When the Texas Rangers lured Kevin Millwood to the hitter-friendly confines of Ameriquest Field in December of this last year, it was heralded as the big ticket signing the Rangers desperately needed to help overhaul a pitching staff lacking in star power. It took a five year, $60 million contract to get him, but eventually Millwood agreed to pitch in the “American League Coors Field.” We're going to take a close look at Millwood and see if he stacks up to the hype and can make a real difference in Texas in 2006 and beyond.
After an injury-filled season in 2004 with the Philadelphia Phillies during which he posted a 4.85 ERA, the Cleveland Indians decided to take a gamble and sign Millwood to a one year deal for 2005. Boy, did that gamble pay off as Millwood had one of the best seasons of his career:
| G | GS | IP | W | L | SV | ERA | H/9 | HR/9 | BB/9 | K/9 |
|---|---|---|---|---|---|---|---|---|---|---|
| 30 | 30 | 192 | 9 | 11 | 0 | 2.86 | 8.5 | 0.9 | 2.4 | 6.8 |
Obviously, the Indians weren't willing to pony up the cash to keep Millwood beyond 2005, so for the second season in a row Millwood would be testing the open market. Fortunately for him, the Texas Rangers were willing to give him the kind of money every player dreams of. Obviously Texas is taking a rather large gamble that Millwood can have many more seasons like he did in 2005 as a Ranger. This doesn't look to be a good gamble on the surface though, since Millwood will be 31 years old for the 2006 season, and by the time his contract is up he will be 35. Maybe for the first few years of the contract Millwood can maintain that production, but the back end of the contract could be a different story.
Let's focus on the near-future though and try and project Millwood's first season in the Lone Star state. To get the ball rolling, let's take a look at each season of Millwood's career. Included are some predictive metrics such as DERA, DIPS%, BABIP, and GB/FB ratio:
| Year | Team | G | GS | IP | W | L | ERA | DERA | DIPS% | H/9 | HR/9 | BB/9 | K/9 | BABIP | GB/FB |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 1997 | ATL | 12 | 8 | 51.3 | 5 | 3 | 4.03 | 3.09 | 0.77 | 9.6 | 0.2 | 3.7 | 7.4 | 0.325 | 1.21 |
| 1998 | ATL | 31 | 29 | 174.3 | 17 | 8 | 4.08 | 3.41 | 0.84 | 9.0 | 0.9 | 2.9 | 8.4 | 0.304 | 1.08 |
| 1999 | ATL | 33 | 33 | 228 | 18 | 7 | 2.68 | 3.43 | 1.28 | 6.6 | 0.9 | 2.3 | 8.1 | 0.231 | 0.88 |
| 2000 | ATL | 36 | 35 | 212.7 | 10 | 13 | 4.66 | 3.86 | 0.83 | 9.0 | 1.1 | 2.6 | 7.1 | 0.285 | 0.71 |
| 2001 | ATL | 21 | 21 | 121 | 7 | 7 | 4.31 | 4.44 | 1.03 | 9.0 | 1.5 | 3.0 | 6.2 | 0.266 | 0.88 |
| 2002 | ATL | 35 | 34 | 217 | 18 | 8 | 3.24 | 3.16 | 0.98 | 7.7 | 0.7 | 2.7 | 7.4 | 0.264 | 1.22 |
| 2003 | PHI | 35 | 35 | 222 | 14 | 12 | 4.01 | 3.44 | 0.86 | 8.5 | 0.8 | 2.8 | 6.9 | 0.278 | 1.04 |
| 2004 | PHI | 25 | 25 | 141 | 9 | 6 | 4.85 | 3.49 | 0.72 | 9.9 | 0.9 | 3.3 | 8.0 | 0.321 | 1.10 |
| 2005 | CLE | 30 | 30 | 192 | 9 | 11 | 2.86 | 3.87 | 1.35 | 8.5 | 0.9 | 2.4 | 6.8 | 0.274 | 1.34 |
| Career | 258 | 250 | 1559.3 | 107 | 75 | 3.76 | 3.58 | 0.95 | 8.5 | 0.9 | 2.7 | 7.4 | 0.278 | 1.01 |
***DIPS% is DERA / Actual ERA. A number higher than one indicates that a pitcher's actual ERA was better than his DERA. A number lower than one indicates that his actual ERA was worse than his DERA.***
To start, let's take a look at the Millwood's K/9. Here is Millwood's K/9 progression chart:
![[Millwood's K/9]](sbimages/MillwoodKs.jpg)
