The Baseball Economist: The Real Game Exposed (9 page)

However, such an experiment isn’t feasible. Major League Baseball isn’t going to allow some young economist to shuffle coaches from team to team over the course of several seasons just to see which coaches are better than others. But we don’t have to give up the quest to disentangle coaching and player contributions. What we can do is look at how individual players perform as they move from coach to coach. The frequent movement of players from team to team over their careers makes this possible. Using the right statistical tools, we can adjust for other factors that influence achievement, to see how players perform with and without a particular coach.

Specifically, I’m interested in charting the success of one man who has never even managed a major-league team: Leo Mazzone. The current pitching coach of the Baltimore Orioles made his name in baseball with my beloved Atlanta Braves during the mid-1990-through-2005 seasons. Though he never got higher than Double-A in the minor leagues as a pitcher, he was able to become a big-league pitching coach. Mazzone had the good fortune, as he sees it, to hook up with the legendary Johnny Sain.

Sain was a very good major-league pitcher who was the subject of a baseball poem about the 1948 Boston Braves:

First we’ll use Spahn, then we’ll use Sain.
Then an off day, followed by rain.
Back will come Spahn, followed by Sain.
And followed, we hope, of two days of rain.
²⁴

After finishing his successful eleven-year career as a player, he became a well-traveled pitching coach. Although his staffs were successful, he had a reputation for being a bit of a rebel, leading to his transient employment. When Leo had the opportunity to coach alongside Sain in the Braves’ minor-league system, he listened and learned. Mazzone writes in his autobiography,

He was so helpful to my thought process and approach. He was the smartest son of a gun and the nicest guy I’d ever talked to. He had all these things to offer. Believe it or not, there were some who wouldn’t listen to him. I think sometimes he was so far ahead of his time that those other people feared his knowledge, so, therefore, they turned it off. Whatever their reason was, they were stupid. I went the exact opposite direction. I was going to jump all over it and listen to everything he had to say. . . .

He didn’t give the usual clichés when it came to what he was teaching me. He taught me everything from throwing programs to proper spins on a baseball to strategies of baseball to dealing with the front office. Everything he talked about and taught me, I’ve seen unfold from the top down. It was a tremendous education.
²⁵

Tom Glavine, a two-time Cy Young Award–winner who pitched twelve seasons for Mazzone, describes this unique approach:

You get guys to throw as much as they want to, so they can develop feel on their pitches and understanding of their mechanics. In terms of getting guys to throw more often, I think a lot of people don’t want to do that simply because they are scared to death somebody’s going to get hurt. Coaches are going to err on the side of caution and not encourage guys to throw as much as Leo does.
²⁶

Four-time Cy Young–winner Greg Maddux elaborates:

There are no parachutes on your back, no cones to run around, no ten different meetings talking about something that doesn’t concern you. All the other stuff, you don’t partake in. So you spend less time doing nothing, and you spend all your time doing what it is you have to do to get better on the mound.
²⁷

Mazzone embraced a novel method others eschewed. He made his pitchers focus on pitching, with less concern for throwing and and general fitness drills. So if Mazzone’s success with pitchers is more than a myth, then his pitchers should excel compared to other pitching coaches who rely on traditional methods.

As the Braves emerged from cellar-dwellers in the 1980s to perennial division champs in the early 1990s, Mazzone began to gain some notoriety for the consistently good pitching staffs he put together year after year. Clearly, the Braves’ winning ways in the 1990s and 2000s had a lot to do with pitching. Table 8 lists the average league rank and ERA for each team in Major League Baseball since 1991, Mazzone’s first full year as the Braves’ pitching coach.

There is no doubt that the Braves had the best pitching staff in baseball during Mazzone’s tenure as the organization’s big-league coach. But early in his career, the media gave Mazzone little credit for the Braves’ success. He was seen as an oddity: a little man who rocked back and forth on the bench as he nervously monitored his pitchers. Though the Braves pitching was always good, the praise went to the players. Leo Mazzone was just an unsuccessful minor leaguer who’d inherited some good players. Steve Avery, Tom Glavine, Greg Maddux, and John Smoltz were some of the best pitchers in the game. Even Mazzone has graciously acknowledged that he has been fortunate to work with some of the game’s elite pitchers.

But after many of these pitchers moved on or missed time due to injuries, Mazzone continued to field good staffs. Soon people began
to notice that the Braves had a habit of getting good seasons out of pitchers that no one else wanted. People began to discuss whether or not the coach deserved more credit.

John Burkett was waived by the Tampa Bay Devil Rays in 2000—a season in which they won sixty-nine games—when the Braves signed him. In the following two seasons with the Braves, coached by Mazzone, Burkett pitched 354 innings with an ERA of 3.74 and a strikeout-to-walk ratio of 2.45.

Chris Hammond was a career journeyman, having had two above-average seasons with the Florida Marlins in 1994 and 1995, but was otherwise unspectacular. His skills declining, Hammond retired from the game in 1998. In 2002, he returned to the majors with the Braves at the age of thirty-six, where he posted an ERA of 0.95 as a reliever. During that season, Hammond had two strikeouts for every batter he walked, while surrendering only one home run during seventy-six innings of work.

