Platoon the lineup
Posted by disappointmentzone on 27 July 2006
Indians GM Mark Shapiro in the last eight months has made a number of changes to the Indians roster. To Chicago went set-up man Bob Howry (3-3, 3.54 ERA this season) and in came Guillermo Mota (1-3, 6.21 ERA). The Indians received nothing in return for 2B Brandon Philips, who is playing for the Reds and having a borderline All-Star season. To make matters worse, the Indians will now have to re-sign Ronnie Belliard in the offseason or risk brining in an even worse 2B. So the Philips trade has bitten Shapiro twice. Coco Crisp went to Boston, but dividends of that trade have only started to emerge: Kelly Shoppach is now catching a couple games per week, Andy Marte is still in Buffalo. Then there is Jason Johnson, signed in the offseason and then traded in June to Boston, only to find himself a few weeks later learning how to pitch with their single A affiliate. Paul Byrd has turned out OK. The Jason Michaels for Arthur Rhodes trade has worked out about as well as the Byrd trade. Neither player has set the world on fire, but neither was expected to, either. The only move that was an unqualified success was signing aging free agent Eduardo Perez to platoon at 1B with the erratic Ben Broussard. Through late May the Indians had the second most productive first basemen in the league, behind only Albert Pujols and whoever spells him. Benuardo was such a success, in fact, that Seattle has unloaded two of the top players in their farm system to acquire the duo to serve as their DH (which is smart since neither plays 1B particularly well). In return we got a young middle infielder who’s a wiz in the field and a powerful corner outfielder who’s fast with a strong arm.
The outfielder, Shin-Soo Choo, is a left-handed hitter who doesn’t hit left-handed pitching very well. He will be platooning in right field with Casey Blake, a right-handed hitter who doesn’t hit left-handed pitching very well. Next season Choo will likely serve as a platoon partner for LF Jason Michaels, a right-handed hitter who doesn’t hit right-handed pitching. And in between will be Grady Sizemore, a left-handed hitter who doesn’t hit left-handed pitching.
As a rule, right-handed hitters don’t hit right-handed pitchers and left-handed hitters don’t hit left-handed pitchers. Casey Blake is a rare exception to this rule. But even he falls into the same problem as the rest: there is a handedness he just doesn’t hit as well. The most rare–and most valuable–players are those who hit both left- and right-handed pitchers equally well, such as Travis Hafner. Even switch hitters tend to fair better against one handedness of pitcher and not the other.
With this in mind Shapiro perused the free agent market last season to find a right-handed 1B to play with Ben Broussard, a player who’s always had great success against right-handed pitchers and something less than great success against lefties. His man was Eduardo Perez, a player not highly touted due to his inability to hit right-handed pitching. Perez was essentially the mirror of Broussard, a perfect compliment to Broussard’s deficiencies.
Why don’t managers platoon all of their positions? There are two primary reasons. First, a manager is only permitted to have 25 men on his active roster (bumped to 40 men in September). Usually half of the roster is allotted to pitchers, leaving only 11-13 spots on the roster for field players. With eight position players in the field, that leaves only about one player to back up each group of positions: one backup catcher, infielder, and outfielder. In other words, there isn’t space on the roster to platoon for every position. Second, defense matters, and it’s hard to find average players for most positions. There just isn’t a left-handed Jhonny Peralta on the market, nor is there a left-handed Aaron Boone or right-handed Grady Sizemore. The more demanding the position, the harder it is to find average players, let alone a surplus. 1B is one of the easiest fielding positions. It wasn’t difficult to find a platoon player for Broussard because there are a number of average to slightly below average 1B on the market. And with limited defensive responsibilities comes greater gains in batting. In other words, a power-hitting 1B that can bat .275 is much more common than a power-hitting SS that can bat .275. This is why the most common platoons in baseball are corner outfielders and first base, the three least-demanding positions.
Managers do, however, platoon their pitchers, although it’s not usually referred to as such. There is not a single roster in baseball in which all of the pitchers are of the same handedness. Most rosters have at least one left-handed starter and one left-handed reliever, if not two lefty starters and two lefty relievers. When the Indians sent down RP Rafael Perez it left them without a left-handed pitcher in the bullpen. This is highly unusual. The reason managers platoon their pitching is obvious: it forces the opposing teams’ rosters to be diverse–full of lefties and righties–which in turn leaves the roster in a weaker position than it would be if it could be composed of all right-handed hitters to face left-handed pitchers, for example. The reason we don’t think of pitching staffs as been a group of complimenting platoon players is because we usually think of platooning as a reactive measure. That is, managers platoon players because they there will be times when the handedness of the pitcher won’t favor one hitter, so they react by putting in a different-handed hitter. Platooning the pitching staff is a sort of preemptive platoon and, as such, it is much more effective in that it dictates further action (who to bat against which pitcher) rather than being dictated to (who to pitch against which lineup).
While managers are unable to platoon for each position (and would certainly do so if there were a surplus of talent to go along with increased roster size) there is an alternative that has never, as far as I know, been broadly implemented.
Platoon the lineup.
Let’s take the Indians lineup. Here are two lineups Eric Wedge has used recently that are fairly representative of the lineups he’s used this season, if not by player certainly by design. One against a lefty, one against a righty.
