Thursday, April 28, 2011

Dave Duncan's Magic Formula (and Kyle Lohse's Ridiculous April)

One of the hardest things for more statistically minded baseball fans like myself is attempting to determine the influence of a coaching staff. Sure, we can make guesses, see which team’s players outperformed expectations, look at who did the best job of optimizing lineups and using their best relievers at the most critical points in a game, but in the end, it’s very difficult to come to many conclusions about which coaches are truly the best of the best, and how their influence affects a team. However, we don’t have it down to a science, in the way that we do for evaluating pitchers and hitters.

However, I think it would be tough to say Dave Duncan hasn’t been as influential as any other non-manager on any coaching staff in the last two decades. The former Oakland and current (as of 1995) St. Louis pitching coach has spent the better part of the last quarter of a century turning low-cost and low-expectation starters into productive quasi-aces, King Midas armed with a bullpen phone.

Duncan’s record is staggering. A quick look through the résumés of pitchers under Duncan’s tutelage in the last decade or so shows that a huge percentage have experienced complete turnarounds after moving from other staffs to work with Duncan and the Cardinals. In recent years, Kyle Lohse, Jake Westbrook, and Brad Penny have experienced resurgences under Duncan. Going back a little further reveals others who have benefited from Duncan’s influence, including Darryl Kile, Woody Williams, Jason Marquis and current Cardinals’ ace Chris Carpenter. Duncan was working his magic even before the turn of the century, coaxing the four best seasons of Dave Stewart’s career out of the starter from 1987-1990 in Oakland.

Duncan’s latest reclamation projects have all followed a similar formula for success. Lohse (98 FIP- in the season before joining the Cardinals, 93 in his first season in St. Louis), Westbrook (113/91), and Penny (99/88) all found new life under Duncan. As you might have guessed, Lohse’s early season returns from 2011 are the impetus for this article. Thus far, Lohse has registered a 2.53 FIP, 31% better than a league average pitcher this year.

Duncan tweaked each of these pitchers’ approaches in different ways, but in each case was effective in making the most of that pitcher’s repertoire. Penny dropped his walk rate under Duncan, from a 2009 2.65 BB/9 and career 2.88 to a 2010 1.46 mark with the Cardinals. Penny also increased his groundball rate from 43.7% in 2009 to a 2010 52.8%, the first time he had been above 50% GB after 10 years ranging between 42-48%. The changes were successful, as Penny put up a 3.40 FIP and 3.23 ERA in his year with the Cards, both more than a full point below his 2009 mark. Duncan’s largest influence, especially in the increase in groundball rate, was a huge increase in the number of splitters Penny threw. Penny began throwing his splitter in 2008, using the pitch 10.1% of the time over the course of that campaign. In 2009, he pulled his use of the pitch back slightly, throwing it only 6.5% of the time. Then after moving to St. Louis, Penny’s use of the pitch more than quadrupled, as he used it 28.4% of the time in 2010. Penny’s dropped the frequency with which he throws his straight fastball and replaced a large portion of these fastballs with splitters. At 47.2% fastballs, he used the pitch 20% less than he had in any season since 2003 last season, and early this year has continued to throw his splitter more than 23% of the time and his fastball only on 56.5% of offerings. After not putting up a season below 60% fastballs until last year, this represents a huge philosophical change in Penny’s pitching, and the strategy certainly seemed to work last year.

Though Jake Westbrook made a different change in his repertoire after moving to St. Louis from Cleveland in the middle of the 2010 season, this change brought about a very similar result. After his midseason trade that sent Westbrook to the Cards, Ryan Ludwick to San Diego and prospect Corey Kluber to Cleveland, Westbrook increased his fastball frequency more than 10%, from 57.4% in Cleveland to 67.6% under Duncan. To compensate, Westbrook decreased his use of his cutter from 12.9% to 4.7%. The change worked similarly to Penny’s, increasing his groundball rate from 53.3% to 62% once he reached St. Louis. This increase coincided with a 2.9% drop in line drive rate, allowing Westbrook to drop his FIP from 4.64 to 3.52 after the trade. It seems as though Duncan’s main goal is to increase his pitchers’ rates of inducing groundballs, changing their repertoires as necessary to achieve that end.

