Lots of sliders seemed like the obvious adjustment, but he’s doing something even more tried and true instead
Sandwiched around being sidelined with COVID for a couple of weeks, Twins starter Dylan Bundy has been mostly solid at the back of the rotation. After last Wednesday’s strong start against Detroit (5.2 IP, 1 ER, 6K), Bundy is carrying a 4.54 ERA over 33.2 innings pitched (7 starts) while averaging almost a strikeout per inning and limiting walks and home runs at near career-best levels.
No, that ERA mark does not look impressive in comparison to the overall league average mark of 3.92 (through Sunday, May 29). And Bundy had back-to-back clunkers against Tampa Bay and Baltimore (15 runs over 9.2 innings combined) before his injured list stint. But, outside of those two rocky outings, he’s only allowed two total runs over 24 innings in his other five starts and has given the Twins great chances to win more often than not. That’s about all that can be asked from the back of the rotation.
Of course, 33-plus innings is not a large sample, even if Bundy striking batters out and limiting walks and home runs are good signs. All of the mainstream ERA estimators look favorably on Bundy’s work so far:
- FIP: 3.92
- xFIP: 3.65
- SIERA: 3.69
- xERA: 2.86
The numbers above suggest Bundy has pitched at a league-average level and are reasons to believe what he’s delivered so far is not only real but that he’s pitched better than his ERA might suggest.
The Other Side of the Coin
Despite that data, Bundy makes for a tough case to truly buy into because of the low quality of his “stuff.” The velocity on all his pitches, which was already well below average, is down demonstrably this season. He’s averaging just 89.0 miles per hour on his fastballs (7th percentile across all of MLB), easily a career-worst, and about a mile-and-a-half below his average from last season.
In conjunction with the diminished velocity, Bundy’s raw spin rates are also significantly lower than last season. A strong four-seam fastball spin rate that helped the pitch play up is part of how he’s survived in recent seasons despite well below-average velocity. He still has above average spin on the four-seamer, but it’s grading in the 66th percentile instead of in the 80s as it had.
Fortunately, the lost velocity and spin have not changed the movement on Bundy’s pitches very much. His four-seamer is holding steady around 14 inches of vertical drop and 4.5 to 5 inches of horizontal break like last year. His slider is dropping about two inches less this season, but that brings it back in line with seasons past when it was clearly his best pitch. His sinker and curveball are almost identical to last year. If anything, the lost speed and spin have his changeup dropping about 4 inches more this season.
By Stuff+, a comprehensive metric developed by Eno Sarris and Max Bay for The Athletic that measures the quality of a pitcher’s pitches using their physical characteristics — like velocity, movement, spin, release points, and the differentials between all of those — and the outcomes they get, Bundy’s raw arsenal is about 15% below the major league average. That’s a major decline from last season when his stuff was more like 7% below average.
The run prevention stats combined with the underlying pitch characteristic data make Bundy a confounding case. In terms of the things we believe pitchers control, he’s looking strong. In terms of the eye test and data of what we know makes a pitcher hard to hit, things don’t look so good. So, how is he doing this?
The (Obvious) Adjustment That Wasn’t
When Bundy was signed back in November, he was viewed as a potential bounceback project for Wes Johnson to restore. Bundy had finished 9th in the AL Cy Young Voting for his strong 2020 campaign (3.02 ERA, 2.0 fWAR over 65.2 innings). Still, his performance fell off a cliff last season (6.06 ERA/5.51 FIP over 90.2 innings) thanks to a combination of gopher balls and injuries, including a season-ending shoulder injury.
Coming off the bad season, Bundy was not seen as the kind of front-line starter many Twins faithful were hoping for in free agency, but there were valid reasons to believe he’d perform better going forward than he did last season.
A clean bill of a health would make for a nice start. So would evening out some of the bad fortunes he had with allowing home runs (20 allowed vs. 16.7 expected per Statcast). If he could keep the ball in the yard, pitching in front of Byron Buxton and Max Kepler instead of the Angels’ near league-worst outfield defense (28th OAA in 2021) was yet another check in the pro column.
