Glossary

Every metric used across Blindside Alpha — plain-English definitions, fantasy context, usage guidance, and caveats.

92 metrics·7 categories·Updated for 2026

Start Here

The core metrics behind every research surface — click to jump.

Core Fantasy

FPTSpoints

Basicfpts

Total season fantasy points.

Cumulative PPR fantasy points across the current season.

Why it matters

Useful for dynasty value and overall seasonal production context.

How to use

Use for season-long context; prefer PPG for week-to-week projections.

Watch out: Penalizes players who missed games – a player with 10 GP at 22 PPG beats 16 GP at 14 PPG.

e.g. 221 FPTS in 12 games

used in:Player ProfilesStats Leaders

src: player_stats_season

Basicgp

Games played in the season.

Number of games the player has appeared in this season – drives the PPG denominator.

Why it matters

Durability is a skill. Games played is a baseline availability signal.

How to use

Filter out low-GP samples before ranking by PPG.

used in:Player ProfilesStats Leaders

src: player_stats_season

PPGpoints / game

Basicppg

Fantasy points per game (PPR scoring).

Average fantasy points scored per game played. Uses PPR (1 point per reception).

formula:Total Fantasy Points ÷ Games Played

Why it matters

Better predictor of future output than season totals – normalizes for games played.

How to use

Rank players within their position group. Pair with GP to check consistency.

Watch out: Can overrate players on touchdown-dependent scoring runs. Check floor/ceiling spread.

e.g. 18.4 PPG in 12 games

used in:Player ProfilesStats LeadersDraft CenterWeekly Projections

src: player_stats_season

Usage

aDOTyards

Intermediateadot

Average depth of target.

Average air yards per target. Measures how far downfield a receiver is being targeted. High aDOT = downfield role; low aDOT = underneath/slot role.

formula:Total Air Yards ÷ Total Targets

Why it matters

Receivers with high aDOT carry more big-play upside but inconsistency. Low-aDOT players have high floors in PPR.

How to use

Combine with YPRR and target share. A high-aDOT receiver with declining target share is a concern.

Watch out: aDOT alone can overrate downfield receivers who rarely convert. Pair with catch rate and CPOE.

e.g. 12.4 aDOT (deep receiver) vs. 4.1 aDOT (slot receiver)

used in:Player ProfilesTape RoomStats Leaders

src: player_season_metrics

Air Yardsyards

IntermediateairYards

Total yards in the air on pass attempts to this player.

Air yards = yards the ball traveled in the air from the line of scrimmage to the intended target. Measures downfield targeting, not YAC or actual receiving yards.

Why it matters

Air yards predict target quality. A 6-catch, 80-yard game means different things at 2 avg air yards vs. 14 avg air yards (aDOT).

How to use

Use aDOT for per-target efficiency. Total air yards for volume.

e.g. 280 air yards in 4 games

used in:Player ProfilesTape Room

src: player_season_metrics

ATT

pAtt

Pass attempts (for QBs).

Total pass attempts. Includes incompletions and sacks taken in some sources.

Why it matters

Volume for QBs. High pass-attempt teams boost all receivers.

How to use

Use team-level ATT to contextualize receiver target share.

src: player_stats_season

AY%%

airYardsShare

Share of team air yards.

Sum of intended yards on a player's targets divided by team total air yards. Captures downfield workload.

Why it matters

Downfield air yards convert to big plays and TDs. Share above 30% is true X-receiver volume.

How to use

Use to identify deep threats whose ceiling exceeds catch-volume scoring.

src: player_season_metrics

CAR

carries

Rushing attempts.

Count of rushing attempts (designed runs + scrambles for QBs).

Why it matters

Volume is the floor for RBs. More carries = more scoring chances.

How to use

Combine with snap share to see if a player gets their carries in critical game situations.

src: player_stats_season

I5 (Inside-5)count

rz_i5

Opportunities from inside the opponent's 5-yard line — the highest-value red zone zone.

I5 counts plays where the player was the intended target or ball carrier with the line of scrimmage inside the opponent's 5-yard line. This is the highest-leverage zone in football: touchdown conversion rates from inside the 5 are 3–4× higher than from the 6–20 yard range. The I5-weighted expected-TD model assigns a position-specific league average conversion rate to these plays separately.

Why it matters

A player who leads their team in I5 looks is the closest thing to a guaranteed TD machine in fantasy. Even a player with mediocre overall RZ volume can be a top-10 scorer if they hoard I5 looks.

How to use

Sort by I5 absolute count first, then RZ share second. A player with 8+ I5 looks in 10 games has a clear path to 8+ TDs. Use as a tiebreaker in DFS for stacked game environments.

Watch out: I5 is volatile week-to-week (a player can go 3 games without a goal-line look). It is a season-level signal, not a weekly edge.

e.g. 12 I5 opps → 4 TDs scored (expected: ~5.5 for a QB, ~4.8 for a RB)

src: v_redzone_player: count of plays where field_position ≥ 95 (5 yds to goal)

Leverage Score Basis (Within-Pos vs. Overall)0–100 percentile (two separate windows)

leverage_basis

Whether Leverage Score is percentile-ranked against positional peers or all offensive skill players.

The Leverage Score model computes the same raw composite (GL + RZ + C/L + 2M + LD weighted sum) for every player, then exposes two percentile rankings: within-position (comparing the player to other RBs or WRs or TEs) and overall (comparing against all skill position players together). Within-pos is the default because usage patterns differ structurally by position — RBs dominate GL by design. The overall basis is useful when evaluating roster construction across positions.

Why it matters

A TE at 80 within-pos %ile has an elite role for a tight end. The same raw composite for an RB may only be 55 within-pos because RBs as a class receive far more high-leverage touches. Using the right basis avoids misreading cross-position comparisons.

How to use

Default to within-pos for all single-position decisions. Switch to overall when comparing across positions for the same flex/UTIL slot. The basis toggle is in the page header.

Watch out: The overall basis can make slot receivers look weak (they have fewer GL looks structurally) and goal-line RBs look dominant. Within-pos is almost always the more actionable view.

src: v_leverage_score: two separate percent_rank() windows

Leverage Weightsweights: GL 30%, RZ 25%, C/L 20%, 2M 15%, LD 10%

leverage_weights

The five situational usage categories and their weights in the Leverage Score model.

Leverage Score is a weighted composite of five high-value usage situations: Goal-Line (GL, 30%), Red Zone (RZ, 25%), Close-and-Late (C/L, 20%), Two-Minute (2M, 15%), Late-Down (LD, 10%). Each component measures the player's share of team touches in that situation, then the weighted sum is percentile-ranked within position to produce the 0–100 Leverage Score. Weights reflect the relative TD/scoring importance of each zone.

Why it matters

Not all snaps are equal. A player who dominates goal-line and two-minute drill usage is worth far more in fantasy than their snap share alone suggests — they are disproportionately present on the plays that produce points.

How to use

Use the composition bar to identify which components drive a player's score. A player with 90% GL contribution but near-zero RZ is a goal-line-only back — dependent on short-yardage situations. A player with balanced contribution across all five has the most robust role.

Watch out: Leverage Score is a ROLE metric, not a raw-production metric. A team with a low-scoring offense can produce a BELL-COW player whose ceiling is capped by the team's scoring volume.

src: v_leverage_score: weighted composite, /methodology#leverage

Route %%

IntermediaterouteShare

Share of routes run out of total team pass routes.

