In the midst of trying to put together the projected Wins Above Replacement (W.A.R.) for the 2009 Nats, a couple of questions have been tickling the back of my mind. The first is one that LOUtheMetFan brings up every so often: do those crazy W.A.R. calculations work? The other is the great unknown from last season: what if the Nats hadn't all broken?
Well, I decided to take the stats from the 2008 season and see what the trusty WAR spreadsheet said. 1). Given the individual playing time and performance of the 2008 roster, does the spreadsheet spit out 59 projected wins? 2). Given the actual performance of the 2008 roster, except with more playing time for the Opening Day starters, what would were our projected wins?
Long story short: 1). we should have won 63 games in 2008, meaning that either the WAR model is a bit optimistic, or that the Nats managed to underperform their injury-plagued roster's talent level (possibly a bit of both). 2). Ah, but if that roster had been healthy? A full season of NJ, The Chief, and SooperDooks against extra playing time for PLoD and Johnny Estrada (and no Flores) should have meant 70 wins.
Details after the jump.
(With apologies to Herman Wouk)
"Predicting" the past: 2008
One way to check a projection model is to see if it can "predict" the past. I'm actually cheating a little bit, since I'm not limiting myself to the data and projections that were available prior to the 2008 season. Instead, I'm taking the actual playing time and stats from the 2008 season, and seeing how close the W.A.R. math gets us to the actual record. In effect, I'm testing this hypothesis: "If your projections of playing time and playing performance are correct, then your projected record will be correct."
The 2008 roster: musical chairs
We all remember what a disaster 2008 was in terms of injuries and constant lineup and position shuffles. Here's a quick, position-by-position recap with # innings at each position (courtesy baseball-reference.com):
Catcher: Flores (673), Nieves (449), PLoD (161), Estrada (94), Montz (56)
1B: Boone (342), NJ (300), DY (290), Belliard (201), Casto (175), PLoD (118)
2B: FLop (622), Boney (325), Belliard (229), A-Hands (138), TAWH (86), Orr (31)
3B: Zimmy (910), Belliard (215), Boone (113), Casto (103), TAWH (44), Orr (37)
SS: Guz (1174), Gonzalez (93), FLop (54), Orr (43), Belliard (33), A-Hands (24), TAWH (13)
LF: TAWH (562), WMP (408), Langerhans (141), R-Mac (73), Dukes (57), Bernadina (56), Casto (53), FLop (49), PLoD (23)
CF: Lasto (1185), TAWH (131), Bernadina (108), Dukes (9)
RF: AK (734), Dukes (602), Langerhans (74)
Ouch! Remember the pain of PLoD at 1B and LF? Or the unexpected joy of TAWH at every position except 1B and C? By comparison, our rotation was steady as a rock! The main starters in 2008 (with innings pitched) were:
Tim Redding (182), John Lannan (182), Odalis Perez (159.2), Jason Bergmann (139.2), Collin Balester (80), Shawn Hill (63.1), Matt Chico (48).
Matt Chico? It's like another era. The bullpen was headed up by:
Jon Rauch (48.1), Saul Rivera (84), Joel Hanrahan (84.1), Luis Ayala (57.2), Jesus Colome (71), Steven Shell (50), Charlie Manning (42), Garret Mock (41)
Ah, who doesn't miss Rauch? And who can forget Ayala's many disastrous bridge collapses?
Okay, so to turn this into WAR, I filled out the spreadsheet by dividing each position's PA among the 3 or 4 (or 5... or 6...) players who played most at each position. I filled in the hitting stats from the 2008 season, with fielding adjustments from the 2009 CHONE projections (mainly because I didn't have anything else handy, and I don't know enough about defensive stats to figure out something myself). I split up IP to each pitcher by the innings they actually threw, filling in their 2008 ERAs and assigning any leftover innings to "Scrubs" with an ERA of 6 (to represent outings by Mike O'Connor or Levale Speigner or the like that I didn't include explicitly). I gave most of the PH to the scrubbiest of scrubs, players like Rob Mackowiak, Pete Orr, and Kory Casto. And the magic number?
REALLY? Well, that's not so far off. The Nats' expected "Pythagorean" win value (a Bill James invention based on the number of runs scored versus runs allowed) was 62 wins in 2008--so they're already unlucky compared to what you'd expect for the season. And the 63 wins is an average--about half of time they should win more, half the time they should win less. According to the numbers, they had a roughly 2/3's chance of winning more than 59 games.
