Methodology

How we read the numbers.

Most products show you stats. The edge is in reading them. We separate anchors, stable, predictive inputs like opportunity, volume, accuracy, and quality of contact, from noisy outputs like efficiency, results, and touchdown rate that regress to the mean. An accurate projection leans on the anchors and treats the noise as a sample, not a skill. Here's every signal we use, and exactly how we read it.

14 metrics tracked8 anchors12 live · 2 queued

Coverage Map

Stats we have that closed-box products do not show

Only clearance-approved rows render here. If a source cannot clear commercial derived analytics, the row stays out of the public claim surface.

1 cleared1 tier 11 sources0 withheld

Opp-adjusted EPA (our DVOA)

Tier 1: open play-by-play derived

cleared · priced=false

Closed-box gap: Proprietary efficiency ratings (DVOA-style, opponent-adjusted) are paywalled and their methodology is closed.
Transparent equivalent: Solve epa ≈ leagueMean + offense + defense by iterative coordinate descent over open play-by-play, re-centred each pass: a transparent, reproducible opponent adjustment.
Attribution: Data from nflverse (https://github.com/nflverse), CC-BY-4.0

Receiving

Weighted Opportunity RatingWOPR

Anchor: stable & predictive

live

1.5 × target share + 0.7 × air-yards share

What it is: A single number for how much of the passing game runs through a receiver: volume and depth combined.
How we read it: The leading indicator of receiver fantasy points. Opportunity is sticky week to week, so high WOPR forecasts production even before the box score catches up.
Commonly misread: Chasing last week's yards or touchdowns, outputs that bounce around, instead of the role that generates them.
Our edge: We rank by WOPR and flag receivers whose opportunity outruns their production as buy-lows before the market reprices them.

See it live →

Target Share

Anchor: stable & predictive

live

What it is: The percentage of a team's targets that go to a player.
How we read it: The cleanest claim on volume. A rising target share is a role change, the most actionable waiver signal there is.
Commonly misread: Treating a one-week target spike as a trend instead of waiting for the role to stabilize.
Our edge: We weight share by recency and pair it with air-yards share (WOPR) so a checkdown machine and a field-stretcher aren't valued the same.

See it live →

Air-Yards Share / aDOTAY%

Anchor: stable & predictive

live

What it is: A player's share of the team's intended air yards; aDOT is the average depth of his targets.
How we read it: Depth = upside. Two receivers with equal targets but different air-yards share have very different ceilings and TD equity.
Commonly misread: Ignoring depth entirely and ranking on catches, which over-values low-aDOT possession roles.
Our edge: Air-yards share is half of our WOPR and the tiebreaker between a safe floor and a league-winning ceiling.

See it live →

Receiver Air Conversion RatioRACR

Noisy: regresses; read as a sample

live

receiving yards ÷ air yards

What it is: How many yards a receiver turns each intended air yard into (YAC + accuracy + luck).
How we read it: An efficiency descriptor, not a forecast. Extreme RACR regresses hard, so we read it as a sample of context, not a repeatable skill.
Commonly misread: Extrapolating a hot RACR as if it'll continue. It's the single biggest source of false breakouts.
Our edge: We use RACR to explain a result, never to project one; the opportunity metrics carry the forecast.

See it live →

Rushing

Rush Yards Over Expected / AttRYOE

Noisy: regresses; read as a sample

live

What it is: Yards a back gains per carry above what a tracking model expects from the blocking and box.
How we read it: The ceiling, not the floor. Per-carry efficiency is real talent but regresses fast, so we treat hot RYOE as regression-prone unless it's earned against loaded boxes.
Commonly misread: Drafting last year's efficiency leaders and expecting it to repeat. RB efficiency is one of the least sticky stats there is.
Our edge: We separate efficiency from volume and only trust RYOE that's earned vs stacked fronts; otherwise volume carries the projection.

See it live →

Rush Volume / Touches

Anchor: stable & predictive

live

What it is: Carries (and total touches) a back actually gets.
How we read it: The RB floor. Volume is coach-driven and sticky. It's the single most predictive RB input, full stop.
Commonly misread: Falling for an efficient backup over a plodding bell-cow. Points come from touches, not highlight runs.
Our edge: For RBs we lead with volume and treat efficiency as the swing factor, the inverse of how we read receivers.

