The Minnesota Twins Saved Runs By Going All In

[Parker Hageman]

When it comes to run prevention, the Minnesota Twins have discovered that marginal improvements across a variety of areas has led to substantial gains.

 

They have experimented with catching set-ups, even hiring umpires in spring camp to test these strategies against. They reexamined their infield play, adding new warm-up techniques to their routines in order to improve the defense. They added biomechanic systems to measure posture, direction and angles, hoping to improve velocity and command with the slightest tweaks. And so on.

 

Recently they found a new use for an old trick.Admittedly, pulling the infield in is nothing new. With no exact origin story, the alignment likely began when the first runner reached third base at Elysian Fields.

 

As an organization, the Minnesota Twins have hired people to run their baseball operations who question the game’s status quo -- Why do pitchers need to establish their fastballs when a breaking ball might be their best pitch? Why do catchers need to squat on two legs? Playing the infield in early in the game is another traditionally held no-no, but why?

 

Part of the long-standing rationale is the tendency to surrender extra hits. Cheap hits at that.

 

Baseball Info Solution recently looked at all of their infield positioning data going back to 2015. According to their research, in the face of the infield drawn in, hitters posted a batting average 70 points higher on ground balls and low liners as those balls found more seams and carried just out of reach of the defenders. From 2015 through 2019, hitters posted a .296 average with a normal infield alignment while they hit .366 average with the infield in.

 

Like the defense shift, nothing seems to agitate traditional baseball people more than when a weak grounder squibs through an opening where a defender should be.

 

But here’s the rub: according to BIS’s study, runs scored declined significantly with a drawn in infield. Under normal conditions, that runner on third would score 63% of the time yet with the infield playing in, the runner scored just 49% of the time. That’s a significant swing.

 

The Twins are not alone in pushing the boundaries of the game, especially in this regard. In 2017, depending on the situation, the Boston Red Sox looked to cut off the runner at third early in the game.

 

“We try to tell our guys defensively, ‘Let’s not give away easy runs. Let’s make the other team execute. If they don’t execute, let’s knock down lead runners,’” Red Sox third base coach Brian Butterfield told the Providence Journal’s Tim Britton. “It changes a guy’s hitting approach. With all the infielders back, he just has to stand in the middle and play pepper with the second baseman or shortstop and you’re guaranteed a run.”

 

Twins manager Rocco Baldelli echoed this sentiment in this season when he told reporters that the teams’ decision is based on “trying to cut down runs and giving yourself an opportunity to cut down runs as opposed to giving them up and simply playing the infield back.”

 

As BIS’s research previously showed, the odds indicate that teams would indeed save runs. So why not take advantage of them?

 

A runner on third will score on a hit to the outfield no matter what. While a normal depth infield might increase the range of where an out can be made, it also means that a run would likely score regardless. Because the Twins have confidence in their batted ball data and their starting pitchers’ ability to execute pitches, they believe they can maximize coverage with the infield pulled in.

 

When the Red Sox were determining whether or not to move everyone in, there were several factors at play in their decision-making process. For instance, is the person hitting fast and capable of stealing a base? Sacrificing a cheap hit and a run might result in a runner in scoring position quickly. Is there a runner on second? If that runner is fast, a cheap hit through the infield might score two. Under those circumstances, it might be better to play it conservatively.

 

Data suggests that Baldelli tended to be more aggressive in 2020 when it came to those situations.

 

There is no publicly available data that shows exactly how many times the Twins opted to play the infield in early in the game but if you were to examine Statcast’s infield starting position, you can see that there was a big effort in moving their infield in when runners are on third and less than 2 outs. Looking at just the middle infield positions, both fielders started on average over 10 feet closer to the plate than they did in 2019 (and more so than 2018 and 2017).

 

Download attachment: MIN Infield Alignment 2017-2020.png

Compared to the rest of baseball this season, the Twins were second in the depth their infield started in those situations.

 

In the first three innings, only the forward-thinking Tampa Bay Rays positioned their infield closer on average in runner-on-third/less than 2 out situations. Their shortstop would begin precariously close to a hitter at 113 feet. Meanwhile, the Twins would start Jorge Polanco, et al at 121 feet, the next closest in.

 

In all, the Twins had 21 instances where there was a runner on third and less than 2 outs. A review of those plays shows that among those 21 balls in play, there were only three plays that 1) the infield was in and 2) managed to keep the runner from scoring where a normal defense would have conceded the run. A more robust analysis would be required to determine how many of those bleeder/cheap hits gained during the shift resulted in extended innings. That being said, saving three runs in a shortened season is not nothing.

 

Focusing on reducing easy runs might be one of the reasons why the Rays (67) and the Twins (69) were second and third in baseball at allowing the fewest runs in the first three innings in 2020.

 

On its own, bringing the infield in to cut off a handful of runs might not seem that significant but if you look at it in the aggregate -- combined with the other elements -- it becomes another piece of overall run prevention.

 

Tiny improvements. Big gains. That’s the secret sauce.

 

Click here to view the article

[h2oface]

I have always hated letting a run score from third early in the game, and going for a double play, instead. I still do. Stop the run. Never concede a run. Runs matter most. Outs not as much. I would always rather have the score 0-0 and runners on first and second and one out, than behind 1-0 and no runners on and 2 outs. No matter what inning. But a traditionalist will fansplain every time that that is how you play baseball.... you give up the run early in the game and take the 2 outs to prevent a big inning. That single run could end up being the game, and you just conceded it, on purpose, for a measly out. 