Mike Remlinger had bounced between starting and relief pitching roles, mostly with the Cincinnati Reds, when the Braves acquired him in a trade prior to the 1999 season. Before joining the Braves, he had a career ERA of 4.63 and a strikeout-to-walk ratio of 1.62. In his four seasons with the Braves as pure reliever he improved his ERA by two runs and his strikeout-to-walk ratio by one.

Jorge Sosa was acquired from the Tampa Bay Devil Rays for a light-hitting utility infielder, and then Sosa had a career year under Mazzone. As a fifth-starter/long-reliever, Sosa had a career ERA of 5.17 and a habit of giving up too many long balls (1.29 per nine innings). He began the 2005 season as the Braves’ mop-up reliever, but soon became a starter, posting an ERA of 2.55 and cutting his home runs by approximately 40 percent (0.8 per nine innings).

Jaret Wright was once a promising young starter for the Cleveland Indians, but he was failing as reliever for the woeful San Diego Padres when the Braves picked him up off of waivers in 2003. In 2004, he became the number one starter in the Braves rotation, posting an ERA of 3.21 with a 2.27 strikeout-to-walk ratio.

After rebuilding their careers with a little help from Mazzone, these players, except for Sosa, left for new teams as big-name free agents. The Braves didn’t have to hold on to these guys, because they could count on getting similar performances from pitchers whom the rest of the league found to be unworthy.

Mazzone’s success is not without warts. Some pitchers who didn’t pitch so well under Mazzone have had otherwise decent careers, like Jason Marquis, Odalis Perez, and Dan Kolb. Maybe we’re just remembering the success stories, and some of the guys who were successful with the Braves just got a bit lucky during their tenure. Even if the tendency to succeed under Mazzone is real, there could be other contributing factors, such as age and defense. Hence, the success of Braves pitchers might be a product of selective memory or other factors that have nothing to do with coaching.

Given the rate at which pitchers move on and off the Braves roster, we can take a large group of pitchers and analyze how much better (or worse) they were with Mazzone than without. All we need are the right statistical tools.

Let’s look at pitchers who pitched regularly for Mazzone and for other pitching coaches in the league over their careers. The sample includes every pitcher who pitched at least one full year for Mazzone, and at least one full year for another pitching coach. Let’s throw out pitchers who pitched for more than one team in a single season, to ensure the transition isn’t affecting pitching performance. Also, the pitcher needs to throw a significant number of innings for the statistics to be useful. The cutoff for inclusion in the study is thirty innings pitched in a year. And because pitchers have various roles—starters throw many innings, while relievers pitch very few over the course of the season—I also compare performances according to pitching role.

We can compare seasonal pitching performances with and without Mazzone, but even with this suitable sample of pitchers, it is still complicated. Many other factors change as pitchers change teams; therefore, we need control for other potential factors. I employ multiple regression analysis to weight different influences, including Mazzone’s oversight, on pitcher performance. What are they?

First, pitchers change over time. As pitchers mature they tend to get better, but then decline after a certain point—typically in their late twenties or early thirties. Including pitchers’ ages when they post their performances allows us to adjust for this impact.

Another factor is that pitchers rely heavily on their defenses to help them, so we need to account for the quality of the defense on the team. I use the percentage of balls hit into the field of play that the pitcher’s team turned into outs to proxy defensive quality. Good defenses will have a higher percentage of outs on balls in play.

Also, the run environment in which the pitcher pitches can heavily influence his performance. Demographic and architectural features of ballparks can help or hurt pitchers. Atlanta, Mazzone’s employer during this study, had two different ballparks during the sample. Until 1997, the Braves played in Atlanta-Fulton County Stadium, so loved by batters that it was nicknamed the “launching pad.” Since 1997, the Braves have played in Turner Field, which slightly favors pitchers. We need to make sure that pitchers pitching for the Braves and for other teams are not punished or rewarded for playing in parks that favor or hinder hitters. Using a run-impact factor that measures the influence of a pitcher’s home park on scoring, we can adjust the number of earned runs a pitcher gives up.
²⁸

Additionally, the era in which a pitcher plays influences the typical runs allowed. As I discuss in a later chapter, these differences can be quite extreme. Therefore, I look at the average ERA of the league for every year that a pitcher pitches, so that he will not be punished for pitching relatively well in a high run-scoring season, or rewarded for pitching relatively poorly in a low run-scoring season.

Finally, the quality of the pitcher himself is very important. If the Braves are simply bringing in very good pitchers, then noticing that Leo Mazzone’s pitchers always pitch well doesn’t tell us much about the coach. I include the career ERA of the pitcher—corrected for the run environment of all the parks in which he pitched—so that we can observe how well pitchers perform relative to the way that we expect them to perform. That is, we can see both good and bad pitchers improve or decline under Mazzone’s guidance.

Defining a pitcher as a starter or a reliever is a bit tricky, because some pitchers do a little of both over the course of the season. After some toying around with the data, I settled on the following definitions. A starter is a pitcher who pitches at least seventy-five innings and starts 75 percent or more of the games played in that season. A reliever must pitch at least thirty innings and start in only 25 percent or less of
the games in which he played that season. I experimented with several other starter/reliever measures, but they yielded results that were not meaningfully different from the ones I present in this chapter.