“Lance McCullers has not looked like himself lately.” It rings true this season, but this is not the first time we’ve heard this.
Throughout his career with the Houston Astros, McCullers has always had some streakiness in him. It’s a function of his play style - he attacks hitters aggressively, which has always yielded high strikeout and walk numbers.
As detailed in a piece by Ben Lindbergh of the Ringer, Lance McCullers makes a living avoiding the strike zone with heavy usage of his curveball, which has in turn made him a lethal strikeout pitcher, but also gets him into trouble.
His 10.42 career K/9 rate shows why McCullers is so fun to watch when he’s in the zone, but his 3.61 career BB/9 rate helps explain why McCuller’s production seems so volatile at times. It also offers an easy explanation for why McCullers average Major League start has gone less than six innings this far into his young career, as he consistently finds himself in high-count showdowns with opposing batters.
McCullers relies on his high powered curveball more than any other Major League starting pitcher relies on a secondary pitch - we’ve even seen McCullers buck tradition and turn the curveball into his #1, like when he famously turned to it 24 straight times against the Yankees to close out the AL Penant.
McCullers’ curveball has a nasty 18.1% career swing-and-miss rate, and has held batters to a paltry 45 wRC+ (league average is 100) since McCullers joined the Majors in 2015. McCullers has continued the cutting edge trend in 2018, as he has gone to his curveball on over half his pitches, but this season something is fundamentally different about the pitch. He is throwing it faster and with less spin than usual - and the results have been a mixed bag so far.
In 2018, Lance McCuller’s average curveball velocity has increased from a career average of 85.5 to 87.8. It could be because its early in the season, but for the first time in McCuller’s career his curveball is topping off in the lower 90’s. On the other hand, he is putting less spin on the ball, as the pitch is moving about half as much as it did last season. It seems the scheme is to make the curveball and fastball less distinguishable, and so far the pitch has yielded the same impressive swing and miss statistics as before.
On the other hand, batters are making contact with the pitch at the highest rate in McCuller’s career, and as a result opposing batters have a 116 wRC+ against the pitch. I don’t expect this to continue as this season progresses, but it may be worth monitoring - while I don’t believe the fundamental changes to McCuller’s curveball should be a longterm problem, it is possible that he could be over relying on the pitch’s usage.
If you are sabermetric savvy, you may have noticed that throughout Lance McCuller’s career he has routinely underperformed his peripherals. If you are unfamiliar with baseball analytics, a player’s “peripherals” are stats that are considered to be indicative of a player’s “true” talent level. As a player’s career progresses his numbers should gradually gravitate, or ‘regress’, to his ‘peripheral’ stats.
For example, Lance McCullers has a career ERA of 3.72, but his career xFIP, a stat that attempts to strip fielding, sequencing, and other ‘luck’ variables from the equation, is 3.19. Based on that, we might assume that 3.19 is a more accurate barometer of McCuller’s talent level than his 3.72 ERA, and going forward, we should expect his ERA to be closer than his xFIP than his cumulative career ERA tally to date. Yet we’ve been waiting for McCuller’s ‘traditional’ stats to suit with his ‘advanced’ stats for quiet a while so far, and so far we’ve been kept busy.
There are always exceptions to the rules, and when someone like McCullers has a history of under performing his peripherals throughout his career, it’s worth a closer examination.
One of the strangest thing that jumps out about Lance McCuller’s career numbers is his BABIP, or batting average on balls in play. The average ball put in play falls for a hit roughly 30% of the time, so pitchers with BABIP’s above .300 are traditionally considered to be getting ‘unlucky’, as pitchers have little control of what happens to a ball after it is hit in play. On the flip side, pitchers with a BABIP below the line are generally considered to be getting somewhat lucky, and their traditional statistics may be inflated as a result.
According to the laws of the BABIP gods, Lance McCuller’s career mark of .328 is abnormally high. One would expect this number to regress, but the opposite has held true - since his rookie season, McCullers has had a BABIP over .330 in every season, and has a .381 BABIP 27 innings into the 2018 season. Is this something the Astros should worry about?
It’s true that some pitchers can outperform the league average BABIP over the course of their career - Kershaw has a .271 career BABIP, something that has held up considerably over the seasons. On the other hand, it’s possible that a pitcher can have a true BABIP higher than the .300 baseline.
According to a baseball researcher that goes by the name “Pizza Cutter”, we can expect pitcher’s BABIP to stabilize to statistical reliability around 3,800 balls in play, and at that point the pitcher’s career BABIP is more predictive than the league average of .300. McCullers isn’t quiet there with just under 1,000 batted balls put into play, but it’s still a big enough sample size to take a closer look at.
There are several reasons that you cannot just take a pitcher’s BABIP as an obvious sign of regression - first of all, not all balls hit into play are created equally, and not all balls have the same chances to fall for a hit. Hard-hit balls and line-drives fall for hits far more often than other batted balls, and pitcher’s with a pension for giving up hard contact can have career BABIP’s well over the expected .300 line.
McCuller’s batted ball numbers do not pop off the page as being too concerning - his soft-medium-hard and line-drive rates are slightly better than league average, but also not elite.
The other indicator of a BABIP problem, pitch count advantage, is much more applicable to McCullers’s situation. Balls put in play in a batter’s count are far more likely to fall for hits than balls put in play on batter’s counts, which could be a problem for someone with control issues like McCullers.
McCullers has led off with a first pitch strike just 56.9% of the time throughout his career, while the average pitcher gets ahead in the pitch count in 59% of at-bats. This struggle with counts and control has been costly for McCullers, and will need to be addressed if he wants to become a truly elite starting pitcher.
Yet McCuller’s approach is partly by design - he avoids the strike zone at all costs on purpose, which yields him plenty of strikeout pitches. But if McCuller’s wants to have a long career as a starting pitcher, he may want to ease up on this approach - if he works to get ahead in counts and walk less batters, he’ll have better success on the mound, his BABIP should regress to a normal level, and he’ll finally be able to go deeper into games.
McCuller’s gun blazing, breaking ball heavy style has usually been more conducive to the bullpen, but McCullers is adamant that he sees himself as a long-term starter in this league, and I see him as one as well. But his current approach - while entertaining as hell - has become somewhat of a problem.
His curveball usage has become more predictable, particularly the second and third times through the lineups where he particularly struggles. Furthermore, as someone who has had an injury history as concerning as McCullers, it may be better for his long-term health to not beat the death out of his curveball.
It’s easy to look at McCuller’s peripherals and conclude that he is simply getting unlucky and is due to turn it around, but this is a lazy analysis. Advanced stats are more complex than that, and McCuller’s has yet to have the ‘regression breakthrough’ many have predicted over the years - and it seems we know why.
In the last ten years just two pitcher have had a seasons with walk-rates above 12% and an xFIP below 3.50 - Clayton Kershaw in 2009 and Lance McCullers in 2016. That’s rare company, and if McCullers keeps up his 12% walk-rate this season, he should not be expecting a repeat.
Overall, control needs to be the name of the game for McCullers. The addition of an effective changeup should help, and the changeup has looked better this year (it is down from a career 89 to 87.4), but the surest bet for McCullers is to work counts. If he can get his curveball control down, great - but if not, it might make sense to ease up on its use, both for his short-term production and long-term health.