[chpettit19]

 

I have always hated letting a run score from third early in the game, and going for a double play, instead. I still do. Stop the run. Never concede a run. Runs matter most. Outs not as much. I would always rather have the score 0-0 and runners on first and second and one out, than behind 1-0 and no runners on and 2 outs. No matter what inning. But a traditionalist will fansplain every time that that is how you play baseball.... you give up the run early in the game and take the 2 outs to prevent a big inning. That single run could end up being the game, and you just conceded it, on purpose, for a measly out. 

The interesting part of this is that the Twins would also tell you sacrificing an out at any time, on purpose, is not a way to win games. No out is measly when you're hitting. Big innings is the best way to win games. So on a very basic level it is a bit of a contradiction. That's where it really shows how detailed the Twins, Dodgers, Astros, Rays, etc. are in trying to get any fractional advantage in run scoring and run prevention. The difference is the "guaranteed" run scoring to get an out vs improving your chances to score a run by sacrificing an out on offense.

 

The idea of "guaranteed runs" is also why HRs are the most desirable offensive result. On the surface that seems super obvious (and it is), but there are dozens and dozens of comments on this site of Twins fans being real upset with the perceived "HR or Bust!" approach of the lineup. They've completely failed to be successful in the postseason with that approach and it's lead fans to become frustrated and claim the approach has to change. But the teams winning in the playoffs have the same approach. HRs are guaranteed runs. They're not an improved chance at a run (like base hits, doubles, bunting, stealing, sacrificing, etc.), they are runs. The analytics era has changed the game and I'm in the boat that feels the game isn't quite the same, and, frankly, not quite as enjoyable to watch, with all the Ks and shifting (I've worked in baseball analytics and acknowledge my hypocrisy in adding to the changes to the game while saying they've made it slightly less enjoyable to watch).

 

Teams are looking for even the slightest advantage in saving runs and scoring runs. The casual fan's brain would explode if they got a peak behind the curtain at the incredible amount of information these teams are collecting and algorithms they are running. Rocco catches a lot of heat for his in game "decisions," and rightfully so. He makes a lot of money to take the heat. Many of those "decisions" were made pre-game by spreadsheets and I think the Twins need to better balance pre-game and in-game decision making, but they are all made in an effort to sneak out miniscule percentage points of advantage over opponents. The most glaringly obvious spreadsheet data points are guaranteed runs and outs. HRs are guaranteed. The rest can just get close. That's when the algorithms really kick in (runner speed, arm strength of fielders, K rate of hitter, K rate of pitcher, speed of hitter, etc. etc. etc.) and baseball becomes more about the 4400ish outs you get, or need to get, during a season and not about the 1 you're worried about in the moment. Fans are worried about that 1 play in that 1 game. FO are worried about all the games of all the seasons. They know they're not going to get all the 1 play in 1 game situations right, but if they can get more right than they get wrong they're that much closer to winning.

[rdehring]

Interesting topic that drove me over the edge this year. Have no problem with their bringing the infield in with a runner on third with less than two outs and no on else on base. But with one out, why not play the middle infield deep and go for the inning ending double play. Too often saw them not do that and the runner scored on a dribbler thru the infield.

 

[chaderic20]

 

Interesting topic that drove me over the edge this year. Have no problem with their bringing the infield in with a runner on third with less than two outs and no on else on base. But with one out, why not play the middle infield deep and go for the inning ending double play. Too often saw them not do that and the runner scored on a dribbler thru the infield.

This is the situation I struggle with as well. Multiple times this year there were runners 1st and 3rd with one out and we played the infield in, a weak grounder was hit through the pulled-in fielders that, had we been at normal depth, would have been an inning-ending double play. So instead of no runs and inning over, we allowed the run, got zero outs, and there's still runners on first and second/third. I'm fine with pulling the infield in in other scenarios, but this one specifically bothered me.

[Parker Hageman]

Multiple times this year there were runners 1st and 3rd with one out and we played the infield in, a weak grounder was hit through the pulled-in fielders that, had we been at normal depth, would have been an inning-ending double play.

 

 

According to my video search, there was one instance where a team had a runner on third and first, less than 2 outs and hit a ground ball that could have been construed as a potential double play ball. 

 

https://baseballsavant.mlb.com/sporty-videos?playId=8aea6ebf-3690-4e6a-8f88-2c49b7b74d82

 

You never can assume two outs but with the slow-footed Ravelo running, it's probably a higher probability of turning it.

 

That said, it's fine that the Twins rolled the dice. It's not always going to land in their favor but over time it has the odds of reducing runs. 

[rdehring]

 

According to my video search, there was one instance where a team had a runner on third and first, less than 2 outs and hit a ground ball that could have been construed as a potential double play ball. 

 

https://baseballsavant.mlb.com/sporty-videos?playId=8aea6ebf-3690-4e6a-8f88-2c49b7b74d82

 

You never can assume two outs but with the slow-footed Ravelo running, it's probably a higher probability of turning it.

 

That said, it's fine that the Twins rolled the dice. It's not always going to land in their favor but over time it has the odds of reducing runs. 

I'm not going to quibble with anyone taking the time to research every inning this year, Parker, but my memory tells me there were more than just one.

 

As for the one you showed, that was the beginning of the end for Dobnak. I seem to recall several runs scored on weak ground balls. Were they all in that same disastrous inning? Some may have been on their crazy shifts which also drive me nuts.

 

I expect their advanced data tells them the shifts work more than not...but not real certain that holds when you have a ground ball machine like Dobnak on the mound. But varying from the data sheet would require the manager to make a decision against what I suspect the organization has rules that are cast in stone.

[joefish]

Always hate losing those low scoring games where the runs score in the first inning, then nothing more.

But I suppose that is better than giving up 11 in the first inning.