Lohse followed a similar pattern after signing with the Cardinals in 2008. As was the case with Penny, Lohse’s walk rate fell from 2.66 BB/9 in 2007 (and 2.74 career) to 2.21 in his first year in St. Louis. Like both Penny and Westbrook, Lohse began to induce worm-burners at a rate higher than at any other point in his career. Lohse averaged a 40.8% groundball rate over the first six years of his career, including a career-worst 36.9% in 2007. After moving to St. Louis and beginning to work under Duncan, Lohse put up a 45.8% GB rate, the highest of his career so far, and during his time in St. Louis has averaged a 45.1% rate over four seasons. As a result, Lohse logged a career-best 3.78 ERA and 3.89 FIP in 2008, after managing only a 4.62 ERA and 4.53 FIP the year before.

Duncan’s successes are not limited to these recent starters. In 2002, Chris Carpenter was shut down early with a torn labrum, finishing the season with a 5.28 ERA and 4.95 FIP. Carpenter missed the 2003 season rehabilitating, and expectations were fairly low when he signed with the Cardinals. Though Carpenter had been a highly regarded prospect when he was promoted by the Blue Jays in 1997, he had spent six years with the team logging unimpressive stats including a 4.83 ERA and 4.42 FIP. Upon returning from injury, Carpenter was a different pitcher, dropping his walk rate from 3.31 BB/9 in 2002 to 1.88 in 2004, increasing his K/9 by two per nine innings, and dropping his ERA and FIP to career-bests. As is becoming something of a pattern here, his groundball rate also increased from 41.6% in 2002 to a post-injury 52.2%. Since moving to the Cards, Carpenter has won a Cy Young and been an All-Star three times. Working under Duncan, Carpenter evolved from post-hype prospect to a yearly contender for the title of the best pitcher in the NL.

This formula has worked for others as well. When Jason Marquis moved to the Cardinals in 2004, he increased his groundball and strikeout rate and decreased his walk rate, resulting in career-best numbers in ERA and xFIP. The late Darryl Kile halved his walk rate and added two strikeouts per nine, turning his career around after arriving in St. Louis in 2000. Woody Williams came to St. Louis in the middle of the 2001 season, and dropped his ERA and FIP from 4.97 and 5.02 in San Diego to 2.28 and 3.76 after the trade. Williams continued to perform at a high level in the several seasons following his move, his 3.59 and 3.72 FIPs in 2002 and 2003 representing the two top seasons of his career. Even Dave Stewart, who moved from the Phillies to Duncan’s staff on the A’s early in the 1986 season, immediately put up career years in the next four seasons at ages 30-34. One would expect Stewart to begin a decline after age 30, but uniting him with Duncan turned him into a staff ace on an A’s team that would lead the AL in ERA in 1988, 1989, and 1990 and lead the club to a World Series championship in ’89. Though Fangraphs doesn’t show data on batted ball types before 2002, I wouldn’t be surprised in the least to learn that these earlier pitchers benefited from increases in groundball rate similar to the ones that have become a pattern for the more recent starters.

It’s unclear whether this is all Duncan’s influence. Maybe it’s more of an organizational philosophy, and maybe some of these pitchers’ moves to working under Duncan simply coincided with them reaching the best years of their career by chance. However, it seems as though Duncan has developed a method to turn ordinary pitchers into gold, and his employers have benefited immensely because of it. Duncan isolates the best weapon each pitcher has for inducing ground balls and helps tweak that pitcher’s repertoire to maximize the rate at which these pitchers get hitters to put the ball on the ground. His pitchers reduce their walk rates, allowing fewer free passes and forcing hitters to earn their place at first base. The formula works, and Duncan has turned many a pitcher’s career around as a result.

As for Lohse, he’s not going to keep this up, at least at the level he has up to this point. His BABIP of .202 is sure to regress, as is his nearly 80% strand rate and 3.1% HR/FB. However, he does have the underlying stats that suggest this could be another career year under Duncan. Lohse’s 1.17 BB/9 would represent a career low, and his 48.6% groundball rate would be a career high mark. Following the Duncan formula, Lohse is once again finding success, as he and so many others have over the past quarter of a century.


  1. You mentioned in detail what Duncan suggested to Westbrook & Penny to induce more GB, what about Lohse?

  2. Lohse's mix of pitches didn't change a ton as he moved to St. Louis, but his curveball got a ton better (although he was using it with roughly the same frequency) from 2007 to 2008. His O-Contact also jumped 7%, so to me the most likely change was towards using his curveball more effectively by throwing it below the zone more consistently and using it to induce grounders. That's not as clear a trend as in Westbrook and Penny's case, however, and since it's more speculation on my part I didn't feel comfortable enough calling that a cause-effect relationship to put it in the piece. Thanks for reading!