For the first seven years of his MLB career, Marco Estrada was treated like a swingman. But after he earned some MVP votes in 2015 and made the All-Star team in 2016, Estrada entered last season as an entrenched member of the Blue Jays’ rotation and fantasy-relevant starter. Then it all fell apart. He finished last season 4th-worst of 58 qualified starters with a 4.98 ERA. Sabermetrically-minded analysts like Paul Sporer had entered the year with concerns that Estrada had outperformed his peripherals in the 2015-16 seasons, and his reversal from FIP overachiever to FIP underachiever looked like a confirmation.
In an interview on MLB Network Radio’s Inside Pitch earlier this week, Estrada offered an alternative explanation for what happened.
I was having issues with sleeping—stressing out, stressing about being traded and a few little things. I’d been struggling. It all kind of snowballed together. I stopped sleeping. But once I knocked all that stuff out, I was able to get back on track.
Estrada’s assertion is that anxiety over trade rumors and other issues prevented him for sleeping well, and that sleep deprivation led to his decline in performance on the mound. That explanation struck a chord in me because I had previously read about Mike Napoli’s sleep apnea issues and of the growing interest in sleep research among MLB players and teams.
A company called Fatigue Science synthesized a number of research studies that demonstrated the impact sleep or a lack of sleep could have on athletes, such as improved reaction times and reduced rates of injury. That included a research study by Cheri Mah of Stanford that showed that college basketball players improved their shooting accuracy on both free throws and three-point attempts after two months of extra sleep, and that seemed like a particularly strong parallel to the mechanical sequence a pitcher has to go through to execute a pitch.
Estrada’s explanation matches up with his period of worst performance in 2017, too. Jon Paul Morosi’s article pointed out that Estrada pitched to an 8.87 ERA in his 10 starts prior to the end of the trade deadline, when he then bounced back to his typical level of performance over the rest of the season. At the time, I chalked that period up to the random fluctuations that every pitcher experiences around their true talent level. But Estrada’s interview brought me back to question I expect everyone who reads FanGraphs obsesses over: how much of what we call randomness in statistical analysis would be explainable with better information?
The more I investigated, the more confident I grew that Estrada’s slump in 2017 was not simple randomness. I think the easiest way to illustrate that is by pulling the worst 10-game stretches of all of his seasons, which I labeled as such based on his strikeout minus walk rates (K-BB%).
|2012||7/30 – 9/24||.329||22.1%||5.6%||16.5%|
|2013||4/8 – 6/3||.305||19.5%||6.9%||12.6%|
|2014||5/23 – 7/19||.270||16.2%||9.4%||6.8%|
|2015||8/4 – 9/28||.171||14.7%||7.8%||6.9%|
|2016||4/10 – 6/5||.181||20.7%||10.2%||10.5%|
|2017||5/27 – 7/26||.370||19.6%||14.7%||4.9%|
Last season’s 4.9 percent K-BB% 10-game stretch was easily the lowest of any 10-game stretch in his career. In particular, Estrada’s walk rate and BABIP ballooned to 14.7 percent and .370, respectively. Obviously BABIP is going to be volatile over small samples like these, but consider that Estrada had been a consistent BABIP limiter in his career prior to 2017. In fact, his .250 BABIP allowed from 2012-16 was 12 points lower than any other pitcher with 500 or more innings.
To me, the changes in Estrada’s production over the 10 games leading into last year’s trade deadline look like evidence a marked loss of command, and based on what I now understand about sleep’s impact on athlete performance, I think anxiety over a potential trade neatly explains why that happened.
If you buy into the narrative like I do, then the next question becomes what one can expect from Estrada next season. I don’t think it’s fair to throw out his worst 10 starts last season and then compare his performance otherwise to that in his previous seasons. Like all pitchers, Estrada will continue to endure slumps, hopefully just not ones exacerbated by insomnia. Instead, I decided to put together Estrada’s seasonal numbers excluding his worst 10-game stretches each season. That overstates his general effectiveness as a pitcher but provides a better basis for comparing his seasons without sleep issues to his season with sleep issues in 2017.
With that perspective, Estrada’s 2017 season looks awfully similar to his 2016 season. Both his K-BB% and BABIP allowed are within 2 points of each other, and he pitched in at least 19 games apart from the 10-game stretches I excluded in both seasons. That level of performance in 2016 made Estrada a top-40 starter, and while issues like his age, lack of a consistent track record, and the possibility this narrative is incorrect should drive his draft position down, his ADP as the No. 74 starter and outside the overall top-300 provides a lot of profit potential. In shallow formats, I would look his way over some of the lower-ceiling, higher-floor starters in front of him like Rick Porcello and Jake Odorizzi.
Meanwhile, look for sleep/rest to become a more prominent area of baseball research. Just last week, Mariners GM Jerry Dipoto discussed on his podcast that the Mariners ran a sleep study and then built a sleep room to help deal with the added fatigue their players have to deal with because of Seattle’s abnormally big travel demands being in the northwest corner of the country (jump to 24:56 of Episode 7). Teams clearly believe this is an issue, and hopefully that will lead to more stories like Estrada’s and Napoli’s that will help us better understand just how much it can impact player performance.
Scott Spratt is a fantasy sports writer for FanGraphs and Pro Football Focus. He is a Sloan Sports Conference Research Paper Competition and FSWA award winner. Feel free to ask him questions on Twitter – @Scott_Spratt