Let's Hit Our Cutoffs!

(flip to horizontal if reading on phone to see the videos a little easier)

October 22, 2024

Something about fielding has always drawn me to the game of baseball. Probably because in my much younger years of playing, that is what I was best at. I am not sure I hit any over the fence homeruns in my life, but I had my share of "web-gems" (haven't heard that saying in years), as many as a seven- to twelve-year-old could have had.

I was thinking about the ways that both fielding and baserunning come together in one play, which obviously happens all the time. But as I watch more games and clips, I am curious as to why teams are not cutting off more throws from the outfield. I know that there is a certain expectancy that teams have on likelihood of throwing a runner out given a ton of variables. Hit distance, hang time, arm strength of the fielder, speed of the runner, and game situation are all factors. That is why the fact that all teams are giving up extra bases is surprising to me. Let's dive in:

Below, I created a table of all teams in 2024 who were able to throw runners out at the plate but also who surrendered extra bases to the batter on some throws or fielding errors. I thought it would be more insightful to look at the difference between the two rather than looking at them alone. I also know that an out at the plate is more valuable than a runner moving to an extra base, so I gave that twice the importance (will adjust that to a more context dependent number in future work). The top three teams are the Rangers, Yankees, and Royals. The bottom three are the Pirates, Angels, and White Sox. At the very end of this I look at a couple of specific teams and go through some video to see where they may be able to improve.

I want to show some examples of plays where not hitting the cutoff man would have saved an extra base:

wallner1.mp4

Matt Wallner, who has an average arm strength of 97.5 mph (99th percentile in my data) throws home to a runner who is in standing up and gives up second base.

oneillmisscut.mp4

O'Neill Cruz, who has an average arm strength of 99.1 mph (100th percentile) throws low towards home, first baseman cannot come up with the ball, and runner advances (which he likely would have anyways).

correa.mp4

Andrew McCutchen, who has an average arm strength of 81.0 mph (0th percentile) makes a relatively good throw home and Correa advances to second.

profar.mp4

Spencer Steer, who has an average arm strength of 88.4 mph (50th percentile) throws the ball home with...~barely a cutoff man in place and Profar takes second. Goes to show that it's not always on the fielder!

Problem Framing:

Now that you have a little bit of insight into what I am thinking, I decided to look at the 2024 season on plays specifically where the batter "singles" but reaches second base on the same play. My only efficient way to do this was to match strings of descriptions in Baseball Savant's data.

Data Collection:

I used Baseball Savant's "Search" function and used a bunch of filters in my first request. I set the batted ball location to only included the three outfield positions, the batted ball type to be anything except for pop up (even though fly balls are almost always outs), and filtered for when runners were on base. I also used Savant's Arm Strength leaderboards for the last four seasons and used the average arm strength for all players to merge into one data set. I thought about including sprint speed of both the runners on base and of the batter but I decided not to for two reasons. Like I stated in the intro, the runners on base (and the third base coach for that matter) are calculating a number of different variables when deciding to advance a base and sprint speed is surely important in that, but it may have added an extra layer that I did not want to deal with at the moment. For the speed of the batter, again it is important, but I feel like a lot of these plays are based on savvy and the skill of baserunning. Reading a ball in the air as you round first base and knowing that you have the chance to get to second is not purely speed related.

Some Visuals to Chew On:

Below we can see a distribution of arm strengths by position, exit velocities on balls where runners advanced, the number of outs in the inning when the batter stepped to the plate, and the top 10 leaders in 2024 for bases advanced on singles.
We have a density plot of hit locations where batters are more likely to take an extra base. It is about the distribution that I expected. Obviously, some of the hit locations are not precisely where the fielder retrieved the ball, and the outline of the field is generic for both the infield and outfield.

