Mike Matheny as a sabermetric lineup constructor

On Monday, in a Cardinals intrasquad game, Kolten Wong batted 1st in the lineup and Matt Carpenter batted 3rd. On Wednesday, in the team's first Spring Training game against Florida Atlantic University, Charlie Tilson, who produced a 107 wRC+ in AA Springfield in 2015, batted leadoff, while Anthony Garcia, who produced a 149 wRC+ and a .400 on-base percentage in Springfield, batted sixth.

Virtually nothing predictive can be derived from Spring Training batting orders, particularly ones this early. But these arrangements of the lineup are so different from what we think we know about the Cardinals' philosophies that it got me thinking about the recent history of Cardinals batting orders.

Over the course of the season, it is generally agreed that the win differential between an optimally constructed lineup and the most poorly constructed lineup in baseball could be counted on one hand. But there is something so instinctive about the optimal lineup that it is difficult for some fans (myself included) to ignore the shortcomings of many managers in lineup construction despite its relative insignificance.

For much of baseball history, the leadoff spot was where teams put the fast guy. It's where the Whiteyball Cardinals batted Vince Coleman, and it was unquestioned baseball orthodoxy. The #2 spot was for a somewhat fast guy who could bunt or get a hit or...it wasn't well-defined, but #2 hitters had a distinct "feel". Certainly, the two-spot wasn't where you put your best overall hitter. And you wouldn't put somebody as flat-footed as Matt Carpenter at leadoff.

Yet if you met somebody unfamiliar with baseball, it probably wouldn't take too long to explain the logic of sabermetric lineup theory.

"Basically, the idea here is that you want to give your best hitters more plate appearances. The guy who bats the most should be the guy who avoids getting out, and the guys who come up immediately after should be the guys who have the best chance to create runs the most often, since they will also bat a lot."

And baseball has largely accepted the logic here. In 2015, two of the three best hitters in baseball, Mike Trout and Joey Votto, each logged significant time in the 2-hole. For those who have not read the many articles explaining why a team should bat its best hitter second, here's one of them.

In 2015, the most common two-hitter by plate appearances on the Cardinals was Matt Carpenter, who by most measures was the team's best hitter last season. To credit Mike Matheny as a convert to sabermetric lineup optimization would be ignoring context, however: Carpenter was also the team's most frequent leadoff hitter, and it was only after a series of injuries that Carpenter became a mainstay in the two-hole.

In 2014, Kolten Wong was the most common #2 hitter. And this was in a season in which he at one point was demoted to AAA Memphis in favor of Daniel Descalso.

Although the 112 plate appearances in the two spot from Matt Holliday and Jhonny Peralta were mostly aligned with the idea of putting a premium on that spot in the lineup, the top three in plate appearances, to varying degrees, suggest a certain arbitrariness to the selection process. Here is a breakdown of the top five in 2nd-spot plate appearances for the Cardinals in 2014.

Player	PA Batting 2nd	wRC+	Team Rank, PA	wRC+ Rank Among Players with as many PA
Kolten Wong	300	81	7	7
Jon Jay	129	115	5	5
Randal Grichuk	67	90	14	8
Matt Holliday	57	132	2	1
Jhonny Peralta	55	120	3	2

Was Matheny relying more on "feel" than any tangible evidence of efficacy? Is he relying on "feel" if he utilizes the relatively fast, relatively poor OBP Kolten Wong at leadoff while putting perhaps the team's best offensive threat in a spot, third, where the most devout sabermetricians believe you ought to put your 5th best overall hitter? When Matheny, say, bats Tommy Pham first, as he did yesterday, is it a reflection of a belief that Pham profiles in the long run as a high-OBP bat, or a reflection of a belief that Pham is fast?

There is almost certainly a psychological element to Matheny's construction. For instance, he has been reluctant, if not entirely unwilling, to bat Matt Holliday anywhere but third, a spot which has conventionally been where the team's best hitter has gone, even if saber-partisans would disagree that it should be, with notable examples of the last two decades being Albert Pujols and Mark McGwire.

Would moving Holliday, say, to leadoff, as has been suggested on this site, be considered an insult? Maybe. Would it affect Holliday's mental approach? Perhaps. Is the cost of theoretically mediocre decision-making worth it for a happier player? I can buy that. But does it reflect negatively on a manager if he needs to placate player emotions at the cost of team production?

One interesting way to look at the evolution of baseball is comparing Matheny's track record to that of his predecessor, Tony LaRussa. Prior to the aforementioned Carpenter and Wong, Matheny's most common two hitter in 2012 and 2013 was Carlos Beltran, who was among the team's best hitters.

Arguably, Beltran fit the two-hitter "feel" because he was once a fast player during his days with the Royals, Astros, and Mets. Maybe Matheny doesn't make him the two-hitter if 2012 Beltran had always run as poorly as 2012 Beltran, but regardless, Matheny did make the right, or at least a sabermetrically defensive, move, even if by happenstance.

Here is a breakdown of the primary #2 hitters of the Matheny years.

	PA Batting 2nd	wRC+	Team Rank, PA	wRC+ Rank Among Players with as many PA
Carlos Beltran, 2012	185	124	2	2
Carlos Beltran, 2013	369	131	4	3
Kolten Wong, 2014	300	90	7	7
Matt Carpenter, 2015	298	139	1	1 (1st among all qualified hitters)

In the final four years of Tony LaRussa's tenure, the primary two-hitters by season were...

	PA Batting 2nd	wRC+	Team Rank, PA	wRC+ Rank Among Players with as many PA
Aaron Miles, 2008	163	102	8	6
Colby Rasmus, 2009	326	89	5	5
Ryan Ludwick, 2010	217	125	9	4
Jon Jay, 2011	311	115	5	5

It appears at a glance that Matheny has prioritized the two-hitter a little bit more than LaRussa. In three of his four seasons, Matheny has opted for one of the team's big bats in the 2-hole. The occasion in which LaRussa's most frequently used #2 hitter was the most sabermetrically idealized was Ryan Ludwick, who was traded in late July to open a spot for Jon Jay, a slightly above-average hitter who, while not the worst option as a two-hitter, almost certainly was not the best.

Tony LaRussa was generally considered ahead of his time statistically: he was a forefather of the modern bullpen and was the first manager in 19 years to bat his pitcher eighth, a tactic recently favored by Cubs manager Joe Maddon. But that does not mean LaRussa was incapable of blind spots. He won a World Series in 2011 with his most common leadoff hitter being Ryan Theriot, and in the greatest game in Cardinals history, LaRussa's #2 hitter was Skip Schumaker.

Mike Matheny has been far from perfect as a manager, and this applies to his lineup construction as well as many other facets of the game. And if, when the games start to count in the standings, he bats Kolten Wong leadoff while Matt Carpenter bats third, many questions will (and should) be asked.

But Matheny has shown at least some willingness to continue baseball's trend towards optimized lineups. On a purely statistical level, his lineups do represent an improvement from yesteryear. It is impossible to know if this is a coincidence or the inevitable result of baseball's stat-heavy modernization. But hopefully Matheny can build upon contemporary understanding of maximizing production while still maintaining clubhouse morale.