2017’s Most Aggressive Base Stealing Teams

In June of 2017, Baseball Savant introduced a new Statcast metric for public consumption. They defined sprint speed as “‘feet per second in a player’s fastest one-second window.’ The Major League average on a ‘max effort’ play is 27 ft/sec, and the max effort range is roughly from 23 ft/sec (poor) to 30 ft/sec (elite). A player must have at least 10 max effort runs to qualify for this leaderboard.”

Now that we have a year’s worth of data, and we are trapped in the middle of a cold, dark offseason, I thought it would be an opportune time to play around with it. Over the last few weeks I sliced, diced, and minced the numbers. One area explored was team sprint speed and its relationship to stolen base attempts. Sprint speed was pulled from the Baseball Savant Leaderboard and stolen base data was downloaded from Baseball Reference.

Team Att.SBO Plus

As a proxy for team aggressiveness I added team stolen bases to caught stealing and divided by stolen base opportunities (plate appearances through which a runner was on first or second with the next base open). SB Att% = (SB+CS)/SBO

The table above shows teams that sent runners most often in their SBOs each of the past three seasons. Light green represents one standard deviation above average and dark green denotes two standard deviations above average. One has to question how much these totals were the result of team aggressiveness and how much they were the result of team personnel. To check I took a weighted average of each player’s sprint speed by respective player’s plate appearances and then added to find team totals. SS=speed in the next table. If you’re curious to see the exact figures I attached my sheet here: SS Weighted Avs

Att.SBO vs SS Rank

Surprisingly, the 2017 Angels attempted steals at the highest rate despite being the second slowest team! The Brewers and Rangers were also quite aggressive despite their below average collective sprint speeds. The implication is these three teams had some sort of organizational edict to test defenses and attempt more steals. On the other hand, the Reds were the tenth fastest team. It’s also worth noting they were the only team to attempt steals at a plus rate (1 STD > AVE) each of the past three seasons.

21uqhv

To make more sense of the team-level numbers, I decided to get granular and look at them on a player level. First, I wanted to see how many players on each team attempted steals at a higher rate than the league average (5.17%), which you can see in the first column of the table below. Then, I wanted to see how many “not fast” players on each team attempted steals more than the league average rate. I did this by introducing a second filter, player sprint speed. The second and third columns show how many players per team were non-elite & below average runners but attempted steals more than the average rate.

For the purpose of this post I defined above average sprint speed as greater than 27 ft/sec and less than 28.3 ft/second. In the sample of 451 players on Baseball Savant’s Sprint Speed Leaderboard, 28.3 ft/sec was one standard deviation greater than average and 27 ft/sec was average. The middle column below shows eight Rangers, six Angels, and six Brewers attempted steals at a greater than average rate while having above average or worse speed. In other words, eight Rangers, six Angles, and six Brewers were non-elite runners but stole at a rate greater than league average anyway.

# Plyrs per team with att per SBO over 5.17

And they were not alone! My efforts monkeying around with filters and subtotals in Excel yielded this table. What does it tell us? Well for one, the Reds were not as aggressive sending runners as their overall SB Att% would suggest. Additionally, you could argue the Red Sox and Diamondbacks belong in the same conversation as the Angels, Brewers, and Rangers. They were willing to send a comparable number of non-elite runners at greater than league average rates.

While useful, this table is imperfect. Notably, it fails to account for magnitude of the SB Att%. In other words, it treats players who barely clear the league average threshold the same as players who are well above average.

The screenshots below display SB Att% by player for the five aforementioned teams plus Cincinnati. I highlighted SB Att% and sprint speed columns to add some meaning to the figures. It is intended to help visualize whether players attempted steals at a rate commensurate with their speed. The darker the green, the further from the mean. For example, Ryan Braun’s sprint speed cell is white (below average), but his SB Att% is green (over a standard deviation greater than average) so he attempted more steals than his sprint speed would suggest.

This slideshow requires JavaScript.

Color Coding Explained

For those curious I attached my full spreadsheet. There are two tabs which show SS vs SB Att%: One for a sample of 451 and another for a sample of 491. The second tab adds extra lines for players who played for multiple teams. Sprint Speed vs SB Att%


I made observations on each of the six teams from the slideshow in the blurbs below:

Arizona Diamondbacks – Micro-level, Ketel Marte only attempted steals in 3.85% of his SBOs, a below average rate, despite ranking 29/451 in sprint speed. He should see an increase in his stolen base attempts to at least the above average range next season. Macro-level, the team was less aggressive than I thought. They were the sixth fastest team on my weighted average sprint speed sheet. And they had five players who ran less than their sprint speed may imply. The large number of Diamondbacks attempting to steal at a high rate seems to be due to their roster construction rather than organizational tendencies. I.e. their roster was littered with above average sprint speed players.

Boston Red Sox – I would classify the Red Sox as somewhat aggressive. They were willing to send three above average runners, Benintendi, Betts, and Nunez at well above average rates. They were also “sneaky aggressive” in the sense they sent below average runners Vazquez and Chris Young at above average rates. They should consider sending Bogaerts more often. Not only does his speed indicate he is capable of it, he was only caught stealing once in 16 attempts (94% success rate).

