July 14, 2005
Glass Half EmptySometimes, the data look like a crystal ball. If you stare at information long enough, patterns seem to emerge, only to vanish into the fog. Since Major League Baseball ceased publishing the Redbook a couple seasons ago, there has been a complete lack of data, both public and private, about the types and trends of injuries in baseball. Over the last three seasons, we've been attempting to keep the data ourselves.
There are problems with the data, mostly due to two factors: collection and transparency. We collect our information from public and private sources, attempting to get the most accurate view of injuries via the disabled list. It remains possible that we will miss some underreported transaction, though the use of multiple sources plus our double- and triple-check procedures should reduce this doubt. Using the disabled list has its own problems. A recent discussion with a major league trainer last week highlights this. The trainer bemoaned the use of the DL by teams to "hide players" or make roster moves and its effect on teams' stats. Another trainer advocated a system used in the minor leagues, where each trainer lists each player on his team as either "able" or "unable" to play. This system would give much more accurate views on minor injuries that do not necessitate disabling, yet would require a co-operation and co-ordination unheard of in MLB, especially with an independent outside organization.
That said, what do the data tell us about what we have seen in 2005 thus far? Quite a bit actually. There's been a perception that injuries are up this year. Viewed in a year-over-year manner, though, injuries are actually down a bit overall using days lost. Total salary dollars lost are up, however. The big-dollar injuries to players such as Barry Bonds, Jim Thome and Kerry Wood create a monetary imbalance--a month of Bonds costs a team as much as 10 full seasons of a player at the minimum salary. They also create a public perception of increased injury. There are some interesting reductions in injuries that we see in position, age, service, and contract groupings. Our first chart, days lost to the DL, makes it very clear why the Mariners, Reds, and Dodgers are struggling.
Of course, for every team like the Reds or Dodgers, there is a Nationals team that is exceeding every expectation in the face of injury loss. There's very little correlation between the ranking on this chart and where teams find themselves in the standings. The 2004 Red Sox overcame better than 1000 days lost to the DL by amazing depth and a reduction in key value injuries. The Nationals have had a significant number of injuries, but have mostly seen the injuries occur to replaceable players like Claudio Vargas (now with Arizona) and Jon Rauch. Do dollars lost tell us more?
Teams spending money have, very simply, more money to lose. It's hardly a one-to-one relationship, clearly showing that there are some stark differences in the ways even large-market teams deal with injuries. Last season's best training staff, Tampa Bay, is experiencing an increase in days lost. Of course, Rocco Baldelli's off-season, off-field knee injury makes up a significant portion of that increase, something that can hardly be blamed on Ken Crenshaw and his staff. With Roger Caplinger's Brewers and George Poulis' Blue Jays competing for the honor in 2005, it will be interesting to see if the relative youth of the Brewers or Roy Halladay's shin will determine the winner.
We continue to see the "big five" injuries--elbow, shoulder, back, groin, and ankle--making up a majority of the time lost to injury. Ankles, however, have seen a distinct reduction in both severity, incidence, and lost time/dollar/value. The flip side of that is a great spike in finger/hand injuries. There is no logical explanation for these changes. There is neither a change in the way that major league medical staffs deal with these nor any significant change in equipment or technique that would explain the data.
Smaller-revenue teams show a significant advantage in dollars lost, simply for the reason that their miniscule payrolls do not allow for many dollars to be lost. In some cases, percentage of Opening Day payroll gives a better indication of the actual opportunity cost of injury to a specific team. It can also disguise the great job being done by a team. Houston and Atlanta have both had few injuries, though the injuries they have had are significant, high-dollar ones. Take the case of Jeff Bagwell, the Astros first baseman lost for the season after shoulder surgery. He makes up better than half of the dollars lost, despite this being a long-standing injury that could actually call into question the system of "crediting" any chronic injury to a certain season. The Astros medical staff should be praised for keeping Bagwell in the game as long as he did, rather than singed for the loss of his services during the 2005 season.
Tommy John surgery is another interesting case, where a number of players are in various stages of recovery from the all-too-common surgery. Injuries that occurred in previous seasons and in some cases under different medical staffs or even with different teams are being credited to the current season. Scott Williamson is currently being credited to the Cubs despite the injury occurring last season while he was a Red Sox. We're still struggling to find a way to come up with a more accurate and fair system without making the system retroactive unnecessarily.
Is there more to find in this data? Absolutely. As we continue to collect, experiment and learn from our data, there will be new and interesting findings that could help the game of baseball or help a team gain some advantage. A reduction of just 10% of value lost to injuries by the average team last season would have effectively expanded their roster by one player. A greater reduction in time and dollars lost would help keep the 25 most talented players on the field, rather than seeing the 35th or 40th players in some instances. That could mean the difference between playing baseball or golf come October.