« Palmeiro hits 3,000 | Main | Judge to Comcast: Yer out! »

2005 midseason analysis, part 3: injury accounting

“Guess what? I got a fever! And the only prescription is more cowbell!”

—fictional record producer Bruce Dickinson (as played by Christopher Walken) on a Saturday Night Live sketch that first aired April 8, 2000

With 70 games left in the regular season, Baltimore finds itself in the thick of the playoff race. But in the past month or so, the team has not been winning consistently the way it did early in the year. Injuries have thrown the Birds' formerly smooth-running outfit out of rhythm. Will the Orioles find their figurative cowbell in time to fuel their drive for the playoffs?

From the top (down)

If you haven't read parts one and two of this series, I suggest that you skim them before proceeding. The first article looked at the Orioles' place in the standings at the All-Star break; the second one broke down the Birds' pre-break statistics to identify strengths and weaknesses in their team performance. This article and the next one will focus mainly on specific solutions to the Orioles' problems.

To put it in medical terms, the first two articles consisted mostly of high-level diagnosis along with some tentative prognosis. The team-wide statistics used in those articles represented a macroscopic view of the Orioles' performance. Now, however, the analysis will move toward a lower level of diagnosis by looking closely at particular symptoms, the better to prescribe appropriate treatments for the team's peculiar ailments. (My diagnoses are provided pro bono. The treatments, though, could be costly. I hope the O's have a good insurance plan.)

That's just the way the story goes

The first half of the season demonstrated that the Orioles' roster, when healthy, has the ability to play at a high level—perhaps even a championship level. But the last two months in particular have shown that the organization is not deep enough for the O's to continue to speed along when key players are injured.

The slide in the team's performance coincided roughly with the period when Erik Bedard, Javy López, Brian Roberts, Melvin Mora, and Luis Matos were injured. Although many sportswriters attempt to quantify the impact of an injured player by comparing the team's won-lost records with and without him in the lineup, that information is flawed because too many other variables—such as other injuries and quality of competition—exist that could confound the analysis. Not only is this slapdash method scientifically unsound, but it's also unnecessary given the many more precise tools available.

Here's a different—and I would argue better—way to look at the injury effect:

  1. Determine what the injured starters have contributed to the Orioles when healthy.
  2. Determine what their replacements contributed when the starters were injured.
  3. Extrapolate the starters' healthy performances over the time they missed to injury.
  4. Subtract item 2 from item 3 to determine the lost production.

To measure individual contributions, I'll use games played and Runs Above Replacement (RAR) data from the Davenport Translations at the Baseball Prospectus web site.

In the tables below, the line labeled "[Starter's name], if healthy" corresponds to item 3 above—that is, the starter's healthy performance projected over the period he missed to injury. "Est. production lost to injury" corresponds to item 4, showing the hypothetical difference between the starter and his replacements at that position during the time that he missed. Starters' names and statistics are in bold.

Starting Pitcher
Player GS RAR
Erik Bedard 9 25
Hayden Penn 7 -2
Another Oriole SP*
1 1
Bedard, if healthy 8 22
Est. production lost to injury 8 23

* I assumed Bedard would have made one more start after July 1, when Penn was demoted. I assigned that start to the quartet of remaining starters, Rodrigo López, Sidney Ponson, Bruce Chen, and Daniel Cabrera, who collectively have added about 1 RAR/G.

Player AdjG RAR
Javy López 29 15
Gerónimo Gil 28 3
Sal Fasano 29 3
Eli Whiteside 1 0
All C replacements 58 6
J. López, if healthy† 29 15
Est. production lost to injury 29 12

† I estimated that López would have played two-thirds of the games he missed, essentially taking over Fasano's playing time. Note that numbers for Gil and Fasano are for the entire season, not just the period in which López was injured (it was too inconvenient to calculate RAR splits over a given date range).

Second Base
Player AdjG RAR
Brian Roberts 77 59
Chris Gómez 10 4
Roberts, if healthy 10 8
Est. production lost to injury 10 4

Third Base
Player AdjG RAR
Melvin Mora 75 36
Chris Gómez 11 2
Mora, if healthy 12 6
Est. production lost to injury 12 4

The AdjG discrepancy is due to rounding error.

Center Field
Player AdjG RAR
Luis Matos 47 16
David Newhan 17 4
Jeff Fiorentino 11 3
Larry Bigbie 5 1
Eli Marrero 3 1
Ramón Nivar 3 1
Keith Reed 1 0
All CF replacements 40 10
Matos, if healthy 40 14
Est. production lost to injury 40 4

Totals for C, 2B, 3B, and CF
Player AdjG RAR
Starters 228 126
Replacements 92 20
Starters, if healthy 92 44
Est. production lost to injury 92 24

Notes: For pitchers, RAR = PRAR + FRAR. For position players, RAR = BRAR + FRAR. Numbers cover the period up to the All-Star break. Full team DT's are available at BaseballProspectus.com.