How often a receiver runs a route when the team passes. Routes run per game divided by team passing plays. Separates receivers who are truly in the passing game from those who block on pass plays.

Why it matters

Filters out 'snap share illusion' – a receiver in the backfield or blocking on pass plays won't generate YPRR opportunities.

How to use

Pair with YPRR. Route Share × YPRR = opportunity-weighted efficiency.

Watch out: Low-volume receivers can post elite YPRR rates on 20 routes/game. Route share exposes the ceiling.

e.g. 78% route share

used in:Player ProfilesTape Room

src: player_season_metrics

RZ Opportunitycount

Intermediaterz_opportunity

Total red zone looks (targets + carries) inside the opponent's 20-yard line.

RZ Opportunity aggregates all plays where the player was the intended recipient (target or handoff) with the ball spotted inside the opponent's 20-yard line. Includes the full RZ from the 20 to the goal line, with separate breakdowns for inside-10 and inside-5 zones.

Why it matters

Red zone opportunity is the primary driver of touchdown probability. Players who receive consistent RZ looks have a repeatable path to 6-point scoring plays regardless of game script.

How to use

Compare RZ opportunity share — not raw counts — to correct for team offense volume. Pair with rzOppRank (%ile within position) to see whether volume is elite for the position.

Watch out: High raw RZ opportunity counts can come from low-efficiency offenses that need many chances to score. Always check conversion rate (TD%) alongside volume.

e.g. 35 RZ opps in 14 games = 2.5/game — top-tier RB volume

used in:ModelsPlayer Profiles

src: v_redzone_player: sum of rz_targets + rz_carries where field_position ≥ 80

Snap%%

BasicsnapShare

Percentage of offensive snaps the player was on the field.

Snap share = player snaps / total team offensive snaps. A proxy for role stability and usage ceiling.

formula:Player Snaps ÷ Total Team Offensive Snaps × 100

Why it matters

Players who are on the field can only score if they're on the field. Snap share is the on/off switch.

How to use

Look for players trending up in snap share – usage often leads points. Trend direction matters more than a single week.

Watch out: High snap share doesn't mean high target share. A blocker can have 90% snaps and 0 targets.

e.g. 85% snap share

used in:Player ProfilesOffensive LineTape Room

src: player_season_metrics

TGT

Basictargets

Times targeted by a pass attempt.

Count of pass attempts intended for this player. Strong leading indicator of receiver value.

Why it matters

Targets are the raw opportunity signal. High-target players have more paths to big games.

How to use

Pair with target share – raw targets vary by team pass volume.

Watch out: Target share must be paired with team pass volume. A 30% share on a team that throws 20 times is not equivalent to 30% on a team that throws 40 times.

e.g. 9.2 targets / game

used in:Player ProfilesStats LeadersTape Room

src: player_stats_season

TGT%%

BasictargetShare

Share of team pass targets.

Targets divided by team pass attempts. Captures opportunity independent of team volume.

formula:Player Targets ÷ Team Pass Attempts × 100

Why it matters

Target share is the cleanest pass-game opportunity stat. 25%+ is elite WR1/TE1 territory.

How to use

Pair with team pass volume – share of a high-volume offense beats share of a low-volume one.

e.g. 29% – clear WR1 target

used in:Player ProfilesStats LeadersTape Room

src: player_season_metrics

TPRRratio

tprr

Targets per route run.

Targets divided by route participations. Pure per-route efficiency.

Why it matters

Filters out players who only got targets because they ran more routes. 0.25+ is elite per-route conversion.

How to use

Use to find efficient pass-catchers buried in low-route roles.

src: player_season_metrics

Verdict: BELL-COWverdict label (categorical)

verdict_bell_cow

Elite positional leverage — dominates the highest-value usage situations within their position group.

BELL-COW fires in the Leverage Score model when a player's within-position leverage percentile is ≥90. They are among the top 10% of their position in the weighted composite of goal-line, red-zone, close-and-late, two-minute, and late-down usage. This is the rarest and most valuable role in fantasy: a player who receives elite volume on the plays that produce points.

Why it matters

Bell-cows have first-round fantasy value independent of how their team scores. Even in a mediocre offense, a true bell-cow RB will see the ball at the goal line and in two-minute drills. Their floor is the floor of the team's scoring.

How to use

Draft bell-cows as foundational pieces. In DFS, bell-cows have high safe upside — price in the range of their ceiling, not their floor. In dynasty, bell-cow status is the single most important role signal.

Watch out: Bell-cow status can change overnight with an injury, trade, or depth chart addition. Monitor weekly snap share and carry share to confirm the role persists.

src: Derived verdict in leverage-score ModelLeaderboard config: leverageScore ≥ 90

WOPR0–1+

Intermediatewopr

Weighted Opportunity Rating – 1.5×TGT% + 0.7×AY%.

Combines target share and air-yards share into one opportunity score (Josh Hermsmeyer formula). Standardized so 0.6+ is elite.

formula:1.5 × TGT% + 0.7 × AY%

Why it matters

Best single number to measure receiver opportunity – ahead of target share alone.

How to use

Rank pass-catchers by WOPR within position; gaps reveal opportunity tiers.

used in:Player ProfilesStats Leaders

src: player_season_metrics

Efficiency

%ile (Within-Pos)0–100 percentile

pctile_within_pos

A player's percentile rank among players at the same position.

Within-position percentile rank (0–100). Ranks the player relative to all players at their position who meet the minimum game threshold. Eliminates cross-position distortion so a RB's TD dependency is measured against other RBs, not against all QBs who have naturally different TD rates.

Why it matters

Raw metrics like TD% or Sharpe ratio cluster by position — QBs naturally have lower raw TD% than RBs and TEs. Within-pos percentile makes comparisons fair and surfaces actionable signals across all positions simultaneously.

How to use

Use the within-pos toggle as the default for all research. Switch to overall only when comparing players across positions (e.g., evaluating a TE vs. WR for the same roster slot).

Watch out: A 90th-percentile RB is elite among RBs — but may trail many WRs on the same raw metric. Always anchor to the same basis when comparing across positions.

e.g. TD Dependency %ile = 85 means more TD-dependent than 85% of positional peers

src: Computed via percent_rank() OVER (PARTITION BY position) in DB views

CATCH%%

BasiccatchPct

Catch rate — receptions divided by targets.

Catch% = receptions / targets. The fraction of pass attempts intended for a receiver that result in a completed catch. Measures reliability as a pass-catcher, though catch rate is heavily influenced by throw depth — deeper routes naturally produce lower catch rates.

formula:Receptions ÷ Targets × 100

Why it matters

Catch rate is a floor signal for PPR scoring. A receiver who drops or allows targets to go incomplete sacrifices the per-reception point that PPR leagues reward. High catch rate at high volume is the most durable PPR floor combination. Declining catch rate can flag poor target quality or deteriorating hands.

How to use

Always pair with aDOT. A slot receiver at 82% catch rate and 4.2 aDOT is expected to catch most targets — that is not elite accuracy, it is scheme design. A downfield WR at 70% catch rate and 14.5 aDOT is converting deep balls at an above-average rate.

Watch out: Raw catch rate penalizes deep routes by design. Use CPOE-equivalent receiver metrics when available to separate scheme-driven rates from genuine hands/separation quality.

e.g. 76% catch rate (deep WR, efficient) vs. 61% (slot, likely drops issue)

used in:Player ProfilesStats Leaders

src: player_stats_season: receptions / targets

CMP%%

BasiccompPct

Completion percentage — completions divided by pass attempts.

