Floor and Ceiling Projections: Measuring Upside and Downside Risk
A single projected point total tells only half the story. Floor and ceiling projections capture the range of likely outcomes around that midpoint — the low-end baseline a player is expected to reach in an unfavorable game, and the high-end explosion in a favorable one. Understanding both numbers changes how a manager drafts, starts, and builds lineups across redraft, DFS, and best-ball formats.
Definition and scope
A floor projection represents the lower boundary of a player's realistic outcome distribution — not the absolute worst case, but the performance level that should occur in roughly 10–20% of the worst scenarios without invoking injury or a completely anomalous game. A ceiling projection sits at the opposite end: the production achievable in the top decile of outcomes, where usage spikes, touchdowns accumulate, and nothing goes wrong.
These numbers are distinct from a median projection. A running back with a 14-point median, an 8-point floor, and a 28-point ceiling has a different risk profile than a receiver posting identical median projections but an 11-point floor and a 22-point ceiling. The first player offers lottery-ticket upside; the second offers consistency. Both can be correct simultaneously — they just answer different questions.
Floor and ceiling values are closely related to the concept of projection confidence intervals, which express the same underlying uncertainty in probabilistic terms rather than labeled endpoints.
How it works
Projection systems derive floors and ceilings by modeling outcome variance alongside the point estimate. The mechanics typically follow a three-step structure:
-
Build a distribution of inputs. Historical carry totals, target shares, snap counts, and scoring rates are assembled across comparable game contexts. Snap count and target share data is foundational — a receiver who sees 6 targets in a bad week and 14 in a great one has a wider natural spread than one who sees 9 almost every game.
-
Simulate across scenarios. Monte Carlo simulation or percentile-based modeling runs thousands of game outcomes, each drawing from realistic ranges of the key inputs. A quarterback's ceiling week might combine above-average completion rate, 2 rushing yards that accidentally become a scramble score, and a game script that forces trailing comebacks. The simulation captures those co-occurrences.
-
Extract the percentile endpoints. Floor is often pulled from the 10th or 15th percentile of simulated outcomes; ceiling from the 85th or 90th. The precise cutoffs vary by system — comparing projection systems reveals meaningful differences in how aggressively endpoints are set.
Usage rate adjustments in projections directly affect the width of the resulting spread. High-usage players who touch the ball 25 times per game have a narrower floor-to-ceiling gap than boom-or-bust receivers who might see 3 targets or 11 depending on game flow.
Common scenarios
Best-ball formats are where ceiling projections dominate decision-making. Because lineups are set automatically using the highest scorers each week, the median is less valuable than the probability of a top-end explosion. A slot receiver who averages 11 points but occasionally reaches 32 is structurally more valuable in best-ball than a tight end who reliably posts 13. Best-ball projections are built around ceiling weighting for exactly this reason.
Cash game DFS inverts the logic entirely. In head-to-head contests and 50/50s, a player posting 30 points when a manager needed 22 provides no additional value. The floor becomes the operative number — the minimum acceptable output that keeps the entry in the money without requiring a ceiling performance. A strong floor player in a $10 cash game is worth more than a high-variance play who cracks 40 points one night and 6 points the next.
Injury-risk positions present a floor compression problem. Running backs behind thin offensive lines can post an 8-point floor not because of usage floor but because a single heavy hit resets their involvement mid-game. Injury adjustments in projections attempt to factor this in, but it's one reason running back floors are structurally lower than positional medians might imply.
Weather-affected games can simultaneously compress ceilings for passing games while elevating rushing floors. A wide receiver facing a 30 mph wind in a December road game may see his ceiling collapse by 40% while a workhorse back in the same game sees his floor actually rise. Weather impact on fantasy projections quantifies these adjustments by game-time conditions.
Decision boundaries
The floor-ceiling gap creates three actionable thresholds for lineup construction:
- Narrow gap (under 12 points): Indicates a high-floor, low-variance player — ideal for cash formats and safe tournament fills when a roster already has enough upside elsewhere.
- Moderate gap (12–20 points): The most common range for starters. Useful as a baseline and requires evaluating the shape of the distribution — whether outcomes cluster near the median or skew toward one end.
- Wide gap (20+ points): Classic boom-or-bust profile. These players are tournament weapons in DFS and best-ball, but represent meaningful roster risk in season-long weekly starts when a floor outcome likely costs a matchup.
The choice between a safe floor and a high ceiling isn't just a personality preference — it's a function of format, roster construction, and what scoring format impact on projections means for that specific league's point values. A PPR league compresses variance for volume receivers; a big-play bonus league widens ceilings for explosive athletes. The same player, two different risk profiles.
When building a complete picture of projection methodology, the Fantasy Projection Lab home situates floor and ceiling analysis within the broader framework of how point estimates are generated, validated, and applied across formats and positions.