A Baseball Weblog

Friday, November 20, 2009

A look at Phil Hughes through PITCHf/x

Ask anybody who followed the Yankees in 2009, and they’d tell you that one of the most important decisions the Yankees made in 2009 was moving Phil Hughes to the bullpen. Hughes had begun the year as a starting pitcher at Triple A Scranton before getting called up by the Yankees in late April. He pitched well enough in his 8 starts, save for one terrible outing against Baltimore in May that put his ERA at 5.45 through the end of May. When the historically bad Chien-Ming Wang made his much anticipated return to the rotation on June 4th, the Yankees were left with this important decision of sending him back to Scranton, where he would’ve been a starter, or moving him to the bullpen, where he would attempt to shore up a relief corps that was not one of the Yankees’ strengths over the first two months. The Yankees decided to keep him on the active roster and make him a reliever, and it would be an understatement to say that that move paid off. All Hughes did was rack up 65 strikeouts and a 5.00 K/BB ratio over 51 1/3 relief innings to go along with a 1.40 ERA, 1.83 FIP, 0.86 WHIP, and .176 opposing batting average. He worked primarily out of a middle relief role in June before dethroning Brian Bruney in July as the Yankees’ primary setup man. For the final three months of the regular season, Hughes was about as good as it gets. He was the Yankees' most valuable reliever behind Mariano Rivera in 2009 and ranked 1st among major league relievers in ERA, FIP, and WHIP. Out of the bullpen, his velocity spiked and he missed more bats, which led to greater success. Let’s take a closer look at his 2009 numbers through PITCHf/x.

Phil Hughes - Starter
Pitch Data

PitchAverage Speed (mph)Max Speedpfx_xpfx_zspin_dir

Hughes primarily used a four-seam fastball, a cutter, and a curveball, mixing in a few two-seamers and changeups along the way (the two-seam clusters probably aren't perfect, since his two kinds of fastballs seemed to "blend together" in terms of spin movement). Here are how his pitches fared in the rotation.

Pitch Results

PitchPitch#Pitch%Swing%Whiff%Wide Zone%Chase%Watch%SLGCONRSvRSv/100


The high SLGCON and run value numbers show that Hughes wasn’t that dominant in his starts. Sifting through the numbers, something that struck me was the 17% whiff rate on fastballs, which is noticeably above the league average of 14%. A look at the strikeout breakdown below shows that he was about as comfortable picking up strikeouts on the fastball as he was on the curveball.

Strikeout Breakdown



Here are your graphs and tables from Hughes’ bullpen stint, which began on June 8th.

Phil Hughes - Reliever
Pitch Data

PitchAverage Speed (mph)Max Speedpfx_xpfx_zspin_angle

Pitch Results

PitchPitch#Pitch%Swing%Whiff%Wide Zone%Chase%Watch%SLGCONRSvRSv/100


Just about everything here is greatly improved. The velocity on the four-seamer and cutter was up about 2 mph on average, and the cutter had more horizontal and vertical break. The whiff rates for both pitches also rose drastically. Notice also that Hughes ditched the two-seamer and changeup out of the bullpen (the two two-seamers and one changeup picked up by PITCHf/x are from his long relief appearance against Boston in early June). What’s strange is that Hughes didn’t like the curveball that much out of the bullpen. Here’s the strikeout breakdown:

Strikeout Breakdown



That’s a lot of fastballs for strike three. His “K pitch” out of the bullpen was far and away the swinging strike on the four-seamer. In addition, Hughes began to throw more fastballs as the season wore on, as evidenced by this cumulative pitch selection chart.

So, what does all of this say about Hughes’ strange 2009 season? A switch to the bullpen turned him from a decent, low 90s fastball/curveball guy into a strikeout pitcher who loves his 95 mph heater. And what does it say about his future? The Yankees see him as a starter in the long run, so it would not surprise me to see some of these numbers (particularly the fastball velocity) decline. Ideally for Hughes, the successful bullpen experiment will have given him enough confidence in his arsenal to be able to consistently attack hitters as a starter.

Gameday PITCHf/x data is from MLB Advanced Media; it can be easily accessed via this tool. Other statistics are from Fangraphs and Baseball-Reference.

More reading: River Ave. Blues did a great piece last month on Hughes and his pitch selection.


  1. What happened to Hughes in the playoffs?

  2. He didn't do too well ... I haven't looked at the numbers, but it looked like he stopped attacking hitters as he was in the regular season --- trying to hit corners, "getting too fine" with the strikezone, etc. I can't say whether or not this caused him to fall behind in the count and thus get hit hard in "hitters' counts," but it seems like a reasonable theory.

  3. I'd be interested in a thorough analysis of his playoff performance. I'm assuming that the pressure/stress of performing in that situation caused some changes in a measurable aspect of his pitching. It would be interesting to find out what specifically gets affected with the nerves get the better of you!

  4. Yeah, I'd love to take a look at that within the coming weeks. I'll post a "part 2" to this and look at his postseason.

  5. By the way, I'm working on the follow-up and it should be ready sometime next week.