I’m not a doctor

image
Does it hurt when you do that? (Icon/SMI)

As the season finishes, there are the usual injury question marks lingering around prominent players for next year. Grady Sizemore, Jake Peavy, Brandon Webb and half of the Mets are each keeper candidates whose values are hazy because of injuries. In the past, if you gambled on Albert Pujols‘ elbow or Chase Utley‘s hip, you won. If you bet on Justin Duchscherer or Ervin Santana this year, you lost. Since sizable discounts are often applied to players with injury risks, any informational advantages would be very valuable. Sabermetrics to the rescue?

I’m not a medical doctor, but I know a few things about statistics. And yet, when it comes to my own valuations for players with injuries, I often base my forecast for games missed due to injury on innuendo, rumor, and now personal experience. “Sabermetrics” (i.e. the use of the Law of Large Numbers) is not very useful because, when it comes to injury forecasts, the numbers aren’t large enough. Why? Here are several reasons followed by something that I’ve learned through personal experience.

1) There are too many different kinds injuries and players. (Or alternatively, not enough players getting injured). If all injured players only had a sprained ankle, we’d have a large sample of past histories to use. We could look at how that injury affects speed, power and pitching statistically using averages. Unfortunately, there are lots of different injuries (and lots of grades of injuries—tear, partial tear, sprain, etc…). This complicates things.

2) Historical data isn’t that helpful. Simply put: Tommy John‘s post-surgical recuperation was a lot different than players undergoing “his” surgery these days. Only the very recent past is helpful. For instance, Utley’s post-surgery performance was the main data point for forecasting Alex Rodriguez‘s post-hip surgery performance. It doesn’t help that we’re trying to forecast both recovery time and post-recovery performance.

3) Information is sketchy. Utley’s surgery was helpful for forecasting A-Rod’s because we knew exactly what A-Rod’s problem was and exactly what procedure was going to be performed. I’m a bit iffy on Sizemore’s lower abdomin. There are often players who limp into the offseason (all puns intended)—Alfonso Soriano this year is one likely example. These types of players are big candidates for a surprise spring training visit to the hospital.

This isn’t to say that we can’t take some averages—for instance, I think it is safe to say that most injuries affect pitchers more than hitters. But it does mean that we should be prepared for a lot of variance and, therefore I think, many temptations to make “eyeball” adjustments based on hearsay and personal opinion. Which isn’t to say personal opinion is useless.

This summer, I’ve been recovering from broken ankle for the past four months and I’ve learned a lot from my sample size of one. I’ve learned (from my orthopedist) that physical therapy speeds up recovery time but doesn’t change the end result—no amount of legwork is going to make my ankle 100% again (though I’m not quite sure what 95% of an ankle means). I’ve learned that healing takes a long time and getting old stinks. If I was a baseball player, I’d still be in my “prime years,” but I felt the effects of this surgery a lot more than I did eight years ago when I had my last one.

All this means that I’m still going to use a rule-of-thumb adjustment to the values of players with injury concerns. Only this time around, due to the, perhaps excessive, coloring of my recent experience, I’ll be more conservative than I have been in the past.

Print Friendly
 Share on Facebook0Tweet about this on Twitter0Share on Google+0Share on Reddit0Email this to someone
« Previous: Batted balls and cheese
Next: Welcome Ian Desmond »

Comments

  1. carter richard carter said...

    very cool article, maybe for the future you could please tell us what your thoughts are on the big stars and their future prognoses. thanks. best wishes.

Leave a Reply

Your email address will not be published. Required fields are marked *

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <strike> <strong>