Summary: the hypothetical production lost to Bedard's injury was approximately 23 RAR spread over eight starts. The production lost to injuries at catcher, second base, third base, and center field was about 24 RAR—and that's just before the All-Star break. The toll is still increasing because Javy López has not returned to the lineup yet. The starters at those four positions missed about 90 games prior to the break.

I want you back

It is obvious that the seven-week loss of Bedard from the rotation to a knee ailment hurt the Orioles far more than any other injury. Bedard's temporary replacement in the rotation, Hayden Penn, pitched at a barely acceptable level in his seven big-league starts. If Penn had replaced a marginal pitcher such as Sidney Ponson, that would not have been a problem. However, Bedard was so dominant when he was healthy that Penn's contribution in Bedard's absence amounted to a significant drop in performance. To put it in perspective, Bedard's loss meant about as much to the Orioles in terms of lost RAR as the combined losses of López, Roberts, Mora, and Matos. If the round approximation of 10 RAR = 1 win is applied, then those 23 RAR lost to Bedard's injury represent the equivalent of two to three games in the standings.

Fortunately, Bedard returned from the disabled list Monday. It's unlikely that he will continue to pitch as well as he did before his injury, when he saved an amazing 2.8 runs per game compared to a replacement starting pitcher. But if he declines just a little—say to 2.0 RAR/G—in the approximately 15 starts he has left, he still could contribute another 30 RAR, or about three wins, to the team. And that would be a significant boost: for instance, it could turn a 35–35 team into a 38–32 one.

The next most costly injury to the Orioles has been the broken metacarpal of Javy López. López went on the disabled list one day before Bedard did, and he is about to begin a minor-league rehab stint that will take about a week, so his DL stay will be about two months when complete. His replacements, Gil and Fasano, have been adequate at best. Assuming that López would have started about two-thirds of the games he missed, his absence has already resulted in a loss of about 12 RAR, or about one win, to the Orioles. If established trends continue, his absence will total about 15 RAR down the drain. But if he comes back and hits as well as he did early on, he could add another two wins to the team down the stretch. López's return should be another significant addition to the lineup.

Injuries to starters at the other three positions were less significant, adding up to about one more loss in the standings. Mora and Roberts only missed about two weeks each, and Gómez did a decent job in their stead, so those absences were not terribly painful. Although Matos missed about six weeks, he hasn't been all that valuable when healthy—his pre-ASB batting line of .282/.353/.393 was about average for a center fielder—so his weak-hitting replacements did not constitute a huge step down in production. (Upgrading to a better center fielder such as Mike Cameron would be nice, but even if such a player were available, the price—say, Matos or Larry Bigbie plus one of the organization's top prospects—could be substantial enough to make a general manager think twice.)

Can't live if living is without you

Oriole management and journalists covering the team have frequently brought up the effect of injuries as an explanation for the team's flagging performance over the past month or two. And that talk is not mere claptrap—injuries to several starters, in particular Bedard and Javy López, took a serious bite out of the Birds' competitiveness, costing them in the neighborhood of five wins before the All-Star break. Adding Bedard and López to the regular lineup will make the team significantly better down the stretch.

It is foolish to think that all of the Orioles' regulars will stay healthy the rest of the season. But not all injuries are equal: it's clear that losing a star player hurts a team much more than losing an average or below-average player, simply because the star's production cannot be as easily replaced. Hence, any prescription for Baltimore's success has to include keeping the team's most irreplaceable front-line players as injury-free as possible. Such players—Miguel Tejada, Roberts, Mora, and Javy López, and on the pitching side, Bedard, B.J. Ryan, and to a lesser extent Bruce Chen—are just too valuable to lose for an extended period of time. And given the lack of ready minor-league talent in the Birds' system, even lesser players such as Matos, Jay Gibbons, Rafael Palmeiro, and Rodrigo López are important to this team.

Good health is critical to the success of any ballclub, the wealthy ones not excepted. That's a truism that many people (including yours truly) fail to grasp concretely at times.

Next: Proposed changes to personnel and usage patterns.

Comments (1)

Neal Dorfman:

one down, but plenty more to go...

with ponson headed to the beaches of san diego (where he'll survive on truckloads of fish tacos and tequila), that only leaves the beattie-flanagan wrecking crew with the task of getting rid of the bums still remaining on the orioles' 40-man roster...

...including, but not limited to:
- eric dubose
- sal fasano
- jay gibbons
- geronimo gil
- jorge julio
- eli marrero
- john parrish
- sammy sosa

and why aren't they starting eli whiteside? he's already proven that he can field, hit, run AND throw better than fasano...

i don't hate maz, but i've been questioning quite a few of his managerial decisions throughout the year...


This page contains a single entry from the blog posted on July 19, 2005 10:44 AM.

The previous post in this blog was Palmeiro hits 3,000.

The next post in this blog is Judge to Comcast: Yer out!.

Many more can be found on the main index page or by looking through the archives.

Powered by
Movable Type 3.33