Tuesday, October 12, 2004

Fire Jim Tracy?

Checking out the comments over at Dodger Thoughts, I discovered a new baseball blog, Steve Haskins' Fire Jim Tracy. I don't like to do hit pieces, but in this case it's fairly warranted and I hope this post is taken in the spirit of thoughtful debate.

Haskins' first entry criticizes Tracy's decision to let Wilson Alvarez face Albert Pujols with two out and two on in yesterday's game, a decision which resulted in Pujols launching the go-ahead 3-run home run. Haskins writes:
Why is Wilson Alvarez pitching to Albert Pujols? Since I can already guess at your answers, I'll take them ahead of time.
He then proceeds to list four potential reasons: because Alvarez is attempting to walk Pujols, because Alvarez is superior to Duaner Sanchez, because loading the bases by issuing a walk is disadvantageous, and because the bullpen couldn't be overtaxed because of game 5, and to each potential response Haskins presents a rebuttal. I agree with him on his fourth point, but his other arguments don't quite work.

First, Alvarez was pitching pretty carefully to Pujols. Alvarez was behind 3-1 and Mayne set up pretty far inside on the fifth pitch. Alvarez missed his location on the pitch and left the ball out over the plate, causing the home run. Alvarez certainly wan't being overly aggressive, and the pitch that Mayne set up for was not "very little reward and a lot of risk."

Second, Alvarez was a good option at this point. Haskins asserts that the stats say Duaner Sanchez is better; I'd have to disagree. Sanchez has a better ERA, but that's not a very good indicator of performance. Duaner Sanchez' fair run average using the BPro reliever reports is 3.82. Alvarez' is 2.61. Alvarez' ERA just as a reliever was 2.44 and his opponents average as a reliever was .201 with 8.57 K/9, far superior to Sanchez' 3.38/.266/4.95. If you use Alvarez' overall numbers, it's closer but Alvarez still comes out on top: 7.61 K/9, 2.31 BB/9, 0.90 HR/9, for Alvarez; 4.95, 3.04, 1.01 for Sanchez, all of which translates to a 3.57 to 4.58 edge for Alvarez in Fielding Independent Pitching ERA. If it's splits you're after, Sanchez' line vs. RHB is .260/.333/.370; Alvarez' is .222/.269/.315. This one's pretty open and shut. Haskins also makes the argument that Yhency Brazoban could have been brought in. That just doesn't seem like a very good idea. For one, There were still another 5 innings in the game, and counting on Gagne and Brazoban to get through all of them is unrealistic. Alvarez is probably a better option than anyone besides Gagne and Brazoban; there could be an argument for Carrara, but his performance against right-handed hitters this year hasn't been close to Alvarez'. Bringing in one of the relief aces in Brazoban or Gagne is a bad idea in this spot since the pitcher's spot was due up fourth in the bottom of the inning, meaning that if Bradley and Cora got on the Dodgers would likely have to waste an out on Brazoban (or end up utilizing him for only one out as a pitcher). This is a situation where any value added by turning to a reliever is nullified by the likelihood that that reliever won't be available in the fifth inning, so it only makes sense to replace Alvarez if a less valuable pitcher (Dessens or Sanchez, in this case) is more likely to retire Pujols.

Third, it is not particularly logical to walk Pujols to get to Rolen. With runners on first and third and two out, a team has .5219 expected runs. That jumps to .8082 with the bases loaded and two out. While Rolen might not have had a hot hand, there's not sufficient evidence to suggest that his injury made it impossible or unlikely for him to deliver a hit. Let's break this down statistically. Albert Pujols hit a home run every 12.9 at bats this year, got a single every 6.1 at bats, and got a double or triple every 11.2 at bats, and he was retired once every 1.49 at bats. Let's assert that Alvarez is roughly equal to the competition that Pujols accumulated those numbers against (which is being generous to Pujols when you look at the opponents quality figures at Baseball Prospectus-- Pujols' opposition yields .252/.332/.408 whil Alvarez has yielded .244/.291/.367, which might be artificially low because of Dodger Stadium but since this game is at Chavez Ravine that's irrelevant). If this is the case, walking Pujols to get at Rolen is a bad idea. Using run expectancy data from this year as well as each player's outcome rates for this year, I made a chart of how many runs to expect facing Pujols with runners at the corners versus Rolen with the bases loaded with two out:


Total RE Prob
WRE



Pujols Rolen Pujols Rolen Pujols Rolen Rolen 80% Rolen 65%
Out 0 0 0.6689 0.596 0 0 0 0
1B 1.4605 2.461 0.1639 0.148 0.2393 0.365 0.291739 0.237038
2B 2.3359 3.336 0.0895 0.061 0.2091 0.205 0.1636694 0.132981
HR 3.1135 4.114 0.0777 0.058 0.2419 0.238 0.1906085 0.154869
BB/HBP
1.808
0.136
0.246 0.1971462 0.160181