Lessons for 2009?
There are two ways to look at it: 1). the WAR spreadsheet overestimated the 08 Nats' performance, even given exact individual stats or 2). the 2008 Nats managed to underperform even the injury-, has-been-, and newbie-riddled roster they were dealt. Frankly, I'd lean towards explanation #2, although #1 may have contributed a bit. I think I have a tendency to neglect the inevitable number of "scrub" PA and IP that crop up over the course of the season, even if your main starters and bench players are mostly healthy. We're going to see Bernadina or Mike O'Connor or someone else Not Ready for Prime Time this season, and I'm not sure how to fix that in the 2009 WAR spreadsheet. Perhaps I should give a chunk of the PH PA to a generic "scrub" with a 200/250/300 batting line, likewise with a chunk of IP to a "scrub" with a 6 ERA.
Happier Times: the semi-mythical Healthy 2008 Nats
Now for something a little more fun! What if the 2008 Nats had been mostly healthy? I know, we would have been saddled with a lot more Paul LoDuca. Still, remember how optimistic we all were after Zimmy's walk-off on Opening Day? Were we all high on crack, or was any of that warranted? Let's plug some numbers into the magic spreadsheet and have a guess.
A Healthy 2008 lineup
Here's how I'm going to play this: the opening day lineup gets the majority of the PA at each position, and I plug in their 2008 stats (as above). Here's how I see the playing time shaking out, in PA at each position:
C: PLoD (350), Estrada (150), Nieves (195): I assume Johnny Estrada's craptastic-ness becomes apparent by the AS break and that "Walk-off" Wil ends up sharing time behind the dish. Sorry, fans, healthy PLoD = No Jesus (BLAME!). At least he's not at first in this version.
1B: NJ (450), DY(150), Boone(95).
2B: FLop (300), Belliard (300), TAWH (95). FLop is still gone at the AS break (...slacker), but we don't trade for Bonifacio.
3B: Zimmy (590), Boone (50), TAWH (45)
SS: Guz (590), Gonzalez (50), TAWH (45)
LF: Dukes (450), WMP (150), TAWH (95). WMP stills gets his extended tryout before turning out to have been broken prior to 2008, but Dukes is Sooper! Let's just forget about Rob Mackowiak.
CF: Milledge (590), TAWH (105)
RF: Kearns (590), Ryan Langerhans (105)
I'm less comfortable fiddling with the rotation, since they were mostly healthy (Shawn Hill aside...), so I'll leave that as it was from above. But I am going to make a big change to the bullpen: The Chief is healthy, and we don't trade Rauch. Ayala still stinks, but the 7-8-9 is Saul/Rauch/Chief, so it doesn't matter as much.
Could we have been a contender?
Well, no. The spreadsheet says an average of 70 wins. Why not more? For one thing, most people's 2008 numbers were probably skewed downward by going on and off the DL and having to spend a bunch of PA figuring out that they were hurt or finding their timing after returning (AK being the prime example, and SooperDooks being the notable exception--at least, after overcoming his godawful opening slump). For another, PLoD and Estrada at catcher are like millstones around the neck of the offense. If we assume that JimBo has a revelation and never signs those two has-beens, letting Flores start at the big leagues with Nieves for relief, we pick up a bit over 2.5 wins. Likewise, if WMP never got a single PA, giving them to Dukes and TAWH instead, we pick up another 2.5 wins. Of course, this starts going from "everyone is healthy" to "everyone is healthy, plus JimBo's Segway is also a time machine" in terms of what-ifs.
Lessons for 2009?
The Nats' 2008 roster had the talent to get 75 wins, at least at the start of the season. Injuries and mismanagement (or so we can say with the benefit of hindsight) smacked that down by about a dozen wins, with bad luck probably adding 2 or 3 more. How does the 2009 roster compare? I think it's equal to or better than the 2008 "healthy" lineup in terms of position players, although the bullpen (especially) and rotation are shakier with a bunch of unproven or high-risk arms. But... if the Injury Fairy returns? The OF is solid enough, with 5 good players. The IF is a bit shakier--a mix of fragile veterans, feckless youths, and TAWH (when he's not in the OF). It could go all go south pretty damned fast (TAWH may be able to play every position, but not at the same time... at least not all at the same time).
That said, I remain cautiously optimistic: 70 wins for sure, with a decent shot at 75.