OC Register: Hoornstra: New sign stealing report corroborates evidence against Astros, Red Sox

[AngelsWin.com]

This is a story about applying data to the game of baseball, one you haven’t read before. It’s about a science experiment. It’s about cheating. It’s about the sign-stealing scandals you know about, the (potentially illegal) sign-stealing you don’t know about, and all the would-be sign stealers of the future.

Like any science experiment, this one began with a question. If we look closely at the swing habits of two teams Major League Baseball recently penalized for cheating – the 2018 Red Sox and 2017 Astros – does something look different? Is evidence of sign-stealing born out in the numbers?

The answer, Jason Wilson believes, is yes.

Wilson is a statistics professor at Biola University in La Mirada. He is the co-creator of a baseball metric, Quality of Pitch, that attempts to separate good pitches from bad using a variety of components. He’s used Quality of Pitch to grade pitches, pitcher starts, and pitcher seasons. For this experiment, he wanted to use Quality of Pitch to discern cheating.

More specifically, Wilson wanted to know if Astros and Red Sox hitters were laying off high-quality pitches that other hitters could not. His initial report, co-authored with Brian Zarske, Jason Lane and Wayne Greiner, was finalized Monday and shared privately with the Southern California News Group. It’s 29 pages long. The relevant analytical details are many; some are less wonky than others. Here are a few useful things to know:

• The report doesn’t analyze every hitter on the 2018 Astros and Red Sox, only six of them. For his comparison, Wilson looked at every team’s top six hitters by pitches seen – with one notable exception.

• The six 2017 Astros chosen for the report were those identified by SignStealingScandal.com as having benefited the most from the so-called “Banging Scheme”: Marwin Gonzalez, George Springer, Alex Bregman, Carlos Beltran, Yuli Gurriel and Carlos Correa. The 2017 Astros’ top six hitters by pitches seen were Springer, Jose Altuve, Bregman, Gonzalez, Josh Reddick and Gurriel. Beltran ranked seventh. Correa ranked eighth. Swapping Beltran and Correa for Reddick and Altuve makes an important difference, as you’ll see.

• The Red Sox and Astros were found by MLB to have cheated during home games, but their methods diverged. While the Astros banged and whistled, the Red Sox relied on a runner on second base to relay signs illegally, using information taken from a real-time video feed.

• The Astros’ data looked at every high-quality pitch they saw at home in 2017. This was compared to every high-quality pitch they saw on the road, and also to every high-quality pitch every other team saw at home. The Red Sox’s 2018 data includes only high-quality pitches they saw at home with a runner on second base after the third inning. This was compared to similar data for every other team from 2013-18.

Wilson’s data claims to affirm what we already know: the 2017 Astros and 2018 Red Sox didn’t swing at pitches that their opponents did, at a rate too high to be considered random. That’s great, but the league already punished both teams. Why should we care now?

Glad you asked.

Wilson’s data also offers corroborating evidence for not cheating.

Remember that Altuve and Reddick were not included in the 2017 Astros data that amounted to a statistical smoking gun. That’s because they were largely exonerated by SignStealingScandal.com, a website launched by Astros fan Tony Adams in January. There’s evidence of a banging trash can in 19 of Reddick’s plate appearances in 2017 and, in most of those, the bang was audible when it should not have been. In other words, his teammates did a poor job stealing or relaying the opponent’s signs. Altuve made only 22 similar plate appearances, and the “correct bang rate” was nearly as poor as it was for Reddick.

Adams used video-based evidence to unearth the sound of a banging trash can, but he noted that the Astros used other means to steal signs. What about the mysterious “buzzer” allegedly hidden under Altuve’s jersey?

If that buzzer existed during the 2017 regular season, it probably didn’t help Altuve. When Wilson added Altuve and Reddick to his data set, and swapped out Beltran and Correa, the statistical signature for cheating disappeared. The Astros’ take rate on high-quality pitches was “not statistically distinctive because at least one team is this extreme every year,” the report claims.

Other teams have been accused of cheating in recent years, but none were punished so roundly by MLB – or accused by Wilson’s data.

The New York Post reported in June that the Yankees were recently accused of using a replay room and a dugout phone to relay signs to batters. Wilson analyzed the Yankees during three relevant seasons (2015-17). He looked specifically at players who became Yankees during this time period, and compared their swing rates before and after joining the team. He found nothing of significance.

To be clear, Wilson’s report doesn’t exonerate the entire Yankee organization during that three-year period. It merely claims a subset of their hitters didn’t benefit from stolen signs like the 2017 Astros or 2018 Red Sox, if the Yankees stole signs during these years at all.

More broadly, the report is incapable of convicting or exonerating any team or individual. Dozens of teams could have stolen signs illegally but, if their top six hitters (by pitches seen) failed to take advantage of the information, it would not have appeared in Wilson’s data. If only a portion of their top six hitters took advantage – even four of the six, like the 2017 Astros – the data might not signal a team-wide pattern using Wilson’s method.

In a telephone interview, Wilson acknowledged these limitations. He described his report as a “first pass.”

“We tried to focus on groups of players because we’re looking for team-wide behavior,” Wilson said. “That seems to be a safer approach. So I want to be careful about how I say this: in principle, you could pull out and look at individual players. That said, individual players have the ability to make larger changes, say from season to season. There could be physical issues that are happening that could affect performance. Maybe they get some kind of private coaching or something. We might not have information about that.”

Perhaps most critically, the report can’t detect illegal sign-stealing. Baseball bans teams from using electronic equipment to steal signs in real time, a crime for which the 2017 Astros and 2018 Red Sox were punished. By identifying the one trait these teams had in common, under the only conditions their hitters were known to receive signs, Wilson’s data reveals a statistical signature that can be applied retroactively, after every season to come.

It can also be applied retroactively to seasons past. That’s where the report gets really interesting.

Related Articles

• Dodgers extend win streak to 7 games as Gonsolin shuts down Mariners
• Inside the Dodgers: Maintaining a prospect pipeline during a pandemic
• Dodgers’ Dave Roberts says baseball’s unwritten rules ‘have changed, should change’
• Seager brothers each hit home run but Dodgers win matchup with Mariners
• Brother vs. brother: Seagers face off for first time as Dodgers face Mariners

Using the same method Wilson used to identify the supernatural patience of the 2018 Red Sox, four teams met his threshold of statistical significance from 2013-18. The Red Sox weren’t even the most patient club in the face of high-quality pitches, at home, with a runner on second base. That honor belongs to the 2013 Cincinnati Reds.

That year, the Reds’ top six hitters by pitches seen were Joey Votto, Shin-Soo Choo, Jay Bruce, Brandon Phillips, Todd Frazier and Zack Cozart. Votto and Choo are two of the league’s most patient hitters. Each was in his prime in 2013. They finished first and second in baseball, respectively, in bases on balls. Votto missed most of the 2014 season with an injury, while Choo left for Texas as a free agent. Perhaps their absence affected the Reds’ data profoundly enough to give the statistical appearance of sign stealing. I spoke to one league source who was not aware of any sign-stealing allegations against the 2013 Reds.

Cincinnati replaced manager Dusty Baker with Bryan Price after the 2013 season. The Reds haven’t reached the postseason since. Wilson’s report can’t tell us whether their last playoff berth was achieved legally or illegally, but it can tell us where to look for sign stealing in the future – no banging trash cans or anonymous allegations necessary.

View the full article