September 6, 2003

Prospectus Today

Survival of the Fittest, Relievers

by Joe Sheehan

You can read Wednesday's column for a longer discussion of the principles behind this research, but the general idea is that you have to dig deeper than just looking at seasonal stats to see what a team has going for it down the stretch. That piece addressed the quality of rotations; this one looks at the quality of bullpens.

As I wrote Thursday, I orginally researched this using Adjusted Runs Prevented (ARP). I had to throw out that data, though, because using a value stat distorted the results. I went back and used Runs Responsible Average--the rate equivalent of ARP--to generate the lists below. I determined each team's current core relievers by looking at game logs; most teams had six, but some had as few as five or as many as eight. The point is to pull out the pitchers whose performances show up in the seasonal numbers but who are no longer pitching, or pitching much, for their teams. This correction should give us a better idea of who has the best bullpens in baseball for the stretch drive.

Here are the data; I'll run some notes following the chart.

Through September 4

RRA+: RRA of current core relievers
Season RRA: Overall team RRA

                 RRA+     Rank      Season RRA       Rank
----------------------------------------------------------
Mariners         2.48       1          3.46            4
Dodgers          2.55       2          2.91            1
Angels           2.71       3          3.11            2
Astros           2.86       4          3.38            3
White Sox        2.99       5          4.43           13
Rockies          3.12       6          4.15            9
A's              3.14       7          3.68            5
Indians          3.25       8          3.88            7
Twins            3.75       9          4.09            8
Marlins          3.80      10          4.92           17
Brewers          3.81      11          4.54           14
Diamondbacks     3.83      12          3.83            6
Yankees          3.96      13          5.02           19
Cubs             4.01      14          4.01           16
Reds             4.21      15          4.21           13
Giants           4.22      16          4.43           10
Mets             4.23      17          5.06           20
Blue Jays        4.30      18          5.10           21
Red Sox          4.33      19          5.44           26
Devil Rays       4.40      20          4.62           15
Rangers          4.50      21          5.13           22
Phillies         4.69      22          4.45           12
Padres           4.79      23          5.91           30
Pirates          4.81      24          5.16           23
Royals           4.82      25          5.36           25
Cardinals        4.85      26          5.77           29
Tigers           4.88      27          5.68           28
Orioles          5.14      28          5.26           24
Expos            5.43      29          5.63           27
Braves           5.51      30          4.96           18

• The first thing that jumps out is the Braves, who currently have the worst bullpen in baseball by this metric. That's mostly because John Smoltz is on the Disabled List, and Smoltz is the only Braves reliever with more than token value. They're also missing the average pitching of Darren Holmes and Kevin Gryboski, both of whom went to the DL in August.

• Just two teams other than the Braves don't look better when you tease out the relievers who no longer have jobs or are unavailable: the Phillies, who are currently without Terry Adams, and the Diamondbacks. The Snakes are interesting in that they have almost no pitchers who blew up in a handful of innings, which is how their non-core relievers can have virtually the same RRA (3.82) as their core staff (3.83).

• Moving back to the top of the list, we see that the Mariners leapfrog three teams to claim the title of best bullpen in baseball. This is largely because they're getting amazing work out of Rafael Soriano and Julio Mateo, and disposed of Giovanni Carrara and Jeff Nelson. If the M's can reach the post-season, they're going to be scary for any team that trails in the sixth inning.

• The big jump is by the White Sox, for whom I did not count just-activated Billy Koch as a core reliever. He hasn't pitched in nearly a month, and with Scott Sullivan now around and Jerry Manuel comfortable with his relief rotation, Koch is unlikely to pitch innings of any importance until 2004.

• The Marlins also look a lot better this way. Armando Almanza being out for the year with a shoulder injury helps, but the big difference is that the 40 nightmare innings Blaine Neal and Kevin Olsen provided are removed from the ledger.

  The Fish are a great example of why you need to do this type of analysis. Their aggregate ARP and RRA make it appear that their bullpen isn't very good. Right now, however, it is, and that's what matters in the Wild Card race, not that three pitchers who have been sent to Elba were awful.

  Teams that had a lot of problems filling the back of the bullpen, or who have relievers returning from injuries, look better in this analysis. Here are what we might call the top five most improved bullpens in September:

  
                   RRA+     Season RRA     Improvement
  ----------------------------------------------------
  White Sox        2.99        4.43           1.44
  Marlins          3.80        4.92           1.12
  Padres           4.79        5.91           1.12
  Red Sox          4.33        5.44           1.11
  Yankees          3.96        5.02           1.06
  
  

• You know who impresses me? The Twins, who have promoted two relievers to their rotation and still have the ninth-best pen in the game.

There's a lot of other information you can glean from this, and from the information I presented on starting pitchers. The key point, though, is that seasonal data can hide as much as it reveals. Particularly for teams in a pennant race, you want to make sure that you're evaluating the right players, and not allowing the work of people who aren't going to have an influence on the race to have an influence on your analysis.

This will be especially important next month, when we're looking at playoff matchups. I've always used a team's seasonal totals and ranks in those analyses, and I'll have to remember that what I want to know is how the pitchers who will actually pitch in October stack up, not how Juan Acevedo or Mark Mulder or Kurt Ainsworth helped their team prevent runs.

Joe Sheehan is an author of Baseball Prospectus. 
Click here to see Joe's other articles. You can contact Joe by clicking here