Diving Into the Twins’ Futility with the Bases Loaded

It’s become a running gag on these electronic pages that the Twins’ opponents have the Minnesota nine right where they want them when they let them load the bases. The best gags are often rooted in truth and this one is no exception. The Twins have performed terribly with the bases loaded this season, hitting a combined .143/.171/.143 (.142 wOBA) over 35 bases-loaded plate appearances, easily the worst in the majors thus far. They’ve managed to hit just four singles (no extra-base hits) and have drawn two walks in such situations.

Now, thirty-five opportunities is a tiny sample, but it’s emblematic of what’s been a theme with this team for much of the past five seasons. In 2017 and 2018, the first two seasons under this front-office decision-making regime, the Twins ranked 8th and 3rd in MLB in hitting with the bases loaded (by weighted on-base average, wOBA).

Surprisingly, for all of its offensive production and prowess, the 2019 Bomba Squad Twins sank to 29th in the league when the bases were full, hitting just .217/.231/.348 (.233 wOBA) and beginning a trend that has largely persisted until now:

Data via FanGraphs

You can see in the table, cumulatively since 2019, the Twins have been the worst-performing team in baseball with the bases loaded. Not only is their overall production (by wOBA) terrible, they have the worst walk rate (4.0%), 9th-highest strikeout rate (25.1%), 3rd-most double plays hit into (31), and 4th-fewest runs scored (320).

Don’t Call Clutch a Skill

Research has shown repeatedly that there is no evidence to suggest that coming through in the clutch is a skill that players can have or lack. Despite the stories we tell about larger-than-life legends that had the “heart” and “will to win” to come through in big spots, performance in big spots tends to just track with hitters’ overall talent levels over time. Good hitters are good hitters and bad hitters are bad hitters, in all situations.

Certainly, players will come through in big spots (and we’ll probably remember those well), and they’ll fail to come through in even more big spots (and we might not remember those quite as well). After all, more than 7 in 10 plate appearances ends in failure, so even the clutchiest of clutch hitters are not going to come through significantly more often than they deliver.

Photo by Brace Hemmelgarn/Minnesota Twins/Getty Images

As a whole from 2012 through last season, Major League Baseball batted .251/.318/.409 (.315 wOBA) in all base-out situations. With the bases loaded, the league hit .273/.311/.441 (.313 wOBA), which is more or less the same level of production.

So given that, there is little reason to believe that the Twins are lacking any kind of special talent for being clutch or handling pressure situations. But, we’re now going on five seasons of poor performance with the bases loaded, and we expect things to move toward the mean over time.

Why hasn’t it?

Signal vs. Noise in Small Samples

Before we dive into that, it’s worth a quick detour about signal and noise in small samples. Go back and look at the table above again. While this poor performance has been going on for parts of five seasons, we’re still only dealing with a little more than three-quarters of a season’s worth of plate appearances. In that light, it’s a pretty small sample that is spread over 49 different players, which means there is a lot of noise in this data. The most bases-loaded plate appearances from any one Twins player over this time span is Jorge Polanco’s 44.

Thanks to prior research about reliability and statistic stabilization by Russell Carleton and others, we know the number of plate appearances and batted ball events required to have some confidence that there is more signal than noise in a sample. If all of these bases-loaded plate appearances were by one player, we’d be in a pretty good spot. For example, strikeout rate becomes more signal than noise at about 60 plate appearances and on-base percentage requires 460 PAs to do the same.

Photo by Brace Hemmelgarn/Minnesota Twins/Getty Images

But we aren’t talking about a single player. We’re looking across a whole team, over multiple seasons. So, the simplest explanation for the Twins’ woes with the bases loaded is that it is mostly just random variation (i.e., noise) that hasn’t been in the Twins’ favor.

That’s probably not a satisfactory answer, nor does it make for an interesting article, but it’s a caveat that needs to be made. With that out of the way, let’s dive in deeper to see if there are any patterns in the Twins’ collective approach with the bases loaded that might be potential signals.

To Swing or Take?

Perhaps the most straightforward way to look into hitters’ approaches is by evaluating how often they swing the bat. Minnesota hitters collectively, since 2019, have swung the bat against 47.4% of the more than 85,000 pitches they have seen, in any base-out situation. In bases-loaded situations, almost 1,900 pitches, that rate of swinging has leaped up more than 5 points to about 52.8%.

Photo by Mike Stobe/Getty Images

We should expect hitters to be more aggressive in bases-loaded spots because pitchers have no margin for error for a walk or wild pitch and should throw pitches aggressively in the strike zone. Hitters are also likely frothing at the mouth with the opportunity to drive in multiple runs and that all should lead to swinging the bat more often.

