Diving Into Dempster’s Season

In a season rife with unexpected twists and turns, Ryan Dempster’s transformation from sufficient reliever to rotation stalwart might have been the biggest surprise. Prior to 2008, Dempster had not made more than twenty starts since his last season in Cincinnati (2003), and the results were disastrous back then: A 6.54 ERA, 134 hits and 14 home runs surrendered in 115.2 innings pitched. Dempster’s lowest Fielding Independent ERA (FIP ERA) as a starter was 4.35, all the way back in 2000 as a member of the Florida Marlins. And, to top it all off, Dempster had posted a FIP ERA over four as a reliever for the Cubs over the 2006 and 2007 seasons combined, and failed to crack a 2/1 K/BB ratio in any of his four seasons in the Chicago bullpen between 2004 and 2007.

Despite the laundry list of reasons as to why “Dempster the starter” seemed like a dubious proposition, the Cubs went ahead with the plan, and were rewarded with one of the better pitching performances in the National League. In 206.2 innings, Dempster struck out 8.14 batters per nine innings, and kept his walks at an adequate level (3.31 BB/9), something he has struggled to do in the past (career BB/9 of 4.47). Dempster’s 2.46 K/BB ratio was the highest of his career, and was actually only the second time that he has surpassed that 2/1 mark (2000 being the other year).

However, his 2.96 ERA overstates his case by about a half a run, as Dempster’s FIP ERA was 3.41. That dichotomy is the product of a low home run/flyball rate (HR/FB%) of 7.7%. HR/FB rates tend to stabilize around 11-12% for pitchers, so expect that number to regress going forward. As that HR/FB rate ticks up, so will Dempster’s 0.61 HR/9.

Using Expected Fielding Independent ERA (XFIP) from The Hardball Times, we can get a more accurate read on Dempster’s performance. XFIP is perhaps the best way to gauge of a pitcher’s controllable skills, as it uses strikeouts, walks and an average HR/FB rate to calculate ERA. By doing so, XFIP eliminates lucky or unlucky performances on flyballs. So, Dempster’s auspicious HR/FB rate is regressed, giving us a better indication of his actual performance. If we adjust for the home run luck, Dempster’s XFIP ERA comes in at 3.94. That figure ranked 11th among NL starters.

Is Dempster’s 2008 showing a sign of things to come, or just an aberrant, blip-on-the-radar season? As we have seen, Dempster’s performance was quite good, but not sub-three ERA good. But can Dempster be expected to keep an ERA in the high three’s to low four’s during the course of his new four-year, $52 million contract?

Ordinarily, one would have every reason to be highly skeptical of a 31 year-old with a mild track record of success, a pitcher who last performed well in the rotation at the advent of the new millennium. However, Dempster does have some things working for him on a scouting level, as his repertoire played differently in the starting five than it did out of the bullpen.

Using Josh Kalk’s pitch F/X blog, we can compare Dempster’s stuff out of the ‘pen in 2007 to what he threw as a starter in 2008.

(X is horizontal movement. A negative X number means that the pitch is moving in toward a right-handed hitter, while a positive X means that the pitch is moving away from a righty hitter (in to a lefty). Z is vertical movement- the lower the Z number, the more the pitch “drops” in the strike zone.)

2007

Fastball: -7.21 X, 8.15 Z
Slider: 0.16 X, 0.81 Z
Changeup: -7.13 X, 5.71 Z

Dempster’s fastball had a good deal of tailing movement in on right-handers (-7.21 X), but his slider and changeup weren’t especially reliable. His slider did not register much horizontal break away from righties. His changeup did a nice job of mirroring his fastball in terms of horizontal movement, but the difference in vertical movement between the two pitches was just 2.44 inches. That’s not a whole lot, especially considering that his fastball (92 MPH) and change (83.1) had less than a 9 MPH dichotomy in velocity.

2008

Fastball: -6.72 X, 8.26 Z
Slider: 1.23 X, 0.08 Z
Changeup: -6.95 X, 3.75 Z

Dempster’s fastball registered similar horizontal and vertical break, but the quality of his secondary pitches increased. Dempster’s slider broke over an inch more away from righties this past season, while also showing a little more dropping action in the zone. Dempster’s changeup continued to mirror his heater in terms of horizontal movement, but the difference in vertical break between the two pitches improved. Dempster’s change dropped in the zone 4.51 inches more than his fastball did, almost doubling the difference from 2007. When commentators talk about a pitcher “pulling the string” on a changeup, this is what they are referring to. With similar horizontal break but a pronounced drop in vertical movement, hitters gear up for the fastball, get the changeup, and swing over top of the pitch.

Will Dempster’s improved secondary stuff allow him to keep most of the gains that he made during the 2008 season? The Marcel projection system seems to think so, forecasting Dempster for a 3.70 ERA, 7.77 K/9 and 3.58 BB/9 in 2009. It’s important to keep in mind that Dempster outperformed his peripherals this past season, and we would expect some regression in his ERA even if he repeated his strikeout and walk levels. If Dempster keeps his walk rate in check and continues to display a nasty slider/changeup combo, he could meet Marcel’s projection.