CMP% = completions / pass attempts. The most basic QB accuracy metric: the fraction of thrown balls that result in a caught pass. Does not account for throw depth, receiver separation, or pressure context.

formula:Completions ÷ Pass Attempts × 100

Why it matters

A QB with a high completion rate sustains drives and keeps receivers in rhythm. Decline in CMP% can signal deteriorating accuracy, scheme change, or increased pressure. For fantasy, a QB who completes passes at a high rate tends to be more consistent week-to-week.

How to use

Pair with CPOE (Completion % Over Expected) to separate scheme-driven high rates from genuine accuracy. A QB checking down in a quick-game system can post 70%+ CMP% with no actual accuracy edge.

Watch out: CMP% rewards short throws and penalizes aggressive downfield passing. A QB with 58% CMP% and a 14 Y/A is often more valuable than one with 72% CMP% and a 6 Y/A.

e.g. 68.4% CMP% — above average; 55% — low, check throw depth

used in:Player ProfilesStats Leaders

src: player_stats_season: completions / pass_attempts

CPOEpercentage points

cpoe

Completion % Over Expected.

Difference between a QB's actual completion rate and the rate expected given throw depth, location, pressure, etc. Above zero = beating model expectation.

Why it matters

CPOE isolates QB decision quality from receiver influence and scheme. A consistent +CPOE QB sustains receiver production.

How to use

Identify QBs who consistently beat expectation – their receivers benefit. Negative CPOE teams suppress receiver value.

Watch out: CPOE is not purely a receiver metric. Offensive line quality and scheme influence it heavily.

e.g. +4.1 CPOE

src: player_season_metrics

EPAexpected points / play

Advancedepa

Expected Points Added per play.

Average change in expected scoring outcome attributable to this player on each play. Positive = better than baseline. For QBs: EPA/dropback; RBs: EPA/rush; pass-catchers: EPA/target.

Why it matters

EPA is the gold-standard efficiency stat – it accounts for situation (down/distance/field position) that raw stats ignore.

How to use

Use for ranking within position group. Combine with volume metrics. An EPA leader with low volume may be scheme-dependent.

Watch out: EPA can be noisy in small samples. Requires 50+ snaps to stabilize. Early-season EPA for low-usage players is often misleading.

e.g. 0.22 EPA/target

used in:Player ProfilesTape RoomModels

src: player_season_metrics

RACR

racr

Receiver Air Conversion Ratio.

Receiving yards divided by air yards. RACR > 1 means the player turns intended yards into more than expected via YAC and contested catches.

Why it matters

Identifies YAC monsters and contested-catch specialists who produce beyond the throw.

How to use

Use to separate air-yards-dependent receivers from self-creators. RACR > 1.2 consistently = excellent run-after-catch skill.

src: player_season_metrics

Success %%

successRate

Share of plays that gain ≥ 50% / 70% / 100% of needed yards.

Percent of plays that succeed per the EPA framework: 50% of needed yards on 1st down, 70% on 2nd, or 100% on 3rd/4th. Position-agnostic efficiency measure.

Why it matters

Consistency metric – separates players with sustainable production from boom/bust players.

How to use

High success rate + high EPA = elite, consistent player. High EPA + low success rate = boom/bust.

src: play_participation + EPA framework

TD Share%

Intermediatetd_share

Percentage of a player's fantasy points sourced from touchdowns.

TD Share = (TD points) / (total fantasy points) × 100. A high TD share means the player relies heavily on scoring touchdowns to generate their fantasy output, which is volatile week-to-week.

formula:TD Points ÷ Total Fantasy Points × 100

Why it matters

Touchdown production is the least stable component of fantasy scoring. High TD share players carry regression risk; low TD share players have a more durable, volume-backed floor.

How to use

Compare within position group. A WR with 40%+ TD share is an over-performer relative to their target volume — fade in DFS. A WR with sub-15% TD share posting strong PPG has a sustainable floor.

Watch out: Don't conflate low TD share with low upside. Some players have low shares because they are volume monsters — TDs are a small fraction of a very large pie.

e.g. 38% TD share — fragile, regress-candidate; 12% TD share — volume-backed floor

used in:ModelsPlayer Profiles

src: v_td_dependency: td_pts / total_pts × 100

TD± (vs Expected)touchdowns (signed)

rz_td_diff

Actual red zone TDs minus the I5-weighted expected TDs from the opportunity model.

TD± = rz_td − expected_td. expected_td is computed from the I5-weighted baseline: expected_td = (rz_i5 × i5_rate[pos]) + ((rz_opps − rz_i5) × outer_rate[pos]), where i5_rate and outer_rate are position-specific league averages. A positive TD± means the player is scoring more TDs than their opportunity profile predicts (SELL-HIGH — likely to regress). A negative TD± means they are under-converting (BUY-LOW — TDs are due).

Why it matters

Raw TD counts are noisy. TD± normalizes for opportunity quality, isolating conversion luck from volume. This is the single best indicator of whether a player's TD pace will revert.

How to use

TD± > +1.5 = meaningful over-conversion, consider selling. TD± < -1.5 = meaningful under-conversion, buy before the reversion. Pair with rz_opp_rank to ensure the player has enough volume for the signal to be actionable.

Watch out: Small-sample TD± swings wildly. Require 8+ RZ opps before treating the signal as meaningful. A player with 2 RZ looks can be +2.0 on a single fluky score.

e.g. TD± = -2.1 → scored 2.1 fewer TDs than expected given opportunity → BUY-LOW

src: v_redzone_player: rz_td − expected_td (I5-weighted model per /methodology#redzone-expected)

True YPRRyards / route (adj.)

trueYprr

YPRR adjusted for target depth and team pass volume.

YPRR normalized for average depth of target (aDOT) and the team's pass volume per game. Removes the advantage of running deep routes in a high-volume passing offense.

Why it matters

Compares receivers across different schemes and route trees on a level playing field.

How to use

Use when comparing a deep WR1 in a high-volume offense to a slot receiver in a run-heavy scheme.

src: internal model from player_season_metrics

TTTseconds

ttt

Avg time to throw (seconds).

Average seconds from snap to release for a QB. Lower TTT = quick-game scheme.

Why it matters

Quick-release offenses benefit slot receivers and underneath targets. Deep offenses boost high-aDOT WRs.

How to use

Pair with aDOT to understand scheme context. Low TTT + low aDOT = volume-first slot play.

e.g. 2.4s (quick-game) vs. 2.9s (timing offense)

src: nextgen_player_week.avg_time_to_throw

Y/Ayards / attempt

Basicypa

Yards per pass attempt — total passing yards divided by attempts.

Y/A = passing yards / pass attempts. Measures the average yardage gained on each throw, including incompletions. The simplest single-number measure of passing efficiency and downfield aggression.

formula:Passing Yards ÷ Pass Attempts

Why it matters

High Y/A signals a QB operating downfield or in a scheme that creates big plays, which boosts all receiver ceilings. For fantasy, a high-Y/A QB with a good pass-catching TE or WR1 elevates the entire stack. Low Y/A often signals a check-down-heavy system where yardage accrues slowly.

How to use

Pair with CMP% to distinguish aggressive-accurate QBs from wild downfield throwers. The best combination is high Y/A + above-average CMP% — that quarterback is generating yards efficiently, not just bombing it.

