The ADP vs. Projections Debate
Fantasy baseball owners have long debated the merits of two primary ranking tools: average draft position, or ADP, which reflects the collective consensus of thousands of participants, and player projections, which are derived from statistical models that forecast future performance. While ADP captures the wisdom of the crowd, it is also shaped by human biases, injury concerns and short‑term trends that can skew perception.
To evaluate the reliability of each method, the analysis examined year‑to‑date player value across three popular scoring formats — standard 5×5 roto, ESPN points and Yahoo points — using data sourced from FantasyPros for ADP and from FanGraphs for projections. Spearman rank correlations were calculated to measure how closely each ranking system matched actual performance, and Steiger’s Z‑Test confirmed that the correlation differences were statistically significant.
Statistical Evidence Supports Projections
The results were clear: projections demonstrated higher correlation coefficients with real‑world value in every setting, and the advantage grew stronger in the later portions of drafts. In particular, the model identified a larger share of effective relievers than ADP, highlighting a niche where projection algorithms excel.
These findings suggest that relying on projections can help managers uncover undervalued assets, especially when the pool of elite talent thins out. By integrating model‑based forecasts with traditional draft strategies, participants can reduce the influence of anecdotal bias and make more informed decisions that maximize roster productivity.
Beyond the correlation metrics, the study visualized the relationships through scatterplots that plotted ADP‑derived rankings against projected rankings, revealing tighter clustering around the diagonal for projection‑based rankings. This visual confirmation reinforced the numerical evidence that projections are not merely a supplement but a more predictive foundation for draft planning.
The implications extend beyond individual roster construction; they also inform broader strategy discussions about when to prioritize power hitters versus reliable relievers, and how to allocate budget in auction formats. As fantasy leagues continue to evolve, the data‑driven approach championed by projection models is likely to become an even more central component of competitive play.