You can see that Millwood has always been a little above average in this category, but last season he actually dropped below his career average (7.4) after posting a 8.0 K/9 in Philadelphia in 2004. If you take a look at his career stats above you will see that last season Millwood's K/9 was the lowest it has been since 2001 when he posted a 6.2 K/9 in an injury-shortened season in Atlanta. Yet, Millwood had the second best season of his career, how could this be? I believe his 2005 GB/FB ratio serves as an important clue. He set a career high with a 1.34 mark last season, and this came with a drop in his BABIP. Normally, groundball pitchers are more prone to giving up in play hits (although the difference is commonly small) not less, so Millwood possessed a rare combination of the benefits of ground balls (double plays and a lack of extra base hits), without the trade-off of a larger number of base hits (it's easier to make plays on fly balls than it is to field grounders and you don't run the risk of giving up infield hits to fast base runners).
Getting hitters to put the ball on the ground doesn't always equate to success for all pitchers, but for Millwood, it has generally helped. His ERAs in his two best seasons according to GB/FB ratio, 2005 and 2002, were 2.86 and 3.24 respectively. Whatever Millwood is doing, he seems he has the most success when getting hitters to put the ball on the ground. This career study also makes Millwood's career season (1999) very interesting and "out of place." Millwood posted his best ERA and most IP that season, yet his GB/FB ratio was very bad for him (0.88). Anyways, the point to be taken is that for the most part Millwood pitches best when letting hitters put the ball on the ground letting his defense make the out, as opposed to having to strike batters out, although perhaps he is smart enough to find success no matter how his pitches are working (in 1999 he compensated for more fly balls by dropping his walk rate).
The reason for the increased number of groundouts in 2005 was a change in Millwood's delivery last season. He threw more straight up and drove to the plate a little more effectively. It is well known that when you get on top of the ball and use your back to help with your drive the ball will sink more and have more weight, and observation of Millwood made it clear that this is what was going on. His two-seam fastball and sinker had much more late break and hitters had a lot more trouble elevating the ball. Obviously this worked for him, and because of that I will not be predicting any rebound in Millwood's K/9 in 2006. If getting groundouts in favor of strikeouts worked in 2005, why would he stop trying that same formula again in 2006? For the Rangers next season Millwood's K/9 will be 6.9.
Here's the problem though: Millwood's new delivery induces more groundouts, but his GB/FB ratio still isn't as high as one might prefer for a pitcher having to endure "The Bandbox in Arlington." Obviously, when pitching in Texas, you'd like to keep the ball on the ground as much as possible because when a ball is on the ground it cannot turn into a fluky homerun. Even if Millwood's GB/FB ratio continues to increase it will most likely still be below average (1.59), and some of the "extra" fly balls he will give up will become "fluky" HRs. Due to this I am projecting an increase in Millwood's HR/9 from his 2005 figure (0.9) to 1.3.
Now let's take look at Millwood's control (BB/9). This is an area of the game where Millwood has always been exceptional and his career BB/9 proves that. Here is his progression chart in this category:
![[Millwood's BB/9]](sbimages/MillwoodBBs.jpg)
Millwood posted his second-lowest BB/9 in 2005, and I have to believe this was a result of Millwood's new delivery as well. Obviously his new approach is not only helping him keep the ball down, but it's improving his mechanical consistency as well (many instructors preach "drop-and-drive" pitching for this very reason. It gives you mechanical clarity that improves control as well as limiting home runs and insolating you against injury). With this newfound control a projection of 2.5 BB/9 sounds just right for 2006.
