Best of BP: Banging scheme hurt Astros just by helping Astros – MLB Sports

Image Credit: Thomas B. Shea-USA TODAY Sports

This article was originally published on January 30th.

Some of the biggest protracted questions in the Astros signature stealing controversy circulate around its effectiveness.Excellent report by Ken Rosenthal An investigation by Evan Drerich and MLB proved that Houston had cheated beyond reasonable suspicion, but was exposed to the darkness of whether the cheating was useful. So far, every attempt to study the Astros bang plan has been plagued by questions about when and where the system was adopted. Therefore, comparing splits is a waste of work.

The new data collected by Tony Adams shows that when the Astros scheme was accurate it was a great help to the batter, but when it was wrong it was a powerful stumbling block. Overall, there seems to be little net impact of cheating. The downside completely offset the upside, and the Astros were almost as good from the start.

In November, when the news about the Astros bang plan was first reported, many people (including myself) jumped into the various old Houston games stored on YouTube for clues about when the Astros were using them. I tried to identify it exactly. He pointed out that spectrograms (essentially digital fingerprints of active sound frequencies) can be used to separate the distinctive Twin Peaks that Astros normally uses to signal non-fastballs along the way. After investigating 12 games of audio using the spectrogram, I noticed that Houston uses this scheme frequently and accurately.

The next logical step was to digitize valuable audio throughout the season and relentlessly hunt it using either a computerized approach or a painful manual hunt. The former path would have involved teaching a computer to identify sounds (very difficult to do), listening hard to the entire valuable audio of the second half of the season (81 games x 1.5 hours per game). = Because I don’t have time). Without rushing to be tortured for at least dozens of hours on either route, I set aside the idea of ​​collecting late data for weeks off-season.

enter Tony Adams, A self-proclaimed Astros fan who took on the task. Adams used an ear-eye combination (looking at the spectrogram) to listen to all 58 Astros home games in available videos, and whether he heard the trash can hit or not, all pitches (more than 8,000). There was) was marked. or not. He estimates that it took 40 or 50 hours to collect and listen to the information. Accelerated playback Use the custom application he created.He provided this data to the public for free On his site..

Adams data can be used to give a more thorough analysis of how the Astros bang affected the team’s attack. Instead of assuming it’s perfectly accurate after the Astros start beating, you can know exactly when they broke the sign and when they turned out to be inaccurate.

In previous analysis of audio, if there was a trash can bang (arbitrary amount, but often two hits), the code breaker in the tunnel thought the next pitch was not a fastball. Conversely, when the can is silent, it could be a two-seam or four-seamer, but both methods were hot. In general, the Astros tended to get the signal correctly, but it wasn’t perfect.

Non-fastball Fastball
bangs 93% 7%
No bangs 35% 65%

They were most accurate when they hit the can: when they did, non-fastballs (curves, sliders, changes, etc.) were 93% likely to be on the way and 7% likely to be wrong. However, the lack of bangs did not always predict fastballs. At least sometimes, the player may have known that he hadn’t stolen the sign yet, or that the code was unreliable on the plate.

Even more surprising is the fact that most of the Astros faced plate appearances that seemed to lack intense information. With the exception of the turn at bat, where only fastballs were seen, batters acquired at least one pitch of bangs in just 70% of the number of hits. Most of the bangs in the turn at bat that were missing were in the second half of the match against relief. Relief seems to have debuted the new code when it came in.

Based on Adams data, the Astros used the trash can frequently and were relatively accurate, more accurate when hit than when there was no sound. Let’s look at whether the bang plan actually helped the Astros.

Here it is important to split the data based on whether the hit was correct. They hit the impressive .282 / .346 / .478 throughout the year and .289 / .346 / .503 with a correctly identified pitch (+ .007 / + .000 / + .025). For reference, in 2019, the difference between the whole team hitting: Francisco Lindeau Or Trea Turner Like hitting Franmil Reyes Or Christian walker.. A 25-point slugging is an important boost, especially if you’re straddling nine batters in the Astros lineup.

But this ignores an important part of Adams’ data. The Astros either mistakenly signaled the breaking ball when the fastball came in, or were silent when the breaking ball was in the middle. When this happened, the batter suffered a serious disadvantage. They thought they knew the curve would go into the soil, but in reality they rang a fastball. These misconceptions did more damage to Houston’s fate than the correct identification helped.They hit .237 / .294 / .385 when the signal was wrong, which is about the same as 2019 Chris Davis Or Leury Garcia And certainly, it’s a lot worse than the incredibly talented 2017 Astros lineup could have been a hit.

Therefore, when the fix was made, the Astros received a substantial benefit. But when the player in the tunnel thought he had cracked the code, but found that he didn’t, it did more harm to the batter on the plate than he knew that the incoming pitch helped them. I did. Given that the trash can is relatively accurate and has been used for at least two-thirds of the time, how does everything shake? Did the Astros win or lose by stealing the signature?

The ultimate test of effectiveness should be valued for the fact that it is one of Astros’ most skilled batters. Jose Altuve, Wasn’t banged during his turn at bat. You also need to consider the opponents you faced when the bang was on or off, and the fact that all the bangs happened within Minute Maid. It is possible to slice and dice the data to address all of these confounders, but instead of capturing the data in a small sample size area, a statistical model (generalized linear model, hitter and instead of all) Measure the result with a single-pitch linear weight value of. This approach not only tells you how much the bang has had an overall effect, but also how confident the model was. I can do it.

And if you do it all, the net effect of the bang is surprisingly close to zero. There is nothing. Statistically, for all the work and effort spent on the fraud scheme, at least its spectacular results measured in this way proved to be completely non-performing.

This is not the last word of the Astros bang plan. That’s not possible: MLB research states that it used other methods like whistling, including the 2017 postseason. None of these alternatives are represented in this dataset, only the bang scheme is shown. As a result, I do not consider these conclusions definitive. However, it does point to major issues regarding the effectiveness of sign stealing systems. The penalty for bad signals outweighs the benefits of accurately relaying the incoming pitch. As a result, every method must be very accurate in order to actually increase the productivity of the batter.

If the other signaling schemes were as inaccurate as the trash can destruction, they may not have had a significant impact on the Astros’ success as a whole. It could have been a big advantage if the Astros refined their plans or read better in the postseason than in the regular season.If your team is smarter at their tricks and better at switching signs in the middle of the game, you may have won the championship. in spite of Swipe, not for that.

And there’s no way to tell if a timely bang, whistle, or scream shook the decisive playoff series. There is nothing. Not all forensic voice analysis in the world is sufficient to rule out all the ways Astros may have signaled the plate. It is due to an electromagnetic signal sent directly to a bandage called Boon or a batter’s neuron. But at this point, the loud bang plan in which the Astros were captured seems to have done at least as much harm as it did.

thank you for reading

This is a free article. If you enjoyed it, consider subscribing to Baseball Prospectus. Subscriptions support ongoing public baseball research and analysis in an increasingly exclusive environment.

Subscribe now

– MLB Sports

Best of BP: Banging scheme hurt Astros just by helping Astros Best of BP: Banging scheme hurt Astros just by helping Astros

Back to top button