The PECOTA projection system, run by Baseball Prospectus, famously projected the Royals to finish the 2016 season with a disappointing 76-86 record, a year after winning the World Series. In 2015, PECOTA thought the Royals would go 72-90. The Royals have used PECOTA’s pessimism as a source of motivation, as Rustin Dodd reported for the Kansas City Star in February.
Royals manager Ned Yost spoke more about statistics not giving the Royals enough credit. Via Jorge L. Ortiz of USA TODAY, Yost said, “They just don’t understand who we are. They can’t quantify who we are. The things we’re so good at, they’re not stat-able.”
Ortiz cites the Royals’ “intangibles such as chemistry and teamwork” but then goes on to cite statistics illustrating the club’s actual on-field skill, as well as the front office’s slant in roster composition. The Royals are good not because of juju or devil magic; it’s because their players are talented.
Last year, the Royals had the third-highest batting average in baseball at .269, 15 points higher than the league average. They stole 104 bases, the fifth-most in the game. They struck out at a 15.9 percent rate, the lowest in baseball by a wide margin. Their bullpen ERA of 2.72 was second-best in baseball behind the Pirates. The Royals’ defense, as a whole, was about 57 runs above average according to FanGraphs, by far the best in the league.
Why, then, is PECOTA so pessimistic? A handful of Royals, including first baseman Eric Hosmer and third baseman Mike Moustakas, had career years. PECOTA doesn’t believe that either player will replicate that performance.That’s not a bad assumption — most players regress after a highly productive season. Hosmer and Moustakas belong to the group “major league players”, from which we can glean what to expect in the future. This doesn’t mean that every major league player will regress after a really good season; in fact, some go on to set even higher career bests. But if you’re a betting person — or, in PECOTA’s case, a betting algorithm — you’re best off wagering on lower numbers the following year.
PECOTA is also not too fond of the Royals’ starting rotation, and it’s not really wrong for thinking that way. Yordano Ventura is the only pitcher projected to compile an ERA below 4.00. The other four have spotty histories, and the most likely expectation is that they post an ERA in the 4.00-4.75 range. For instance, Edinson Volquez posted ERA’s of 3.55 and 3.04 in each of the previous two seasons, but had been between 4.31 and 5.71 in the five seasons prior and has an overall career ERA of 4.26. PECOTA projects a 4.24 ERA for him in 2016.
Some cancer patients receive a “projection” of how much longer they have left to live. Sometimes those patients live much longer than expected. This doesn’t mean that the practice of estimating the time left for cancer patients is a fruitless endeavor. Some of those patients just got lucky; others have resilient immune systems. Still others may have unique body characteristics that allowed them to hang on longer. Some may have tried risky or unproven treatment methods. That these outliers exist is completely expected.
Similarly, PECOTA expects outliers. It projects the 50th percentile outcome, but just based on how the universe works, there are going to be many 90, 95, 99th percentile outcomes during the season. As Baseball Prospectus refines PECOTA year after year, adding more and more pertinent information it has learned in prior years, it will become better and more accurate. If the Royals defy PECOTA for yet another season, they will actually be providing yet another valuable data point — much more important than its myriad correct projections.