As a reporter, when your hometown baseball team wins the World Series there is a tremendous urge to write about it. When your beat is the compliance business, the challenge is finding a way to bridge two very different worlds.

Compliance Week, as many of you may know, is based in Boston, home of the 2018 World Champion Red Sox. As such, we have been bombarded with coverage and analysis of the team’s October run. It was within the inevitable critiques and comparisons of Red Sox manager Alex Cora and L.A. Dodgers manager Dave Roberts that our mind turned to, as it often does, matters of compliance. Specifically: analytics.

A big trend in baseball in recent years is the application of data analysis. Number-crunching has always been part of the game, with dutifully kept records of hits, batting average, runs, errors, strikeouts, and the like. Aided by technology, these calculations are now increasingly granular and gleaned for actionable insights. Which batter has the best chance against a specific pitcher, and vice versa? When should a defensive shift be utilized? Whom should the team turn to when bullpen relief is needed? Which batting order offers the best odds? Who should get the nod as a pinch hitter or runner?

A storyline of the recently completed World Series, at least among sports writers and pundits, concerns how Cora and Roberts diverged in their approach to pregame and in-game data analytics.

Post-SOX (pun intended) and since the Financial Crisis, companies are increasingly turning to new and emerging technologies to improve and automate their compliance functions. The goal is to move away from manual processes, like old-fashioned spreadsheets, and let “machines” do the grunt work.

Both teams have well-developed data operations. Both rely on number-crunching and statistical comparisons to formulate strategy and, ideally, maximize success. To a keen observer, however, Roberts stuck rigidly close to the guidance this data provided him. There was little divergence or improvisation from the mathematically derived playbook handed to him. Decisions that were very likely data-fueled—limiting star pitcher Walker Buehler to only his scheduled start with nary a hint of an inning or two of relief, benching first baseman Cody Bellinger for three of the five games against lefthanded starting pitchers—may have ultimately backfired.

Cora, in contrast, appears to have ingested the same sort of analysis, but added a human touch to it. What might have once been considered “hunches” on, for example, who to pinch-hit, proved to be data-supported moves that were supplemented with his knowledge of the psychology and motivational behaviors of his players. Those sorts of decisions, more spot-on than not throughout the series, may indeed have given the Red Sox their needed edge.

Back to compliance.

Post-SOX (pun intended) and since the Financial Crisis, companies are increasingly turning to new and emerging technologies to improve and automate their compliance functions. The goal is to move away from manual processes, like old-fashioned spreadsheets, and let “machines” do the grunt work. Artificial intelligence and machine learning are among the solutions that can help parse customer and supply chain data, separate the “wheat from the chaff,” and advance red flags or other matters requiring further action and inspection to the true brains of the outfit: the CCO and his or her staff.

Success lies in not blindly “trusting” technology with your success, but using it to supplement the decision making that well-trained humans still excel at.

People alone are not the answer. Technology alone won’t cut it. It is important to always maintain a hybrid approach to data analytics, one that weighs knowledge, experience—and yes, even hunches—into the process. From inception to implementation, all new technology investments must be treated as just a tool in the compliance toolbox. As amazing as some of these data-crunching tools are, they are not miracle solutions; there needs to be a skilled person “behind the curtain” who knows how to integrate all this valuable knowledge into their personal workflow. Artificial intelligence will always need flesh and blood intelligence to ultimately call the shots.

That was the difference between Cora and Roberts—and the former now has a trophy to hoist as proof of the approach. A similar approach is what can help make a CCO an organizational MVP.