Did the Houston Astros gain competitive advantage from knowing pitch signs?

With the 2019 baseball season officially closed, rumors about the Houston Astros cheating began to spread. The allegations say that the Astros used a camera in the outfield to read the catcher’s pitch signs to gain a batting advantage. For example, when an off speed pitch such as a change up, or curve-ball were signed by the catcher, someone in the Astros dugout would hit a trash can 5 times to alert the batter of an incoming off speed pitch. The allegations quickly spread around the baseball community and is now one of the biggest hot topics. But no one has looked into the Astros statistically to determine wether they gained a competitive advantage by knowing the signs, or did it not really matter?

To look into this deeper, I determined a few possible areas that I thought might be of interest to look at: Home vs. away batting averages, pitch outcomes, winning percentages, and through different averages and counts over the years. I wanted to focus on home vs. away statistics because the current allegations state that they were only able to do this surveillance at their own stadium, so the Astros should be performing better home instead of away because they know the pitch signs.

Firstly, let’s look at how the 2015 season compares to the 2017 season. The 2015 season does not have cheating allegations against the Astros and should be a good baseline.

This graph above shows the sum of outcomes of At-Bats for the Houston Astros at home. As you can see, the 2015 season and 2017 season had the same results besides for a few switches in order such as doubles and triples. Although one may look at the 2017 season and say that the numbers are much better, it’s more about looking at the change in outcome quantities, from 2015 to 2017. If the Houston Astros were cheating and gained a competitive advantage we should see a much smaller gap between Strikeouts and all other fields, but they remain pretty similar. Now let’s compare it to when the Houston Astros are away.

Not huge differences once again, with the gaps staying pretty similar, and numbers staying almost consistent home vs. away in each season.

Now let’s look into how Houston’s batters performed against certain pitches in the 2017 season.

The graph for home is on the left, and the graph for away is on the right. As you can see they’re also very similar and there are not really any statistical differences in the outcomes of pitches in the home vs. away games.

Just to reiterate, if the Houston Astros DID gain competitive advantage from knowing the pitch signs, they should be batting lower strikeouts and higher walks, and on base percentages.

So let’s look into the on base percentages for the different seasons. For the 2015 season, the Astros batted on average a .315 for the entire season. And for the 2017 season, the Astros batted on average a .282 which is significantly lower, if they gain an advantage this number should be higher. Now lets split it up by home vs. away games in the 2017 season:

So while their home batting average is better than away, it’s not significantly larger since its less than .01%. The increase at home could be explained by home field advantage. Next let’s take a look at on base percentages.

The Astros in 2015 average on base percentage was .315, and the 2017 average was .346. This is important to note because this one of the only metrics that reflects some sort of improvement from 2015–2017, but I still don’t think it was a large enough difference to conclude some sort of advantage. Now let’s split up the 2017 season by home and away games again.

As you can see, the Astros actually got on base more when they played away games, not by a lot, but still more. This shouldn’t be the case if the Astros had an advantage by knowing the pitch signs at home. If they had an advantage at home, their on base percentages for home games would be significantly higher. But this isn’t the case.

Next, let’s take a look at how batters did with each pitch, did they get a strike, ball, or did the players hit it? Here is the visualization for when the Astros are home:

and here is the same visualization but for when the Astros were away:

So while there is a difference between season totals, the gaps are similar again. I think if the Astros had an advantage we would see the number for strikes be lower than the number for balls for certain off-speed pitches like cutters, sinkers, curveballs, and changeups.

So what does this all mean for the cheating scandal? Well that can be answered a few ways. Well, first off, the Astros did cheat, but knowing the pitch signs at a professional level doesn’t correlate to batting better. This could be explained by split second decisions by batter and pitches, and how each pitch is pitched differently, the batter still needs to make a decision on where the pitch is going to land and calculate their swing to meet that location. So while knowing that a pitch could be coming in slower, there are still a few more factors that decide whether a batter does hit the ball. If we look at a scatterplot of how pitches came across the strike zone it might paint you a better picture.

As you can tell from the visualization above, pitches can come in at many different locations with different speeds, so it gets very complicated when trying to tell a batter where and when a weird pitch might be coming. The Astros tried to gain an advantage by predicting some factors such as pitch type, but unfortunately for them they were not able to get sizable results.

Although, I do not think they gain an advantage, they still made it to the world series in 2017. I don’t think predicting the pitches alone would get a team to the world series as there are is a whole other half to the game the Astros need to play on the field. But it may have provided them help in key situations and games where it was very important for the batter to know pitch information, and that level of detail would be very hard to pick up through basic statistics.