Fantasy drafters systematically overpay for last year's stats and underpay for age-adjusted projections. The gap between average draft position and actual end-of-season value is predictable, exploitable, and wide open every March.
Across 12-team standard 5x5 rotisserie leagues from 2019 through 2024, the average drafter leaves $47 of auction value on the table by overdrafting players whose prior-year stats exceed their projected current-year value. That $47 comes from paying premium prices for names attached to seasons that already happened.
The mispricing works in both directions. Overvalued players (ADP higher than projected value warrants) cost fantasy managers an average of $31 in wasted auction dollars. Undervalued players (ADP lower than projected value warrants) represent $16 in uncaptured upside that goes to managers who draft on projections rather than reputation.
The pattern repeats every year. Drafters anchor to last season's stat line, and the market consistently misprice players whose true talent level has shifted due to age, injury recovery, or role changes. The information exists in public projections. The market ignores it.
Three player archetypes are consistently overdrafted relative to their projection-based value. The most expensive mistake is the aging slugger coming off a career year. Hitters aged 32+ who posted top-20 fantasy seasons get drafted an average of 18 spots higher than their age-adjusted projections suggest. Their batting average, power, and counting stats regress toward career norms, and the aging curve accelerates the decline.
The second archetype is the reliever who posted a sub-2.00 ERA. Closers and setup men with extreme ERA seasons get overdrafted by an average of 22 spots. Reliever ERA is one of the least stable stats in baseball. The year-to-year correlation for reliever ERA is 0.21, compared to 0.58 for starter ERA. Drafters pay for an outlier that has less than a 1-in-4 chance of repeating.
The third is the late-blooming starter aged 30+ who posts a breakout season. Pitchers who have their first elite year after 30 get overdrafted by 15 spots on average. The breakout is often driven by a mechanical change or pitch-mix adjustment that opposing hitters haven't yet solved. By the following season, hitters have made the adjustment. The breakout regresses.
The mirror image of the overvalued list reveals equally consistent patterns. The most undervalued archetype is the young hitter entering his age-26 or age-27 season. Hitters in this age range who posted solid but unspectacular prior seasons get drafted an average of 14 spots lower than their projections warrant. The aging curve peaks for hitters between 26 and 28. Drafters who focus on last year's stats miss the upward trajectory.
| Archetype | Avg ADP Discount | Why Undervalued |
|---|---|---|
| Hitter, age 26-27 | -14 spots | Approaching peak years; prior season understates trajectory |
| SP post-injury return | -11 spots | Injury discount persists after full recovery; stuff metrics stable |
| Team-change hitter | -9 spots | Park factor upgrade not reflected in ADP; new environment underpriced |
| Low-BABIP hitter (26-30) | -8 spots | Bad luck on balls in play suppressed prior stats; xBA suggests rebound |
Pitchers returning from injury are the second-most undervalued group. Starters who missed significant time (60+ days on the IL) but whose stuff metrics (velocity, spin rate, extension) returned to pre-injury levels get drafted 11 spots lower than projections suggest. The market overweights the injury history and underweights the recovery data.
Hitters who changed teams and moved into a better park factor environment are the third. A hitter moving from a pitcher's park to a hitter's park gains 5-8% in expected counting stats from the park factor alone. ADP reflects the player's name and last stat line. It rarely adjusts for where he'll be hitting 81 home games.
To measure the mispricing gap, we compared ADP-implied player rankings against Steamer projection-implied rankings for every season from 2019 through 2024. Then we measured which ranking system better predicted actual end-of-season fantasy value (total rotisserie dollars earned).
Correlation between average draft position and actual end-of-season fantasy value. Decent, but leaves 32% of the variance unexplained. The crowd gets the general shape right and the details wrong.
Correlation between Steamer preseason projections and actual end-of-season value. An 11% improvement over ADP. The gap comes from age adjustments, regression to the mean, and park factor modeling.
