by 
The art of the stolen base is a dying one. There are several reasons for the shift away from it, which I covered in part one of this series. The Reds are one of many teams stealing fewer bases per game, but are they at least making the most of their attempts? That’s what I’ll dive into in part two.
The Reds have followed the rest of the league’s lead and stopped attempting steals with such frequency. Still, they rank 10th in baseball in steal attempts per game this season. Only two teams (Royals and Rangers) are trying more than one theft per game.
Despite attempting more steals than the average team, the Reds haven’t been efficient. Their success rate is an abysmal 64.7%, which ranks 26th in MLB. Remember, the rough success rate required to provide positive value is 75%.
Weighted Stolen Base Runs (wSB) estimates how many runs a player or team adds by stealing bases. It works exactly like other weighted stats in that it compares a player or team to the league average, which equals 0 in this case. Unsurprisingly, the Reds also rank in the bottom five in this metric:
- Giants: -4.5 wSB
- Twins: -4.3
- Rockies: -4.0
- Cubs: -3.8
- Reds: -3.3
There’s no denying the Reds have cost themselves on the bases with these unsuccessful attempts. How much damage has it done from a run expectancy standpoint?
As I covered in part one, an unsuccessful steal in certain situations is less harmful than others. If a runner leads off an inning with a double and gets caught stealing third, it costs his team 0.85 expected runs. For the most part, stealing second base is the safest bet in terms of run expectancy. The least risky situation is with a runner on first and two outs. That requires about two-and-a-half successful attempts to outweigh an unsuccessful one. Every other situation requires over three successes.
Here’s the breakdown for situations with a runner on first base only:
- 0 outs: ~3.0 steals to equal the cost of one caught stealing
- 1 out: ~2.9 SB per CS
- 2 outs: ~2.4 SB per CS
Largely, the Reds have attempted to steal bases in the low-risk scenarios. About 35% of their attempts have come with a runner on first and two outs. Despite maintaining a so-so 70.8% success rate in these situations, they have basically broken even, losing only -.0007 expected runs.
Stealing bases in the other two situations has proven costlier. The Reds have been terrible at stealing second with less than two outs. They have a 55.6% success rate with no outs and 47.6% with one out. They have lost almost five expected runs by stealing in those situations—not a huge total in the grand scheme but not entirely insignificant, either.
Trying to swipe third base is even riskier.
On the surface, stealing second seems easier than stealing third or home because the catcher has to throw a longer distance (127 feet). However, steals of third have about an 81% success rate this season versus about 71% for steals of second. That’s because runners can get a larger lead since a pickoff throw to second takes longer for the pitcher. Thus, the runner can often get further off the base.
Does that mean the Reds should attempt more steals of third base? Not so fast.
While it may be easier to steal third, it’s not always safer. Consider potential cost in these situations when a runner may break for third base:
- Runner on 2nd, 0 outs: 4.0 SB per CS
- Runner on 2nd: 1 out: 2.5 SB per CS
- Runner on 2nd: 2 outs: 6.4 SB per CS
Stealing third with a runner on second is generally the only time it’s worth taking the risk. With no outs, there are numerous ways for a runner to move up to third and score. Making the last out at third eliminates the chances of scoring. Thus, attempting a steal with one out is the least risky in terms of run expectancy. However, the Reds have not attempted that this season. Double steals need fewer successes because even a failed attempt will usually still result in having a runner in scoring position.
The Reds have played it conservatively with runners in scoring position; only eight of their 68 steal attempts have been of third base. They’ve attempted three double steals, succeeding twice. Given the known risks of stealing third, it’s unlikely this will change any time soon.
To sum it all up, the Reds have gained about six expected runs with their successful steals and lost about 10 with their unsuccessful ones. It adds up to a net total of minus-4.25 expected runs this season. While that’s only a portion of the team’s poor baserunning, it plays a key role.
On an individual level, not many Reds stand out either. Here are the leaders in stolen base percentage:
Only five players are at or above the 75% threshold, and only two of those five have attempted more than five steals. Yasiel Puig has the most stolen bases on the team and a success rate just above 75%; however, that has equated to about -0.1 runs expected based on the situations in which he’s attempted to steal.
Jose Peraza (-1.6 RE) and Jose Iglesias (-1.0 RE) have hurt the Reds the most with their inefficiency. Michael Lorenzen has provided the largest boost in expected runs (0.4 RE), but he has only tried three steals.
The running game is a clear weakness across the board. How can the Reds increase their efficiency?
