Monday, February 1, 2010

2010 Baltimore Birds Nest Projection, Part I

Hard as it may be to believe, spring training is barely two weeks away. With that in mind, let’s take a look at how the Orioles roster is shaping up for 2010.

By my count, 23 of the 25 roster spots are spoken for. We can assume the Orioles will keep 13 position players and 12 pitchers. First, the position players:

CA: Matt Wieters, Craig Tatum
IF: Brian Roberts, Cesar Izturis, Miguel Tejada, Garrett Atkins, Ty Wigginton, Robert Andino
OF: Adam Jones, Nick Markakis, Nolan Reimold, Felix Pie
DH: Luke Scott

I’m giving the 13th spot to Robert Andino since he is the only back-up available to play shortstop and second base (except maybe Justin Turner), meaning that both Michael Aubrey and Luis Montanez are likely off the roster. At this point, I think it is once again fair to scratch our heads at the Garrett Atkins signing. Josh Bell, Brandon Snyder and Justin Turner are dark horses to make the team out of spring training, but each could make his Baltimore debut sometime in 2010.

On to the pitchers:

SP: Kevin Millwood, Jeremy Guthrie, Brad Bergesen, Brian Matusz, Chris Tillman
RP: Mike Gonzalez, Koji Uehara, Jim Johnson, Cla Meredith, Mark Hendrickson

Two relievers will come from this group: David Hernandez, Alberto Castillo, Matt Albers, Jason Berken, Kam Mickolio, Armando Gabino, and Wilfrido Perez. Waiting in the wings, the Orioles have Jake Arrietta, Brandon Erbe and Troy Patton who, while currently starting pitchers, could easily fill bullpen roles if no spots open up in the rotation.

In 2009, the Orioles allowed 876 (rank: 14, league average: 771) runs while scoring 741 (rank: 11, league average: 781) runs for a Pythagorean record of 69-93, five games better than their actual record of 64-98. Thanks to Dave Cameron, I know we can’t stop there. According to WAR, the Orioles received 15.8 WAR from their offense (including pitcher hitting) and 7.5 WAR from their pitchers. Since replacement level in 2009 was about 46 wins, the Orioles are actually about a 69 win team according to WAR. Thus, since the two methods agree, I’m very comfortable working with a baseline performance of 69 wins from the 2009 Orioles.

Thus, in order to make a run at .500 in 2010, the Orioles need to improve by about 120 or 130 runs. The fastest way for this team to make up those runs will be to improve the pitching staff, and my early analysis shows that they will do just that. Last season, Orioles starting pitchers threw 877.2 innings, compiling a cringe inducing 5.37 ERA and 1.52 WHIP. According to FIP, that ERA was no fluke either. The starters allowed 154 homers, walked 299 and struck out just 533; that’s a FIP of 5.38. Fortunately, the starting pitching projects to be much improved for 2010.

(Note: For each of the pitchers below, I’m taking a weighted average of three readily available projection systems: Bill James, PECOTA, and CHONE, except for Brian Matusz, who does not have a Bill James projection.)

Kevin Millwood: 174.0 IP, 194.7 H, 21.0 HR, 58.3 BB, 5.0 HBP, 113.3 K, 4.56 FIP
Jeremy Guthrie: 192.7 IP, 200.3 H, 27.0 HR, 62.0 BB, 5.7 HBP, 115.7 K, 4.87 FIP
Brad Bergesen: 148.7 IP, 164.7 H, 17.7 HR, 42.7 BB, 8 HBP, 74.3 K, 4.77 FIP
Brian Matusz: 109.5 IP, 117.5 H, 15.0 HR, 42.0 BB, 1.0 HBP, 85.0 K, 4.61 FIP
Chris Tillman: 137.0 IP, 143.7 H, 19.3 HR, 58.7 BB, 8.3 HBP, 112.0 K, 4.87 FIP
Combined: 761.8 IP, 820.8 H, 100.0 HR, 263.6 BB, 28.0 HBP, 500.3 K, 4.74 FIP

