“The Book” on Lineups

Most of us who find ourselves at this site, also find ourselves considered by friends and family as “baseball fans”. Baseball fans views of the game, their favorite team and their favorite team’s efforts to put together a winner fall in many camps along several different continuum: Traditional or Sabermetric? Prospects or Proven talent? Offense or Defense first? The list goes on and on.

Suffice to say, anyone who likes to engage in these types of debates is a true fan and, at the very least, has a great understanding of their view and good enough of an understanding to argue against the other side! Regardless of which camp you happen to sit in, it is likely that you have at least heard of the famous (infamous) book by Tom Tango, appropriately called: “The Book”.

In “The Book” many time-honored beliefs about the game are rigorously studied from a statistical standpoint. One such study is related to the oft-debated topic of lineup construction. The basic question is simple: given a set of nine players with varying skills, what is the optimal way to order them in order to score the most runs over the course of a season?

The basic premise of the study is to analyze how often a certain position in the batting order gets up to bat and when that position does get up to bat, what is the frequency and type of base-runner/out combination that they inherit? Given those scenarios, certain offensive outcomes (walk, single, double, fly ball, home run, etc) are worth more than they might be in other situations. The final layer is obvious, but needs to be stated: the #1 hitter gets up to bat more than the #2 hitter. The #2 hitter gets to hit more if the #1 hitter makes fewer outs (read: gets on-base more). So on and so on down the lineup….

Intuitively, we know the above is true. Lineups matter because a batter who hits 20 homers over a season in the #9 spot is less likely to have multi-run home runs than someone who hits 20 homers out of the #3 spot. The same way that a guy who hits .300 and gets on base 40% of the time but never hits a home run is really valuable as a leadoff hitter, but will be less effective as a cleanup hitter than someone who hits lots of doubles and home runs.

Ok, if you didn’t understand what we are talking about before reading this, and you kind of understood after the small lead-in, you definitely understand the example above all making the point that “The Book” is trying to make. Certain skills belonging to players have relatively more value than others given the situations those players face.

Indians in 2013

Ok, horse sufficiently beaten; what does this mean for the Indians in 2013? The question of lineups is especially pertinent when the Indians have recently acquired a player who is an extreme example of possessing one talent at the expense of several others. Mark Reynolds is just such a player.

Mark Reynolds hits home runs and does little else. His defensive liabilities will be masked by moving him to first base and having him as the designated hitter a fair amount of the time, but we are talking offense my friends. Mark Reynolds has hit .221 the past two seasons, which is significantly below league average, while posting an OBP right at .330 which is merely slightly below league average and has struck out in over 33% of his at bats which competes with Adam Dunn for worst in the Majors. However, like Adam Dunn, Mark Reynolds hits a lot of home runs. In fact, he hit 60 in the last two seasons despite whiffing so frequently.

Where does a player like Reynolds best fit into a lineup and how can the Indians fit their current pieces into the best possible lineup? Obviously, we want the few times that Reynolds does make contact to come with men on base, but also understand that he will almost never “set the table” for hitters behind him. Basically, he clears the bases, he never loads them.

Using Bill James 2013 projections courtesy of www.fangraphs.com, I have placed them in a “traditional” batting order that one might expect Terry Francona to utilize this upcoming season (*barring any future trades, which is funny because I have run this analysis twice already given the Stubbs trade shortly after the Reynolds acquisition!):

(Note: I am assuming that Aviles, Canzler, McGuinness, Carrera, etc, etc ultimately end up filling the last two positions in some capacity at a 250/310/400 level. Feel free to assume away as much as you’d like and be sure to add it in your comments!)

If the above is our traditional view, and we are aware of a “non-traditional” view that exists; how best do we fit the Indians current roster into our new form of thinking? Utilizing the lineup optimization tool from www.baseballmusings.com, we are presented with a different possible lineup to trot out each day. I merely propose the following, but am really looking for others to chime in and see what the community thinks.

According to “The Book’s” way of thinking: here is the Indians optimal lineup:

This is a gain of about 0.07 runs per game, which if extrapolated over the course of a season is worth about one win. That isn’t that important when you are talking about going from 68 to 69 wins, but is extremely important when you start getting in the mid-80s range and considering how many seasons have come down to just a game or two, can make all the difference in the world.

As an aside, if anyone was wondering about potential player change scenarios facing the Indians:

Swisher signs: Projection: .356/.362/.458

Santana, Swisher, Cabrera, Reynolds, Kipnis, Chisenhall, Stubbs, DH, Brantley

Expected Runs Per Game: 5.072

Cabrera Traded, Swisher Signs: No Major League Position players acquired, Aviles Starts 

Santana, Swisher, Stubbs, Reynolds, Kipnis, Chisenhall, Aviles, DH, Brantley

Expected Runs Per Game: 4.996

Cabrera Traded:

Santana, Reynolds, Stubbs, Chisenhall, Kipnis, Aviles, LF, DH, Brantley

Expected Runs Per Game: 4.815

This last scenario could be interpreted to support the case of dealing Cabrera for young pitching. The basic reasoning is: you don’t hurt the team THAT much offensively by starting Aviles (0.1 runs/game) and you get the dual benefits of shedding $16.5 million in payroll over the next two years that could be used to entice Swisher (showing a 0.9 run increase over the initial lineup) while simultaneously acquiring the young arms that will be required to make all of this lineup tinkering worthwhile in the first place and actually get the Indians in the marginal area of 85+ wins where +/- 1 game actually matters.

What does it all mean?

As we can see, as the offensive talent drains from the big league club, the ‘optimal’ lineups change greatly. Reynolds as a #2 hitter is completely counter-intuitive until you realize that beyond Kipnis, nearly every hitter is barely above replacement level given their current projections. Reynolds elite power wins out over his low OBP and is given the opportunity to bat a few more times over the course of the season and hopefully get another home run or two by batting that much more often.

In the end, lineups make a difference at the margins and the real key is to have great players at every position, not arrange mediocre players in some magical order. I sincerely believe that both Chisenhall and Brantley will perform significantly better than their predictions this season and will justify their likely batting spots in the top-five of the order. I am optimistic on Drew Stubbs to outperform both last year and the .250/.320/.390 projection assigned to him.

As a final note, please remember: this discussion is meant to be academic and showcase the value of getting differently skilled players onto the same roster in order to maximize their talents and overall have a better team.  After all, no possible lineup construction will make the starting rotation any better. Though Trevor Bauer very likely will!