At the league level, that’s played out. In all situations over this time span, hitters have swung 47.4% of the time, and in bases-loaded spots, that has crept up to 49.9%. That number shows that the Twins take bases-loaded swing aggressiveness to more extremes than the rest of the league. Their 52.8% swing rate with the bases loaded is the 3rd-highest rate in MLB, trailing only Detroit and the White Sox, who rank 20th and 25th in bases-loaded productivity since 2019.

Swinging is Overrated

On the other side of the bases-loaded swing aggressiveness spectrum are the Dodgers and the Rays who have swung at only about 46.5% and 46.6% of their bases-loaded pitches, respectively. They are followed by the Brewers, Yankees, Mariners, and Astros. I don’t know about you, but when that group of teams is hanging out together on a list of strategic approaches, I tend to pay attention.

There have been several studies in recent years that have shown that hitters probably swing too often and would be better served by swinging less. The Dodgers’ rookie infielder Miguel Vargas made this case during Spring Training when he was medically prohibited from swinging because of a broken finger. Vargas took his plate appearances anyways, to track pitches and get mental reps to prepare for the season, and despite everyone knowing he wouldn’t swing he still drew 4 walks in 11 plate appearances.

The math is clear that, on per pitch basis, takes are far more valuable than swings, and have been throughout baseball history. Even the best hitters in history have produced far less value with their swings than their takes, and only a handful of the best ever even produced positive total value with their swings:

So, Just Swing Less?

Perhaps just taking more pitches in bases-loaded situations is the answer. That’s easier said than done, of course, because pitchers have significant influence over the flow of a plate appearance via the ball-strike count. And that influences production more than any other factor in the game.

I wondered if the Twins were swinging more with the bases loaded as a response to the ball-strike count distribution. Perhaps if they were ahead or behind in the count significantly more often than they usually are, they’d swing more often. Here’s how that breaks down:

Data via Baseball Savant

You can see in the table above that they do operate from behind in the count a little more often with the bases loaded than in other situations. Maybe feeling the need to protect the plate could be part of why they swing more often.

Next, let’s look at how often the Twins are swinging in these different count situations with the bases loaded and how that compares to what the rest of the league does:

Data via Baseball Savant

You can see the Twins' overall swing rate tracks pretty closely to the league average in all count situations (2nd and 4th columns). It’s also clear that the Twins’ extra bases-loaded aggressiveness is true regardless of the ball-strike count situation. They swing a lot more when behind in the count, as I suspected, but they are also swinging a lot more when they are ahead and when the count is even. They just swing more often when the bases are loaded.

We already know that approach isn’t yielding results, and here’s how the Twins bases loaded production breaks down by count situation:

Data via Baseball Savant

While none of those numbers are particularly strong, I want to focus in on when they are ahead and in even counts. I think this data paints a picture that shows how their expansive swing approach can be detrimental. Too often they are swinging at borderline pitches and hamstringing their own production.

Ahead in the count with the bases loaded, the Twins swing at about 75% of pitches located in the heart of the plate. Those are the ones you want to swing at. But they also swing at more than 62% ahead in the count pitches located in the shadow zone, on the edges of the strike zone, where, ostensibly, it’s harder to do damage. Shadow zone pitches are, by definition, 50-50 to be called strikes or balls. Those kinds of pitches are rarely the kind a hitter should be hunting for when they are ahead in the count.

Because of the power of the ball-strike count, plate appearances often hinge on what happens in 1-1 and 2-2 counts. Get ahead 2-1 by taking a close ball and you’ll be in better shape than falling behind 1-2 because you fouled off a borderline pitch you really couldn’t drive anyways.

In even counts with the bases loaded, the Twins are similarly aggressive against pitches in the heart of the plate, swinging almost 72% of the time. They’ve gotten little to show for it, though, producing a league-worst .197 wOBA on those pitches. They also swing at almost 54% of even count pitches located in the shadow zone and have produced a 27th-ranked .227 wOBA against them.

We’re really getting down in the weeds now and the sample sizes are way too small to declare anything conclusive. But I think this data suggests how often they swing with the bases loaded could be a potential signal in the noise. (Not for nothing, the Twins swung at less than half the bases-loaded pitches they saw in 2017 and 2018 when their bases-loaded production was quite good.)

Whether it is the organizational hitting philosophy to get into a favorable count and then let it fly in search of extra-base damage or players pressing to do too much in big spots and expanding their approaches (they are humans with all the ups and downs, emotions, and inconsistencies that come with them, after all), the Twins might be well served to try to swing a little less with the bases loaded. That, and wait for the random variation of the game to turn in their favor.

---

John is a writer for Twinkie Town and Pitcher List with an emphasis on analysis. He is a lifelong Twins fan and former college pitcher. You can follow him on Twitter @JohnFoley_21.