Can The M’s Fix Charlie Furbush?
As Jeff Sullivan’s explained, Charlie Furbush is an intriguing talent – a better strikeout rate than new-Nat Edwin Jackson, a better than average contact rate, and a solidly above-average fastball velocity from the left side. Coupled with a deceptive delivery and three off-speed pitches (curve, slider, change) and you’d have a guy you’d feel comfortable entrusting the 4th or 5th rotation spot to. Or, you would, if it wasn’t for the home run problem.
Matthew Carruth speculated that Furbush allows more pulled contact than most pitchers; when hitters pull the ball, it tends to go a lot further. Is there anything in Furbush’s pitch mix or location that might explain why hitters square the ball up so well?
Well, no, not really. I don’t want to deceive you, reader: I don’t have any answers here. A closer look at the pitch FX data show a number of moderately interesting things, but I’m not going to sell them as explanations for Charlie’s elevated HR rate. What I take away from the following is pretty much what I believed going in, which is basically the dictionary definition of confirmation bias – so you tell me: what should the M’s do with Furbush?
First off, I’m using the new pitch FX cards from BrooksBaseball, which include manual categorization of every pitch in 2011 thanks to the tireless work of Harry Pavlidis and Lucas Apostoleris. We’ve known for a long time that the MLB Advanced Media/Pitch Fx pitch categorization algorithm is good, but has its flaws. We discussed this last April, but Furbush has a change-up that he throws fairly often that’s mislabeled a slider or a two-seam fastball. Not only does this give an improper view of Furbush’s arsenal, it also drags down his average fastball velocity. Thanks to the new data, we now see that Furbush used five pitches pretty often – each of them more accounted for more than 10% of his total pitches.
Looking at Furbush’s 2011, three things stand out following his arrival in Seattle. First, his HR rate went up significantly. It was fairly high in the Detroit system, but it was remarkably high with the M’s. The sample size is quite small, but hey, it stands out. His HR rate was 3.6% with Detroit and 3.3% with Toledo, but jumped to 4.7% with the M’s. Second, the M’s apparently altered his release point, moving it more towards the first-base bag. Take a look at this graph of his horizontal release point over the course of the 2011 season (up=more towards first base):
Theoretically, this might help him against lefties, but leave him vulnerable to righties; moving his release point further from straight-over-the-top and more towards LOOGY-Sidearmer should increase his platoon splits. Indeed, he gave up 10 HRs to righties with the M’s and only 1 to a lefty (to Robinson Cano). Third, his pitch mix changed markedly. With the Tigers, he mainly threw a four-seamer, with a curve and a slider as his breaking balls. He mixed in change-ups and two-seamers, but was, by and large, a three-pitch pitcher. He threw 34 two-seamers, in total, in a Tigers uniform. With the M’s, he almost immediately began throwing more sinkers/two-seamers – in the month of September, he threw 138 of them. With Detroit, he used the pitch 5-7% of the time, but with the M’s it was around 27%. This isn’t a bad idea in theory – a sinker/two-seamer generally gets ground-balls, and grounders don’t generally go for home runs. The downside is that it’s got the highest platoon splits of any pitch.
With samples this small, we can’t say *anything* definitively. 100% of Furbush’s HR problem could be terrible luck. However, the M’s made two moves that seem perfectly reasonable, but which might make Furbush more vulnerable to righties, and his FIP against righties was nearly 6. On the other hand, his xFIP improved over the course of the season as he trimmed his walk rate while maintaining a respectable K rate in a starting role. The two-seamer did get more grounders than his four-seamer, and his change-up (which he used more with the M’s as well) had a GB% just a step behind his sinker.
So, to sum up (this is probably more than you wanted to read about Charlie Furbush):
– He moved his release and changed his pitch mix upon joining the M’s. The release point and the use of a sinker hurt him against righties*, but using more – change-ups instead of sliders helped him a bit against righties.
– Right-handed hitters hit 10 HRs off of him in a vanishingly short sample.
– Furbush’s K:BB ratio with the Mariners was 41:15.
– Furbush’s fastball velocity dipped a bit at the end of the year, but still stood at 91-92 MPH – below the Sabathia/Price/Lester level, but similar to that of John Danks/Cole Hamels/Madison Bumgarner/Cliff Lee/CJ Wilson.
– The sample’s small, but Furbush was mighty tough against lefties – holding them to a 3.3 FIP/3.2 xFIP amidst a flurry of ground balls.
I said at the beginning that I don’t have an answer here. It’s not as simple as ditching the two-seamer or moving his release point. He gave up HRs like it was 1999 in Colorado, but his other peripherals improved. So: what would YOU do? If you’re the M’s pitching coach or a player development staffer, what would your recommendation be?
1: Change nothing and wait for his HR rate to regress to the mean?
2: Move the release point back to where it was in Detroit and limit two-seamers to righties; keep it against lefties, and go four-seamer/change against righties?
3: Move Furbush to the bullpen, ready to assume the situational lefty role if Hong-Chih Kuo falters/gets hurt?
4: Attempt to cut bait and trade Furbush to San Diego or another team that might be able to manage the kind of contact he yields?
As for me, I’m somewhere between 1 and 2. His numbers against righties will regress, but it’s doubtful that they should regress towards the overall population mean. A part of his struggles can be chalked up to his delivery/arsenal, too. I’m not certain that it’s the two-seamers that “caused” this problem, and it’s not clear that his change is good enough (at least right now) to be a primary weapon against righties. But I think it’s worth exploring, and I think Furbush is talented enough to warrant a second, third and fourth chance. What this means is that I wouldn’t have him as the long man/7th arm in the M’s bullpen – I’d have him start in Tacoma. That way he can work on a number of changes to release or pitch-mix in the only minor league that yields HRs at equivalent (or more than equivalent) rates to MLB. The M’s shouldn’t need a lefty long reliever, especially if they carry Rule 5 guy Lucas Luetge, and Furbush may hit upon a formula that allows him to maintain his more-than-solid K:BB ratio while limiting mistakes.
But what do you think? Was this all luck? Or is the only way he can avoid gopherballs nibbling at the corners, resulting in a slightly different form of bad pitching?
* – This is the most Furbush plate-appearance ever. In his last start, 9/25/11 in Arlington, Furbush started Ian Kinsler with a two-seamer away. Here’s a graph of the pitch’s location. Kinsler’s a righty. Take a wild guess as to where this pitch ended up. OK, got it? Here’s where it actually went. Yes, it was “only” 89.5 MPH, but it was reasonably close to the outside edge. And Kinsler pulled it 408 feet.