The underappreciation of the manager in today's analytics-driven MLB

San Francisco Giants manager Bruce Bochy has been one of the best in the game at getting unexpected production out of his players. Jake Roth/USA TODAY Sports

Baseball's analytics revolution is giving us the Incredible Shrinking Manager. Efficiency-minded, value-seeking front offices are hiring pliable, if not disposable, underlings in place of forceful field generals. As a result, a third of teams now have skippers with two years of experience or less.

I understand why GMs might think a manager's most important job is simply to stay out of the way of big data. Sabermetrics showed a long time ago that one-run strategies tend to hurt offenses. Research has debunked the idea that we can predict clutch performance. And now Statcast can analyze every batted ball. To the modern eye, tactics such as bunting, playing the "hot hand" and positioning fielders by instinct are just counterproductive meddling.

But today's baseball experts are making the classic mistake of overvaluing what they can measure most easily. A manager's job, after all, isn't limited to small-bore decisions.

I thought of this recently when the great Frank Robinson passed away. For all the plaudits Robinson garnered as an inner-circle Hall of Famer and racial pioneer, he was underappreciated as a manager. Time and again, Robinson led terrible clubs on tight budgets into pennant races. (The 1982 Giants, 1989 Orioles and 2002 Expos were all-time unexpected -- and fun -- contenders.) Athletes seemed to play at their upper limits for Robinson, who commanded respect and inspired loyalty. And his teams declined after he left. Sabermetrics, however, is stuck evaluating Robinson through the narrow lens of the small ball his aggressiveness favored. For example, in "Up, Up, and Away," an otherwise excellent 2014 history of the Expos, Jonah Keri writes: "Robinson got much of the credit for inspiring the troops. ... In reality, Robinson was a poor tactician who bunted far too often." Yet in his next paragraphs, Keri relates how Vladimir Guerrero, Brad Wilkerson and Tomo Ohka (among others) improved under Robinson!

Truth is, many of the game's greatest managers have had a knack for getting the most out of their players, using widely varying methods. Joe McCarthy instilled professionalism among his charges. Earl Weaver mixed and matched lineups using multiple platoon systems to maximize the number of men his teams put on base. Bobby Cox (with Leo Mazzone) conditioned pitchers to throw more often while exerting themselves less and developed remarkably healthy starters. It's crucial to study what such managers did and figure out how to apply their best practices.

As far as I can tell, one researcher has made serious headway on this subject: Chris Jaffe, who published Evaluating Baseball's Managers in 2010. Jaffe consulted a database compiled by sabermetrician Phil Birnbaum that estimated how players should have performed in each year since 1896. (Basically, Babe Ruth's expectation for 1930 is a weighted average of his actual numbers in 1928, 1929, 1931 and 1932.) Jaffe added up how much the players on each team in every season surpassed, or fell short of, their projections and credited the aggregate difference to the manager. He found that over time, managers' totals tend to grow bigger or smaller, rather than regress toward zero, which is what you'd expect if luck alone drove the results.

I draw three big lessons from Jaffe's work. For one thing, over the long haul, handling players is more important than strategy: According to his analysis, the best and worst managers at coaxing unexpected seasons out of players earned (or cost) their teams more runs over their careers than the best and worst at getting their teams to outplay their underlying statistics.

For another, MLB is about to say goodbye to the best in the game at getting players to overperform: Bruce Bochy, who will retire at the end of 2019. Jaffe rated Bochy 27th all time at extracting high-end years from players based on his tenure in San Diego, where players such as Ryan Klesko, Mark Loretta and Phil Nevin improved on his watch. That has accelerated with the Giants, where another slew of players in their 30s (Pat Burrell, Aubrey Huff, Marco Scutaro) surged, while Bochy was also developing Buster Posey and a flock of stud starters. Clearly, Bochy is good at making veterans comfortable and putting players in spots where they feel they can succeed. The man has won three titles because he's a great coach.

Finally, Jaffe's analysis is excellent, but his database is now a decade old and needs an update. Especially with so many managers getting fired and hired, we need to isolate the aspects of their work that actually matter and evaluate them by a stat like OPS or WAR -- before the job disappears altogether.