Projections vs. Rankings: What to Trust on Draft Day

Two of the most common tools in any drafter's toolkit — projections and rankings — are often used interchangeably, as though they're saying the same thing in different fonts. They're not. Understanding how each one is built, what it can and cannot tell you, and when to override one with the other is the difference between a draft that holds up in Week 10 and one that falls apart by Week 4.

Definition and scope

A projection is a statistical forecast: a specific predicted output, usually expressed as fantasy points or counting stats (rushing yards, strikeouts, goals). Projection systems use historical data, snap counts, target share, park factors, line combinations, or any number of measurable inputs to estimate what a player will produce over a season or a given week.

A ranking is an ordered list. It takes projections — or expert judgment, or both — and converts them into a priority sequence: Player A before Player B before Player C. The ranking may also factor in positional scarcity, injury risk, variance, and draft context in ways that raw projections do not.

The scope of both tools covers every major fantasy sport. Platforms like ESPN, Yahoo, and NFL.com publish their own rankings, while dedicated projection systems — FantasyPros aggregates consensus projections across more than 100 analysts — translate those numbers into tiered consensus rankings. The two outputs answer different questions: projections ask how much, rankings ask who first.

How it works

Projection models vary considerably in their methodology. The most widely referenced systems in fantasy football, including those aggregated by FantasyPros, combine multiple analyst models and weight them by historical accuracy. In fantasy baseball, Steamer and ZiPS (hosted at FanGraphs) use regression toward the mean, aging curves, and batted-ball data.

Rankings are downstream of projections, but they don't stop there. A player projected for 220 fantasy points might be ranked lower than a player projected for 210 points because of injury history, offensive line concerns, or the fact that the first player is the 14th-best quarterback in a 12-team league — meaning his surplus value is minimal. That context is the work rankings are doing.

The relationship between projections and rankings can be summarized in three operational steps:

1. Build the point estimate — a projection system outputs expected fantasy points per position.
2. Apply scarcity adjustments — positional scarcity determines how much a player's projected value is worth relative to the replacement-level player available on the waiver wire.
3. Rank by adjusted value — the final ranking reflects both projected output and draft-context factors like ADP, roster construction, and format settings.

For a deeper look at how ADP and consensus rankings interact during the live draft process, that relationship becomes particularly visible in the middle rounds.

Common scenarios

Trusting projections over rankings: In auction drafts, raw projections often matter more than ranked order because every player is theoretically available simultaneously. A projection showing a running back at 1,400 rushing yards is a concrete number to anchor a bid against. Rankings, which presume a draft sequence, are less structurally relevant in the auction draft format.

Trusting rankings over projections: In the final 5 rounds of a snake draft, roster need and positional depth matter more than whether a player's projection is 80 or 95 fantasy points. Rankings that account for positional depth — particularly for streaming-friendly positions — are more actionable than a raw number. The draft-day cheat sheet format is designed exactly for this: filtered, position-aware priority lists rather than raw stat lines.

When they conflict: A player projected for strong numbers but ranked 15 spots lower than projection implies is usually flagged for injury risk, offensive role uncertainty, or an inflated market price (ADP far above true value). The ranking is often encoding information the projection hasn't fully priced in.

Decision boundaries

The practical rule is straightforward: use projections to evaluate players in isolation, use rankings to decide between players in context.

A quarterback projected for 320 passing yards per game is meaningfully better than one projected for 260 — that's a projection doing its job. But whether to take that quarterback in Round 5 versus a wide receiver with a similar point projection depends on roster construction, positional scarcity, and what the back half of the draft looks like. That's a ranking question.

Four boundaries where the distinction becomes operationally significant:

1. Format matters — Half-PPR and full-PPR settings shift rankings meaningfully for pass-catching backs and slot receivers without changing raw projections by the same degree.
2. Sample size limits projections — First-year players, trade acquisitions, and injury returnees have thin historical data. Rankings often encode analyst judgment that projection models can't capture.
3. Variance is invisible in projections — A wide receiver with a 150-point projection and high weekly variance is a different asset than one with the same projection and consistent target volume. Rankings sometimes reflect this; projections rarely do.
4. Late-round strategy shifts the calculus — The late-round draft strategy framework treats upside and volatility differently than a projected point total alone can express.

The Draft Day Authority home page provides the broader framework for how these tools fit into a complete draft preparation workflow, from initial research through live-draft decisions.

Projections are the evidence. Rankings are the argument. The best drafters know which one they're consulting at any given moment — and why.