The lead-back rule is useful because it forces the right first question: who actually owns the work? It is not useful when a column answers that question without a depth chart, injury report, or usage row.
Data limit: the evidence ledger for this post contains no rows. There is no player projection, market quote, injury status, or consensus rank to compare, so this article publishes no named start or bench.
The lead-back rule is a filter, not a result
A running back moves toward the lineup when his role is clear across early downs, passing downs, and scoring-area work. A player moves away from the lineup when those jobs are split, uncertain, or limited by health. That is the framework. The current player still has to be identified with current evidence.
“Lead back” should mean more than the first name on a preseason depth chart. Check who is active, who stays on the field when the offense can pass or run, and who receives the high-value touches. A player can lead a committee in nominal carries and still lose the work that matters most for fantasy scoring.
The RB rankings can apply the rule once live rows exist. This module does not backfill them from reputation.
Game script comes after workload
A favored team may have more chances to run late. An underdog may throw more. Those are reasonable game-level hypotheses, not automatic player outcomes. A back with receiving work can survive a trailing script; a nominal starter can disappear in a win if the committee takes the valuable snaps.
That is why the order matters. Verify the role. Then inspect the market. Then ask whether the likely script strengthens or weakens the specific workload. Starting with the spread and assigning touches afterward is backwards.
The odds board supplies current team context. It does not supply the backfield split.
Consensus is not a stand-in for missing evidence
The old version of this article claimed gaps between “our projection” and consensus without publishing either row. That language is gone. A comparison needs two named values drawn from two named sources. Otherwise the gap is an adjective.
Consensus can be useful as a map of market expectations. It can show where a player’s reputation is carrying the rank. But disagreement is not automatically an edge. The independent projection still has to be built, calibrated, and available before the lineup locks.
The model-versus-consensus guide explains the comparison. This page does not pretend the comparison happened when the evidence ledger is empty.
How to make the actual start/sit call
Open the current roster and compare two real players. Check active status and injury limitations. Confirm the backfield role. Account for league scoring, especially receiving points. Review the team market and likely pace. Then choose the player with the stronger combination of opportunity and floor.
When the inputs conflict, name the conflict. A healthy committee back on a strong offense may still trail a receiving back on a weaker team. A starter returning from injury may carry a famous role and a limited workload. The framework is supposed to expose those tradeoffs, not hide them behind a confidence label.
The start-sit tool is where those current rows belong. The fantasy desk is the final check before lock.
What would change this empty state
A current player feed would. The minimum useful row needs the player, team, opponent, availability, role evidence, scoring format, and market timestamp. A consensus comparison also needs a named consensus rank or projection from the same decision window.
Until those inputs arrive, the honest Week 1 running-back advice is procedural: verify who owns the work, use game script second, and refuse to bench or start a name the data never supplied.
This article is decision support, not a player recommendation. The empty evidence state is intentional.
DFS outcome leverage versus recorded ownership
This chart remains empty until a verified source binds ownership projections to settled lineup outcomes for the same contests.
Prop hit rate versus recorded line distance
This chart remains empty until a verified source binds a player projection distribution, the offered prop line, and the settled result.
