The 2017 Houston Astros: Cheaters or Champions?
There are a ton of unwritten rules in baseball. It’s part of what gives the game its charm. “Sign-Stealing” (stealing the signs the catcher is giving to the pitcher) is generally considered good gamesmanship and despite the reference to thievery, is normally not in violation of the oft-debated unwritten rules of baseball. However, sign stealing with the aid of technology, is against all rules of baseball, written or otherwise.
Sign-stealing with the aid of technology during their 2017 World Series Championship run is what the Houston Astros are currently being investigated for by Major League Baseball. From the evidence the public has so far, it appears that they used a camera aimed at the opposing team’s catcher to steal and relay the sign to the batter. Using state-of-art technology that involved beating the crap out of a trashcan in the dugout, they were able to alert the batter to the nature of the incoming pitch. Here is the system in action (please excuse the language) :
To those unfamiliar with baseball; knowing the type of pitch that’s being thrown to a batter would be like a quarterback knowing the opposing team’s play before the snap. Does it make the offense unstoppable? No. But it tips the scales in the cheater’s favor significantly.
We’ll be exploring the likelihood that the Astros and their batters had a significantly greater home-field advantage (the allegations state that they were only able to use their system at home) than the rest of the MLB during the 2017 regular season.
Let’s compare the difference between the home and away batting averages of teams across the MLB. In order to control for a few other variables, I limited my data to situations where the bases were empty (no way to legally steal a sign) and high and medium leverage situations (evidence suggests that the system was used in the more pivotal moments of a game). If the allegations are true, the Astros should have been much more successful batting at home compared to the rest of the league in these scenarios.
The Astros did perform better in these situations at home. But, it appears that they are well within a range we would expect based on the rest of the league. They must be innocent! Right?
The evidence suggests that not EVERY Astros hitter took advantage of the system. This could explain how the allegations could be true, despite what we observed in the graph above. Let’s take a look at how successful the Astro’s hitters were at home in these situations compared to on the road:
This graph shows that there are a handful of players with larger home/away splits than the league as whole (MLB). One of these players, Jake Marisnick, stands out in particular. The difference between his split and the league’s is certainly worth looking into.
In 2017, Jake Marisnick batted .308 at home and .187 on the road. To put this into perspective; a player averaging .308 for their entire career would be considered one of the best hitters of their generation and could potentially make the Hall of Fame. A player that bats .187 for too long would have a hard time keeping their job.
To explore the significance of this observation, I’ll use a 1 sample t-test to determine the likelihood that Marisnick’s home-field advantage at the plate differed from the rest of the league’s during the 2017 regular season.
Null Hypothesis: Marisnick had the same home-field advantage as the rest of the league.
Alternative Hypothesis: Marisnick had a different (possibly greater) home-field advantage than rest of the league.
With a p-value of less than 1%, we can reject the null hypothesis in favor of the alternative hypothesis. To put this in simple terms: We can be 99.3% sure that there was something going on, other than general randomness, that gave Marisnick a greater advantage over the rest of the league when batting at home during the 2017 season.
Is this proof that at least one Astros player cheated? Not quite. All this test reveals is that there was likely SOMETHING going on. It could be as trivial as Marisnick having a particularly hard time sleeping while on the road. But, given the allegations and the mounting evidence that suggest the Astros used technology to implement an organized sign-stealing system during the 2017 season, it can certainly be used to make the case for the allegations being true.
Notes:
- No other Astros players’ splits tested for significance at the 95% confidence level. This doesn’t mean that they weren’t cheating, it just means that the differences we observed weren’t enough to rule out general randomness as an explanation for the differences.
- In the interest of greater sample size, I used all home and away data rather than just using bases empty, high/medium leverage scenarios for the t-test. Larger sample sizes lead to more conclusive results so I accepted the possibility of confounding variables skewing the data. Legal sign-stealing with runners on 2nd in opposing team’s stadiums was the particular scenario I was worried about.
- Batting AVG is not necessarily the best indicator for whether or not the batter has an extra advantage. It’s pretty common to get a hit on poorly hit balls and even more common to not get one on well hit balls. I plan on investigating further using strikeouts (it’s difficult to make batters miss when they know what pitch is coming) and contact quality (it’s much easier to hit the ball hard when you know what pitch is coming) Stay tuned for updates!