Lastly, Bundy was an obvious candidate to make the en vogue pitch mix tweak to feature his excellent slider (-7 Statcast run value, .138 BA allowed, 50% whiff rate in 2020) more frequently. Nearly every Twins blog post breaking down the addition of Bundy forecast that the Twins would bring him on board and set him loose on the Maeda plan of using his slider as his primary pitch. Despite that offering long being Bundy’s best-performing pitch by numerous metrics, he had been featuring his fastballs more than half the time.
So far, most of those thoughts have played out. His arm health has been good. He’s suppressing home runs at a more tolerable rate (1.34 home runs allowed per 9 innings). The Twins’ outfield defense is tied for tops in baseball with +9 OAA this season.
But, the popular idea of featuring his slider has not happened, as the yellow line in the chart below shows:
Bundy’s pitch mix this year is mostly the same as it was last season. There are some slight changes like a few more changeups and fewer curveballs, but the general profile is right in line with the recent past, and his four-seamer, sinker, and slider mix is nearly identical.
Pitch selection is heavily influenced by batter-handedness, so I also took a look at the breakdowns of Bundy’s pitch usage against right-handed and left-handed batters but did not find much in the way of change. He’s leaned a bit more onto his sinker-slider combo against same-sided bats now than he did last year, at the expense of four-seamers and curveballs, but not in a dramatic way. Against lefty-swingers, Bundy scraps his sinker and slider and attacks with four-seamers, changeups, and curveballs. He has made a noticeable shift to feature his four-seamer more often while still using a mix of changeups and curveballs as his go-to secondaries against lefties.
Perhaps part of the reason he has not leaned more heavily on his slider as expected is that it has surprisingly been his worst-performing pitch by Statcast run value so far this season (Remember, negative values are good for pitchers):
Despite those poor results in the early going, his slider remains the pitch that gives him the best chance to get a swing and a miss and it’s his most chosen putaway pitch (27.1% of his two-strike sliders result in a strikeout). Its 44.3% whiff rate is an improvement over last year’s 36.3%.
The Crafty Veteran
So, he’s mostly pitching the same way, but with diminished stuff. But, he’s getting better results. The ERA estimators do not suggest he’s been getting lucky, as I showed earlier. In fact, his .303 batting average allowed on balls in play is about 30 points worse than last season (and 10 points worse than his career average). He’s arguably had poorer luck than you would expect.
The big adjustment that was predicted has not happened. Instead, an adjustment he has made is one that’s exactly what you would want to see a veteran with declining stuff make. Take full advantage of the biggest weapon a pitcher has – the ball-strike count.
Among the 135 pitchers with at least 30 innings pitched this season, Bundy ranks 4th in first-pitch strike percentage. His 69.9% rate of first-pitch strikes is a career-high by a significant margin and 12.5 percentage points greater than last season. Not only is Bundy getting ahead, but he’s also doing so with all his pitches all around the zone. He’s not just pouring in first-pitch fastballs. In 0-0 counts, he’s used his four-seamer 40%, his curveball 21%, his slider 15%, his sinker 15%, and his changeup 9%. Then he follows up with another strike in 0-1 counts at the 4th-highest rate.
It is an inarguable fact that hitters perform far worse when they are behind in the count. This season, at the league level, batters are producing .418 wOBA when they’ve gotten ahead and .220 wOBA when they’ve fallen behind. For Bundy specifically, those numbers are .380 (batter ahead) and .236 (Bundy ahead). Given that, it’s advantageous that he’s gotten in positions to throw a career-high rate of his pitches with the count advantage (34.7%, ranked 15th-highest across MLB, and up from 29.3% last season).
So, he’s making it work with the tried and true method of getting ahead, staying ahead, and mixing his pitches, speeds, and locations. It’s a fine line to walk with the kind of stuff Bundy has right now and it requires exquisite execution to work. How well he can continue to execute over time is an open and valid question. Opposing teams will no doubt adjust and start attacking those early count pitches in the zone. What Bundy does to counter will be critical.
But, so far, so good. This approach has enabled him to be effective. Hopefully, that will continue while also buying time for Bundy to regain the feel for his slider and/or recover a tick or two of velocity as he gets farther away from the shoulder injury issues and the weather warms up. Either of those happening would add a little margin for error to this formula and give us stronger reasons to believe in Bundy’s performance.