June 29, 2011

Painting the Black

Profiling the Nats' New Skipper

by R.J. Anderson

When filmmaker Guillermo del Toro is creating a monster for a movie, he follows a four-step process. First, he focuses on the silhouette—the broad, basic shape that imparts general information about the creature. From there, del Toro works on the creature’s movement, then its color, and finally the rest of the details necessary to flesh out a convincing creation. Building an analysis of a baseball skipper can follow the same sequence of steps, and with Davey Johnson hired over the weekend as the Nationals’ new pilot, the time is ripe to discover whether Washington has a managerial monster at the helm.

The Silhouette

Since we’re beginning Johnson’s professional biography with the broad strokes, we can start with his win-loss records in his previous stints as skipper. Looking at a team’s records under a particular manager as well as just before and after his arrival and departure offers an overview without providing much context. We’ll get to the context in time, but first, here’s how Johnson’s teams fared with him:

Season	Team	Wins	Losses	Win %
1984	NYM	90	72	.556
1985	NYM	98	64	.605
1986	NYM	108	54	.667
1987	NYM	92	70	.568
1988	NYM	100	60	.625
1989	NYM	87	75	.537
1990	NYM	20	22	.476
1993	CIN	53	65	.449
1994	CIN	66	48	.579
1995	CIN	85	59	.590
1996	BAL	88	74	.543
1997	BAL	98	64	.605
1999	LAD	77	85	.475
2000	LAD	86	76	.531

And now, in a cheap With or Without You (WOWY)-style analysis, how those teams fared in the seasons before and after Johnson’s employment. The Johnson Effect—if there is one—is noted by the winning percentage differential year-to-year. If Johnson took over a team that won 90 games every season but won only 60 the season prior because of a rash of injuries, then saw the team recover to win 90 again with perfect health, this method gives him credit for improving the team, regardless of his actual effect.

B/A Johnson	Season	Team	Wins	Losses	Win Rate	Johnson Effect
Before	1983	NYM	68	94	.420	Improved
After	1991	NYM	77	84	.478	Declined
Before	1992	CIN	90	72	.556	Declined
After	1996	CIN	81	81	.500	Declined
Before	1995	BAL	71	73	.493	Improved
After	1998	BAL	79	83	.488	Declined
Before	1998	LAD	83	79	.512	Declined
After	2001	LAD	86	76	.531	Neither

Johnson has won more than half his games as manager, distributed across four teams and more than 10 seasons. Those results are admirable, even without knowing the processes behind them. That’s the next step; to remix Winston Churchill’s quote, “However beautiful the results, you should occasionally look at the strategy.”

The Movement and Color

Reactions to Johnson’s laid-back demeanor serve as a litmus test. If folks are okay with it, then Johnson’s team is likely winning ballgames. If folks are up in arms, then Johnson’s team is likely underperforming (or winning but located in New York). Johnson is known for dabbling in statistical analysis and even tried delivering the gospel of stats to the venerable Earl Weaver way back when. Stories like that are amusing, but they also have the tendency to buy managers free passes in sectors of the sabermetrics community.

In the relatively brief time that Baseball Prospectus and Johnson coexisted in the baseball world, much of our analysis was positive, but it rarely encroached on what made Johnson unique or worthwhile. That’s not a fault of the authors—it’s still quite difficult to peg down what makes a manager good—but there was a notable absence of strategy talk. The most illuminating observation might stem from a sentence in this article from 2000:

Davey Johnson called for more steals of third base, and more double steals, than any other manager last season.

The three areas in which the manager tends to have the most control are a team’s stolen base attempts, sacrifice hits, and intentional walks issued. Comparing those statistics across the entire league can be tricky, as the National League’s barbaric mandate to have pitchers bat means that intentional walk and sacrifice hit totals can be skewed upward. To adjust for this, I measured Johnson’s teams against their own league only, as the graph below shows. The multi-colored line below the x-axis is used to differentiate between Johnson’s tenures with each team.

Johnson seemingly adjusts his stolen base and bunting strategies based on the teams he is handed, which seems intuitive, but there is something to be said for a willingness to adapt, a quality not every skipper shares. Regardless of the team or league, Johnson seems hesitant ever to issue a free pass, a point in his favor. It’s hard to analyze the numbers too much beyond the surface, as not every bunt, intentional walk, and stolen base attempt is created equal. Generally, though, the fewer outs you give up and the fewer extra bases you allow, the better, and Johnson has largely succeeded in keeping his teams’ counterproductive practices to a minimum. 

The Details

Now that we’ve established the outlines of Johnson’s career, we can fill them in with details taken from his previous work as a manager:

• Leadoff hitters:  Over Johnson’s career, his team’s leadoff hitters have posted higher on-base percentages than the league’s average leadoff hitter in six of his 14 seasons. The degree of over- or under-performance has varied, though, as Johnson has seen his leadoff men outperform the league-average by more than 20 points in four seasons and underperform by more than 20 points in three others. This is about personnel as much as it is about lineup decisions, since Johnson tended to shuffle through a handful of options during his worst seasons (in 2000, for instance, he tried Todd Hollandsworth, Devon White, Tom Goodwin, and F.P. Santangelo). 
• Closer usage: In Johnson’s Mets days, he used Jesse Orosco and Roger McDowell for more than an inning apiece and alternated who closed. From 1985 through 1987, McDowell led the Mets in saves but still recorded less than 50 percent of the team’s sum. After that, Johnson seemingly bought into having a set closer (as did the rest of baseball): Jeff Brantley in 1994-1995; Randy Myers (whom he also had in New York) in 1996-1997; and Jeff Shaw in 1999-2000. Only once in Johnson’s first 10 seasons did his leading save-getter average less than an inning pitched per appearance, but that was not the case in three of his last four seasons.
• Platoons: How often did Johnson give his batters and pitchers the platoon advantage at the plate? This is perhaps the most difficult aspect of his performance to judge, as it can be skewed by a lefty- or switch-hitter-heavy lineup. With the Mets, Johnson’s batters consistently held the platoon advantage more often than the league-average team, but with the Reds, the opposite was true. His time with the Dodgers and Orioles brought mixed results. The same is true of Johnson’s pitching staffs. There is enough noise in the data to prevent me from making a definitive claim.

The Conclusion

That’s a lot of information to digest, but what is known about Johnson can be boiled down to a concise summary: he has won many ballgames with teams of various talent levels and compositions, hates giving away outs and bases, and seemingly adjusts his tactics based on his teams’ strengths and weaknesses. Johnson has served with Washington in an advisory capacity over the last several seasons, so it stands to reason that Mike Rizzo and his associates are fans of the qualitative abilities that Johnson brings to the table, too. No one can predict how the latest Johnson administration in Washington will work out, but there’s little reason to conclude that it will be as capable of inspiring horror as one of del Toro's creations.

R.J. Anderson is an author of Baseball Prospectus. Follow @r_j_anderson
Click here to see R.J.'s other articles. You can contact R.J. by clicking here