Watch out: Y/A can spike in garbage time when a trailing team throws downfield repeatedly. Filter to competitive-script situations before drawing conclusions about scheme.

e.g. 8.2 Y/A — aggressive downfield attack; 5.8 Y/A — short-to-intermediate scheme

used in:Player ProfilesStats Leaders

src: player_stats_season: passing_yards / pass_attempts

Y/Cyards / carry

Basicypc

Yards per carry — rushing yards divided by carries.

Y/C = rushing yards / carries. The fundamental per-attempt efficiency metric for rushers. Captures average yardage gained on each designed run or quarterback scramble, before accounting for volume.

formula:Rushing Yards ÷ Carries

Why it matters

Y/C separates volume-driven RB production from efficient carry usage. A back averaging 5.2 Y/C on 15 carries per game has a more sustainable efficiency profile than one at 3.8 Y/C on the same volume. For dynasty or trade evaluation, Y/C is a key indicator of whether a back is thriving or just surviving on volume.

How to use

Pair with carries per game to assess whether the Y/C is built on a sufficient sample. Isolate non-garbage-time carries when comparing scheme-influenced backs. Compare within the same season — league-wide Y/C varies by era and rules.

Watch out: Y/C can be inflated by a small number of long breakaway runs, which are not predictable. Median carry (not average) is a better stability metric, though it is rarely surfaced publicly.

e.g. 4.8 Y/C — solid efficiency; 3.6 Y/C — below league average; 6.2 Y/C — elite

used in:Player ProfilesStats Leaders

src: player_stats_season: rushing_yards / carries

Y/Ryards / reception

Basicypr

Yards per reception — receiving yards divided by receptions.

Y/R = receiving yards / receptions. Average yardage gained on each completed catch, combining air yards delivered at the catch point with yards after the catch (YAC). High Y/R signals a receiver targeted deep or one with elite YAC ability — but a low Y/R does not imply poor value in PPR formats.

formula:Receiving Yards ÷ Receptions

Why it matters

Y/R captures the upside of each individual catch. A receiver averaging 17.5 Y/R is generating big plays when targeted. For DFS, high Y/R + a matchup that allows big plays sets up ceiling upside. For redraft, it identifies volume receivers who may outpace their catch total in yardage bonuses.

How to use

Pair with receptions per game and aDOT. A player at 18 Y/R and 3 rec/game has a high ceiling but a low PPR floor. A player at 9 Y/R and 8 rec/game is the opposite. The combo drives expected weekly PPR output more reliably than either in isolation.

Watch out: Y/R does not normalize for throw depth. A receiver at 20 Y/R almost entirely due to being targeted on go routes 15+ yards downfield is not generating the same YAC ability as one reaching that figure from 7-yard routes and great separation.

e.g. 17.2 Y/R — downfield role with big-play upside; 8.4 Y/R — volume slot target

used in:Player ProfilesStats Leaders

src: player_stats_season: receiving_yards / receptions

YPRRyards / route

Intermediateyprr

Yards per route run.

Receiving yards divided by routes run. Best single-number receiver efficiency stat – neutralizes opportunity differences and surfaces per-route production.

formula:Receiving Yards ÷ Routes Run

Why it matters

YPRR reveals efficiency regardless of scheme or volume. The best way to identify elite receivers in pass-light offenses.

How to use

Pair with route share to get the full picture. Elite YPRR with high route share = top-tier WR.

Watch out: YPRR needs route volume to stabilize. Players with fewer than 40 routes have volatile numbers. Can overrate low-volume players.

e.g. 2.4 YPRR (elite); 1.0 YPRR (below avg)

used in:Player ProfilesStats LeadersTape Room

src: yprr_player_season + player_season_metrics fallback

Projections

ADPpick #

Basicadp

Average Draft Position across redraft leagues.

Average Draft Position is the mean pick number where a player is selected across many drafts. Lower = drafted earlier.

Why it matters

ADP is the public's expected price. Players whose projections clearly beat their ADP are draft-day value.

How to use

Pair ADP with projected rank. Sort by `proj rank – ADP` to find value (and reaches).

e.g. ADP 14.2 = late 1st round, 12-team

used in:Draft CenterADP Tracker

src: consensus aggregator (multiple sources)

Earned vs. Sell-Highbinary classification based on pos_week_rate ≥ 0.55

gap_earned_vs_sellhigh

Whether a positive xFP gap reflects real efficiency or unsustainable variance.

Among players with a positive xFP gap (scoring above expectation), the EARNED verdict applies when pos_week_rate ≥ 0.55 — meaning they've beaten their expected value in at least 55% of their weeks. This separates players whose outperformance recurs (genuine efficiency edge, don't sell) from players who posted one or two volatile outlier games (SELL-HIGH — reversion is likely).

Why it matters

Not all positive gaps are sell signals. A player who consistently beats their opportunity model week-over-week is doing something the model doesn't capture — elite route running, leveraging coverage weaknesses, or scheme targeting. That player should be held, not sold.

How to use

When a player has a large positive gap, check pos_week_rate. ≥0.55 = EARNED (hold or buy more). <0.55 = SELL-HIGH (fade or trade). The threshold is conservative by design.

Watch out: Early in the season, pos_week_rate has limited sample. A 4-for-6 rate (67%) with only 6 games is encouraging but not definitive. Weight the signal more after week 10.

src: v_xfp_gap.pos_week_rate: fraction of weeks player's actual > xFP

ECRrank

ecr

Expert Consensus Ranking – the position rank.

Expert Consensus Ranking aggregates dozens of analyst rankings into a single positional rank (e.g. WR7, RB12). Lower is better. Updated weekly.

Why it matters

ECR is the market price for a player – use it to find places your view diverges from consensus.

How to use

Compare to your own projection rank. Big gaps either way flag a real disagreement worth investigating.

Watch out: ECR moves slowly; injury news and depth-chart changes lag the market by a few hours.

e.g. WR7 – top-tier WR1

src: consensus aggregator

ECR Δranks

ecrDelta

Weekly change in consensus ranking.

Difference between this week's expert consensus rank and last week's. Negative = ranking improved.

Why it matters

Quickly spots risers and fallers in analyst sentiment.

How to use

Use as a sentiment indicator – not a direct value signal.

src: consensus aggregator

Floor / Ceilingpoints

floorCeiling

10th and 90th percentile fantasy outcomes.

P10 (floor) and P90 (ceiling) projected outcomes from the player's distribution. Width = volatility; midpoint anchors to base projection.

Why it matters

Shows the range of realistic outcomes, not just the midpoint prediction.

How to use

High floor = safe start. Wide range = upside play for showdowns/tournaments.

src: internal projections

PROJ (wk)points

projWeek

This week's projected fantasy points.

Model projection for the upcoming game in PPR scoring. Built from base ROS projection, matchup adjustment, weather, snap-share trend, and recent EPA. Implied, not guaranteed.

Why it matters

Starting point for weekly lineup decisions. Always pair with floor/ceiling.

How to use

Sort by projection then filter by floor to find safe floors. Use ceiling for upside plays.

e.g. 18.2 projected, 10/28 floor/ceiling

src: internal projections

Rk Spreadranks

rkSpread

Disagreement across ranking sources.

Standard deviation of position rank across analyst sources. High spread = volatile or divisive player.

Why it matters

High spread means market uncertainty – which can be exploited on either side.

How to use

Wide spread + your model has a firm view = a real edge. Agree = press it.

src: consensus aggregator

Tier1–5

tier

Cluster grouping of comparable players.

Tiers group players whose expected output is statistically indistinguishable. Within a tier, draft for need; across tiers, draft for value.

