The Yankees’ historic success with lineup volatility

Ballplayers are creatures of habit, right? There’s consternation about whether a starter can adapt to the trend of openers, and of course, the closer is the closer because he’s the closer. Part of this belief in the habitual tenancies of baseball players comes in discussions around lineups too – as fans, we tend to think that a player should have a stable, consistent spot in a lineup. If a guy is a cleanup hitter, he should be a cleanup hitter as often as possible.

The benefits therein are anecdotal and speculative, often revolving around players being mentally offput by batting in different spots, or just nebulous worries that inconsistency will throw a guy off.

The Yankees have been nothing but inconsistent with regards to playing time and lineup construction this year, mostly because of their historical run of injuries. Gio Ushela has hit as high as third and as low as ninth this year, and Brett Gardner has hit in at least six different lineup spots. All in all, the Yankees have not had an identical lineup more than twice this year, and have used 117 lineups in 121 games as of this writing.

And yet, they’re one of the best teams in baseball, perhaps challenging the notion that players need consistency. They’re on pace to use 157 different lineups over 162 games, and I think that they have a fair shake at hitting that number, given the team is still waiting on injured returnees like Giancarlo Stanton, Luke Voit and Edwin Encarnacion.

157 lineups sounds like a lot, right? Fortunately, a few weeks ago, Effectively Wild published a list of the most lineups used in a single season, and the Yankees are on pace to be the 128th team to use at least 157 lineups per 162 games. We can use this to see if volatility has a relationship with success or not, and that’s exactly what we’re about to do.

Obviously, most of this list is dominated by National League teams - if the Yankees were to reach 157 lineups, it would be the first time since Cleveland in 1971 that an AL team fielded that many distinct batting orders. The list is divided almost exactly in half, with the dividing line being the year 2000 - the entirety of the twentieth century produced as many 157+ lineup seasons as the first 19 years of the twenty-first, which perhaps highlights that teams don’t take lineup consistency as seriously as they used to.

Our correlation coefficient for lineups/162 and winning percentage is -0.0169, which is effectively zero, showing that there is no mathematical case to support the benefits of a consistent lineup. The ranges of winning percentages, broken into lineup buckets, is below, and does show something pretty amazing about this year’s Yankee squad:

See that dot above the 157 range? The 2019 Yankees are the outlier of outliers in this entire data set, far and away the most successful team to field at least 157 lineups in a season - or at least, are on pace to be. Turns out, the key to having a good lineup is fill it with good players.

One more notable thing about this data set:

Since our data halves so nicely in the year 2000, we can look and see that there’s actually a much stronger positive correlation between lineup volatility and winning percentage as we grow closer to the current day. I think this highlights how much more important depth, development and rest has become in the eyes of front offices. The Dodgers are famously deep, and the Yankees of this year have shown how important identifying players that you can develop in-house are. Mike Tauchman and Urshela have helped save the season after being virtual castoffs until a few months ago.

This has been a remarkable season for the New York Yankees, and regardless of where the team ends up in October, they are one of the most memorable squads of my lifetime. If they keep pace with their inconsistent lineups, they may very well end up as the best team in baseball history, if only by this one esoteric way of looking at it.