February 9, 2006
Team Health Reports
2006 IntroIt's time to take our annual look at team health. Since starting "Under The Knife" five years ago, the holy grail of injury analysis has always been a combination of injury prevention and injury management. That prevention starts with as objective an assessment of risk as possible. The Team Health Reports have been my attempt to use the tools available, both objective and subjective, to make an accurate assessment of risk.
Teams have long done this, either in an organized manner or on an ad hoc basis. With each signing and spring training physical, teams take a look at their situation and determine value, their lineup construction, and weaknesses. Where some teams avoid risk, some contain it and some take the gamble knowing that it's their best chance of winning.
More and more teams are now taking an organized approach to the process. Teams from top to bottom have noted that putting a relatively small amount of money into medical personnel, facilities, and preventative practices is paying big dividends on the field. It's one thing that the Brewers won some award; it's quite another that an AGM-level executive is in charge of the program. Across the league, teams are noticing the success of people like Ken Crenshaw, Stan Conte, and Dave Labossiere.
As we head into the season, teams are more or less at a point where they're going to hand the talent they have off to the field and medical staffs, hoping that the teams they envisioned, built and paid for are written into as many lineup cards as possible. Some teams will have that happen because they understand the risks. Some won't. Some, like the Dodgers, just can't be as bad as they were last year. Lineup construction is not only a series of what-ifs and understanding the personnel available; it also has to take into account a large measure of injury risk. A team like the Reds can only handle having Ken Griffey Jr. in the lineup if they have a good fourth outfielder on hand. On the other hand, there's no replacement possible for someone like an A.J. Burnett or, as we saw last year, a Curt Schilling or Randy Johnson.
Injury analysis is still in its infancy. There are few in the field doing serious work with a few more around the periphery. Whether it's the historical look at injuries provided in Gary Gillette's encyclopedia or Sig Mejdal's work for BIS, there are some things moving in the field. It's no different here at BP. I don't pretend to be a great researcher, but with concepts like Tom Gorman's Injury Accounting system (you'll read more about this in BP 2006) and the database work by Mike Groopman and Bill Burke that truly power my work at UTK, it's an exciting space to be in. If you add in the organizations like the Commissioner's Office, insurance and worker's compensation carriers, and risk management consultants, there's a lot of people working on this problem, some with amazing tools and data.
The Team Health Reports are one of my favorite toys and this year, we're doing things a little bit differently. I focused on improving the system under the hood this season and have worked hard on getting the best data both into the system and making sure that the new data we have for adjustments is as precise as possible. Like PECOTA (but much less deadly accurate), the THR system got an overhaul. Extreme Makeover? No, not that much--it's still based on eight categories that adjust the base age and position risk that serves as the baseline. Those scores are then banded into three categories--green, yellow, and red--that serve as a quick shorthand for risk. We even adjusted the icons to make sure that our color-blind friends have an easier time.
The biggest change is a new base age and position risk table, created with the help of an outside agency that works closely with Baseball on insurance issues. This new table, the base for all players, should immediately make our risk projections more accurate. Instead of the previous versions' three-year average, this is based on a five-year weighted average.
I also have a better handle on what adjustments to this base risk should be made and how to weight them more accurately. We have more data on player histories and how teams handle injuries and we've started work on skill-specific injuries--an injury to a pitcher's arm or a speedy player's leg is more important to them than others. I'm also giving durability a bigger part in the system; playing 150 or more games on a regular basis is a valuable and predictable skill. I've begun making adjustments based on position rather than the simple player/pitcher split used in previous THRs. We have long known that just as there is a defensive spectrum, there is also a risk spectrum. It's been used as a factor when players change positions; now it is being used for each projection.
Finally, the most apparent change will be seen in the by-line at the top of each THR. Instead of writing them all myself, I've teamed up with three great young writers. Mike Groopman (Easts), Dave Haller (Centrals), and Tom Gorman (Wests) will be handling two divisions each. This not only keeps me from getting in a rut, but it'll give the THRs a fresh pair of eyes that will look at each team and each division a bit differently.
On Monday, we'll start the fourth set of Team Health Reports at BP, which is quickly shaping up to be the best set yet. Who knows who this year's Phil Nevin or Ben Sheets is going to be? I hope you enjoy them and get the information you need from them. As I'd say over at The Black & Blue Report, let's get to it!