Why it matters

Tier breaks – not individual ranks – are where the real positional cliffs live.

How to use

When the last player in a tier is about to be drafted, prioritize that position.

src: internal cluster model

Verdict: BUY-LOWverdict label (categorical)

verdict_buy_low

The opportunity is real but the production hasn't arrived yet — acquire now.

A BUY-LOW verdict fires when a player has high opportunity rank but is under-converting relative to expectation. In xFP Gap: negative Gap/G (actual < xFP). In Red Zone: high rzOppRank + tdDiff < 0. The argument: the volume is there; variance has suppressed the points. Positive regression is statistically likely. The market is mispricing them relative to their forward opportunity.

Why it matters

Buy-lows are the core mechanism of fantasy value extraction. If you identify players being suppressed by small-sample variance before the market reprices them, you win trades and draft picks.

How to use

In redraft/DFS: target in trades or off the waiver wire. In DFS: buy on days when ownership is depressed by a prior bad game. Confirm the opportunity metrics are stable (target share / carry share not declining) before buying.

Watch out: Not every negative gap is a buy-low. A player with declining target share who is under-converting has a fundamental problem, not a variance problem. Check the opportunity trend before buying.

src: Derived verdict in ModelLeaderboard configs (xfp-gap, red-zone)

Verdict: EARNEDverdict label (categorical)

verdict_earned

The player consistently beats their expected output — this is real efficiency, not luck.

EARNED fires when a player has a positive xFP gap AND a high pos_week_rate (≥ 0.55 — beaten expected points in 55%+ of weeks). This pattern indicates the player is doing something the opportunity model doesn't fully capture: elite route separation, target quality above their depth-of-target, contested-catch skill, or favorable matchup exploitation. The outperformance recurs, which makes it real.

Why it matters

Most positive gaps regress (see SELL-HIGH). EARNED players are exceptions — they are systematically extracting more points per opportunity than the model predicts. Holding or buying EARNED players is correct, not a fade opportunity.

How to use

Hold all EARNED players. Do not sell on a positive gap alone — check pos_week_rate first. In dynasty, buy EARNED young players (their efficiency is becoming an established trait). In DFS, EARNED players are safer upside plays than SELL-HIGH players at similar ownership.

Watch out: EARNED has a sample requirement. 6–8 games minimum before trusting the pos_week_rate. Early-season efficiency that looks like EARNED is often just a hot streak.

src: Derived verdict in xfp-gap ModelLeaderboard config: positive gap + posWeekRate ≥ 0.55

Verdict: REGRESSION RISKverdict label (categorical)

verdict_regression_risk

High TD dependency relative to positional peers — scoring is fragile and likely to regress.

REGRESSION RISK fires when a player's within-pos TD dependency rank is ≥80 — they are more TD-dependent than 80% of positional peers. Their fantasy points are heavily driven by touchdowns rather than volume and yardage. Since TDs are the most volatile scoring element, high TD-dependent production is historically unreliable over extended stretches.

Why it matters

The most common source of redraft and DFS bust weeks is TD-dependent production that fails to sustain. Identifying REGRESSION RISK players before the market reprices them is the single most reliable way to sell high in trades.

How to use

Sell REGRESSION RISK players when their ADP or trade value is inflated by recent TD production. Fade in cash DFS when their floor (without a TD) would lose you the week. In GPP, they can still be used — but know you are taking on TD-or-bust variance.

Watch out: High TD dependency does not mean the TDs were unearned. A goal-line specialist intentionally has high TD dependency — that's the role. The signal is most actionable for players who are receiving TDs above what their red zone volume predicts.

src: Derived verdict in td-dependency ModelLeaderboard config: tdDepRank ≥ 80

Verdict: SELL-HIGHverdict label (categorical)

verdict_sell_high

The production is ahead of the opportunity — regression is statistically likely.

A SELL-HIGH verdict fires when a player is scoring above their opportunity-based expectation, with low persistence. In xFP Gap: positive Gap/G + posWeekRate < 0.55 (less than 55% of weeks beat xFP — the gap is driven by volatile good weeks, not sustained efficiency). In Red Zone: high rzOppRank + tdDiff > 0 (over-converting TDs beyond the expected rate). The argument: regression to the mean is the most reliable prediction in sports.

Why it matters

The market prices players based on recent output. A player coming off back-to-back 30-point games is likely overvalued — their ADP or trade value is at a seasonal high even if the underlying opportunity is unchanged.

How to use

Sell in trades and DFS — accept the high valuation. In season-long: sell when the player's trade value peaks (usually within 2 weeks of a multi-TD game). In DFS: fade on days when ownership will be high from the hot streak.

Watch out: SELL-HIGH ≠ bad player. It means regression from recent performance is likely, not that the player is a waiver candidate. A SELL-HIGH who still gets 10 targets/game is still startable — you're selling at a premium, not dumping.

src: Derived verdict in ModelLeaderboard configs (xfp-gap, red-zone)

Verdict: SUSTAINABLEverdict label (categorical)

verdict_sustainable

Volume-backed floor — TD dependency is low, meaning production is durable and not TD-reliant.

SUSTAINABLE fires in the TD Dependency model when a player's within-pos TD dependency rank is ≤25 (lower 25th percentile — they are LESS TD-dependent than 75% of positional peers). Their fantasy points come primarily from yardage, receptions, and volume — not touchdowns. This production is more stable and repeatable week-to-week.

Why it matters

Touchdowns are the least predictable component of fantasy scoring. Players who generate their points through volume metrics (targets, catches, yards) have a more reliable weekly floor and are less prone to zero-TD variance weeks.

How to use

Prefer SUSTAINABLE players in cash formats and dynasty. These players are less exciting but more consistent. In dynasty leagues, SUSTAINABLE young players are better holds than REGRESSION-RISK players at the same overall point total.

Watch out: SUSTAINABLE does not mean high upside. A player with 5 targets and no TDs has low TD dependency but also limited ceiling. Check the overall volume (targets, PPG) to confirm the floor is meaningful, not just TD-less.

src: Derived verdict in td-dependency ModelLeaderboard config: tdDepRank ≤ 25

xFP (Expected Fantasy Points)fantasy points

Advancedxfp_expected

Model-derived expected fantasy points based on opportunity volume and field position.

xFP (Expected Fantasy Points) is computed from the opportunity model: each snap, target, or carry a player receives is assigned an expected point value based on its field position, down-and-distance, and personnel context. The sum of these per-play expected values is the player's xFP for the week or season. Unlike raw PPG, xFP isolates what the OPPORTUNITY was worth — regardless of outcome variance.

Why it matters

xFP separates luck from skill. A player scoring above xFP on a sustainable basis (high pos_week_rate) is demonstrably efficient. A player scoring below xFP despite good opportunity is either unlucky (buy-low) or scheme-limited.

How to use

Compare actual fantasy points to xFP. Negative gap (actual < expected) is a buy-low signal — the opportunity is there but hasn't converted yet. See /methodology#xfp-expected for the full model specification.

Watch out: xFP does not predict the future directly — it describes the past opportunity. An opportunity model can only be as predictive as its input volume metrics (target share, carry share, etc.).

e.g. xFP = 18.4, actual = 14.1 → gap = -4.3 (under-converted; buy-low signal)

used in:ModelsPlayer Profiles

src: ff_opportunity_player_week: opportunity model via volume + field_position + game_state

xFP Gapfantasy points

xfp_gap

Actual fantasy points minus expected fantasy points (xFP).

xFP Gap = actual_fpts − xFP. Positive gap means the player is outperforming their opportunity; negative gap means the opportunity hasn't fully converted to points. The gap is evaluated both in total and per game (Gap/G) to normalize for games played.