Against Mike Mussina (righty) on July 5th:
Against Johan Santana (lefty) on July 15th:
The three leaders in home runs per at bat on the Indians roster (at the time of these lineups) are Hafner, Sizemore, and Broussard. The spacing of the home run hitters is curious. The two biggest threats–Hafner and Broussard–bat fourth and sixth in one lineup and fourth and seventh in the other. The next best home run hitter bats first in both lineups. Traditionally the most powerful section of a lineup (the “meat of the order”) is the third through fifth hitters. But only one of the power hitters bats here. Of the five players with the highest on-base percentage (OBP) and at least 250 at bats, only one, Sizemore, bats in the first three spots of the lineup. Sizemore, third highest in OBP and with the third highest home run rate, bats first and therefore is usually in a position to drive in the bottom of the order–those who are least likely to get on base. If Sizemore manages to get on base, the player most likely to drive him in via a home run, Hafner, bats fourth, which is as it should be. However, the two players following Sizemore in the lineup–Michaels and either Belliard or Peralta–are not as likely as other Indians players to get on base and, therefore, reduce both the likelihood of Hafner batting with Sizemore on base and the likelihood of Hafner batting with multiple men on base.
Managers ought to recognize that where a player bats in a lineup should be mostly a function of the handedness of the starting pitcher, the power of the batter, the likelihood of that player getting on base, and the ability of the lineup to adjust if the handedness of the pitcher changes during the game—in that order. These two lineups address these factors—if at all—in something approximating reverse order. And the most egregious offense is the complete reversal of the most and least important factors.
The top of the order goes left, right, right, left, switch. If a manager’s focus is on how to best suit his lineup to face relief pitching late in games, when opposing managers often switch pitchers based on the handedness of the hitter, he would construct something like this, in which the handedness of the hitters is not the same for more than two consecutive hitters, thus forcing the opposing manager to either burn through his corps of relief pitchers quickly (adjusting to each new batter) or leave a pitcher in the game to face batters who are suited to get hits off of him (usually batters of the opposite handedness). As the manager of the hitting team, you want the left-handed relief pitcher to stay in the game to face your right-handed batters. By alternating between handedness you are pushing the opposing manager to a decision: leave the left-hander in to face the righty or expend another relief pitcher. Since roster size is limited–managers can’t put in new relief pitchers for each batter–this sort of decision can have a large impact on the game.
Which is exactly why managers should focus on the handedness of the starting pitcher when constructing a lineup. The primary conceit managers make in the late innings is that the handedness of the pitcher matters. These managers hope that the other team decides to save their left-handed relief pitcher for the eighth inning when there are two on and two out in the seventh and a lefty is coming to bat. Managers are right. The handedness of pitchers matters a great deal. So why not construct a lineup to mimick the two on and two out situation, but in the first inning of the game?
In these two lineups from Eric Wedge (like most managers) there is practically no consideration for the handedness of the starting pitcher (Sizemore always bats first, followed by two right-handed batters, then Hafner, and then Martinez). A lineup should reflect the handedness of the best pitchers it will face as well as the handedness of the pitchers it’ll most likely face the longest–this is crucial. I’ll explain why in a moment. Since starters are the best pitchers (usually) and since starters pitch most of the innings, lineups should adjust according to the handedness of the starting pitcher.
What of the crucial point? Few runs are scored by home runs. The reason why is that home runs are very hard to hit. Most runs are scored by singles and doubles and drawing walks. Of the many reasons why far fewer runs are scored in baseball than in cricket, the primary reason is how many outs a team gets per inning. In cricket each team is given 10 outs per inning. In baseball, only three. Fewer outs per inning means the value of an out increases significantly. Imagine if baseball were a game of 27 one-out innings–almost no one would score. Accordingly, the way to score runs in baseball is to string together a series of plays–usually walks, singles, and doubles–in which no more than two outs are surrendered. This is still very hard to do, and the way to maximize the chances of it happening is to put together a lineup in which large chunks of batters are (relatively) unlikely to record outs–the best hitters. To this end a manager doesn’t want to alternate between players who fluctuate significantly between more and less likely to record an out–which is part of the reason the worst hitters in a lineup usually bat at the end rather than being dispersed throughout. But the “best hitters” is a relative term. Sizemore is one of the Indians’ best hitters against righties. He’s one of the worst against lefties.
Managers decide to alternate the handedness of their batters so that in late innings no single relief pitcher is able to neutralize any two consecutive batters. So Eric Wedge, for instance, puts a right-handed batter between Sizemore and Hafner so that late in the game the opposing manager can’t use a left-handed relief pitcher for consequtive batters. Wedge’s bigger concern, however, should be the other instances in the game (usually two or three) when putting that right-handed batter between Sizemore and Hafner costs his team because the right-handed batter is more likely to record an out against the right-handed starting pitcher than the left-handed hitter who could be batting there.
So what does “platoon the lineup” mean? It means that managers should construct lineups based primarily on the handedness of the starting pitcher. If a team is going to face a right-handed starting pitcher, the lineup should reflect that by having large chunks of batters who statistically have a high chance of getting on base; usually left-handed batters. Alternating the handedness of batters throughout the lineup is the antithesis of this method. It is also what most managers do.
To return to the recent trades Shapiro has made, the Indians have, for the moment, a large platoon among catcher, first base, right field, and left field. A lot of this is owed to the versatility of Blake and Martinez, who can both play multiple positions. And the future success of this platoon rests on Choo being about to perform well against right-handed pitchers, something he’s proven himself capable of in the minor leagues.
The Eduardo Perez/Ben Broussard platoon worked very well, but the success wasn’t so much a function of luck as it was the a function of intelligence. Platooning players makes good sense when it’s possible, and so it should be with platooning lineups as well.