Max Scherzer - Starter or Closer?

As we wrap up this painful Diamondbacks season, we're left to look ahead to the future to try to make the best roster for ourselves and our future, and one of the most prevalent and talked-about debates is what to do with Max Scherzer. We all know of his killer heat, but he has very fringy secondary-stuff, with both his slider and change-up being works-in-progress. Everyone knows he has closer's stuff, but the key facet of the debate has been whether better production from a reliever for a reduced amount of innings is more valuable then an increased amount of innings from a starter, but with slightly less-impressive overall results. So, to try to settle this debate, I've decided to pool together the first inning of each of Max's starts to try to get a very rough estimate of how Max does when facing three batters rather than having to go through a lineup multiple times and seeing if any trends appeared that allowed for a decision to be a little clearer.

The result of this is the following table:

*Sacrifice Fly totals were not gathered per inning and are not factored into BABIP, as I do have some sort of social life to maintain. :-P

**Both first-inning HRs against Rockies @ Coors Field - take what you want from this.

The first thing to obviously take note of is that the first-inning totals are with Scherzer having taken a greater beating over the course of the year with a much higher innings total than he would have incurred as a reliever, so one would think that some of his starts later in the year would have poorer outings for a guy just getting his innings up to a rotation-starter level as he wore down. However, his worst month of the season for first-inning performance was actually May, the second month of the season, in which he allowed six runs (five earned) on 12 hits in the respective first innings of six starts, for a 7.50 first-inning ERA (to be referred to as 1ERA). In April, he had a 0.00 1ERA, a 1.80 1ERA in June, a 0.00 1ERA in July, 4.50 1ERA in August, and a 3.60 1ERA in September. In that dreaded May, he gave up three runs (two earned) in the first against the Dodgers, two earned in the first against the Braves, and another first-inning earned run to the Marlins. Perhaps a slight explanation for the somewhat surprising first-inning suckage of May is the fact that he sported a BABIP of .500 for the first-innings of those six starts, a significant factor in his 1BABIP being 21 points higher than his BABIP not including the first-innings of each of his starts. Given the overall small number of starts for 1ERA to be affected, the fact that he had such a significant streak of bad luck resulting in a very bad month of first-innings had a significant impact on the overall outlook of his numbers. And even with this going against him, Scherzer's 1ERA+ of 131 is lower than all but, surprisingly enough, Blaine Boyer's 159 ERA+ (since his arrival in Arizona, overall on the season Boyer's ERA+ stands at 104) amongst the D-Backs' bullpen, and significantly better than closer Chad Qualls' 119 ERA+, Clay Zavada's 127 ERA+, and "replacement closer" Juan Gutierrez's 105 ERA+. Given these numbers alone, Scherzer would instantly fill in the role of closer for the D-Backs, even ahead of Qualls.

The next point that I feel is extremely important to face is who it is that Scherzer faces in the first inning - namely, the top of the order for the other team. This means that Scherzer is forced to go up against three or four of the other team's best hitters, rather than a fairly more random set of batters to face. If one is to trust a fairly rudimentary way of measuring the average position in the lineup faced by the closer in the ninth inning, the MLB-average WHIP for 2009 is 1.390. This, multiplied by 8 innings, means that, on average, 11.12 batters, reduced to 11 for practicality, either record hits or are walked in 8 innings, while collecting 24 outs. So that means that 35 batters, on average, have come up to hit before the closer come into the game. This means that the closer will face the 36th batter of the game in this average situation, which corresponds to the ninth-hitter. So, Scherzer, as a closer, would, on average, likely face the 9th-12th position player on an NL team (assuming a pinch-hitter is used for the pitcher, who is assumed to be batting ninth, and depending on what pinch-hitters have already been used by the team), or the 9th batter on an AL team, who Max is more likely to get out than any other hitter in the lineup. In the course of a 1-2-3 inning for Max, this means he does not have to face the 3-hitter, typically the best hitter in a lineup, and if one batter does not make an out, Max still does not have to face the cleanup-hitter, a position in the lineup that had an OPS split of 1.101 against Max for 2009, the only position over 1.000. The result of this would mean that Max would generally have an easier set of batters to retire as a closer than he does in the first inning as a starter, which would lead to an improvement on his current 1ERA numbers, further indicating that he could be successful as a closer.

So here's the question that we ask - are the extra 140 not first-inning innings that Scherzer gives us as a starter, in which he has compiled an ERA+ of 101 in 2009, worth the possibility of him being a closer over 60 or so innings with an ERA+ around or over 150? Given the commonly-noted correlation between bullpen success and overall team success (if an example is needed, look at the bullpen stats for the D-backs circa-2007), I would argue that Max as a closer would be beneficial to the team, and even possibly give us the added bonus of keeping Max healthy and blowout-free for longer. But yet it's certainly undeniable that starters with 105 ERA+'s don't grow on trees, and the D-backs don't have a plethora of good starters waiting in the wings (especially since it appears we shouldn't depend on Kevin Mulvey to be reliable as a starter in 2010), so this is certainly a debate without a definite answer, however you choose to look at it,