Why it matters

The gap direction and persistence together determine whether a player is a buy-low, sell-high, or earned outperformer. A negative gap with high opportunity is the cleanest fantasy buy-low signal.

How to use

Sort by abs(Gap/G) to surface both extremes. Negative Gap/G = BUY-LOW (opportunity without points). Positive + high pos_week_rate = EARNED (real efficiency). Positive + low persistence = SELL-HIGH (regress).

Watch out: A large positive gap in 3 games is nearly meaningless — variance is high over small samples. Require 6+ games and check pos_week_rate before acting on any gap signal.

e.g. Gap/G = -3.2 → under-converting by 3.2 pts/game → BUY-LOW signal

src: v_xfp_gap: actual_fpts − xfp_total

Wagering

Alpha

alpha

Return above what market pricing implies.

Borrowed from finance: alpha is the excess value generated vs. the expectation embedded in the market price. In fantasy/DFS: how much a player beats their consensus ADP or projection.

Why it matters

Real edge is alpha – consistently finding value the market misses.

How to use

Track hit rate on alpha plays over time. Consistent alpha generation = a genuine edge.

src: internal model

Confidence0–100

confidence

Model's certainty in its own projection, from 0–100.

A composite of sample size, recent consistency, and input data quality. Low confidence means the model has high uncertainty – typically due to small samples, injury clouds, or missing inputs.

Why it matters

Helps you weight projections correctly. High-confidence calls are actionable; low-confidence ones should be tempered.

How to use

Filter for confidence > 60 before acting on any model output. Below 40 is a caution flag.

e.g. Confidence: 72 (solid)

src: internal model

DvPpoints / game allowed

dvp

Defense vs. Position – fantasy points allowed to each position.

How many fantasy points a defense allows to each position (QB, RB, WR, TE) per game, ranked across the league. High DvP = favorable matchup.

Why it matters

Matchup context is one of the strongest weekly adjustments to any projection.

How to use

Use as a tiebreaker between similar players. A player ranked 12th at their position against a top-5 DvP matchup is worth a look.

Watch out: DvP can overrate matchups if the opposing team heavily suppresses a specific sub-role (e.g., tight ends). Check position subtype.

src: player_stats_season + matchup tables

Edge%

Intermediateedge

Model probability minus implied market probability.

Edge = model win probability minus the implied probability derived from market odds. Positive edge means the model projects the outcome as more likely than the market prices it.

formula:Model Probability − Implied Market Probability

Why it matters

The core concept of quantitative sports research. A positive edge does not guarantee a win – it means the price is favorable relative to the estimated true probability.

How to use

Look for edges above +5% after accounting for the vig. Smaller edges are noise; larger ones warrant attention.

Watch out: Edge is only as good as the model generating it. A systematically biased model produces systematically wrong edges.

e.g. +8.4% edge on team total over

used in:Wagering Tools

src: internal model

Implied Prob.%

impliedProbability

The win probability embedded in the market odds.

Derived from American odds: -110 → 52.4% implied probability (includes vig). Removing the vig gives the fair implied probability.

Why it matters

The baseline you're comparing against when calculating edge.

How to use

Convert any line to implied probability, then compare to your model's probability.

Watch out: Raw implied probability includes vig (juice). Use the vig-removed probability for true edge calculation.

e.g. -150 → 60% implied prob (before vig removal)

src: market feed

Market Line

marketLine

The consensus odds price for a given outcome.

The line (spread, total, player prop) available in the sportsbook market. Used as the denominator for edge calculations.

Why it matters

The market aggregates thousands of inputs. Beating it consistently is hard – and valuable.

How to use

Compare model projection to the implied line. The gap = raw edge.

src: market feed

Neutral Script

neutralScript

Projected game state assuming a close, competitive game throughout.

Neutral script adjusts projections by removing expected game-script bias. If a team is favored by 14, the model adjusts for likely run-heavy late-game usage on the winning side and pass-heavy on the losing side.

Why it matters

Game script heavily influences player usage. A neutral-script projection is more predictive when spread uncertainty is high.

How to use

Use neutral-script numbers for close games (+/- 3 points). For large spreads, use the directional script.

src: internal model

Paceplays / game

pace

Plays per game or seconds per play for a team's offense.

Offensive pace = average number of plays per game (or seconds between snaps). High-pace offenses run more plays, creating more opportunities for fantasy production.

Why it matters

Pace multiplies opportunity. Two identical offenses with different pace have very different fantasy floors.

How to use

Filter for high-pace teams when stacking. Pace + high team total = volume upside.

e.g. 70 plays/game (fast) vs. 60 plays/game (slow)

src: play-by-play aggregate

PROEpercentage points above expected

proe

Pass Rate Over Expected.

How much more (or less) a team passes than expected given game situation (score, time, field position). High PROE teams systematically pass more than expected.

Why it matters

PROE is a stronger predictor of target volume than raw target counts alone. Teams with consistently high PROE support high receiver floors.

How to use

Look for receivers on high-PROE teams. Combine with target share for reliable volume plays.

e.g. +8 PROE (frequently passes more than situation calls for)

src: play-by-play aggregate

Signal

signal

Composite indicator of a predictive pattern in the data.

A signal is any data-driven observation that has historically been predictive of future performance. Blindside Alpha surfaces signals as tagged annotations on player cards and the Tape Room.

Why it matters

Signals separate actionable patterns from random noise.

How to use

Signals are starting points, not conclusions. Combine multiple signals to build conviction.

src: internal model

Team Totalpoints

teamTotal

Implied points a team is expected to score.

Derived from the game total and spread: team total = (total ÷ 2) + (spread ÷ 2). The consensus expectation for a team's scoring output.

Why it matters

Team total drives game-script projections. High team totals boost receivers; low totals boost backs in closer games.

How to use

Use as a multiplier on individual projections. A team total of 27+ supports WR upside. Under 20 = tighten receiver projections.

e.g. Team total: 26.5

src: market feed

VORPpoints above replacement

vorp

Value Over Replacement Player.

The fantasy points a player scores above the replacement-level player at the same position (typically the last starter in a standard league). Measures true scarcity premium.

Why it matters

The cleanest measure of a player's actual impact on your team. A WR2 with high VORP is worth more than a WR1 with low VORP.

How to use

Use for trade valuation and waiver priority. Sort by VORP within position – it already accounts for positional scarcity.

e.g. +42 VORP = 42 points above the last WR starter in a 12-team league

src: internal model

DFS

Boom%%

boom_pct

Percentage of weeks the player scored above their position's boom threshold.

Boom% = fraction of games where the player's PPR points exceeded the position boom line. Position boom thresholds: QB 25 pts, RB/WR 20 pts, TE 15 pts. A week above the boom line is a tournament-winning performance at that position.

Why it matters

DFS tournaments are won by hitting boom weeks. Boom% identifies players with the ceiling to deliver tournament-winning scores on any given week — independent of their average.

How to use

In GPP construction, target players with boom% ≥ 25%. Combine with bust% to find true darts (high boom AND high bust). Combine with low bust% to find cash/GPP hybrids.

Watch out: Boom% without context can overrate players on touchdown-driven blowout games. Check if boom weeks are touchdown-dependent (fragile) or volume-driven (sustainable).

e.g. Boom% = 35% → player exceeded boom threshold in more than 1/3 of games

src: v_boom_bust: count(weekly_ppr > boom_line[pos]) / games

Bring-Back

bringBack

An opposing-team player added to hedge a game stack.

