Most Accurate Fantasy Baseball Projections (2025 Results)

With the Super Bowl behind us, it’s officially time for Baseball. Pitchers and Catchers report next week, and fantasy baseball drafts are right around the corner. Compared to other sports, Baseball lends itself particularly well to analytics thanks to its large number of games, relatively predictable playing time, and the 1v1 nature of batting. As a result, Baseball has a very mature and competitive projection industry.

Most Accurate Fantasy Baseball Projections

The ultimate goal of FantasyPros is to help you win your fantasy league. The foundation of a winning strategy is a solid understanding of how you should expect each player to perform. That’s where projections come in. There are many good projection systems, and we’re here to sort out the best. Welcome to the results of our 2025 MLB Fantasy Projection accuracy study. Let’s explore the most accurate fantasy baseball projections of 2025.

Methodology

In the broader baseball context, the purpose of projections is to forecast each player’s contribution to winning real baseball games, and to quantify the value we expect each player to provide their team. But for fantasy, we don’t care about that. We care about their contribution to winning your fantasy league, and as a result, we emphasize different areas than other accuracy studies. Our accuracy methodology can be described as a 3-step process:

1. Select the right data: Since we’re here to win fantasy leagues, we focus on the stats with a direct impact. That is, playing time, and the standard 5×5 categories. For hitters this means R, HR, RBI, SB, AVE and AB; for pitchers W, SV, K, ERA, WHIP and IP. This is rather different from other accuracy analyses which often normalize playing time and tend to ignore stats like W and SV that are reliant on factors external to the player’s talent. Since we explicitly include AB and IP to evaluate playing time projections, other stats are measured on a per-AB/per-IP basis to isolate their effectiveness at predicting quality of play.
2. Evaluate each projection: We care about how each projection system performs relative to the alternatives, so we measure accuracy in that context. This means we measure the error of each projection, and then convert it into a z-score in the context of all the projection systems. In 2025, Kyle Schwarber hit 56 home runs on 604 at-bats, a rate of 9.27%. Suppose you projected him to have just 42 homers on the same number of at-bats, a rate of 8.28%. Your error for that projection is (.0927-.0828) = .0099, but we don’t stop there. The projection systems we looked at recorded an average error of 0.023 with a standard deviation of 0.008. So after converting to a z-score, your error for Schwarber’s HRs is (0.0099-0.023)/0.008 = -1.638. That means you projection was excellent – it wasn’t just better than average, it beat the average by 1.6 standard deviations.
3. Aggregate the scores: Another benefit of converting errors to z-scores is that it puts them all on the same scale. For each position we take the average error z-scores across the whole player pool to get a combined score for each category. Then we average together category scores separately for hitters and pitchers, and finally add together each projection’s system’s hitter and pitcher scores to get an overall score.

Accuracy Results: Most Accurate Fantasy Baseball Projections of 2025

Let’s get to the results! Don’t forget that scores represent error so lower is better, and since they’re based on z-scores, zero is average by definition.

For the first time since 2018, FantasyPros’ own Zeile Consensus Projections have come out on top, with Ariel Cohen’s excellent ATC Projections close behind. These systems are two sides of the same philosophical coin. Zeile is a straight-up average of most of the other projection systems you see here – it’s the wisdom of the crowd, implemented as straightforwardly as possible. ATC takes a different approach to the same idea. It is also a consensus of other projection systems, but it’s not a simple average. Instead, it weights the sources for each category according to historical performance. It’s a “smart” consensus, so to speak. Historically both approaches have been very successful. This year they scored similarly in most categories, but Zeile pulled ahead thanks to having the most accurate projections for Saves and Wins, two categories that many systems don’t focus on, but are crucial for fantasy.

While Zeile and ATC are both aggregate systems, they could not exist without the original systems they aggregate, built from historical data, smart modeling and elbow grease. Derek Carty’s THE BAT is the best of the best original projection systems. Carty’s projections consistently rank as the best original system, and last year even took the #1 spot ahead of the aggregate systems. In recent years he has also published THE BAT X for hitters, which adds another layer of complexity by incorporating statcast data. This has been very successful and is typically an improvement, but this year THE BAT (original flavor) actually came out ahead, thanks to a noticeably better result for Home Runs.

Category Results

The overall results are important, but you wouldn’t be reading this if you weren’t interested in getting nerdy about it. So here’s the breakdown of every system’s score for each of the 12 categories we measure. Starting with hitters:

And for pitchers:

This concludes our presentation of projection accuracy results for 2025. If that’s got you in the fantasy baseball mood, head over to the Mock Draft Simulator to start prepping for the 2026 season!

—

Jacob Herlin is a Data Scientist for FantasyPros. You can find him on Bluesky.