On Lineup Construction: Is Fredi Really to Blame?

There are few topics that come up on a yearly basis that garner as many hot takes from all sides as that of lineup construction. There has long existed a conventional wisdom when it comes to how a lineup should be constructed. You know the one: leadoff hitter is the fast guy, two guy is contact, three is best, cleanup is the masher, and so forth. And more recently, there has come the counter argument for lineup optimization. The thought process behind it is more complex than the surface thinking of the conventional construction. Emphasis here is put on the importance of each position in the lineup as it relates to run production – essentially maximizing run scoring opportunities around the best hitters in the lineup. But one thing remains true about lineups regardless of how it is constructed: You have to have quality pieces in order for it to work.

For decades, the majority of all managers regardless of level have followed the conventional method for piecing together the cogs. Even in this age of advanced metrics and statistical analytics becoming a much larger part of the game, little has changed as far as how the starting nine are written down in ballparks all across the country. With some lineups, it doesn’t matter. If you’re stacked top-to-bottom and have a balanced lineup, how it’s constructed can almost be inconsequential. Sure, it helps to optimize and make sure the pieces are in place to best suit their talents a skill set, but big data has shown that even in the majority of situations an optimized lineup is worth only a couple handfuls of runs, and at most a couple wins, over the course of a 162 game season.

One of the biggest criticisms of Braves manager Fredi Gonzalez in past years has been the topic at hand: lineup construction, specifically how he configures the top half of the lineup, where the theory of optimization focuses the majority of its attention. This season alone, there were only two primary constants: Freddie Freeman’s name was written in at third for all 162 games and Justin Upton hit in the cleanup spot for almost 3/4 of the season. Aside from those two pieces, eight players would hit in each of the top two spots and ten players would fit into the fifth position. In the top two positions alone, often regarded as being the most important in the lineup by most modern standards, twenty different combinations of players were used.

The obvious question is: Why?

Was it a result of Gonzalez simply not being efficient at filling out a lineup card for six months, or was it a result of something else – namely a sheer lack of enough quality pieces to string together even the basic of optimized lineups?


The theory of an optimized lineup came into fashion thanks in great part to the work put into the research of lineup construction done in The Book, the baseball bible for many of us in the analytical circle. I’ll spare you the details, but it lays out the significance of each spot in the order and what the prototypical hitter for each position should be.

Why does all of this matter?

Well, in order for lineup optimization to work, you essentially need a minimum of five quality offensive players of certain skill sets and the rest can be filled in at the bottom. The general basis for lineup optimization is:

Leadoff — high OBP, speed also beneficial
Two-spot — one of three best hitters, high OBP
Three-hole — essentially a second #2, not as important as 1/2/4
Cleanup — most important hitter, best power hitter
Five — second cleanup, good at avoiding outs, equal in importance to #3 spot

From there, six through nine is filled in best to worst, with the wildcard being the pitchers spot/second leadoff hitter ninth debate. I tend to err on the inside of second leadoff hitter, but it all depends on who the hitter is.

The interesting part with the 2013-14 Braves is that essentially all of the pieces were the same from one year to the next.

1B: Freddie Freeman
SS: Andrelton Simmons
3B: Chris Johnson
LF: Justin Upton
CF: BJ Upton
RF: Jason Heyward

The only pieces that changed were catcher (from Brian McCann to Evan Gattis full-time) and second base (from Dan Uggla to just about everyone who was physically capable). Pulling the best option out of the 2B contingent, you get an eight-some of Gattis, Freeman, La Stella, Simmons, Johnson, J. Upton, BJ Upton, and Heyward.

The problem then is determining which players fill which roles. Looking at wRC+ alone to determine which five players fit into the “best five” category, the first thing you notice is of the eight listed, only four were above league average: Freeman (140), Justin (133), Gattis (125), and a Heyward (110). La Stella would be next closest at 84, which is pretty horrible. Johnson would be slightly worse at 82. BJ and Simmons would bring up the rear at 74 and 71, respectively.

The top five includes Freeman, Justin, and Heyward most days. That’s a given. Gattis would start only 90 games, which would only complicate things further given the futility of Gerald Laird as his backup. But for the sake of this exercise, let’s assume Gattis is healthier. So, that’s four of our five. The fifth spot boils down to Tommy La Stella and Chris Johnson, who for all intents and purposes were equal offensively. The difference maker between the two is based of handedness. Having a third lefty in the top five (La Stella) greatly limits how the pieces can be arranged because of the rise in bullpen specialization.

Our pieces become Freeman, Justin, Heyward, Gattis, and Johnson. Now, how do we make it work?

