Milwaukee Brewers Blog - The Junkball Blues : Milwaukee Brewers bullpen

2007 NL Relievers

radio silence — Fri, 16 May 2008 17:13:15 GMT

I've designed an informal index to rank relievers according to the % of leads they converted. The index should be read like ERA+ or OPS+, where 100 = average. I calculate each individual's leads converted % against the league average %. So, in the 2007 NL, where the league average leads converted was .085085, I divide the player's % of leads converted by the league average, and then multiple that number by 100 (i.e., 85085 / 85085 = 1 x 100 = 100 = league average).

Here's a list of relievers with 10 or more leads opportunities from the NL in 2007, arranged according to number of leads opportunities, and then according to their % of leads converted.

***

(An example of the calculation is... Brian Fuentes (28/35, 80%), so 80000 / 85085 = 0.94 x 100 = 94.0. This number is simply an index number used to more accurately compare relievers according to each other and according to league average. It should be read simply like ERA+ or OPS+, where 100 = average. +100 = above league average, -100 = below league average.).

40-54 opportunities

Brandon Lyon, Ari: 37 leads converted (35 hld, 2 sv)           40 opportunities          108.7

Kevin Gregg, Fla: 38 leads converted (32 sv, 6 hld)              42 opportunities          106.3

Takashi Saito, LAD: 40 leads converted (39 sv, 1 hld)          44 opportunities          106.8

Heath Bell, SD: 36 leads converted (34 hld, 2 sv)                 40 opportunities          105.7

Jose Valverde, Ari: 47 leads converted (47 sv)                      54 opportunities          102.3

Francisco Cordero, Mil: 44 leads converted (44 sv)               51 opportunities          101.4

Jon Rauch, Was: 37 leads converted (33 hld, 4 sv)                43 opportunities          101.1

Trevor Hoffman, SD: 42 leads converted (42 sv)                  49 opportunities          100.7

Jonathan Broxton, LAD: 34 leads converted (32 hld, 2 sv)   40 opportunities          99.9

Chad Cordero, Was: 38 leads converted (37 sv, 1 hld)          47 opportunities          95.0

40-54 opportunities: 393 leads converted / 450 opportunities, .87333                       102.6

In this first group, we see a small group of relievers that were able to face a lot of leads opportunities because of their role. The interesting thing about these relievers is that while only two of them were below league average as a whole (Broxton and Chad Cordero), six of the ten relievers were below the group average, meaning that although they were mostly better than a leaguer average NL reliever, most of the members of this group did not do their job as well as the average of the group. This suggests, in some ways, that it is very tough to find a closer that converts leads at 90% or higher (which Lyon, Gregg, Saito, and Bell accomplished). Meanwhile, closers that convert fewer than 90% of their leads are much less valuable because not only are they closer to league average, and therefore more easily replaced, but they are also not converted leads at a % demanded by the role.

According to the 2007 NL, relievers that face 40-54 opportunities should convert leads above 87%.

31-39 opportunities

Jason Isringhausen, StL: 32 leads converted (32 sv)              34 opportunities          110.6

Manuel Corpas, Col: 35 leads converted (19 sv, 16 hld)        38 opportunities          108.3

Derrick Turnbow, Mil: 34 leads converted (33 hld, 1 sv)       37 opportunities          108.0

Matt Capps, Pit: 33 leads converted (18 sv, 15 hld)              36 opportunities          107.7

Tony Pena, Ari: 32 leads converted (30 hld, 2 sv)                 35 opportunities          107.5

Rafael Soriano, Atl: 28 leads converted (19 hld, 9 sv)           31 opportunities          106.2

Ryan Dempster, ChC: 28 leads converted (28 sv)                 31 opportunities          106.2

Bob Howry, ChC: 30 leads converted (22 hld, 8 sv)             34 opportunities          103.7

Billy Wagner, NYM: 34 leads converted (34 sv)                   39 opportunities          102.5

Brad Hennessey, SF: 32 leads converted (19 sv, 13 hld)       37 opportunities          101.6

David Weathers, Cin: 33 leads converted (33 sv)                  39 opportunities          99.4

Ryan Franklin, StL: 26 leads converted (25 hld, 1 sv)          31 opportunities          98.6

Chad Qualls, Hou: 26 leads converted (21 hld, 5 sv)             32 opportunities          95.5

Brian Fuentes, Col: 28 leads converted (20 hld, 8 sv)            35 opportunities          94.0

Brad Lidge, Hou: 26 leads converted (19 sv, 7 hld)              34 opportunities          89.9

31-39 opportunities: 457 leads converted / 523 opportunities, .87380                       102.7
  20-28 opportunities
J.C. Romero, Phi: 22 leads converted (22 hld)                       23 opportunities          112.4

