A lot of pitch f/x data was flying around after Marcum's second playoff start on Monday about his pitch selection, velocity, and results, but it's downright impossible to get any meaning from those numbers without some context. I went through and compiled some information to try to make some sense of it and make an informed decision about his status in the rotation. All data here is pulled from the excellent TexasLeaguers.com database. As usual, I combined the four and two-seam fastballs because I don't trust pitch identification algorithms and Marcum's velocity between the two are very similar. The numbers are close to exact but rounded for some simplicity. Also, pitch f/x isn't perfect so take a one or two game sample of velocity with some skepticism (though the swinging strike rates and pitch selection are going to be accurate for those two samples).
Start of season to September 8
Average fastball velocity: 86.8, 32.6% of the time, 3.7% swinging strike rate
Average cutter velocity: 83.9, 19% of the time, 10.3% swinging strike rate
Average changeup velocity: 79.2, 28% of the time, 19% swinging strike rate
Breaking balls: 20%
September 9 to October 13 (When things began to go downhill)
Average fastball velocity: 86.5, 37% of the time, 5% swinging strike rate
Average cutter velocity: 84.4, 30% of the time, 9.7% swinging strike rate
Average changeup velocity: 80.4, 16% of the time, 17% swinging strike rate
Breaking balls: 17%
9/20 against the Cubs (8 IP, 1 ER, 7 K: 0 BB) (The aberration in the downward trend)
Average fastball velocity: 86.2, 30% of the time, 3.5% swinging strike rate
Average cutter velocity: 83.3, 33% of the time, 25% swinging strike rate
Average changeup velocity: 80.0, 17.5% of the time, 41.2% swinging strike rate
2 Playoff Starts
Average fastball velocity: 87.0, 30% of the time, 4.3% swinging strike rate
Average cutter velocity: 84.9, 42% of the time, 8% swinging strike rate
Average changeup velocity: 81.2, 12% of the time, 0% swinging strike rate
The first thing that jumps out at me, as I think it would to anyone looking over these numbers, is the last line-- in a total of two playoff starts, Marcum has thrown 18 changeups. Batters have swung at 10 of those and made contact with all of them. 1 was fouled off, and the other 9 were put in play. There's clearly something wrong with Marcum's go-to pitch. In the first part of the regular season, batters swung and missed it 19% of the time. Swinging strike percentage is often the best proxy for "nastiness" of a pitch, so it's clear things have gone bad quickly.
The most obvious explanation, it seems to me, is that he's throwing a cutter more and a straight fastball less. His recent fastballs haven't really been outside the range of normal variation for him, they're in the upper 86 to lower 87 range. But it makes sense that if the changeup velocity rises a touch, coupled with throwing more 84 mph cutters, batters are just going to tee off when there's only a 3-mph variation between pitches they're seeing. He's clearly not keeping batters off-balance in the playoffs. Generally a gap of around 8 mph between fastball and changeup is considered to produce the best results. That was Marcum's gap in the early part of the regular season (87 to 79). When it's 87 to 81, or 84 to 81 if he's going to throw a bunch of cutters, it's not going to be anywhere near as effective, and that's been obvious in the results.
So is Shaun Marcum done? Should he be locked up and not throw another pitch in this series? I don't think so. If the team is considering bringing back Gallardo in a game 6 and back to Greinke on 3 days for a game 7, there's an argument to be made that Marcum shouldn't pitch. But if that's not on the table, I think he deserves a shot on an extremely short leash. His velocity hasn't really fallen off, and though it's very possible that his changeup's rising velocity could be linked to fatigue, that seems less plausible considering we haven't seen a drop in fastball speeds from him. I think he needs to sit down with the coaching staff and formulate a game plan, focusing more on fastballs and changeups and staying away from using the cutter too much when it's only a tick off his changeup right now. This looks like a different kind of issue to me than just purely being fatigued. I could easily see him having this stretch in June and arguing to just let him pitch through it.
With that being said, the stakes will most likely be very high in Game 6. If they determine that Gallardo's a go, I'll feel good about that decision. If not, I'd rather not see them panic and use Narveson right away. I'd probably throw Marcum, and basically have Narveson ready to go if he gives up, say, 3 baserunners in the first.
But the larger part of this picture is that if the Brewers are going to advance, and especially if they get to the World Series, they're going to need Marcum to be able to not suck to win. He's been on par with Gallardo and Greinke all year, and there seems to be hope he can get it back. Hopefully he proves capable of making the necessary adjustments.