When every team's W/L record is 0-0, hopes for contention spring eternal That's the nature of baseball in the spring. And perhaps part of the beauty of baseball. We can always point to cases when a team which was projected to to be a lower division team wins a division or gets into the playoffs. And, unless your team happens to be the Yankees or Red Sox, fans and sports writers are fond of pointing out which players have to come through with big years in order for their team to vault into contention. Naturally, those lists are much longer for teams which appear less talented on paper. Sometimes fans take heart in the spring training story lines, like "I'm in the best shape of my life," "I'm out to prove the naysayers wrong," and "I learned a new pitch."
Like a really cold shower, sometimes the projection systems will bring fans back to reality, or what purports to be reality (how do we define reality before reality has actually occurred? let's not get into that). DQ and I prepared a WAR projection which ened up showing a 79 win season for the Astros. DQ correctly pointed out that the projection doesn't mean that the Astros can't win more games than that. The individual player projections in our spreadsheet were based on our best effort at combining projection system forecasts and our own Astros-specific knowledge to arrive at a most probable results. But, as I pointed out in my comments on that article, the error bands around any such projections are large. And, it's really hard to project break out type seasons by individual players, which can and do occur.
I will try to use that same spreadsheet to identify what kinds of individual performances have to occur in order for the Astros to contend. This exercise assumes that 79 wins is the baseline, and I'm looking for reasonably optimistic performances above our projections which can raise that win total. Obviously, a million or so scenarios could be developed based on various combinations. And of course, it's possible that Brian Moehler could go all John Tudor Age 36 on us; however, I doubt that even his mother would forecast that result. Therefore, I will assume that a discrete set of players provide the most reasonable hope for the most significant contributions to added wins. I'm not really asking what kinds of events affecting other NL Central clubs have to happen. I'll just assume that 88-90 wins provides a reasonable shot at the NL Central, over 90 wins probably wins the division, and 86 wins could win the division.
Surprisingly, it wasn't that hard to craft a reasonable-appearing path to contention. Taken in isolation, each performance increase seems plausible and perhaps even probable (if you happen to be bullish on the player). But the probabilities of these events happening in combination is a different story. But I'll leave that for a day without rose colored glasses.
Wandy and Roy
The two players who may be capable of providing the biggest jump in wins are Roy Oswalt and Wandy Rodriguez. The projections for Roy and Wandy include a sizeable dose of regression to mean and/or expectations of age-related decline. Suppose Wandy just repeats his 2009 season ERA and innings pitched. That would add 3 wins to the Astros' record. In other words, Wandy being Wandy of 2009 could be enough to give the Astros a winning record. Suppose that Roy Oswalt bounces back and becomes the typical Roy of old? I used Roy's career average ERA (3.25) and 210 innings pitched (which is near average, and perhaps lower than "typical" for him). This also produced a 3 win addition. Thus, good seasons by Roy Oswalt and Wandy Rodriguez could put the Astros on the cusp of contention, with 85 wins.
When I initially reviewed our spreadsheet projections for Wandy and Roy (3.89 and 3.95 ERA, respectively), I felt that they were somewhat more pessimistic than what I view as a likely result. However, recognizing that some of our other projections on the pitching side might be optimistic, the Roy/Wandy results weren't so high as to cause me to change them. This sensitivity scenario for the two co-aces is reasonably feasible, though my view is that the most probable projections for Roy and Wandy would lie between our spreadsheet projection and the values assumed above.
Berkman and Lee
When Ed Wade is asked whether he is concerned about losing Tejada's offense, he responds by saying that Berkman and Lee have to increase their performance. He is careful to say that both Lee and Berkman had good production last year, but that players of their caliber expect more. If you asked the average fan what needs to happen to the Astros, I think most of them will point to Berkman and Oswalt rebounding. It's worth noting that our spreadsheet projections incorporate some increase in offensive performance (over 2009) for both Lee and Berkman. So, I looked for something which would represent a very good or superlative year or by both Berkman and Lee. I chose the best recent performance season by each player. These aren't quite career years, but they represent the best offense each player has put up in the last three years. This means utilizing the wOBA for Berkman's 2006 season and Lee's 2008 season. In combination, very good years by both players would increase wins by 3.2. So, to some extent, you may be able to understand Wade's thinking. If Wandy, Roy, Berkman, and Lee all have very good (but achievable) seasons, the Astros could be over the 88 win mark which would put them squarely in contention. Again, the probability of these events happening in the same year is a different story. I view this sensitivity scenario involving Berkman and Lee as somewhat less likely than the Wandy/Roy scenario.
Paulino
Some of us are hoping for better results from Felipe Paulino. By most indicators, he had bad luck last season and pitched better than his ERA. Our baseline includes some improvement for Paulino (4.60 ERA). Suppose that Paulino's pitches at the same level as his x-FIP in 2009 (4.05 FIP). That's hopeful, but perhaps not unreasonable if you believe that x-FIP is predictive of future results. This optimistic scenario also assumes a larger progression in innings pitched (180), compared to 145 IP in our baseline. This improvement is worth 1.5 more wins. How likely is this? I don't know. All the caveats about projecting young players apply here.
Pence
A lot of Astros' fan are waiting (or hoping) for the break out season by Hunter Pence. Pence is nearing the typical "peak age" when a break out could occur. Can he move from "good" to "very good"? In my mind, an OPS over .900 would be the mark of a break out. I assumed a .387 wOBA for this scenario, which is slightly higher than the wOBA for Pence's rookie season when he put up a .899 OPS. This scenario would produce 1.5 more wins. As I stated previously, projecting break outs by young players is hard to do. I'm not aware of any projection system which has been proven reliable for forecasting break outs by young players. (See, PECOTA, Re: Weiters, Matt)
Myers
Brett Myers is a tough guy to forecast. It's also not easy to identify what a rebound to "typical" performance might be (since his annual performance is up and down). Our baseline projects Bret Myers at a 4.60 ERA over 135 innings. For a No. 5 starter, that's not bad. But many Astros' fans are expecting a No. 3 starter. For a reasonable "optimistic" case, I used the Bill James projection (which happens to like Myers more than most forecasts) for both innings and ERA: 4.37 ERA, 171 IP. This adds 0.8 wins.
Bourn
The projections I have seen for Michael Bourn span a considerable range. The baseline wOBA of .327 falls in the general range of projections by CHONE, Marcel, and Bill James. This represents a substantial regression from Michael Bourn's actual wOBA of .342 in 2009. What if Bourn can sustain the same wOBA this season? That would add 0.6 wins to the Astros record. I think our baseline projection for Bourn may be slightly low, because Michael Bourn has the characteristics associated with achieving high BABIPs. But a wOBA repeat may not be as likely.
Let's Summarize
Added Wins From Higher Scenarios
Wandy +3
Roy +3
Lee and Berkman +3.2
Paulino +1.5
Pence +1.5
Myers +0.8
Bourn +0.6
Total +13.6
Astros Total: 93 Wins
So clearly I accomplished my goal of finding a path to a win total likely to take the division title. As I have alluded to, several times, the problem with this approach is that, while each individual scenario may not appear unreasonable, the probability of all the high-side scenarios coinciding in the same season is undoubtedly low. Moreover, we haven't taken into account the possibility that other players may not achieve their projection. So, I'm not performing this exercise in order to convince you that the Astros will win 93 games. However, this may give you an idea of the "catch lightning in a bottle" season which would be required to contend. And, perhaps, the impacts of the individual scenarios will give you an idea of the impact of the players' performance on the Astros W/L record.