Pedro Feliciano, NYM: 20 leads comverted (18 hld, 2 sv)    22 opportunities          106.8

Matt Lindstrom, Fla: 19 leads converted (19 hld)                  21 opportunities          106.3

Brett Myers, Phi: 24 leads converted (21 sv, 3 hld)               27 opportunities          104.5

Brian Shouse, Mil: 22 leads converted (21 hld, 1 sv)             25 opportunities          103.4

Saul Rivera, Was: 22 leads converted (19 hld, 3 sv)              25 opportunities          103.4

Justin Miller, Fla: 17 leads converted (17 hld)                       20 opportunities          99.9

Aaron Heilman, NYM: 23 leads converted (22 hld, 1 sv)      28 opportunities          96.5

A. Alfonseca, Phi: 23 leads converted (15 hld, 8 sv)             28 opportunities          96.5

Dan Wheeler, Hou: 17 leads converted (11 sv, 6 hld)            21 opportunities          95.1

Tom Gordon, Phi: 20 leads converted (14 hld, 6 sv)              25 opportunities          94.0

LaTroy Hawkins, Col: 18 leads converted (18 hld)               23 opportunities          91.9

Renyel Pinto, Fla: 17 leads converted (16 hld, 1 sv)              22 opportunities          90.8

Bob Wickman, Atl: 20 leads converted (20 sv)                      26 opportunities          90.4

Jorge Julio, Col: 16 leads converted (16 hld)                         21 opportunities          89.5

Salomon Torres, Pit: 17 leads converted (12 sv, 5 hld)          23 opportunities          86.9

Scott Linebrink, SD: 16 leads converted (15 hld, 1 sv)         22 opportunities          85.5

Shawn Chacon, Pit: 13 leads converted (12 hld, 1 sv)           20 opportunities          76.4

20-28 opportunities: 346 leads converted / 422 opportunities, .81991                       96.3

I put these two groups together because they are a study in contrasts. It seems that the relievers that faced 31-39 opportunities were actually the best relievers in the NL of those that faced 20 or more opportunities. Again, like the higher appearance closers, their role requires a higher lead conversion %, and they need to do their job at a high level. Unlike the higher appearance relievers, however, these relievers are more readily available, and it seems to be easier to find a reliever that faces 31-39 leads and does their job well than a reliever that faces 40-54 leads and does their job well. Perhaps this suggests something about how managers use their bullpen; maybe all relievers should be provided a lower number of appearances with a lead, in the hopes that more members of the bullpen can participate in a leads converted opportunity, and more relievers can do their job well.

By contrast, the 20-28 group was the worst of those relievers that faced 20 or more leads opportunities. As a whole, this group was below league average, meaning that their level of production is very easily replaced on any given team in the NL. This group is full of a range of good middle innings set up men that simply do not face a ton of leads, but do their jobs very well (Romero, Lindstrom, Feliciano, Shouse, and Rivera), and closers that do not face a lot of opportunities, or were bad at their job and were replaced (Myers, Gordon, Wickman).

I think that the lower leads converted% of the 20-28 group reflects managerial usage of relievers more than any other group. What is suggests to me is that by saving their better relievers for later in the game, managers then use relievers that are not as good in the middle innings, and these are the relievers that face 20-28 leads opportunities a year. They're not as good as top closers and set-up men, so they face less high leverage situations to begin with, and they also face fewer appearances because a starter is more likely to make it into the 6th or 7th than into the 8th or 9th. All of these factors taken together create a below average group from 20-28. I'd suggest that this is the aspect of bullpen management that needs to change the most.

10-19 opportunities

Joe Beimel, LAD: 17 leads converted (16 hld, 1 sv)              17 opportunities          117.6

Damaso Marte, Pit: 15 leads converted (15 hld)                    15 opportunities          117.6

Will Ohman, ChC: 13 leads converted (12 hld, 1 sv)             13 opportunities          117.6

Lee Gardner, Fla: 11 leads converted (9 sv, 2 hld)                11 opportunities          117.6

Ray King, Was: 10 leads converted (10 hld)                          10 opportunities          117.6

Doug Brocail, SD: 10 leads converted (10 hld)                     10 opportunities          117.6

Joe Smith, NYM: 10 leads converted (10 hld)                       10 opportunities          117.6

Carlos Marmol, ChC: 17 leads converted (16 hld, 1 sv)        18 opportunities          110.9

Randy Flores, StL: 15 leads converted (14 hld, 1 sv)                        16 opportunities          110.2

Tyler Yates, Atl: 15 leads converted (13 hld, 2 sv)                16 opportunities          110.2

Brian Wilson, SF: 15 leads converted (9 hld, 6 sv)                16 opportunities          110.2

Matt Wise, Mil: 14 leads converted (13 hld, 1 sv)                 15 opportunities          109.7

