The OOPSY projection system made its Major League debut this year. So how did it do?
OOPSY’s methodology mirrors that of the other FanGraphs projection systems, with a few twists – most notably the addition of bat speed for hitters and Stuff+ for pitchers. However, projection systems consist of many different components, including aging curves, Major League equivalencies to account for Minor League and foreign league performance, recency weightsregression to the mean, competition environment and park factors. There are many ways in which projection systems can differentiate or lag behind their peers, beyond just including a certain variable such as the bat’s speed. Projection systems are made up of hundreds of small methodological decisions. Given the sum of the decisions included in OOPSY, has OOPSY held up against its more established peers in 2025?
To assess the projections, this article follows industry best practices outlined by Tom Tango, MLBAM senior data architect. I’ve done this assessment process for jars foras my pitching projections were presented by Eno Sarris in The Athletics since 2023, but this was my first year publishing a full set of striking projections. This review focuses on wOBA for hitters and wOBA against for pitchers (an alternative to ERA, further defined below). These metrics are typically the focus of projection system reviews, the key stats on the batting and pitching speed that allow projection systems to perform well from a “real-life” perspective. Both are catch everything rate statistics which measure a player’s attack and throwing value respectively. The various component projections, for example K% and BB%, are included in these comprehensive measures.
As a guide for anyone looking to replicate the projection review process, I’ll walk you through the steps I took, using my Review 2023.
First, I collected all the 2025 stats from the FanGraphs leaderboards. I removed the batters who pitched from the pitchers data, and vice versa (though I kept Shohei Ohtani in both datasets). I also downloaded all the preseason projections and merged them with the 2025 stat rankings. I left any missing projections blank and came back to them later.
Then me rescaled all projections so that they assumed the same competitive environment. Other prominent projection reviews sometimes skip this, but sabermetricians consider it one non-negotiable step. If you don’t do this, the performance of your projections can be dominated by the assumptions you make about the league environment, which are influenced by decisions outside the typical scope of a projection system—for example, decisions about how the baseball is manufactured. If you’re still curious about the league’s environment, such as seeing which projections best take into account the effects of a new rule change, it’s best to save that for a separate exercise.
For demonstration purposes, I rescaled OOPSY’s wOBA projections as follows. I looked at all 2025 Major League hitters with an OOPSY projection. I averaged their wOBA projections, weighted by their actual number of plate appearances in 2025; for OOPSY it was .316. I then rescaled all projections to assume the same .313 league average wOBA for 2025. To perform this rescaling for OOPSY, I subtracted .316 from each wOBA projection and then added .313. To convert ERA to wOBA against Tango’s preferred As a comprehensive measure for pitchers, I divided each pitcher’s ERA projection by the average ERA of their projection system, took the square root, and then multiplied it by the league’s average wOBA for 2025, .313.
I then had to choose a strategy to engage the missing players. Some projection systems cover a broader group of players (e.g. ZiPS), while others target relatively more established players (e.g. THE BAT X). You can project that all missing players are the league average, the MARCEL approach, or slightly worse than the competition average, which is usually slightly more accurate. I have chosen a combination of these strategies, which I explain below.
Finally, I calculated the root mean square errors (RMSE) for each statistic from each projection system. I could also have calculated the mean absolute error, but RMSE is more common. RMSE represents the standard deviation of the residuals, a measure of the typical distance between a prediction and an actual result. I weighted the 2025 RMSE at-bats for bats and the total number of batters faced for weapons. Based on instructions from a conversation with Tango, I also calculated the difference between each projection and a naive projection, where the naive projection assumes everyone has the same projection. To calculate the difference between a projection and a naive projection, I squared both, subtracted the smaller number from the larger number, and then took the square root.
After these steps, the table below shows the RMSE of the preseason predictions for the full sample of hitters and pitchers in 2025 (a lower RMSE is better). The naive prediction assumes that everyone has a projection of the league average. The composite forecast is the average of the four standalone systems (OOPSY, THE BAT For missing players, I calculated a projection that was either league average, 5% worse than league average, or 10% worse than league average. I chose the option that gave the lowest error for each system. Hitters without Steamer were given a projection of the league average. For the other systems, hitters with missing projections received a projection that was 10% worse than the league average. Pitchers without THE BAT or ATC projections were given a league-average projection. For the other systems, missing pitchers were given a projection that was 10% worse than the league average.
Table 1. Prediction accuracy for the full sample.
Across the entire sample, Steamer was the most accurate standalone projection system for wOBA against pitchers – as is often the case. is. ZiPS was clearly the best system for wOBA for hitters. Steamer’s pitching projections were .0216 wOBA better than the naive projection. ZiPS’ hitting projections were .0238 wOBA better than the naive projection. OOPSY held strong, ranking second among standalone systems for both pitchers and hitters. Notably, all systems were closely grouped for both hitters and pitchers. The composite projection, an average of the four standalone systems, was the runaway winner overall for both arms and bats. This is no surprise, just as the average of the ‘experts’ is typical A win approach in terms of forecast accuracy.
Next, let’s take a look at how the rookie projections performed. Note that for rookies, the naive forecast gives everyone a projection that is 5% worse than the league average.
Table 2. Accuracy of forecasts for rookies.
Among standalone systems, Steamer led the way in projecting rookies in 2025, followed by OOPSY. (OOPSY’s most recent list of top prospects, based on long-term projections, can be found here.) Steamer doesn’t publish long-term projections, but I’d like to see them if they ever decide to do so. Once again, the composite projection is the clear winner overall.
Then we exclude rookies and look only at veterans.
Table 3. Accuracy of forecasts for veterans.
For pitchers, the results are comparable to the full sample. For hitters, ZiPS still ranks first among standalone systems, but THE BAT ATC also improves when it focuses only on veterans. THE BAT THE BAT X was too the best in its class last year for hitters in general. My initial explorations show that THE BAT X may contain variables such as sweet spot and sprint speed that OOPSY does not yet contain.
Finally, as a robustness check, the table below looks at OOPSY against each system but excludes missing players. The RMSEs below focus only on players who had a projection of both systems in each comparison. The results are similar to Table 1.
Table 4. Prediction accuracy for full sample, alternative approach.
Overall, I would have liked the best predictions in terms of accuracy, but it wasn’t a disappointing debut season for OOPSY. I’m happy to join the other FanGraphs projection systems. I’ve had the privilege of talking a little bit with each of these forecasters over the years, and I’ve learned a lot from them. I continue to rely heavily on their systems. There is always more to learn in pursuit of small improvements in forecast accuracy. I still need to look at the accuracy of my component predictions to see where the biggest improvements can be made. I’m already planning to consider additional Statcast metrics to potentially help project hitters sweet place, sprint speed, angle of attackAnd horizontal spray tendency. There is less low hanging fruit for pitchers as my projections there have a longer history, but I can tinker with the recency weights and include pitch-level metrics in the projections. However, I’m not the only one who makes adjustments every season. Regardless of the improvements any of us make, users of projection systems will most likely be best served by simply averaging the predictions from the different systems.
#Review #OOPSYs #debut #season