Takeaways from Distributions and Leaders:

The centerfielders in the data I looked at tended to have a better overall arm strength, but that could be due to using "average" arm strength for all fielders. Some players who were put in center also have experience elsewhere (O'Neill Cruz for example)
The exit velocity distribution shows that players hit the ball hard (we know that) but that there seems to be a sweet spot where fielders are getting to the ball fast enough to think that they have a shot at the lead runner due to high exit velo.
I wanted to look at outs when the batter steps to the plate because I was interested in the second order effects of missing a cutoff. If there are 0 and 1 out, the value of holding the runner at first base is much higher to keep the double play intact. With two outs, selling out to get the lead runner out is reasonable but giving up an extra base without getting an out also causes more pitches and different matchup strategies.
The players who advance the most is not super indicative due to the small sample and numerous different circumstances that lead to a decision to throw towards the lead runner. Still, of those in the top 10, only Chourio is in the top 50 in sprint speed based on Statcast data. Again, a lot of these plays are more than pure speed.

Using The Win Probability Matrix to Frame:

I frequently refer to The Book by Tom Tango in a lot of these random thought exercises. Looking at the win expectancy charts for these different circumstances is always fun, so here is the situation for the O'Neill Cruz video at the beginning of this:

The Pirates are down two runs in the top of the 5th inning with a runner on second and 1 out, their Win Exp before the AB is: 24.3%
If Cruz throws out the runner at home, their Win Exp jumps to 28.1%
If Cruz does not throw out the runner, and the batter stays at first base, the Win Exp is now 17.3% being down three runs
If Cruz does not throw out the runner, and the batter advances to second base on the throw, the Win Exp is now 16.2%

As we can see, the obvious best-case scenario is throwing out the runner and increasing the win expectancy by nearly 4%. The relative difference in letting a runner get to second compared to having him on first is pretty small in that picture. I would want to look at the expected value that is tied with these plays though, especially if we are able to create a model that would have the probability of throwing out a runner. I actually think that having that in place might make the case better for cutting off the throw and holding a runner, but I will have to look deeper into that.

Another fun hypothetical exercise that we can do is to play some of the ensuing scenarios in that game out. If we assume that the run is scored, we can look at these situations:

If the runner stays at first base and the next batter grounds into a double play, the Win Exp goes from 17.3% to 20.4% and the inning ends.
If the runner stays at first base and a different out is recorded, the Win Exp goes to 19%
If the runner stays at first base and the batter singles to make it first and third, the Win Exp goes to 13.5%
If the runner advances to second on the throw home and the next batter walks, the Win Exp goes from 16.2% to 15.3%
If the runner advances to second on the throw home and the next batter singles home a run, the Win Exp goes to 10.5%
If the runner advances to second on the throw home and there is a passed ball / wild pitch in the following at bat, the Win Exp is 14.4%

Obviously, we are splitting hairs and dealing with tiny percentages, but sometimes that is how the game is played and how decisions are made.

Room for Improvement:

There are a couple of factors that I did not have the access to quantify that I wish I could. The only way I could see if a throw was cut off or not was based on the description in the data and video of the play. I would love to have a binary column for "competitive throws" (those with an intent to throw out a runner) that were cut off.
The big thing is trying to quantify an "accuracy" of throws from the outfield in two steps 1) the chance it has to be cut off since some throws are way off line or too high/too low and 2) the accuracy of the throw when it reaches the catcher. If an outfielder has great arm strength but is throwing the ball consistently to the left-handed batter's box, the chance of getting a runner at home is obviously much lower.
With all of these combinations of more variables and cleaner data, I would like to look at percentages of competitive throws that were cutoff or let through and what other factors determined that outcome.
Like a lot of my defensive analysis, I would love to have the distance covered by each player on the play to see how much the outfielder had to move to reach the ball. I'm also very interested in examining the cutoff men themselves to see if we can learn anything from how much they move (or don't move) to get in place to optimize both situations.
This analysis also does not look at sacrifice flys where runners on base also tag up to try to advance to another base. But look at the example of what good things come from hitting a cutoff:

Jazzout.mp4

Conclusion:

I really enjoyed looking into all this and even though I took a very expedited and straight forward path to look at some of the results, I think there is still a lot to be uncovered. The results of these plays may be too situation dependent to grasp a real good understanding of certain datapoints and actionable insights, but I would like to ask coaches about this. Philosophies about run prevention and how to deal with cutoffs would give a lot more detail as to why some players are throwing home at all costs. I know that a run is the most important thing that there is in the game and if you have a chance to prevent one from scoring, then by all means do what you can to do so. But, giving up an extra "free" base with relatively minimal resistance is something that I would circle as something to limit. Especially if there are fewer than two outs in the inning, there is a lot of value in keeping a runner on first to prevent a longer inning. Game situations will dictate decisions from outfielders (and I think there is an ego part from the fielder that shouldn't be left out of the conversation) and they will try to do the best they can to do what is best for their team. Sometimes you miss your cutoff, but if it is consistent, then we may have to address our philosophies and strategy.