See it live →

Stacked-Box RateBox%

Signal: useful with context

live

What it is: Share of a back's carries against eight or more defenders in the box.
How we read it: Context for efficiency. Positive yards over expected against loaded boxes is real; the same number on light boxes is schemed and fades.
Commonly misread: Reading raw yards per carry without the difficulty of the fronts that produced it.
Our edge: We grade efficiency through box difficulty, so we don't reward a back the defense simply wasn't respecting.

See it live →

Passing

Completion % Over ExpectationCPOE

Anchor: stable & predictive

live

What it is: How much more often a QB completes a pass than a tracking model expects, given depth, pressure, and coverage.
How we read it: The most stable public measure of QB accuracy. It's a skill that persists, so it's a real input to QB quality.
Commonly misread: Reading raw completion % (which rewards dink-and-dunk) as accuracy.
Our edge: We triangulate CPOE against results-based QBR and surface when they disagree, rather than trusting either alone.

See it live →

ESPN Total QBRQBR

Signal: useful with context

live

What it is: A results/EPA-weighted 0-100 QB score, play-weighted across the season.
How we read it: A genuinely independent second opinion on the QB, but more outcome-driven (and thus noisier) than CPOE.
Commonly misread: Treating a single composite number as truth and ignoring what's driving it.
Our edge: We never average QBR with CPOE. We percentile both and read the GAP (results-over-accuracy vs the reverse) as the real signal.

See it live →

Pressure Rate & Time to Throw

Signal: useful with context

live

What it is: How often a QB is pressured and how quickly he gets the ball out (PFR / Next Gen).
How we read it: Environment and process. Pressure suppresses QB and pass-catcher value; a fast release can mask a bad line.
Commonly misread: Blaming a QB for a line's failures (or crediting one for a clean pocket) without separating the two.
Our edge: We fold pressure into the QB read so accuracy and results are judged against the conditions that produced them.

See it live →

Usage

Snap Share

Anchor: stable & predictive

live

What it is: The percentage of offensive snaps a player is on the field.
How we read it: The role behind every other stat. Rising snap share precedes rising production. It's the earliest legitimate breakout signal.
Commonly misread: Reacting to fantasy points without checking whether the snaps that produced them are stable or a fluke.
Our edge: We use snap share as the gate on every opportunity read: no role, no projection bump.

See it live →

Availability

Official Availability + Conditions

Signal: useful with context

live

What it is: Public injury designations (Out/Doubtful/Questionable) and game-day weather/surface.
How we read it: A confidence-band modifier, never a body claim. It can only WIDEN uncertainty or move a read to watchlist / no-bet.
Commonly misread: Inventing health percentages or treating a Questionable tag as a coin flip rather than a band-widener.
Our edge: We say 'availability uncertain per public report' and widen the band. We never manufacture certainty about a body.

See it live →

Baseball

Barrel % / Hard-Hit %

Anchor: stable & predictive

queued

What it is: Quality of contact: the share of batted balls in the ideal exit-velocity/launch-angle window (barrels) and hit 95+ mph (hard-hit).
How we read it: The most predictive public power signal in baseball. Quality of contact stabilizes far faster than home runs or batting average and forecasts them.
Commonly misread: Chasing a hitter's home-run total or BABIP-inflated average, outputs that swing wildly on small samples.
Our edge: Quality-of-contact vs results, same as our football reads: barrel% up but production down = buy-low; production up on weak contact = sell-high. (Pending a verified, license-clear Statcast source.)

Strikeout & Whiff RateK%

Anchor: stable & predictive

queued

What it is: How often a hitter strikes out (or a pitcher misses bats).
How we read it: One of the fastest-stabilizing stats in the sport, a near-skill that's usable on tiny samples for both hitters and pitchers.
Commonly misread: Waiting for ERA or batting average to 'come around' when the underlying K profile already told the story.
Our edge: We lean on the fast-stabilizing skills (K%, whiff, xwOBA) over slow, luck-soaked outcomes. (Pending a verified, license-clear Statcast source.)

The doctrine

When two independent signals disagree, we surface the gap. We don't average it into false precision. Opportunity that outruns production is a buy-low; production that outruns opportunity is a sell-high. That single discipline, applied across every position and sport, is how we keep projections honest and close to accurate.