The 0.68 vs. 0.75 gap (an 11% improvement in correlation) translates to real draft value. A drafter who follows Steamer rankings instead of ADP consensus gains an average of 2.3 wins above replacement across their roster over a full season. In a 12-team league, that edge is often the difference between a middle-of-the-pack finish and a top-3 result.
ZiPS and PECOTA show similar advantages over ADP, though slightly smaller (0.73 and 0.72 correlations respectively). All three projection systems share the same core advantage: they regress toward the mean, account for aging, and adjust for park factors. The crowd does none of these things consistently.
The mispricing persists because fantasy drafters are subject to well-documented cognitive biases that projections correct for mechanically. Recency bias is the primary driver. Humans overweight the most recent data point (last season) and underweight base rates (career trends, aging curves, league-wide regression). A player who hit .310 last year feels like a .310 hitter, even if his career average is .275 and his expected batting average was .282.
Availability bias compounds the problem. Players who had visible, memorable seasons (home run milestones, All-Star appearances, postseason heroics) stick in drafters' minds more than players who quietly posted strong underlying numbers. A hitter who hit 40 home runs gets remembered. A hitter who posted a .370 wOBA with a 12% walk rate does not, even if the second player projects for more total fantasy value.
Projection systems are boring. They regress everything toward the mean, adjust for age, and spit out a number with no narrative attached. Humans draft stories. The aging slugger's farewell tour, the breakout pitcher's new slider, the prospect who dominated September. Stories are compelling. Projections are accurate. The two rarely agree in March.
The market also lacks a correction mechanism. In real financial markets, arbitrageurs close mispricings by buying undervalued assets. In fantasy baseball, there is no equivalent. Drafters who exploit the mispricing gain an edge in their own league, but their correct valuations don't move the broader ADP consensus. The information stays private. The mispricing stays open.
The exploitable strategy is straightforward. Before your draft, pull the latest Steamer, ZiPS, or PECOTA projections. Convert them to auction values or rank them by total projected fantasy contribution. Compare those rankings to current ADP. Target players where the projection ranks them significantly higher than ADP. Avoid players where ADP ranks them significantly higher than the projection.
The biggest gains come in the middle rounds. First-round picks are priced efficiently because the top 12 players are genuinely elite and projections agree with ADP. The mispricing concentrates in rounds 5 through 15, where name recognition and recency bias create the widest gaps between ADP and projected value. This is where drafters win or lose their seasons.
Every projection system is free. Steamer runs on FanGraphs. ZiPS is on the same site. PECOTA requires a Baseball Prospectus subscription. The edge costs nothing and requires no proprietary data. The only barrier is the willingness to trust a spreadsheet over your gut. Most drafters can't do it. That's why the mispricing will be there again next March.
Average draft position data from NFBC (National Fantasy Baseball Championship) main event drafts, 2019-2024. NFBC represents the sharpest public fantasy baseball market, with buy-ins ranging from $150 to $1,500. Sample: approximately 18,000 total draft picks per year across 150+ drafts.
End-of-season fantasy values calculated using standard 5x5 rotisserie scoring (R, HR, RBI, SB, AVG for hitters; W, K, ERA, WHIP, SV for pitchers). Dollar values derived using the Standings Gain Points method, which converts each counting stat into its marginal impact on league standings.
Preseason projections from Steamer (FanGraphs), ZiPS (FanGraphs/Dan Szymborski), and PECOTA (Baseball Prospectus). Correlations measured using Pearson's r between preseason projected dollar values and actual end-of-season dollar values. Archetype analysis based on filtering players by age, position, prior-year stats, and team context, then measuring systematic ADP deviation from projection-implied rank.
Aging curve data from FanGraphs' publicly available hitter and pitcher aging studies. Reliever ERA year-to-year correlation from a 2023 analysis of all qualified relievers (50+ IP) from 2015-2022. BABIP regression rates from Tom Tango's foundational work on luck vs. skill in batting statistics.