S. Schoeneweis, NYM: 13 leads converted (11 hld, 2 sv)     14 opportunities          109.1

T. Tankersley, Fla: 17 leads converted (16 hld, 1 sv)             19 opportunities          105.2

Carlos Villanueva, Mil: 17 leads converted (16 hld, 1 sv)      19 opportunities          105.2

Jack Taschner, SF: 13 leads converted (13 hld)                     15 opportunities          101.9

Trever Miller, Hou: 13 leads converted (12 hld, 1 sv)            15 opportunities          101.9

Russ Springer, StL: 11 leads converted (11 hld)                    13 opportunities          99.4

Steve Kline, SF: 11 leads converted (9 hld, 2 sv)                  13 opportunities          99.4

Jesus Colome, Was: 13 leads converted (12 hld, 1 sv)           16 opportunities          95.5

Kevin Correia, SF: 12 leads converted (12 hld)                     15 opportunities          94.0

Jared Burton, Cin: 11 leads converted (11 hld)                      14 opportunities          92.3

Mike Stanton, Cin: 10 leads converted (10 hld)                     13 opportunities          90.4

A. Benitez, Fla: 11 leads converted (11 hld)                          16 opportunities          80.8

Cla Meredith, SD: 10 leads converted (10 hld)                      15 opportunities          78.4

10-19 opportunities: 324 leads converted / 364 opportunities, .89011                       104.6

The relievers that face 10-19 opportunities have the highest percentage of any group of NL relievers because, let's face it, they face fewer opportunities and thus probably have more leverage to succeed. They are probably less likely to work multiple days in the week in this group, and they are also relievers that are used for multiple tasks. Some that snuck into this group, like Wilson, Marmol, and Beimel, have the chance of being better relievers than this group indicates, and simply faced a different opportunity in the bullpen for 2007. They will all likely face different roles throughout 2008.

4/22/08: Observations at Miller Park

radio silence — Wed, 23 Apr 2008 05:20:56 GMT

I am in the middle of job hunting right now, and it is a very stressful time. So, I promised myself yesterday that if I found a few jobs to apply to on Tuesday, I would treat myself to Tuesday’s game. My added incentive was lower level box seats on sale for $22, so I found some job openings, completed my morning appointments, and headed to the stadium to pick up my will call ticket. 118, Row 20, Seat 12.

1. I have not yet sat this close to the action at Miller Park. While the Cleveland Indians were in town during their snow out games, my friends and I sat down the third base line off the dugout. So, those seats were technically “closer” to the field, these seats today placed my right on top of the action.

Today I really captured just how intense every at bat is. Everyone is so focused on each pitch, and after each pitch, the tension explodes, and there is an unfolding of events: the catcher sliding to smother a wild breaking ball, the batter fighting off a sinker, the umpire interacting with the batter, the pitcher following his pitch. I truly appreciate just how difficult this game is for the major leaguers, and the level at which they are playing the game.

2. Manny Parra needs to throw more strikes. And better strikes. Today he finished yet another non-quality start, although it was not an awful start. His start was plagued by nibbling around the strike zone. This disappointed me, because his stuff is clearly good, and he has a very easy time switching speeds, causing deception, and maintaining different velocity levels. Parra has the potential to be one of those quality middle of the rotation (or higher) pitchers that can get by on deception, while having the luxury of sitting on a mid-90s fastball.

Here is an example of some of his velocity sequences:

90-93 (first through second IP)

89-91 (fourth through fifth IP)

73-76 (second through fifth IP)

83-84 (first through fifth IP)

81-82 (third through fifth IP)

I basically found Parra to be working comfortably in the upper 80s and lower 90s, and by the third onward he really established his change up (which I assume to be 81-82). I believe he was also throwing a slider effectively (which I assume to be 83-85). The fact is, by the middle of his start, he was working at least two different mid-80s off speed pitches effectively.

However, his trouble started in the second, where he provided Ludwick with a lead off walk, recorded two fly outs, and then surrendered consecutive singles, including and RBI single to pitcher Kyle Lohse (who now has 3 RBI off of Parra). What could have been an efficient inning took his pitch count to 33, and the top of the order then put Parra through a long third inning, in which Parra through 30 pitches, but got a lot of defensive help from Ryan Braun (who threw out Albert Pujols).

Parra then struggled in the fourth with two outs recorded, allowing two baserunners on two suspect defensive plays (one charged an error, the other a single). He finished the inning without allowing a run, but his pitch count was further damaged.

What I like about Parra is his velocity level, and his mechanics. He has a wide variety of pitches to use, and he has the luxury of that hard fastball in his back pocket. He also seems to vary the fastball velocity at some points, notching different fastball velocity levels in most of his innings pitched. I think that if Parra was more aggressive, he would in turn limit his pitch count, and it seems to me that he needs to simply trust his ability to change speeds.

