The one move
Ask what a pick actually buys. Not "how many points does this player score" — that question has no counterfactual in it and therefore no answer. Ask instead: what happens to my season if I do not take him?
The answer is never "I score nothing at that spot." Your lineup forces you to start a player at every slot every week, so if you skip this one you start someone else. The pick buys the difference between the two. That difference is the only thing that ever changes your record.
Replacement level is the scoring rate of the someone-else — specifically, the best player at that position you could obtain for no draft cost at all, because nobody else wanted him either. Subtract it from a player's projection and you get his real price tag. That subtraction is what VOR does.
The word free is doing all the work. Free means available on the waiver wire, or gettable in the last round nobody cares about — not "the worst starter" and not "zero." A baseline of zero is what raw point totals silently assume, and it is why they rank a position everyone has above a position almost nobody has.
Where the line falls
The baseline is not a judgement. It is arithmetic on the league's own settings: league size × starters per team at that position.
In this league, ten teams each start one tight end, so the tenth-ranked startable tight end is the conservative season-long baseline. Ten teams × two RB slots creates twenty starting RB slots before FLEX; because FLEX absorbs mostly RB/WR, the actual model line is deeper: RB24 and WR26, versus TE10 and QB10. The exact allocation is league-specific, documented in League size and replacement level, not a universal football law.
Two things follow that beginners get wrong in opposite directions.
The baseline is a rank, not a quality. It does not matter whether the players near the line are good at football. It matters only that they are unowned. A position can be full of talented players and still have a high baseline, which makes each of them cheap.
The baseline is per-position, and the positions do not share. There is no single league-wide replacement level. Each slot on your lineup card has its own, and the gaps between them are enormous. This is the whole reason a position that produces the biggest raw numbers can be the last thing you draft.
Change the lineup, move the line
Because the baseline is computed from the lineup, the lineup is what controls it. Nothing about the players changes when the settings change; their value changes anyway.
Add a second starting slot at a position and you employ ten more of them league-wide. The line slides down into thinner air, the free option gets worse, and every player at that position becomes more valuable — including the ones you already own. Take a slot away and the reverse happens instantly.
Grow the league from ten teams to fourteen and the same thing happens at every position at once. In a shallow league the wire is stocked with startable players and almost nobody is worth a premium; in a deep one the wire is a desert and ordinary starters become assets.
So the same player is worth different amounts in different leagues, and this is not a rounding error. A player who clears his baseline by a wide margin in one format can be barely above it in another, with an identical projection. When you read a ranking from the internet, you are reading someone else's lineup settings as much as their opinion about football.
The football-specific twist: FLEX
Here is the piece that does not exist in most people's mental model, and the one that makes a naive baseline wrong in a predictable direction.
The FLEX slot (a lineup spot that accepts a running back, receiver or tight end) does not belong to any one position. Ten teams each fill it every week, and they fill it with whoever is best — which in practice means mostly receivers and running backs, because those two positions produce far more usable bodies than tight end does.
That extra demand is real employment. It is not two running backs per team any more; it is two plus a share of the tenth-of-a-position that FLEX represents. The running back and receiver baselines therefore sit deeper than the raw starter count implies — RB24 and WR26 in this model. Tight end barely moves, because tight ends rarely win a FLEX competition.
The practical effect: FLEX quietly makes the two positions it feeds on scarcer, without appearing anywhere on the lineup card as a slot belonging to either. Miss this and you set both baselines too shallow, which flatters every running back and receiver on your board.
What transfers from basketball, and what does not
The concept transfers completely. You already compare a player to what the wire hands you free, and you already know that the gap is the asset. Nothing about that changes.
Three things do change, and each makes replacement level matter more here rather than less.
Positions are rigid. Multi-position eligibility blurs baselines in basketball — a player who qualifies at two spots is being measured against two different lines. Here nobody is a running back and a receiver. One line each, hard.
One summed number, no categories. There is no category-specific replacement level to reason about, no "replacement-level source of blocks." The baseline is a single scoring rate, which makes the subtraction cleaner and the answer blunter.
The spread between baselines is much wider. In football the distance between the free option at the scarcest position and the free option at the most abundant one is large enough to decide a draft on its own. Ignoring baselines in basketball costs you some efficiency; ignoring them here inverts your board.
Open questions
- The FLEX share between running back and receiver is assumed rather than observed, and it moves both baselines by a few ranks. The direction is certain; the magnitude is not.
- Baselines computed from season-long projections assume every player is available every week, which no position obeys — an injury-prone position's true free option is worse than its nominal one.
- The wire restocks unevenly across positions, so "freely available" is genuinely a weekly quantity at some positions and a preseason one at others. A single baseline flattens that.