Obviously, they can give the green light to their speedy players most often. Their four fastest players as measured by Statcast sprint speed (Nick Senzel, Yasiel Puig, Jose Peraza, Kyle Farmer) have accounted for 65% of the team’s stolen base attempts.
There’s more to it than that, however. Acceleration and jump are also huge factors, as is the pitcher’s pickoff move; unfortunately, there isn’t publicly available data for those parts of the equation (at least as far as I could find). But we can look at some other critical components: lead distance, how quickly the pitcher delivers to the plate, and the catcher’s pop time. Let’s start with the latter since Statcast tracks it.
Pop time is quite literally what it sounds like: the time from glove pop to glove pop. That is, how long it takes from the time the ball enters the catcher’s mitt, gets transferred to the throwing hand, gets thrown, and enters the fielder’s glove at second or third base. A catcher not only has to have a quick transfer, but they need a strong arm to reduce the pop time.
No catcher has a better average pop time than J.T. Realmuto at 1.88 seconds. The Reds may not want to try many steals when they face the Phillies. The league average is two seconds. A catcher above that mark, such as Yasmani Grandal (2.09 seconds), may be marginally easier to steal against. Two-tenths of a second can make the difference between a tag getting applied on time or not.
A pitcher’s delivery to the plate is a little trickier because they can vary it. By using a slide step, they can reduce the time it takes to deliver the ball to the plate. They can also mess with a baserunner’s timing by taking more or less time between pitches or force them to shorten their lead with a pickoff throw.
All first base coaches have stopwatches in their hands to see how quick the pitcher is to the plate. The magic number for base-stealers is generally anything at or above 1.5 seconds unless the catcher has a particularly quick pop time. Realmuto can make up the difference if a pitcher is slow to the plate; catchers with poor pop times, such as Tucker Barnhart and Curt Casali, cannot.
Teams can take this information and combine it with the sprint speed data to give them a good idea of who should steal and when. If we use the league average sprint speed (27.0 ft/s), we can assume a normal baserunner can get to second base in about 3.3 seconds. However, that doesn’t account for the player’s acceleration time. A stolen base attempt starts with the runner more or less standing still; Statcast measures sprint speed in a player’s fastest one-second window.
Split times from home to first provide a more accurate idea for several reasons:
- The batter has to accelerate in the same way they would on a stolen base. They have to speed up from a motionless (or near-motionless) position.
- Statcast only measures split times for maximum-effort sprints, which occur on nearly every stolen base attempt.
- Statcast breaks the split times down into five-foot increments, which can help account for a baserunner’s lead.
- Statcast puts each baserunner on the same scale to prevent skewed data due to the shorter distance from home to first for a left-handed batter.
Let’s use this data and take Nick Senzel as the first example since he’s the Reds’ fastest player. Assume he takes a 10-foot lead off of first base, meaning he has 80 feet left to cover if he breaks for second. His average 80-foot split time is 3.5 seconds. Against Buster Posey, a catcher with an average pop time (two seconds), and a pitcher who takes about 1.5 seconds to deliver the ball to home plate, Senzel should arrive at the base at the same time as the ball. By the time the fielder gets the tag down, Senzel is most likely on the base already even if the throw is perfect.
Against the same pitcher and catcher, assuming the delivery and pop times are the same, a slower runner would need a larger lead to steal the base. Let’s consider Jose Iglesias, a league average runner per sprint speed. He needs 3.6 seconds to hit the 80-foot mark. If he takes a 10-foot lead against Posey and the same pitcher, the ball would get to second baseman or shortstop 0.1 seconds before Iglesias. If the throw is perfect, he’s probably out. But this is a situation where it may be worth sending Iglesias and forcing the catcher to make a good throw. Against a pitcher who delivers the ball to the plate in 1.3 seconds, it gets dicier for Iglesias. Three-tenths of a second isn’t a lot of time, but it’s enough for the second baseman or shortstop to get a tag down even if the throw isn’t right on the money.
All of this, of course, assumes the runners take off at the exact second the pitcher starts his delivery. This rarely happens in reality. We may realistically need to add another tenth of a second or more to this equation to get a real sense of whether the runner has a high chance of success. If the ball gets to the fielder half a second before Iglesias, he’s almost certainly out unless the throw is poor.
These are the decisions coaches have to make as they weigh the risks and rewards of stealing bases. Fortunately, there’s more data than ever to help them make those choices. Many factors go into a decision to steal a base, and teams would be committing malpractice to ignore this information.
The Reds, at least on paper, seem to be embracing the data when it comes to stealing bases. Even with a more cautious approach, they’ve still hurt themselves more than they’ve helped. But they have at least limited the damage by stealing less frequently overall and in less-risky spots.
