Baseball projection systems have a certain amount of built-in caution.
When a player has a breakthrough season, regression will be projected until he does it again — and perhaps again. Players having unusually bad seasons will be projected to bounce back.
And the 60-game season of 2020 is likely to be a less reliable indicator than a 162-game season.
That’s part of the reason for Baseball Prospectus’ PECOTA projections, which have confounded White Sox fans who have championship hopes and surprised Cubs fans fearing the bottom will fall out. Updated last week and reflecting the injury to Sox outfielder Eloy Jimenez, the PECOTA standings at baseballprospectus.com list the Sox at 80.2-81.8 and the Cubs at 85.4-76.6.
Let’s peek at projections for a few key players that help explain the team projections.
o Sox first baseman Jose Abreu: His .987 OPS in 2020 was 117 points above his career average, and his 1.9 WARP — the Baseball Prospectus equivalent of WAR — is equivalent to a career-high-tying 5.1 over a 162-game season. PECOTA projects a return to norms at .870 OPS and 2.3 WARP.
o Sox shortstop Tim Anderson: Batting averages of .335 in 2019 and .322 in 2020 were fueled by batting average on balls in play of .399 and .383. Anderson hits to all fields and can sustain BABiPs above the major-league average, which was .292 last season. But there is an element of chance in whether balls elude fielders. PECOTA projects a .336 BABiP and .271 batting average.
o Cubs shortstop Javy Baez: After a drastic drop to .203/.238/.360 in 2020, PECOTA has Baez rebounding to .246/.295/.458. That .753 OPS and 2.4 WARP are comedowns from .881 and 4.6 in 2018 and .847 and 3.9 in 2019, but not as disastrous as the .598 and 0.9 last season.
o Cubs third baseman Kris Bryant: He played in only 34 games in 2020, but his .206 batting average, .644 OPS and 0.3 WARP were shocks to the system. In five seasons through 2019, his career OPS was .901 and never was below his .834 in 2018. PECOTA expects an upswing to a .245 batting average, .814 OPS and 2.5 WARP.
Past performance, with emphasis on recent seasons, is a prime factor in player projections. So is the performance of players who had similar success through similar ages.
A range of projections for players feed into simulated team seasons. The records reported as ”the PECOTA projection” are among multitudes of simulations. Projections sort into a bell-shaped curve. The top of the curve, the 50th percentile, represents the victory total that comes up most frequently in simulated seasons.
For the Sox, the 90th-percentile projection was 91 victories. That’s if players hold their 2020 improvement and things click. The 10th-percentile projection, with players at the low end of their probabilities, was 77 victories. The Cubs were at 94 victories in the 90th percentile and 77 in the 10th.
PECOTA doesn’t rule out a Sox title with 91-plus victories. But in simulations, the most frequent results are seasons closer to .500.