To try and predict Millwood's 2006 H/9 number we will look at his BABIP allowed. Here is his progression chart:
![[Millwood's BABIP]](sbimages/MillwoodBABIP.jpg)
In 2005 Millwood's BABIP allowed (.274) was right around his career average (.278). Due to his new ball park and its' effects in hit rates though, an upward projection in this category seems appropriate. My first estimate of Millwood's BABIP is .287 (adjusted for park). Before projecting a final H/9 number though I must also account for two more factors: the defense behind Millwood, and the increased number of HRs he is going to give up since these count as hits as well.
In 2005 Millwood pitched in front of the third best team defense in the league (Cleveland) according to Defensive Efficiency. In 2006 he will be pitching in front of a Rangers defense that ranked 26th in baseball in the same statistic. Even though the team defense in Texas may improve with the subtraction of defensive nightmare Alfonso Soriano and the addition of Brad Wilkerson, it is sure not to be as good as the Indians defense was last season. Due to this, I'm going to bump Millwood's 2006 BABIP up to .292.
Now, when you combine that .292 BABIP with his projected 1.3 HR/9 number, you come up with a H/9 number that will be much higher than it was in 2005 (8.5) and much higher than it has been throughout his career (8.5). In 2006, Millwood's H/9 will be 9.6.
To project Millwood's 2006 ERA I will use his career DERA, and adjust it around the peripheral numbers I've already projected, and his tendency to either pitch better or worse than his DERA.
Millwood's career peripherals are as follows: 8.5 H/9–0.9 HR/9–2.7 BB/9–7.4 K/9.
These amount to a 3.58 career DERA against a 3.76 actual career ERA, showing a tendency to "under-pitch" his DERA. But, in 2005 with his new and improved delivery his DIPS% of 1.35 showed that the new delivery helped Millwood "out pitch" his DERA. This was the highest DIPS% that Millwood posted in his career, and this fact will be considered when projecting his actual ERA.
Anyways, his 2006 projected peripherals are: 9.6 H/9–1.3 HR/9–2.5 BB/9–6.9 K/9. These peripherals add up to a projected DERA in the area of 4.05 - 4.20. Using DERA by itself would not be the best way to approach projections as I've discussed in previous articles, so we have to try to account for his DIPS%.
As mentioned above, 2006 marked a huge jump in Millwood's DIPS%, and while some of that was probably luck, and some was the impact of an outstanding team defense, I believe the change is at least partially due to his new delivery. In order to make a projection for Millwood's DIPS% in 2006, I need to know something about how important the Indians' defense was in creating the career high last season. In 2006, the Indians as a team posted a DERA that was a full 14% higher than their actual ERA while the Rangers posted one 6% lower than their actual ERA. If we conservatively estimate that 20% of Millwood's DIPS% comes off right from the start that leaves us with a DIPS% 1.15 mark instead of a 1.35 as it was in 2005.
That 1.15 DIPS% would suggest that if nothing else changed, we would expect Millwood to post an ERA in the area of 3.55 - 3.65 in 2006, but it's also likely that some of the dip in his ERA last season was luck, so we're left with another upward adjustment. In 2006, Millwood's ERA will be near 3.78.
Based on career usage patterns and his place on the Rangers' staff, I estimate that Millwood will get 195 IP next season, and his line will look something like this.
3.78 ERA, 9.6 H/9–1.3 HR/9–2.5 BB/9–6.9 K/9.
It looks like Millwood is going to start his new career as a Ranger off right. Time will tell if, four years from now, the Rangers are still happy with their investment, but in the here and now, Millwood is a step in the right direction.
Thank you for reading. If you have any questions, comments, concerns, or suggestions, please do not hesitate to email me at frnkbndy@yahoo.com.
