Looking Back at the 2022 ZiPS Projections

There is still some unfinished work from 2022 ZiPS projections to do. Calibration is a long-term project. There is improvement to be had in things like calibrating uncertainty and long term data.

The 2022 ZiPS Projections are slightly better than the historical average. The Boston Red Sox, New York Mets, Cleveland Guardians and Baltimore Orioles are the teams that would have improved their accuracy if the system knew the playing time information. The Orioles would be even worse if Zips knew their rotation. It's estimated that 36% of the estimated error in Zi PS is due to incorrect roster construction.

ZiPS got four of the six divisions in the right order. The AL Central error was Cleveland finishing first instead of third. There was no change in year-to-year correlation between errors. Mariners have the highest r-squared in season- to-season errors at 0.02. For hitters and pitchers with at least 300 plate appearances or batters faced, MAE for OPS+ was 16.3 points and for ERA+ 20.9 points.

Albert Pujols is the biggest miss for a player 35 years or older in ZiPS. Michael Harris II's performance in 2022 had a greater impact on his future projections than any of the other top performers. Matt Carpenter was the most unexpected player overall in the 2022. He had only 154 plate appearances, but his projections gave him a one-in-2,300 chance of hitting as well as he did.

The biggest ZiPS misses are the hitters who most underperformed their projections.

The 2022 ZiPS Projections for the Chicago White Sox are not very accurate. Three White Sides made the list. Being a Three True Outcomes player had a slight effect on projection accuracy.

Half of the Dodgers starting rotation made the list. The biggest ZiPS pitcher under-performed. Dallas Keuchel did not perform as expected. Patrick Corbin did well, but he did worse than expected, according to the algorithm I developed. The calibration data for pitchers and hitters is useful. It shows how accurate the models are.