0.6904 1.054 0.8431631 0.68507














































WRE is the weighted run expectancy, or expected runs times the probability of the event. Walking Pujols to get at Rolen is a pretty bad move unless there's reason to believe Rolen is hitting at a level below 2/3 of his regular season performance. His hitless performance in ten at bats to that point in the series is not sufficient evidence, especially since he'd already drawn 5 walks in the series. Walking Pujols is not intelligent strategy unless you believe that the Dodgers goal should have been to prioritize not allowing any runs to score rather than minimizing the amount of runs scored-- in other words, if it's believed that each additional run allowed provides very diminishing returns. With Suppan having issued three walks and allowing a home run in only three innings while only registering one strike out, that's a pretty inane argument; the Dodgers were not, at that point, in a position where preventing one particular run was more important than minimizing the Cardinals' overall run total.

All in all, I think the evidence supports Tracy's decision here.

Haskins then criticizes Glenn Hoffman for waving Alex Cora into third on a play on Saturday. That's a bit of a reach for the "Fire Tracy" campaign. Regardless, Haskins argues that this added no potential battle since Lima was due up next and would bunt. This isn't quite true: a runner on third nullifies the force play at third if the bunt is bad; a runner on third makes whoever fields the bunt hold the runner back, making a successful throw to second less likely; a runner on third creates more options so that the pitcher can swing away since a double play would still score him, as would a sac fly; and a runner on third can score on an Izturis sac fly if Lima strikes out. Is this a tremendous play? No. But it's not exactly evidence that the Dodgers are being overaggressive because of the current field management.

In his second post, Haskins criticizes Tracy for starting Alex Cora in every game of the series. I don't know how familiar Mr. Haskins is with the concept of platoon splits, but all of the Cardinals starters have been right handed. Against RHP, Cora has hit .267/.367/.376 while Hernandez has hit .259/.351/.412. If we use Gross Production Average to evaluate those two lines, Hernandez leads, .261 to .259, a difference which is statistically insignificant. Since there's no dispute that Cora has the better defense, I see no point in this criticism. I might listen to an argument that Cora's production has tailed off recently and that his expected contribution should be lower than his overall line, but that's a nit that's not worthy of picking.

In his third post, Haskins provides a great deal of anecdotal evidence about Tracy not getting optimal performance out of his players. I won't make the argument that Tracy is making his players perform better; he seems to be about average and I think managers tend to have very little effect on individual performances.

In his fifth post, Haskins argues that a closer should always be used in the seventh inning of a tight elimination game if he's available for three innings. I think that's overstated; just because someone is hypothetically available for three innings doesn't mean they'll be able to do that. In this case, however, Gagne would be a good choice since Walker-Pujols-Rolen was due up, and using one's relief ace against this core is a very wise thing to do. However, the pitcher's spot was was due up third in the ensuing inning, and trailing three runs you can't have a relief pitcher hitting. Absent a double switch that puts Gagne in the #5 slot and Choi/Ventura/Grabowski/Hernandez in the #9 slot at first base, I can't think of any way to pull that off, and that means that in the eighth or ninth the Dodgers would have to replace what would be Shawn Green's spot with a pinch hitter, and that simply doesn't make much sense given his production in the series. Considering that Brazoban entered the inning doing well with only 19 pitches under his belt, as Haskins had pointed out in his previous post, I think Tracy made the right move.

Do I think Tracy is a great manager? Hell no. I do think, considered in the context of the set of major league managers, that Tracy is a pretty good manager. I was tearing out my hair about his handling of Hee Seop Choi, but I don't think that was a fireable offense. Choi was struggling and Jayson Werth wasn't. When Choi did play, he was always in a wall of left-handed hitters, meaning his removal from the game in the sixth or seventh against a lefty was nearly inevitable. These are problems. But Tracy seems much, much more receptive to constructive criticism than most managers, and going forward I think having an offseason to talk to Paul DePodesta will make him an improved manager. And do you have any idea who the other candidates are? Check out whose available for the Mariners job. Earl Weaver isn't coming out of retirement, and the only one I think would be an upgrade from Tracy is Larry Dierker. It's easy to be angered by a few managerial decisions. But guess what? Everybody is. Sure, Mike Scioscia, Jack McKeon, Bobby Cox and Joe Torre get a lot of free passes, but a lot of that has to do with the talent they're given. With a choice between Tracy, Felipe Alou, Bruce Bochy, Clint Hurdle, Al Pedrique, Art Howe, Frank Robinson, Larry Bowa, Dusty Baker, Phil Garner, Tony La Russa, Lloyd McClendon, Dave Miley, and Ned Yost, Tracy's at least in the top three or four. It's possible to do better than Tracy, but not probable.

Comments:
A
 
Post a Comment

<< Home

This page is powered by Blogger. Isn't yours?