Yesterday I spent some time building a pitch-count estimator based on the play-by-play data I've accumulated for college baseball. When teams spent $1 million or more on pitching in the first round of the draft, I'd imagine they'd like to know those arms aren't overworked.
The first guy I looked at was our own first-round pick, Eric Arnett. I kind of wish I hadn't. Before we get into the ugly numbers, a bit of background.
I'm not rigid about pitch counts--if someone is throwing well and plowing through the opposing lineup (whether it's Arnett or Suppan--ha!), there's no harm in letting them cross the 100- or 110-pitch mark. On the other hand, there's an awful lot of evidence suggesting that, the farther and more frequently a pitcher goes past those benchmarks, the more likely he is to become injured or ineffective.
Thus, more than a couple of 120-pitch outings out of a 22-year-old should be reason for concern. This, of course, depends on the pitcher, his physiology and his repertoire. If someone throws 150 fastballs, it isn't nearly as bad as 120 or 130 pitches with a mix of curves and splitters. So perhaps the Brewers know something I don't, that there's no reason Arnett couldn't hold up to the workload.
Enough of that. On to the numbers. Each of the rows below shows the date and pitch count of one of Arnett's starts. The one with the asterisk is an estimated pitch count, while the others are the pitch counts reported either by Indiana or their opponent that day.
02/21 89*
03/01 93
03/07 145
03/14 85
03/19 94
03/28 129
04/04 119
04/11 136
04/18 97
04/25 131
05/02 142
05/08 126
05/20 95
05/29 118
I suppose it could be worse, but it's sure not very good. Boyd Nation, who has done a huge amount of work in the last decade on college pitcher workloads, published a report last year with the worst workloads he could find. To compare them, he used a Baseball Prospectus metric called Pitcher Abuse Points (PAP3), which you can read about here.
Given those 14 starts, Arnett racked up 296,316 pitcher abuse points. All PAP numbers are pretty big, but few are that big. That would've placed him in Boyd's top ten last year, and higher than any other prospect in the nation.
In the words of Rambling Al: Oof.
Again, maybe the Brewers know something we don't know about Arnett and his ability to stand up to this kind of workload. They must be aware of the problem. Certainly, I won't be complaining if Arnett doesn't throw very many innings in professional ball this summer.
I'll be taking a broader look at the more problematic workloads in college baseball in my Hardball Times column next week. I've also got some data on supplemental pick Kyle Heckathorn below the jump.
It's clear from Heckathorn's pitch counts that his coach at Kennesaw State is aware of this kind of stuff. Until May, this looks like a major league 4th or 5th starter's workload:
02/20 81*
02/28 96*
03/06 109
03/13 106
03/20 98
03/27 93
04/03 119
04/11 108
04/17 107
04/24 65*
05/01 116
05/08 117
05/14 121
Heckathorn's PAP total for the season is around 25,000, just about the best we can possibly hope for out of a college starter. (It's those 130+ pitch starts that really cause problems.)
That final start on 5/14 was a nine-inning effort in a game K-State eventually won in 11. My pitch count estimator, though it's pretty good, thought he threw well over 140 pitches, but Heckathorn was much more efficient than that. I wish Arnett's workload looked a lot more like this one.