Start/Sit Rankings: How Weekly Decision Tools Are Built
Start/sit rankings answer a narrower question than most fantasy tools — not "who is the best player?" but "who should dress for this specific matchup, this specific week?" That distinction makes them one of the most consequential and most misunderstood tools in fantasy sports. This page breaks down what start/sit rankings actually measure, how the underlying models are constructed, where they diverge from season-long rankings, and where the math genuinely cannot help.
Definition and scope
A start/sit ranking is a weekly, matchup-adjusted ordering of players at a given position that tells a manager which rostered option offers the highest projected output for a single game or slate of games. The scope is deliberately narrow: one week, one lineup decision, one opponent.
This is meaningfully different from the season-long rankings covered in how fantasy rankings work at a foundational level. A player ranked WR12 for the full season might be a WR6 start in Week 10 against a defense allowing the most receptions to receivers, or a WR22 start in Week 14 on a short week against a cornerback running a 92% shadow coverage rate. The season-long number averages across all of that. The start/sit number tries to isolate it.
Most major platforms — ESPN, Yahoo, Sleeper, FantasyPros — publish start/sit rankings by Thursday each week for NFL, with baseball and basketball equivalents updated daily given those sports' compressed schedules.
How it works
Start/sit models layer three categories of input on top of a baseline player projection:
- Opponent defensive ranking at position — How many fantasy points has the opposing defense allowed to quarterbacks, running backs, wide receivers, or tight ends over a trailing window, typically 4 to 6 weeks, weighted toward recent performance.
- Game environment variables — Projected game total (a Vegas over/under of 48.5 signals a shootout; 40.5 signals a slog), implied team total, weather for outdoor stadiums, and spread. A team favored by 10 points will likely run the ball more in the fourth quarter, which compresses passing volume for receivers.
- Usage and opportunity share — Target share, snap rate, air yards allocation, and red zone touches. These inputs come from the kind of granular tracking data explored in target share and snap count rankings. A receiver with a 28% target share on a team with a 34-point implied total is being set up very differently than one with a 12% share on a 21-point team.
The model outputs a projected point total — not a rank — and ranks are derived by sorting those projections within a position group. The rank is a byproduct of the math, not the starting point.
Injury reports interact with this process in a cascading way. A starting running back verified as questionable might receive a probability-weighted projection — say, 65% chance of playing — which deflates his expected points and may elevate his handcuff or a streaming option above him. Injury impact on fantasy rankings covers how that probability adjustment propagates through a depth chart.
Common scenarios
The bye-week pivot. A roster loses its WR1 to a bye, leaving a choice between two lesser options. Start/sit rankings contextualize those options against actual opponents rather than abstract value — a WR3 facing a defense ranked 28th against receivers may project better than a WR2 on a short week against a top-5 cornerback.
The streaming decision. A manager needs a tight end or defense and holds two streaming-grade options. This is where matchup data does the most work, because streaming decisions are almost entirely opponent-driven rather than talent-driven. The waiver wire rankings tool handles the acquisition side; start/sit handles the deployment question once both options are rostered.
The uncomfortable sit. A stud player draws an elite opposing cornerback or a historically stingy run defense. The model may rank him lower than his reputation suggests. Managers routinely override these signals — a top-5 receiver rarely drops off the starting lineup — but the data provides at least a structured basis for the decision.
Decision boundaries
Start/sit rankings are useful. They are not oracular. Three limits define where the tool's reliability drops:
The floor problem. Projected points are means, not guarantees. A running back projected at 14.2 points carries a wide distribution — an early fumble, a game script blowout, or a late-game injury could land him at 4 points. Elite players carry higher floors even against bad matchups, a concept central to tier-based drafting strategy that applies equally to weekly decisions.
Recency weighting versus sample size. A defense that has allowed 40+ fantasy points to receivers over the past 3 weeks may have faced three of the top-5 receiving offenses in sequence. The matchup looks soft; the data may be misleading. Models that weight recent weeks heavily — common in most consumer-facing tools — can overcorrect.
The information edge problem. By the time start/sit rankings publish on Thursday, sophisticated managers have already processed the same injury reports, weather forecasts, and Vegas lines. The edge from publicly available start/sit tools is real for managers who aren't tracking those inputs manually, but it shrinks as the week progresses and information becomes uniform. For a deeper look at how to evaluate whether any rankings tool is actually adding signal, fantasy rankings accuracy and evaluation provides a structured framework.
The full landscape of weekly decision-making — from waiver pickups through playoff scheduling — connects through the fantasy rankings home, where start/sit sits alongside the rest of the decision stack.