The perfect model for Parra occurred in the fifth inning, where he regularly set up his fastball with a slow curveball. I forget the particular batter, but he was unhittable when he attacked with his change up, and also when he used his slow curve to set up his blazing fastball. 93 is that much faster and harder when it follows a 74 curveball, especially when the fastball that follows the low curve is placed high in the zone.

3. Seth McClung was the perfect model of a Brewers pitcher pitching aggressively today. He was placed into a tie game in the 12^th, to face Ludwick, Glaus, and La Rue. He started the inning by testing each of his breaking pitches, which come in at two different speeds. He then turned to working his fastball around the zone, and although he can get the ball up to the upper 90s and tough triple digits, he began working the fastball at slower speeds, working up to a consistent 96 by game’s end (ironically, the GameDay guns are much more consistent in reading McClung’s fastball at very high velocities from the start of the inning)..

By moving the fastball around, and focusing on placement and throwing strikes, McClung was able to strike out Ludwick and La Rue looking; La Rue’s strike out was an especially sweet inside fastball that touched 96.

Glaus was the only batter to put a ball in play during the inning, and he tried to time McClung’s fastball. His effort did not work, and he dribbled out to McClung, who threw to first to make the out.

McClung’s outing was a model for what Brewers’ pitchers need to do: change location, vary speeds, throw strikes. Change the places the batter can look in the zone for strikes, and by being aggressive, it forces the batter to beat the pitcher, rather than allowing the pitcher to defeat himself.

4. Cardinals fans are very nice people and very knowledgeable fans. I had the pleasure of sitting next to a diehard Cards fan, and was in a little pocket of Cards fans for the game. It was an enjoyable day at Miller Park from that perspective.

Adjusted Bullpen Stats in New York

radio silence — Mon, 14 Apr 2008 14:00:00 GMT

The bullpen did a stellar job in NYC this weekend, first eating a bunch of innings in a precautionary weather move on Friday, supporting Sheets on Saturday, and holding up in yesterday's wicked shootout, which saw 7 runs scored by the Mets, but only 1 allowed by our pen in 5 IP of work.

The pen gets a well-deserved break today. Perhaps if the starters get their act together and start pitching deeper into games, the pen can rest some more. Only Sheets and Suppan have pitched a quality start for the Crew thus far, and Sheets is the only starter averaging more than 6.00 IP per start.

If the starters don't get their act together soon, we're going to get a repeat of the bullpen meltdown of 2007. CV, Parra, or Bush must step up and start pitching some quality IP.

Adjusted Bullpen Stats After April 11-13

Adjusted leads converted (6/6 series, 100%, 17/19 season, 89%)

Gagne: 2/2 (two SVO)

Shouse: 2/2 (two holds)

Mota: 1/1 (one hold)

Torres: 1/1 (one tie)

Shared blown saves: 0

Adjusted reliever WHIP (season) (R+IS)/(BB+H+IR)

Shouse: 8% (1+0)/(2+3+8)

Riske: 18% (1+1)/(5+4+2)

Torres:20% (3+0)/(6+8+1)

McClung: 60% (5+1)/(3+6+1)

Series IR/IS: 1/0, Shouse 1/0

Scoreless IP (season):

Torres: 6.3

Riske: 6.0

Mota: 4.0

Shouse: 3.0

Gagne: 3.0

Turnbow: 2.0

McClung: 1.7

Bullpen Distribution

radio silence — Mon, 14 Apr 2008 13:40:00 GMT

Things are getting a bit better, but Yost is still leaning on Torres and Riske pretty hard:

Torres and Riske: 16.7 IP

Gagne, Turnbow, McClung, Shouse, and Mota: 19.7 IP

Adjusted Bullpen Stats After Reds Series

radio silence — Sun, 13 Apr 2008 16:26:00 GMT

Adjusted Bullpen Stats After April 8-10

Adjusted leads converted (series 3/4, 75%; season 11/13, 85%)

Mota: 1/1 (one hold)

Shouse: 1/1 (one hold)

Torres: 1/1 (one tie)

Gagne: 0/1 (one SVO)

Shared blown saves: 0

Adjusted Reliever WHIP (season) (R+IS)/(BB+H+IR)

Shouse: 0% (0+0)/(1+1+7)

Riske: 20% (1+1)/(4+4+2)

Torres: 29% (2+0)/(2+4+1)

McClung: 60% (5+1)/(3+6+1)

Series IR/IS: 6/1, Shouse 5/0, Riske 1/1)

Scoreless IP (season):

Torres: 4.3

Riske: 4.0

Mota: 3.0

Shouse: 2.7

Turnbow: 2.0

McClung: 1.7

Gagne: 1.0