Of the group, Heyward is the best candidate for the leadoff spot. He’s a hitter with a propensity to draw walks, get on base, and has excellent speed and baserunning ability. As much as he doesn’t fit the cliche physical mold of a leadoff hitter, based on the theory of optimization, he’s perfect.

If we are placing the three best hitters into the 1/2/4 matrix, that leaves Justin and Freeman to fit into the two-spot and cleanup. And they actually piece together rather easily. With Freeman being a lefty and arguably the teams most important hitter the last two seasons, he’s your cleanup hitter. This drops Justin in at number two, which is how Fredi used him for a large chunk of 2013, to great success.

And that leaves us with Gattis and Johnson for three and five. One of the determining factors here is the propensity to ground into double plays. Chris Johnson is a contact hitter, and he ended the season by hitting into 23 double plays in 2014. Evan Gattis’s batted ball profile is almost the complete opposite of Johnson, however. He may hit far fewer line drives, but he produces a significantly higher amount of flyballs. This is crucial considering the number of double play situations the three-spot in the lineup faces.

This also presents us with quite the contradiction and the key issue of the 2014 Braves lineup: lack of depth. Gattis profiles as the prototypical five-hole hitter. He has the power of a cleanup hitter and being a strikeout or flyball type hitter, he’s a good hitter at avoiding multiple-out situations. The problem is, the team doesn’t have a player who fits the role of three-hole hitter. Johnson is a contact hitter who doesn’t walk, he’s a horrible fit for really any spot at the top of the order. But, here we are. Despite it bunking all convention thus far, Gattis has more value as the second cleanup hitter than at the three. Ultimately, the determining factor here is the three-hole comes up in more two-out situations than the fifth, lessening the double play severity and increasing the importance of outs from the fifth spot in the lineup.

Which gives us:

(1) Jason Heyward, RF
(2) Justin Upton, LF
(3) Chris Johnson, 3B
(4) Freddie Freeman, 1B
(5) Evan Gattis, C

And the bottom half?

We’re left with the scraps: La Stella, BJ Upton, Simmons, and the pitchers spot.

La Stella, being comparable to Johnson is the obvious choice for sixth.

What of BJ, Simmons, and the pitcher? Well, Simmons and BJ had comparable seasons, so which one hits ahead of the other is essentially inconsequential if we hit them seven/eight, but what if we revisit the second leadoff hitter model?

Despite the horrible OBP, thanks mainly due to an extremely high strikeout rate and simple inability to put the ball in play, BJ Upton has had a decent walk rate and has excellent speed, as well. This actually makes him a perfect candidate for hitting ninth and rolling over the lineup. At least better than the pitcher would.

What does that leave us with now? Well, based on the theories of optimization and the pieces available, the best lineup for the 2014 Braves would have been:

(1) Jason Heyward, RF
(2) Justin Upton, LF
(3) Chris Johnson, 3B
(4) Freddie Freeman, 1B
(5) Evan Gattis, C
(6) Tommy La Stella, 2B
(7) Andrelton Simmons, SS
(8) Starting Pitcher
(9) BJ Upton, CF


One thing Braves fans have routinely complained about over the Fredi Gonzalez era has been his lineup construction. And for the most part, they have had a valid argument. But the flipside is the fact the majority of current managers aren’t any more adept at lineup optimization. There is zero evidence currently that points to there being a shift in this ideology changing at any point in the near future, unfortunately.

Does this mean Fredi is free of criticism simply by proxy of everyone being bad at it? Absolutely not. He is still the manager who only used Heyward as his leadoff hitter a little over 50% of the time and opted instead to use BJ Upton and Emilio Bonifacio in the role for a third of the season. Factor in BJ Upton as the most common number two hitter, and it’s very easy to start figuring out how and why the Braves had issues scoring runs in 2014.

Lineup construction can be a precise science that bears a minimal amount of reward. But even if that reward is a mere one win per season, you have teams willing to pay almost six million dollars for that much on the open market. So there is importance, regardless of how much weight you put into it.

Like everything in the game of baseball, advantages are everything and you have to take them wherever and whenever available. This is no truer than when working with a lineup full of pieces that are of less than average value. Fredi Gonzalez was already working from behind the eight ball in 2014, the problem was compounded by not using his players in the best possible scenarios, and as a result he put himself in an even worse position.

Brandon (@_bg37) is that dude K hates because of his love affair with Little Caesars. As a result of this, the two have been entangled in a years long debate over the sabermetric value of $5 pizza. Brandon‘s path to the Store has been a twisted one, involving studies in statistics and sports marketing and having been a member of the Braves online community since most were wee lads. He’s back in the writing game despite spending his free time chasing around a toddler and telling people not to buy hamsters.

Posted in Baseball, Featured, Sports Tagged with: , , ,

Leave a Reply