Why Tiers Beat Rankings: The Draft Method the Sharps Actually Use

The Problem With a Ranked List

Rankings imply a precision that fantasy football doesn't have. A list says the 14th-ranked receiver is better than the 15th — but it can't say by how much, and that's the only question that matters when you're on the clock. Sometimes two adjacent players are functionally identical; sometimes there's a canyon between them. As RotoWire's tiers primer defines it, a tier is a group of players with roughly interchangeable value — a value cluster rather than an exact order. Athlon's tier-based rankings guide makes the sharper point: rankings alone can lie to you, because back-to-back names on a list are sometimes not remotely close in value. Tiers exist to show you where the real drop-offs live.

Where Tiers Come From

You can build tiers by feel — group players until the next name down feels like a real step back — but the concept has a rigorous pedigree. Data scientist Boris Chen, whose tier charts became a community institution, feeds expert consensus rankings from FantasyPros into a Gaussian mixture model, a clustering algorithm that finds players the experts rank similarly and draws the tier lines where the data naturally breaks. The math matters less than what it proves: the drop-offs are real, detectable structure in player value, not vibes. When dozens of independent experts all leave a gap between the same two players, that gap is information. Modern tier tools like Draft Sharks' annually updated tiers apply the same idea — find the cliffs, then draft around them — and any manager with a spreadsheet and this year's projections can build a serviceable version at home.

How Tiers Change Your Picks

The practical rule is simple: count the players left in the tier, not the names. If your running back tier has five players left and you don't pick for a while, you can take a receiver now and still catch the tail of that tier next round — waiting is free. If the tier has one player left, the cliff is one pick away, and that player is worth taking a little "early" by strict rankings. This is how tiers defuse the two classic draft-room panics. Positional runs stop scaring you: when four receivers go in five picks, tiers tell you whether the run actually cost you anything or just cleared out a tier you'd already priced. And as Lindy's tier guide notes, tiers give you the confidence to let a player go when equal value remains — the discipline that rankings-anchored drafters never find.

Building Your Own in Twenty Minutes

You do not need a clustering algorithm to get most of the benefit. Take any projection set in your league's scoring format, sort one position descending, and put the projected points in a column beside the names. Then add a second column: the difference between each player and the one below him. Almost all of those gaps will be small and similar — a tenth of a point, half a point — and every so often one will be several times larger than its neighbors. Draw a line there. That is a tier boundary, and you have just done by eye what the Gaussian mixture model does formally.

Two refinements make the result genuinely usable. First, do it in your own scoring, not generic PPR — half-PPR moves pass-catching backs down and touchdown-dependent ones up, which relocates the cliffs. Second, do not let a tier grow past about eight players; if a stretch of the board has no meaningful gap for twelve names, the honest read is that the whole stretch is one commodity pool, and the pick to make there is the one that fills a scarcer position instead.

Every Position Has a Different Tier Shape

This is the part that surprises people who build tiers for the first time: the positions do not look alike at all. Running back and receiver boards tend to produce a run of small, tight tiers near the top and then long flat plateaus — many near-identical options, which is why waiting is so often free at those positions. Quarterback and tight end typically produce the opposite shape: one or two very small tiers separated by an enormous cliff, then a plateau that stretches most of the way through the startable pool.

That shape is the entire quarterback and tight end strategy debate, expressed as data rather than opinion. If the cliff after the top tight ends is large, paying up for one is defensible; if this year's board has a small gap and a twelve-deep plateau behind it, the same pick is a waste of premium draft capital, and waiting on the position is not a contrarian stance but simply what the board says. Re-check the shape every season — it changes, and last year's correct answer is not evidence about this year's.

Where Tiers Mislead

Two honest limits. Tiers built from expert consensus inherit consensus blind spots — when the field is collectively wrong about a player, the tiers are confidently wrong in the same place, and the clustering makes that error look like structure. And tier lines are softer than the boxes suggest: a player at the bottom of one tier and a player at the top of the next are usually much closer than two players at opposite ends of the same tier. Use the boundaries as a decision aid, not a fence. The moment you find yourself reaching a full round early because a name sits one row above a line, the tool has started using you.

Tiers Are an Opportunity-Cost Tool

The deepest version of the idea: your best pick is usually the one that saves you from a cliff, even when another position offers a slightly higher-ranked name. Taking the last player in a tight tier at one position, then catching a still-deep tier elsewhere next round, banks value at both spots — that's the whole game. It also pairs naturally with ADP as a price signal: tiers tell you where the cliffs are, ADP tells you when the market will push players off them. A tier ending at pick 40 with three players holding ADPs in the 50s is a gift — you can wait. A tier whose last member goes ten picks early in your league's mocks is a warning. Together they turn drafting from a list-reading exercise into an actual decision framework.

• •Build or borrow tiers for every startable position, in your league's scoring format
• •On every pick, ask one question first: which tier is about to die?
• •Never reach past a cliff — grab the last player in a shrinking tier before it closes
• •When a tier is deep, draft the other position and catch the tier on the way back
• •Re-tier the week of your draft; camp news moves the cliffs