Take Matt Wallner in the first video for example. He is in the 99th percentile of arm strength, he knows that he can gun down runners with an accurate throw if he is in the right spot. But take the game situation in that clip for a second. The Braves are up a run when that ball was hit and will be up two when Olson touches home. A two-run deficit is not impossible to overcome with three more innings of at bats. Since the throw went through to home, Laureano knocks in a run, reaches second and puts himself in scoring position, and breaks up a double play all in one single to right field (about as productive as an outcome as a single can have in that circumstance, especially with no outs). It does not really matter what happened in the inning after that because it is not fair to throw that all on Wallner for wanting to make a play. Maybe it is unfair to even judge what happened in the whole play like I just did but the idea remains.

I think teams looking at this can benefit from looking deeper into their decision making on the defensive side and if they have the proper systems in place to prevent runs from scoring. Maybe teams will put the responsibility on their catchers, who surely make calls to fielders to cut the ball (if they are able to or not is the dealbreaker), to read the play as it develops and move to meet the throw further up the field to prevent a runner advancing for free.

I read once that runs are the currency that MLB teams trade on. Even though one may get by, it likely won't break you. Preventing the second and third ones from scoring and limiting a longer inning is arguably more important.

~~ Thanks for reading and following along! Check out some of the recent work I have done over the past few weeks. Would love to hear any thoughts you have or suggestions for future research on Twitter or via email.

Team Specific Breakdowns:

Since my data collection is still a little tedious and does not offer very good quantifiers for a "competitive" throw, or other aspects of the play that I think would add to its success, I decided to look at the video associated with some of the plays and teams to look deeper. Another reminder and disclaimer is that I am matching the description of the play to find out if the play was made at a certain location, in this case home plate. I could very well be missing many plays that were made or not made but here I will quickly run through a couple. I also add some recommendations based on my ~very limited~ knowledge of anything that is internal with the team...just putting some thoughts together.

Tampa Bay Rays:

The Rays have been credited with 4 outs at home in my data. They also had (at least) two plays where they cut a runner off from going to second. Here are the four outs at home:

rays1.mp4

This out at home came via a rundown, so not directly on a relay or throw from outfield.

rays2.mp4

Arozarena make a great play to gun down the go-ahead runner at the plate in the top of the 9th. Here is an obvious example where the extra base matters much less than the run that may score and Arozarena was great for TB in the outfield. This call was overturned to out.

rays3.mp4

Arozarena again making a great play on one of the fastest runners in the game who chopped a little bit rounding third. No chance for this ball to be cut given the height of the throw. Also, having a runner on first who does not go to third obviously does not allow the batter to go to second.

rays4.mp4

This is not a bread-and-butter relay since it was initially caused by a misplay in the outfield, but great execution and it is rewarded with an out at home.

After looking at the outs, I took a look at some other plays in the data to see if there were any real chances to get a runner out at home. While looking at a lot of these plays, I was struck by the game situation and how that can affect the decisions like I wrote earlier.

rowdy.mp4

Rowdy Tellez, who does not have burning speed, singles to the outfield. Without a great runner rounding first, the wiggle room to throw through to home is much more appealing.

uptheline.mp4

Siri's throw wound up being up the third baseline and it's tough to say if a on target throw would've gotten Vierling out. Báez took second on the throw but with two outs, it might be relatively less crucial. However, we are only in the second inning, so maybe we try to limit the damage early in the game.

bluff.mp4

I love this clip. Josh Lowe aims to throw out Riley Greene but does not make an accurate throw and Pérez makes it to second in the process. Watch closely and see Aranda bluff catching the cutoff and faking a throw to second. It is hard to tell exact timing, but if Aranda gets that ball they likely have an out at second.

tiegame.mp4

Siri camps under this fly ball and makes a really great throw home, but he was just too deep to have a chance. In a tie ball game in the bottom of the 8th, you are absolutely making this throw home. Colt Keith with some good baserunning tagging and taking second on the throw.

