Sean Marshall Is a Rebel
Since a young age, pitchers have had it drilled into them that the best pitch they can throw is strike one. Hitters take a vastly different approach when behind in the count rather than ahead, and so the first pitch of an at-bat is almost always the most important one. Because they want to get ahead in the count, most pitchers throw first pitch fastballs, as it’s the pitch they are most confident that they can put in the strike zone.
Well, Sean Marshall is not most pitchers – at least not anymore. He followed the traditional first pitch fastball model for the first few years of his career, throwing it 64 percent of the time in 2006 and 56 percent of the time in 2007. He was moved to the bullpen in 2008, and dropped his first-pitch fastball usage down to 44 percent, though it was still the pitch he used most often.
The last two years, though, it has lost its prominence, thanks to the addition of the ever popular cutter. Last year, he threw a first-pitch fastball just 23 percent of the time, and this year it’s down to 18 percent, half as often as he throws a first-pitch curveball. He’s also more likely to throw a cutter (27 percent) or a slider (19 percent) on the first pitch of an at-bat.
Despite pitching backwards, Marshall has actually thrown more first pitch strikes than he did when he was featuring his fastball – 57 percent this year. He’s achieving the goal of getting ahead of hitters, but he’s just doing it with off-speed stuff.
Even when he falls behind 1-0, he still doesn’t throw the fastball. He bags the curveball for the most part, but throws a lot of cutters and sliders. Even 2-0, he only throws 26 percent fastballs, sticking with the two softer pitches he feels he can throw for strikes. He finally relents on 3-0, throwing the fastball 86 percent of the time, but given how often hitters have the take sign, he knows its not all that likely to be chased. On 3-1 counts, he rarely throws it.
Besides the 3-0 count, the two times Marshall is most likely to throw a fastball? 0-2 and 1-2, when he’ll throw it 41% and 42% of the time, respectively. He’s still more likely to throw his curve in those counts (he scraps the other two pitches), but he features the high fastball as a strikeout pitch, along with his big curve. Interestingly, however, when the count is 2-2 or 3-2, he again goes away from the fastball.
It’d be interesting to see if he would keep pitching like this if the Cubs moved him back to the rotation next year. He’s defying nearly every established pattern of pitching, and it’s worked, as he’s become one of the game’s best left-handed relievers. Could he pitch backwards for six or seven innings? I’d be curious to see if he could do it.
For lefties with pedestrian fastballs but good secondary stuff, this very well may be the most efficient style of pitching, at least while everyone else in the game is pumping fastballs to start off every at-bat.
Dave is the Managing Editor of FanGraphs.
I remember Greg Maddux once saying that the most important pitch of most PAs (for him, obviously) was the 1-1 pitch. That the difference between 2-1 and 1-2 drastically changed the way he’d approach the rest of the PA more so than any other account.
Just supporting the point that different pitchers will approach counts in different ways, and it’s not necessarily one-size-fits-all.
Great article, Dave. I had taken note of those splits, too. As a ubs fan, Marshall has been fun to watch this year. The tilt he gets on his curve is perhaps unique in the game today.
In 2004, a guy named Bill Felber wrote a book in which he noted that, aside from the increased likelihood of a walk, batters don’t actually get that much better as they get substantially ahead in the count. Felber postulated that most hitters hit best when they make contact on either the first or second pitch. I wonder if Marshall, by using his best pitches–his fastball has always been mostly for show–earlier in the count than most, has found a way to exploit Felber’s findings.
Awesome article. Pitcher’s game theory is fascinating.
I’ve been pretty upset with Cubs management for not giving Marshall more of a chance to start (not saying he’s not been given any chances, but he’s always gone back and forth between roles, and honestly I have more important reasons to be upset with Cubs management), but I’ve been really impressed by how he’s taken to full-time relief this year with little complaint and big results. That said, I hope he gets another chance at starting next year.
I’ve seen Marshall start enough to arrive at the belief that he’s probably worth the same as an 8th inning setup guy as he would be as a 3rd/4th starter, which is what he’d probably be limited to. Much as right-handed sinkerballers occasionally stink up the joint Just Because, soft tossing lefties are prone to the same predicament. This is especially true against power-driven, right-handed lineups, who can tee off against anything less than his best.
Marshall has been worth about 2 WAR this season, having spent one month as a middle man and four months as the 8th inning guy on a losing team. Imagine if the Cubs’ offense and starting staff could have been healthy and/or consistent: he and Marmol might be worth 6+ WAR rather than their current 4.5 total.
Marshall probably can’t close with his kind of stuff. In 59 career starts (most in his first 2 seasons, 2006 and 2007), he averaged 5.1 IP with a 1.43 WHIP and a 6.1 K/9. As a reliever, in 147 games, he has a 1.27 WHIP and a 9.3 K/9. We’re also looking at 0.66 HR/9 as a reliever, versus 1.30 HR/9 as a starter.
He has pitched 311 innings as a starter and 137.2 in relief over 5 seasons. Heading into his final year of arbitration, both the Cubs and Marshall might need to really think of his talents on par with other “true” setup men like Mike Adams, Matt Thornton and Ryan Madson.
This is exactly how J.P. Howell became a great reliever a couple years ago.