After stacking a QB + pass-catcher from Team A, the bring-back is a player from Team B who benefits if Team B scores. Reduces the risk that only one team wins the scoring game.

Why it matters

Bring-backs lower the variance of a game stack while maintaining correlation upside.

How to use

Default to the #1 receiver or a high-usage RB from the opposing team of your QB stack.

src: DFS strategy

Bust%%

bust_pct

Percentage of weeks the player scored below their position's bust threshold.

Bust% = fraction of games where the player's PPR points fell below the position bust line. Position bust thresholds: QB 15 pts, RB/WR 8 pts, TE 5 pts. A week below the bust line is a game-losing score in cash formats.

Why it matters

Cash game survival is about avoiding busts. Bust% is the floor metric — a player with bust% < 10% is almost never a losing start in 50/50 or H2H.

How to use

Cash: eliminate any player with bust% > 25% unless they're the only viable option at their price point. In GPP, high bust% paired with high boom% defines the ideal tournament dart.

Watch out: Bust% can be inflated by injury-shortened games. A player with 3 bust weeks due to mid-game exits has a misleadingly high bust% for healthy projections.

e.g. Bust% = 8% → reliable floor; Bust% = 40% → high-risk tournament gamble only

src: v_boom_bust: count(weekly_ppr < bust_line[pos]) / games

Ceilingpoints

ceiling

P90 projected fantasy outcome – the upside scenario.

The 90th percentile outcome in the player's projection distribution. Used for DFS tournament selection and boom/bust analysis.

Why it matters

Tournament DFS requires ceiling, not floor. Knowing a player's upside shapes roster construction.

How to use

In cash games, weight floor. In GPPs, sort by ceiling. Best plays have both.

src: internal projections

Fantasy Sharperatio (higher = more consistent)

Advancedfantasy_sharpe

PPG divided by weekly standard deviation — risk-adjusted scoring consistency.

Adapted from the financial Sharpe ratio: Fantasy Sharpe = mean(weekly PPR points) / sd(weekly PPR points). Higher = better risk-adjusted output. A player with a Sharpe of 2.5 earns 2.5 units of return per unit of volatility — a highly consistent scorer. A Sharpe of 0.7 means high volatility relative to mean output — boom or bust.

formula:Mean Weekly PPG ÷ Standard Deviation of Weekly PPG

Why it matters

Cash games (50/50, H2H) reward consistency — favor high Sharpe players who reliably hit a floor. GPP tournaments reward ceiling — favor low Sharpe players who occasionally explode. Fantasy Sharpe lets you optimize for format.

How to use

Cash: sort by Sharpe descending. GPP: sort by Sharpe ascending AND boom% descending. The Sharpe + boom% combo identifies optimal tournament darts.

Watch out: Sharpe requires a sample of at least 6 games to be meaningful. Early-season Sharpe is dominated by 1–2 outlier games. Also, a consistently mediocre player has a high Sharpe — check PPG absolute level too.

e.g. Sharpe 2.4 = elite consistency (floor player); Sharpe 0.6 = high-volatility dart

used in:ModelsBoom/Bust Analysis

src: v_boom_bust: mean(weekly_ppr) / stddev(weekly_ppr) across the season

Floorpoints

floor

P10 projected fantasy outcome – the downside scenario.

The 10th percentile outcome in the player's projection distribution. Used for cash game selection.

Why it matters

Cash games (50/50s, head-to-head) are won by not losing. Floor matters more than ceiling.

How to use

In cash games, sort by floor first. A player with a 10-point floor rarely busts you.

src: internal projections

Hit Rate%

hitRate

% of weeks a player scores above their projected point total.

The fraction of games in which a player exceeded their pre-game projection. High hit rate = consistent over-performer relative to market.

Why it matters

Hit rate is the most actionable measure of projection quality for a specific player. Chronic over-performers are buy-low targets.

How to use

Players with > 60% hit rate over 8+ games are systematically undervalued by the market.

Watch out: Small-sample hit rates are meaningless. Require 8+ games.

e.g. 65% hit rate over 13 games

src: internal model vs. historical projections

Iso-Sharpe Contourvisual reference (not a raw metric)

iso_sharpe

Lines of equal Fantasy Sharpe on the PPG vs. Volatility scatter plot.

An iso-Sharpe contour connects all (mean PPG, vol SD) pairs that produce the same Sharpe ratio S. The equation is sd = mean / S. On the Boom/Bust scatter hero, contours for S ∈ {0.5, 1.0, 2.0, 3.0} divide the plane into risk-adjusted quality tiers. Players above the S=2 contour (high mean, low vol) are elite consistent scorers. Players below S=0.5 (high vol relative to mean) are true boom-bust gambles.

Why it matters

Iso-Sharpe contours let you immediately see which format each player is built for without computing anything: upper-left = cash plays; lower-right = GPP darts.

How to use

Read the scatter by quadrant: top-left of a given contour = better risk-adjusted return. Players you'd start in cash sit above the S=1.5 line. Tournament darts sit below S=0.75 with notable boom%.

Watch out: Iso-Sharpe contours assume stationary distributions. A player with 3 early-season bust weeks (injury) and 10 boom weeks has misleadingly poor visual placement. Check games played.

src: Computed from v_boom_bust: sd=mean/S lines for display in model-scatter-hero.tsx

Leverage

leverage

Ownership-adjusted DFS value – how much you own vs. the field.

In DFS, leverage = the difference between your roster percentage and the field's roster percentage. Being the only person in a large contest on a player who overperforms is high leverage.

Why it matters

DFS is a relative game. Scoring 50 points with 80% of the field on your player advances nobody. Same 50 points at 5% ownership = a large move.

How to use

In GPPs, actively look for leverage by fading chalk and targeting low-owned value.

src: ownership data (DFS platforms)

Recurrence %%

recurrenceRate

How often a performance type repeats week over week.

The probability that a specific performance pattern (e.g., 8+ targets, TD + 80+ yards) repeats in the next game given it happened in the prior game. Autocorrelation signal.

Why it matters

Some players have highly correlated weekly outputs. Recurrence lets you ride streaks intelligently.

How to use

After a big week, check recurrence rate before fading the player.

src: internal model

Stack

stack

Rostering players from the same team or game to correlate scores.

A stack is rostering a QB + at least one pass-catcher from the same team. A game stack adds a player from the opposing team. Correlation amplifies upside when a game script plays out high-scoring.

Why it matters

In high-variance GPP formats, stacking is the primary mechanism to differentiate your lineup and achieve the score needed to win.

How to use

In large GPPs, always stack. Use bring-back to hedge the opposing team's scoring.

src: DFS strategy

Verdict: DARTverdict label (categorical)

verdict_dart

High-variance GPP play — boom or bust, but worth the ceiling in tournaments.

DART fires in the Boom/Bust model when a player has a low sharpeRank (≤30, meaning volatile relative to positional peers), high boomPct (≥25%), and high bustPct (≥30%). This identifies players whose weekly distribution is wide — capable of a 40-point game but equally capable of 4 points. In DFS tournaments, that ceiling is what matters; in cash, the floor risk is disqualifying.

Why it matters

GPP DFS is won by differentiated lineups that hit ceilings. A DART player whose ownership is depressed by recent bust weeks is the optimal contrarian play: low ownership + high ceiling + variance in their favor = tournament equity.