As a defensive analyst for the Rays, here is what I would suggest:

We are about middle of the pack in throwing runners out at home and one of our best is no longer on the team. I think we need to bolster our outfield with another big arm thrower in the corner. Siri's arm and overall defense in center is great for us, but a lot of the throws that are coming home are coming from the corners. DeLuca also impressed this year in right. If there is a tiebreaker between two similar players, I'd lean towards the stronger arm. I think we can experiment with Lowe in left field more often and start that in spring training if we want to find a way to keep his bat in the lineup after his uptick in hard hit last season.
We are allowing runners to advance through to second 62.5% of the time with 0 and 1 outs. Game situations are very important, but I think we should lower this, if possible, to give ourselves a better chance to get out of innings. This could even be looked at by throwing lower arcing balls so runners cannot pick up on a noncompetitive cutoff.
Most of the balls we are letting runners get to second on are line drives. Our first baseman or third baseman needs to have some urgency in getting to a good cutoff position to be a presence for either situation. Yandy was out of position a couple of times, just need to coach him and keep it at the top of his mind in these situations, nothing major.

Cincinnati Reds:

The Reds have two outs at home in my data, seen below. They also had (at least) two plays where they cut a runner off from going to second.

reds2out.mp4

Will Benson makes a great throw to get an out at the plate and prevent the tying run from crossing the plate. Gotta throw through here!

reds1out.mp4

Benson again makes a great throw in a much different game situation. Two outs in the inning is a great time to test your guns, literally. Pays off here again.

I loved looking at some of the plays that the Reds made (or didn't). They serve as a great microcosim of what I am hpoing to argue in all this:

arraez1.mp4

Martini makes a really good throw trying to get a speedy runner in Wade out at the plate. The throw is just late and Arraez takes second on the long throw. Two outs in a five-run game makes Martini's throw pretty well justified in my opinion.

belli.mp4

I'll admit I started clapping when I saw this clip because this is exactly what I envisioned when starting this. Bellinger singles but Dansby is held at third with no out in the bottom of the eighth. Obviously, cutting the go-ahead run is the top priority but with an average runner in Swanson at second and no outs, we could think that they would hold him there. Bellinger takes second and now it's a much stickier situation. Like I outlined earlier, if Bellinger stops at first and there is a ground ball double play, sure a run scores but we may limit the damage. If the next batter comes up and hits a single, two could score and it is a much more difficult spot. Not a knock here on Friedl, just trying to show the calculus that goes into decsions like this.

exactly.mp4

I also started clapping on this one as we are in the first inning with one out and a speedy runner in Turner rounding third and coming home. Cutting the throw, giving up the run but preventing Bohm from putting himself in scoring position and shortening the inning? I'll take that with 27 outs left on the offensive side. Love it!

As a defensive analyst for the Reds, here is what I would suggest:

Our arms in the outfield are average to slightly below with a big dip coming in center for really whoever we have in there. It would help us to bolster that internally or via free agency.
I think we could look to move TJ Friedl off of center. He has done okay there but might be able to add more value in the outfield in terms of run prevention if we work with him on approaching balls a little better or finding someone else.
We find ourselves having to throw home a lot in very tight games (tied or +/- 2 runs). Without sacrificing too much on a deep fly ball, we could look into trying to take a few steps in to give ourselves a better chance to gun down a runner. We also need to make sure we are missing to the third base side of the plate if our throw is not right on the money.

Other graphs for viewing:

This is the number of outs that the defensive team makes at second base on these plays. Some of these may be cutting off a ball and throwing to second to stop an advancing runner, or throwing behind a base.

Density plot of the hit distances where outs are being made at the plate. The frequencies are super low which is expected. The sweet spot seems to be around 200 feet, where runners probably think they have a decent chance to score and fielders think they have a chance to get a runner out.

Page updated

Google Sites

Report abuse