of course I am. I just trust the A's read all the smart analysis that is out there, do their own analyses, and have their own system for monitoring pitcher health. Pitching injuries will happen; indeed, the simple act of throwing thousands of pitches per year, and tens of thousands over a professional career makes pitchers especially vulnerable to arm trouble. Unfortunately, there is not one hard-and-fast rule about how to identify those who are susceptible to injury or not.

Great question! Again, please keep the mailbag submissions coming! Remember, it's athleticsnationmailbag [at] gmail [dot] com.