Statcast ushers in new era of player evaluation
Indians finding ways for advanced data analysis to help development
CLEVELAND -- Francisco Lindor knew his batting average was not pretty. As he sat at his locker inside the visitors' clubhouse in Great American Ball Park in mid-July, though, the Indians' rookie spoke of hard-hit balls and bad luck when addressing the offensive slump that was his first month in the Major Leagues.
Lindor referenced conversations with hitting coach Ty Van Burkleo, who had reassured the shortstop that the data indicated that his swing was fine and the results would change soon. Lindor did not need to revamp his mechanics or mindset -- just trust the process.
"His stats are a lot lower than what they could be," Van Burkleo said at the time. "But, he hasn't let it affect him or get him trying to do more. He's just stayed the course and knows it won't keep happening, if he keeps maintaining the swing he's got and the approach he's got."
The statistics that backed that belief were easily found within Statcast™, which told a much different story than could be found in Lindor's early slash line. During the shortstop's first 100 at-bats in the big leagues, an unusual number of outs were occurring on balls consisting of exit velocities in excess of 100 mph. The result was an extremely low batting average on balls in play.
The front office could relay the data to the coaching staff, which could in turn find the most appropriate way to address the issue with Lindor. Rather than having a young player panic at the sight of his batting average on the Major League scoreboard, Lindor could trust that the numbers supported the feeling that he was running into some hard luck.
"It's more important to focus on the process," Indians general manager Mike Chernoff said. "Information like that can show you that, 'Hey, despite that a few of these balls are not falling in, my process is right. I'm doing the right things and I don't need to drastically change what I'm doing.'"
Statcast™ is changing how teams are doing things.
Introduced during the 2015 season, Statcast™ recorded a wealth of information that was not previously available, especially in the public realm. Offensively, exit velocity became a useful and new way to evaluate hitters. On the mound, spin rate -- something many teams already had access to via other tracking systems -- offered a new look at pitchers. Defensively, things like first step, route efficiency and speed were measured.
These elements not only gave fans a new way to view ballplayers, but offered teams new tools for evaluating and comparing.
"Baseball statistics measure outcome," Chernoff said. "There are also ways to measure process, a process prior to the actual outcome; things like strike, ball, velocity, swing and miss, those types of things. They're not exactly an outcome, but somewhere in between. What Statcast™ does is it measures the actual movements of players and not the outcomes of where the ball goes.
"So, take the case of a baserunner. Rather than just seeing how many stolen bases and caught-stealings he has, you actually measure how fast the baserunner is now."
Where Statcast™ currently falls short in some regard is context. Due to the system being so young, teams do not have years of data to pore through to get a true reading on how certain statistics stack up historically. However, the numbers can be used right now as a comparison tool to put one player against another.
For example, as Cleveland searches for outfield options this winter, it can use Statcast™'s defensive metrics to compare possible free-agent targets.
"Yes," said Chernoff, when asked if the Indians have used the data in their offseason evaluations. "One challenge is you don't have a multiyear sample of it yet, and you can't verify it's accuracy quite yet, or just know when it's accurate and when it's not. It takes a lot of data to figure that out. You can use it to do some things.
"This isn't like the answer key, though, that's going to totally revolutionize the way we look at players. Scouts have been measuring tools of players for a long time, and they're really good at it. This is a way to get a slightly more accurate reading on some of those things."
Chernoff believes the addition of Statcast™ is actually a good thing for scouts.
"This should be a huge relief for scouts and a huge complement to what they do," Chernoff said. "[It] allows a scout to process the information rather than just have to transmit information to whoever is going to be reading his report."
One area that has been helped by the recent influx of information has been better measuring a catcher's abilities behind the plate. Statcast™ provides data for first step, release speed and arm strength, among other things. It is also easier to tell when a pitcher is more at fault than a catcher on singular stolen-base events. On top of that, the PITCHf/x numbers can help show which catchers are better at framing pitches.
"We've made a lot of progress over the past several years with that," Chernoff said. "The PITCHf/x data and now the Statcast™ data have enhanced that. I know I sound like a broken record, but really what it's doing is it's breaking down the measurements instead of just saying, 'How many caught-stealings and how many stolen bases against are there for the catcher?' You can now measure the actual tools in the way that a scout would to say how good is he at receiving, how good is his footwork and how strong is his arm."
The key to all the information that continues to pile up is for a team figuring out the best way to filter the data down to the coaching staff and then to the players. Some players want more detailed statistics, while others prefer not to risk being overloaded with numbers.
Lindor, for example, was given a general idea of how the statistics supported that he would eventually turn things around. The shortstop hit .223 with a .568 OPS in his first 103 at-bats, and then he hit .345 (.930 OPS) the rest of the season, finishing as the runner-up for the American League Rookie of the Year Award.
"In the end, there are only certain things that [players] can control," Chernoff said. "And Statcast™ or scout data, or anything like that, doesn't give you how to make it better. It just tells you where you are. So, it's an assessment tool. What our coaches can do is use that to help show a player, 'Hey, here's where you stand. Let's try to work to make it better.' That's a much more complicated process that you can't just use data for."
After only one season, Statcast™ is becoming a part of the process.
"We've gotten very good analytically as an industry inferring a lot from outcome-based data," Chernoff said. "Now, we can actually measure the real movements and not just try to infer them from the outcomes."