Cincinnati Reds – How does a team with the fourth highest stolen base attempt rate (shown in first table) have only three runners attempt steals at an above average rate? Billy f****** Hamilton. That’s how. He attempted steals an absurd 33.33% of the time in his 216 stolen base opportunities last year. Simply silly. I want to know how many of his attempts included pitch outs. Billy (yes we are on a first name basis) and to a lesser degree Jose Peraza were complete anomalies on this list, and their rates heavily skewed the Reds data. Cincinnati as a whole is firmly in the non-aggressive bucket.

Los Angeles Angels – The Angels table was revealing. All of their eight fastest runners attempted more steals than their speed may suggest. None of their 11 slowest runners attempted more steals than expected. This helps to explain how they could simultaneously attempt the steals at the highest rate while having the second slowest roster. In essence, they were selectively aggressive, stealing more often than expected with above average runners and making little to no effort with below average runners. On a player level, I found it surprising that neither Cameron Maybin nor Ben Revere cleared the plus threshold for sprint speed (1 STD > AVE). However! (S.A.S. voice) They attempted steals at an elite rate (3 STD > AVE).

Milwaukee Brewers – This is a run happy team. Seven players ran more than expected based on their sprint speed, two of whom, Ryan Braun and Domingo Santana sported average or worse sprint speed. Jonathan Villar attempted steals three STDs over the league average rate with merely above average sprint speed.

Texas Rangers – Perhaps the most aggressive team in MLB, they sent some below average runners at above average rates, and they sent above average runners at well above average rates. Delino Deshields was their only real burner, but that didn’t stop them from sending eight other players at an above average rate, which was the second most in the league. (Shown in the second table, first column)

For fantasy players it would be wise to keep an eye on offseason acquisitions by the Rangers, Brewers, and Angels. These players are liable to see increase in their stolen bases in 2018. To what degree? This will be tackled in a future post. Also on tap, who were the least aggressive teams and what we can learn from their data?


A few notes on the information from the slide show:

*Many Detailed Statistics are based on play-by-play accounts accumulated by RetroSheet. These totals may be incomplete – (Copied verbatim from Baseball Reference). This is referring to the asterisks next to player names.

**No players in the sprint speed data set are greater than three standard deviations from the mean. That does not mean Buxton, Hamilton etc. are not 80 runners on the scouting scale. It just means the distribution for sprint speeds does not follow a normal curve.

***Not every player was available on the Sprint Speed Leaderboard, limiting our sample to 451, which is why the average SB Att% of 5.01% differs from the overall league at 5.17%.

11/11/17, 11/13/17 AFL Notes

AFL Notes 11/13/17

Kyle Regnault (LHP) New York Mets – Left-handed command specialist is the best four-word phrase I could come up with to describe Regnault. His fastball is not overpowering, but it can be effective due to his array of secondary offerings. He can run the fastball arm side in on the hands of right-handed hitters or straighten it. Regnault’s best pitch is a high 70s curveball with two plane movement and hefty depth. He’ll throw it either at the bottom of the zone or as a chase pitch. The curve gets a lot of swing and miss. Regnault also employs a low 80s slider. It’s not as good as the curve, but it serves as an effective change of pace. The fourth option is a low 80s changeup that has moderate depth and fade. He seemed to reserve this pitch for lefties and would run it away from them. Regnault’s overall arsenal works because he is able to command all four offerings and keep them down in the zone. I think he is polished and ready for a shot at middle relief in the majors.

AFL Notes 11/11/17

Adam Choplick (LHP) Texas Rangers – After losing out to Hafþór Júlíus Björnsson in the casting process for Game of Thrones’ “The Moutain”, Choplick resorted to his backup plan, a baseball career. He is a massive human, listed at 6’9” 250. With his colossal frame and long arms, one might expect natural plus extension, but his use of a high three-quarters arm slot undermines it. On the other hand, his height and arm slot allow Choplick to pitch with significant downhill plane. I can see this having divergent effects on his fastball and curveball. His high 70s curveball already has quality depth, which is augmented by the plane. Hitters should swing over the top or make contact on the top of the ball, meaning lots of ground balls. Alternatively, plane on his 92-95 mph fastball should result in fly balls. I think hitters will gauge its downward angle and respond with an upward-sloping bat path, allowing them to keep their bat in the zone longer. The uppercut path should result in more fly balls.

The curveball has been Choplick’s most-used secondary offering, and he commands it well. I think the command combined with its aforementioned depth make it a plus pitch. His slider was serviceable but below average. Choplick used it inside to jam right-handed hitters. Choplick has posted excellent numbers in the minors but has also been older than league averages. Next season he will be more age-appropriate in the Texas League, which should be revealing. Overall, Choplick looks like a pen piece. I think the stuff falls short of a closer profile, but middle relief or setup are possible outcomes.