Now, normally a team will need about 975 innings out of its starting pitchers (162 * 6 IP = 972 innings), meaning the Orioles have about 200 innings unaccounted for in the above projections. While it would be awfully nice to give 75 of those to Brian Matusz (Matusz threw 157.2 including the minors last season) and Chris Tillman (161.2 combined), let’s instead be conservative and give those 200 innings a replacement level-ish 250 hits, 30 homers, 100 walks, and 130 strikeouts; that works out to a FIP of 5.35. Our new combined starting pitcher line: 961.8 IP, 1070.8 H, 130.0 HR, 363.7 BB, 28.0 HBP, 630.3 K, 4.87 FIP. Some quick subtraction shows that this year’s staff is 0.50 runs per nine innings better than last year’s version (5.38 – 4.87 = 0.51). That’s a savings of about 50 runs over last season. You’ll quickly notice that a 0.10 difference in FIP equates to 10 runs, or one win; keep that in mind when we start to do a bit of sensitivity analysis around these projections.

Moving on to the bullpen, the 2009 ‘pen threw 550.1 innings, posting a 4.83 ERA and 1.53 WHIP (613 H, 64 HR, 232 BB, 400 K, 4.49 FIP). Now, FIP is a bit less reliable for relievers, but it’s the best I can offer; plus, there doesn’t seem to be any clear bias in the data one way or the other. Let’s look at our bullpen projections for this season.

Mike Gonzalez: 60.7 IP, 50.0 H, 6.0 HR, 25.0 BB, 2.3 HBP, 64.0 K, 3.73 FIP
Koji Uehara: 58.3 IP, 60.3 H, 7.0 HR, 15.0 BB, 0.0 HBP, 44.7 K, 4.00 FIP
Jim Johnson: 63.7 IP, 68.7 H, 7.3 HR, 24.7 BB, 2.0 HBP, 41.3 K, 4.66 FIP
Cla Meredith: 65.7 IP, 73.7 H, 6.7 HR, 20.7 BB, 1.7 HBP, 41.3 K, 4.28 FIP
Combined: 248.3 IP, 252.7 H, 27.0 HR, 85.3 BB, 6.0 HBP, 191.3 K, 4.18 FIP

(Note: I'm not using the Mark Hendrickson projections since they seem to be giving him waaay too many innings. I'm just going to roll him into the replacement level below.)

To play a full season, a team needs 1,450 innings from its pitching staff. We’ve now projected 1,210.1 between the starting rotation and the top four relievers, meaning we need 240 more innings. We’ll give those a FIP of 5.00 (240.0 IP, 275 H, 30 HR, 120 BB, 160 K) since we would expect replacement level pitchers in the bullpen to be just a bit better than our replacement pitchers in the rotation (mainly because many of those replacement starters are going to see some time in the bullpen, and they're numbers out of the 'pen should be better than their numbers as starters). The combined relief pitcher line: 488.3 IP, 527.7 H, 57.0 HR, 205.3 BB, 6.0 HBP, 160.0 K, 4.58 FIP. Unfortunately, that’s right about the same level as last year’s bullpen.

Now, that does not mean this year’s ‘pen won’t actually be better than last year’s version. Mike Gonzalez should be a pretty fair approximation for the George Sherrill that was so impressive for the first four months of the season, and I’m very optimistic that Koji Uehara will take nicely to a bullpen role. If that happens, Dave Trembley should be able to use Johnson and Meredith in situations where they face better match-ups, and likewise for the remainder of the bullpen. In addition, last year’s pen actually allowed a bit more runs than their FIP would suggest. I’m comfortable projecting a 0.2 runs per nine innings improvement for the bullpen this year, or about 11 runs better than last season.

Adding up the rotation and the bullpen, here is the combined projected pitching line for 2010: 1,450.2 IP, 1,598.5 H, 187.0 HR, 569.0 BB, 34.0 HBP, 981.7 K, 4.77 FIP. That works out to about 775 earned runs allowed for the Orioles. Add in 35 unearned runs, and I project the club to allow a total of 810 runs for 2010. In other words, the Orioles figure to improve by about six wins on the mound in 2010.

Up next: Part II (The Offense).

No comments: