How many games can Wandy Rodriguez actually win?
It must be incredibly frustrating pitching for a team as bad as the Houston Astros.
After all, on most teams, if you're a starter responsible for 19.6 percent of your club's victories, you're looking at a minimum of 16 wins for a .500 team and as many as 20 if the team in question wins as often as the Philadelphia Phillies did in 2011. For Wandy Rodriguez, that share of the NL Central cellar-dwelling Astros yielded only 11 postgame celebrations.
Yes, last year's Astros finished just 56-106, tied for the fifth-worst record since 1990 and a full 40 games out of first place. Defensively, they were a mess, finishing 14th in the NL with a .981 fielding percentage. Offensively, they fared little better, scoring just 615 runs as a team last season, which translates to an average of just 3.80 runs per game.
That kind of production leaves little room for error from a starting pitcher, even from a consistent pitcher like Rodriguez, who has recorded an ERA of 3.40 since 2008. Last season, in 30 starts, he left the game with the lead only 12 times. Ten other times the Astros were tied or down by no more than two runs. Yet, Houston managed to go just 14-16 in Rodriguez's starts.
On paper, things appear to get no easier in 2012. For example, the pool of potential starting outfielders (Brian Bogusevic, Jordan Schafer, J.D. Martinez) has only 14 career home runs combined. So it should come as no surprise that even though Rodriguez is projected to have a 3.47 ERA, a 1.30 WHIP and an 8.06 K/9 rate, he's being selected as the No. 54 starting pitcher overall (188.4 average draft position) in ESPN live drafts.
Compare that ranking to that of someone with a relatively similar outlook like Ricky Romero (3.36 ERA, 1.26 WHIP, 7.82 K/9). Romero is being taken off the draft board at No. 22 among starters, with an ADP of 89.2, almost 10 full rounds earlier than Rodriguez. The difference is clearly the number of expected wins. Romero "should" win 15, while most pundits would predict that Rodriguez will be lucky to match last season's 11 victories.
But is that assessment fair? Let's assume that we're on board with Rodriguez's statistical projections for 2012. Maybe he's not going to pull off a Steve Carlton-esque 27 wins out of his hat pitching like that, but can't he do better than another .500 season?
One part of the equation is run support. Houston's offense is clearly far from elite. They have just three hitters ranked in the ESPN Top 300: Carlos Lee (No. 155), Jose Altuve (No. 237) and Martinez (No. 260). Run production might well be an issue for this lineup. But how many runs can we expect it to actually score on a regular basis?
To answer this question, we'll use a version of a base runs formula created by David Smyth that estimates the number of runs a lineup is expected to produce: expected runs = A*B/(B + C) + D, where A is the number of runners that reach base, B is a factor of baserunner movement (extra-base hits, steals, walks, double plays), C is outs and D is home runs (which guarantee a run is scored).
We'll spare you the details of the complicated math, but using the stats listed above, which stem from the sum of the ESPN projections for all Houston hitters, we get a result of 687.7 expected runs. That translates to 4.24 runs per game over the course of a 162-game season. To put that in perspective, last season the league average was 4.28 runs per game. So, while still saying that the Astros' offense will be less than average, at a fewer than half a run per game increase over 2011, it's a reasonable number to use going forward in our hypothetical discussion.
Now, given that amount of run support and combined with Rodriguez' 3.47 expected ERA, let's use two different sabermetric methods to estimate how many wins the pitcher "should" get in 2012.
First, we'll use a formula created by Cyril Morong: expected win percentage = .517 + .0899*RS - .0958*RA, where RS is run support and RA is runs allowed. (Keep in mind that we're talking about all runs, not just earned ones, so the RA is going to be slightly higher than Rodriguez's projected ERA.) Of course, there are always going to be outliers, but Morong's research has found that, generally speaking, starting pitchers will end up falling within 1-2 wins of this expectation. Inserting our data so far yields the following result:
Let's take a second pass at this, using Bill James' formula for predicting a team's winning percentage and applying it solely to games where Rodriguez is the starting pitcher. That formula is determined by dividing the square of a team's run scored by that same number after adding the sum of the square of a team's runs allowed to it. In our present example:
Of course, the flaw in thinking that this means Rodriguez should be expected to contend for the major league lead in wins is the fact that these win totals assume a complete game out of him each time he takes to the mound. Considering that in 197 career starts, he's managed only two complete games and not one since 2009, that's an assumption that we're unwilling to make.
So, the bullpen is going to have to get involved in this discussion. Mark Melancon, who secured saves in 20 of 25 opportunities last season, was traded to the Boston Red Sox. Veteran Brett Myers, who last closed games in 2007 for the Philadelphia Phillies, will handle those duties this season. Wilton Lopez and David Carpenter are likely candidates for the late-inning setup jobs. All three have projected ERAs higher than that of Rodriguez, which will result in a slight decrease in the expected win total in games started by Rodriguez.
If we were to use the same formulas as we did above to try to figure out how many games Rodriguez is "expected" to leave with a lead in place we would come up with approximately 15 appearances. If the Astros win only 78.8 percent of those games -- which is how often they won when leading after the seventh in 2011 -- that leaves only 12 victories for him. However, if the Astros hold on to leads at a rate approaching the league average (90.1 percent) in those situations, we're looking at closer to 14 wins.
Of course, there may also be a game or two in which Rodriguez leaves the game tied or behind and the Astros come back to win while he is still the pitcher of record. Even as bad as their winning percentage was last season, the Astros still won 16 games where they were not in the lead after seven innings last season. In fact, two of Rodriguez's wins last season came in games where the Astros took the lead after his departure.
So, let's answer the question. How many wins can we expect Rodriguez to get this year? It's certainly more than 11. Even if it's not going to contend for the playoffs, Houston is very likely to win more games than it did last season as a whole. With a little extra run support, Rodriguez is expected to leave with a lead in place 15 times (up from 12 last year). As no bullpen is perfect, he's bound to lose a couple of those, but he may also be the beneficiary of one or two comeback victories along the way. The upshot likely will be close to a wash.
Quite frankly, a range of 14-16 wins seems quite reasonable to expect this season. When you throw in the X factor of a potential trade to a contending team -- and the probable resulting increase in run support and, ultimately, victories -- as the trade deadline approaches, then the chances of Rodriguez reaching that range would seem to be even more of a legitimate likelihood.
But even if he remains in Houston for the entire season, chances are still pretty good that he'll post a win-loss total that, along with his other solid stat line, would make him look an awful lot like a guy who deserves to be drafted in the top 30 starting pitchers.
If you like the guy wearing that Astros uniform, then pay no attention to the team name that is on it.
Follow AJ Mass on Twitter: @AJMass