The Cubs’ Pitchers Are Making Their Own Luck

Midseason is rapidly approaching, and the Chicago Cubs are still making the rest of Major League Baseball look bad. They’re World Series favorites by a mile according to our Elo predictions — which also have them pegged for 104 wins — and they’ve outscored their opponents by an average of 2.3 runs per game, the most of any team through 70 games since the legendary 1939 Yankees.

The Cubs have excelled on offense, scoring the third-most runs in the majors, but to an even greater degree the team owes its extreme success to run prevention. Chicago’s current 2.73 team ERA would be the lowest full-season¹ figure in the designated hitter era (since 1973), and the lowest relative to the MLB average since World War II.

This Cubs staff is pretty good at making guys miss — it ranks fifth in the majors in K-BB rate — but that alone isn’t enough to explain such a microscopic ERA. Chicago has also allowed a .254 batting average on balls in play, 42 points below the major-league average and 19 points lower than the next-closest team. If the Cubs were to finish the year that far below the norm, their BABIP would also be the lowest relative to average since World War II. As a byproduct, the gap between Chicago’s ERA and its Fielding Independent Pitching (FIP) is 0.64 runs, the widest positive gap in the majors.

We know of three things that could contribute to such a separation. One is flat-out luck, but the others are a good defense and a pitching staff that induces especially fieldable batted balls. Prior to the advent of Statcast, MLB’s new radar-based motion-tracking system, it was almost impossible to separate the latter two elements, parsing out a pitcher’s impact on batted balls from that of his fellow defenders. But now we can start to unravel those relationships and assign partial credit to each possible factor at work.

To do that, we built a couple of models. The first model estimated each MLB pitcher’s effect on the exit velocities and launch angles he allows by comparing his rates to the same hitters’ numbers against all other pitchers.² Then we calculated what would happen if we replaced each team’s actual pitchers with a staff full of generic arms that allowed league-average exit velocities and launch angles. The difference between those actual and generic figures gives us a number of runs attributable to each pitching staff’s contact-management skills, i.e., its tendency to allow batted balls that do less damage.

Next, we modeled fielding on a team-by-team basis by estimating how much each batted ball “should” have been worth (in terms of linear-weight run value) based on its exit velocity and launch angle.³ Then we compared those estimated values to the actual values of the same batted balls.⁴ If a batted ball with an exit velocity and launch angle that would typically produce a single actually yielded an out, the model credited some of the difference to the defense, which we assume prevented the single through some combination of good range, good hands and good positioning.

Finally, we combined those two values into one total figure to see how many runs each team has saved on its balls in play, relative to a team with average contact-management and defense.

There’s a moderate, statistically significant relationship⁵ between a team’s ERA-FIP gap and our estimate of its runs saved from contact management and defense. Add in sequencing (as measured by Left on Base Percentage), and we can explain about 60 percent of the difference between a team’s ERA and its FIP. The rest can be chalked up to random variation, plus a variety of smaller factors⁶ and, admittedly, other unknown elements that we can’t conceive of or are unable to calculate using current data.

According to our models, the Cubs’ defense — aided perhaps by data-driven positioning, if not frequent infield shifting — has been the third-best in baseball, behind the Rangers and Blue Jays. But fielding is a relatively small piece of Chicago’s run-preventing puzzle. Its pitching staff’s collective ability to manage contact leads the next-best team by close to 20 runs. As a group, Cubs pitchers have depressed exit velocity by 0.4 miles per hour and launch angle by almost 2 degrees, relative to average.⁷

That leads to a larger takeaway from our models: Leaguewide, the impact of pitchers’ contact management is more than twice that of defense, which seems to contradict the traditional defense-independent pitching theory that most pitchers have little ability to prevent hits on balls in play. (It’s probably no coincidence that the career leader in Inside Edge’s Soft Contact rate is fabled bat breaker Mariano Rivera.) In other words, much of what appears to be good or bad defense might really be good or bad contact management, which can produce easier (or more difficult) fielding opportunities that make certain fielders look better or worse than they are. In theory, only a Statcast-derived defensive stat could account for this heretofore-camouflaged effect.

Exit velocity is meaningful even over small samples, but at this early stage of the Statcast Era, we still don’t know enough about how pitchers control contact to say whether the Cubs’ BABIP is sustainable, or if it stems from a conscious pitching (or even pitching-acquisition) approach. As with any extreme observation, it seems safe to expect some regression to the mean for Chicago’s pitchers. Still, we can conclude that the Cubs’ historically low BABIP through their first 69 games isn’t merely luck. One way or another, the Cubs have earned a lot of those outs.