QSAO\’s Analytics Mythbusters: Breaking down positional nuances (Part Two)

By Constantine Maragos

As we progress through the seasons of various professional sports, we start to notice trends worth looking into. As I have begun to expand my analytical prowess, I have learned to develop my own questions, and look to answer them through the published work of others – something that QSAO does for the sports community (Look out for all-new content next semester! – shameless plug). But I digress, in Part Two of the latest edition of QSAO’s Analytics Mythbusters, we look into load management in the NBA and how the Houston Astro’s sign-stealing scandal affected team performance.  

1.    What is the best tell for load management effectiveness?


Load management became a prevalent topic in the NBA last year and has taken over headlines this season. The first name that comes to mind when thinking about load management is LA Clippers star, Kawhi LeonardSpanning from the 2016-17 season, Kawhi Leonard has not participated in a back-to-back in the NBA. So far, the NBA has taken somewhat of a stand against load management (due to obvious business-related concerns). The NBA sent out a memo to teams regarding the use of the term “load management.”


Unfortunately, public opinion is divided on the topic. On the one hand, some attribute regular season rest to Kawhi Leonard’s dominant performance in the 2018 NBA Playoffs, on the other, there are objectors of the tactic, citing it unfair to fans of the game. The NBA took a definitive stance against Kawhi and the Clippers’ load management this season. After missing a nationally televised game vs. the Utah Jazz, the League fined the Clippers $50,000 for comments contrary to Leonard’s status (Doc Rivers cited that Leonard was feeling ’great,’ and sitting out is ‘not really a health thing\’).

From an analytical perspective, how can we assess the impact of load management on individual performance? NBA teams have begun to invest increasingly more on player health, and ways to improve player fitness and conditioning. With that being said, teams are “finding ideal games to rest their stars, and they’re doing so because the demands of the game are different today […] [there are] more repetitive stress injuries than [players] are used to,” as per Eno Sarris of The Athletic. 

Looking at the current landscape of the NBA, it is hard to deny the increased physical toll that the game takes out of its players. For example, in 2016, the Memphis Grizzlies opened their season with a gruelling four-in-five schedule to start the season. On top of this, Marc Gasol and Mike Conley, their two top players at the time, had recently returned from season-ending injuries from the previous year. With this in mind, both sat out during the fourth game of the season to little opposition. The Grizzlies medical team cited the rest would significantly improve the team’s remaining 76 games. 

The reasoning behind the rise in NBA load management is again, due to the increased physical element of the game. While there is no specific model open to the public, scaling back individual schedules for NBA athletes is done to protect them from not only further injury but to improve their performance overall. When looking at the game from a ticket-buyers perspective, it is tough to miss out on seeing the best of the best, but in the grand scheme of things, as fans, it is important to recognize that certain players playing every game can be detrimental to team success.

Again, while there is no public information regarding how teams select which games to sit for, and how many, there are various studies in the MLB. As fatigue is not “always obvious […] consider that there are kinematic changes — changes in the movement of muscles — that predict the mechanical changes that we use to spot fatigue before injuries and poor performance occur as a result\” as per Eno Sarris of The Athletic.. Concerning pitchers specifically, former Twins analyst Josh Kalk initially introduced the idea that pitch movement, velocity, and release point to injuries, calling it the Injury Zone.

When looking at load management as a whole, it is hard to argue against it, considering the benefit it provides (injury-preventative, improved performance in the future, etc.). With that being said, load management will become more common in the NBA, and across many other professional sports as well, it is just better for the game overall.

2.    Sign stealing in the MLB – How does this really improve team performance?


In the aftermath of the Washington Nationals’ first-ever World Series win, The Athletic dropped a bomb on the sports world; The Houston Astros were caught electronically sing-stealing. During the 2017 season, where the Astros won their first World Series, team personnel used live-feed video fixated at centre-field and would bang on a trash can to signal witch pitch was coming to the hitter at the plate.

While sign-stealing is not a new phenomenon in the MLB, the Astros went the extra mile, using technology to assist them in the process. Additionally, an Astros executive “asked scouts to spy on opponents’ dugouts leading up to the 2017 postseason, hoping to steal signs and suggesting the potential use of cameras to do so,” as per ESPN’s Jeff Passan. However, this is not the first time a team has been caught electronically sign-stealing. The Boston Red Sox “were fined for reportedly using Apple Watches to relay the catcher’s signs in the dugout,” as reported by Bobby Nightengale of the Cincinnati Enquirer.

Now, looking into the situation as a whole, how much did the Astros’ sign-stealing contribute to team success? FanGraphs’ Jake Mailhot looked into the Astros’ advanced metrics to break it down. In 2016, the Astros’ home strikeout rate was 24.5%, good for 28th in the league. After reportedly implementing their sign-stealing technology, that rate shrunk to 16.7%, the best in the MLB. Mailhot does note that this change is more likely to be coincidental than a direct correlation to their “operational” changes. In the past three years, the Astros have kept their overall strikeout rate below 20%. With that being said, sign-stealing’s effects reach beyond strikeout percentage.

Looking at Houston’s plate discipline numbers, you can see improvements across the board.


Graphic retrieved from FanGraphs

Between 2016 and 2017, the Astros became much better at the plate, but also mind you that there were improvements on the road as well. On top of that, the Astros made significant improvements to their batting rotation, bringing in George Springer, José Altuve, Carlos Correa, Marwin Gonzalez, and Evan Gattis. These five players also saw improvements in plate discipline, posting combined changes of -0.46% O-Swing%, 4.59% Z-Contact%, and -2.09% SwStr% (for more information on plate discipline statistics, check out FanGraphs’ article).

While looking at overall plate metrics can provide insight, analyzing the effect at a per-pitch level serves as a much better tell, given the nature of the Astros’ sign-stealing system. Through Tom Tango’s all-new RE288 (built upon RE24, RE288 is the run expectancy based on the base-out state and the plate count state), we can calculate run values for any pitch thrown “and then aggregate those run values to begin to analyze how a team, say, benefited from specific knowledge about the pitch being thrown” (Mailhot). Looking at the graphic below, we can see interesting changes.


Graphic retrieved from FanGraphs

The Astros saw significant improvements in differentiating breaking balls and fast pitches, which is key in making contact. Further, the Astro’s improved by nearly 75 runs on in-zone pitches at home. However, when assessing Astros Pitch type and overall run values, we can find that they performed better on the road than at home in 2017. When looking at league-wide changes in isolated power vs. strikeout rate, the Astros far and away had the biggest percentage decrease in strikeout rate on the road, as well as at home.


Graphic retrieved from FiveThirtyEight

Leave a Comment

Your email address will not be published. Required fields are marked *