How to use

DFS GPP only. Never start DARTs in cash (50/50, H2H). Best used in games where the player has a favorable matchup that the market may be underrating due to the volatile track record. Stack with their QB (if receiver) to maximize correlation.

Watch out: Not all DARTs are good GPP plays — a player can be volatile without having a meaningful ceiling. Always check that boomPct ≥ 25% before calling a player a DART, not just bustPct.

src: Derived verdict in boom-bust ModelLeaderboard config: sharpeRank ≤ 30, boomPct ≥ 25, bustPct ≥ 30

Verdict: STEADYverdict label (categorical)

verdict_steady

Reliable cash game floor — high Sharpe, low bust risk.

STEADY fires when a player has a high sharpeRank (≥70, elite risk-adjusted output for their position) and a low bustPct (≤25% of weeks below the bust threshold). This is the profile of a dependable cash game starter: weekly production is predictable, floors are high, and the player rarely delivers a game-losing score.

Why it matters

Cash game success is built on not losing. STEADY players make the floor of your lineup reliable, freeing you to take GPP risk elsewhere. A STEADY player is almost always a correct start in 50/50 or H2H.

How to use

Build your cash lineup around STEADY players at every position. In GPP, STEADY players serve as 'safe' lower-ceiling pieces that give you a floor while the rest of the lineup swings for upside. Stack a STEADY player with high-ceiling teammates.

Watch out: STEADY ≠ exciting. These players have suppressed upside by definition (low vol). In large-field GPPs, STEADY players limit your ceiling and often lead to mid-finish results rather than first-place finishes.

src: Derived verdict in boom-bust ModelLeaderboard config: sharpeRank ≥ 70, bustPct ≤ 25

Volatilitystandard deviation (points)

volatility

Standard deviation of weekly fantasy output.

The standard deviation of a player's weekly fantasy scores over the trailing N games. High volatility = boom/bust. Low volatility = consistent floor contributor.

Why it matters

Tournament DFS favors high volatility; cash games punish it.

How to use

Sort by volatility for GPP roster construction. Combine with ceiling to find high-upside plays.

Watch out: Volatility without volume context is misleading. A player with 2 touches/game has inherently higher volatility.

e.g. σ = 9.4 (high volatility)

src: internal model

Other

10-Yd Splitseconds

split10

Time to reach 10 yards from the snap — the best single acceleration proxy.

The first 10 yards of the 40-yard dash measure pure short-area quickness and acceleration, which is more predictive of NFL separation and open-field running than the full 40 time.

Why it matters

Routes start with a release from the line. A WR with an elite 10-yd split can create separation in the first 3 steps regardless of their top-end speed. For RBs, it predicts cut-back burst.

How to use

Lower is better. Elite is sub-1.50s. Sort by 10-split before full 40 for route-running receivers. Compare within position group only.

Watch out: 10-split and 40 time are correlated (~0.85) but not identical. A player can have an elite 10-split with a mediocre top-end 40 — that player is a quick-twitch separator, not a track burner.

e.g. 1.48s — elite acceleration, top-5% for WRs

src: rookie_measurables (split10); seeded for most 2026 players

20-Yd Shuttleseconds

shuttle

Short shuttle drill time — measures lateral quickness and COD speed.

The 20-yard short shuttle (5-10-5) tests lateral acceleration and change-of-direction quickness across a 10-yard span. Paired with 3-cone, it completes the agility profile.

Why it matters

For slot receivers, shuttle time predicts ability to gain separation on quick-game routes. For RBs, it correlates with lateral agility in tight spaces.

How to use

Lower is better. Compare within position. Use 3-cone as the primary agility metric; shuttle as a corroborating signal.

Watch out: Shuttle time is sometimes conflated with 3-cone. They measure different things: shuttle = lateral quickness, 3-cone = multi-directional change-of-direction over a longer course.

e.g. 4.15s — top-10 shuttle for a WR class

src: rookie_measurables (shuttle); seeded for most 2026 players

3-Coneseconds

cone3

Three-cone drill time — the gold-standard route-running agility test.

The L-shaped 3-cone drill measures multi-directional acceleration, deceleration, and change of direction. It correlates more strongly with NFL route-running efficiency than any other combine drill.

Why it matters

Receivers with elite 3-cone times (sub-6.50s for WRs) produce more separation on in-breaking routes. It is the single best combine predictor of YPRR for wide receivers.

How to use

Lower is better. Use for WR and slot receivers primarily. For RBs, it predicts break-tackle ability and open-field agility. Less predictive for TE due to blocking role.

Watch out: Many players skip the agility drills. Missing data (em-dash in the table) does not mean poor agility — it means no measurement. Do not treat null as a red flag.

e.g. 6.45s — elite for a WR, projects to consistent route separator

src: rookie_measurables (cone3); seeded for most 2026 players

Arm Lengthinches

arm_length

Arm length (inches) from shoulder to fingertip, measured at the NFL Combine.

Arm length measures reach, which directly affects blocking leverage for offensive linemen and catch radius for receivers. At WR and TE, longer arms correlate with contested-catch and back-shoulder success.

Why it matters

Short arms at WR (<31") can be mitigated by elite separation skill. Short arms at TE are a bigger concern because the position demands more physicality at the catch point.

How to use

Pair with hand size and contested-catch production. Use as a tiebreaker, not a primary driver. Arm length matters far more for OL/DE evaluation than for skill positions.

Watch out: Arm length in isolation explains relatively little variance in WR production. Use only as a supporting data point.

e.g. 33.5" — above-average arm length for a WR, supports outside-zone role

src: rookie_measurables (arm_in); seeded where official value unavailable

Block Grade0–100

blockGrade

Per-snap blocking grade.

Composite blocking grade derived from per-snap advanced charting (run-block, pass-block, hits/hurries/sacks allowed). Internal model – not a vendor surface.

Why it matters

Critical for evaluating RBs behind their blocking context and TE two-way utility.

How to use

Use to contextualize RB output – elite block grade + low carry share = a committee situation, not talent deficit.

src: internal model from play-by-play participation data

BMIratio

bmi

Body mass index — a proxy for play-strength relative to size.

BMI = weight (lb) × 703 ÷ height (in)². For skill positions it measures functional mass: an RB at 29 BMI is thicker and more likely to absorb contact than one at 26 BMI.

Why it matters

Extreme BMI values signal positional misfit: very low BMI at RB often means a gadget/returner role; very high BMI at WR signals contested-catch upside but route-tree limits.

How to use

Compare within position group only — raw BMI comparisons across QB/RB/WR/TE are meaningless. Look for outliers >1 SD above or below the positional mean.

Watch out: BMI does not capture muscle composition. A player with 33% lean muscle at 28 BMI is far more functional than one with 18% lean muscle at the same BMI.

e.g. WR at 28.3 BMI — dense for a receiver, projects to contested-catch role

src: college_profiles (weight_lb, height_in); seeded pre-combine

Comp FP

competitiveFp

Fantasy points in games decided by ≤ 7 points.

Filters out blowouts and garbage time. Surfaces who delivers when the game is competitive – more predictive of next-game value than raw totals.

Why it matters

Garbage time production is unreliable and one-game-script dependent. Comp FP removes it.

How to use

Use when evaluating players on teams with wide win/loss splits.

src: play-by-play aggregate

Early (Breakout Component)0–100 cohort percentile

Intermediatebreakout_early

Measures how quickly a rookie earned a significant snap role in Weeks 1–8.

Early Playing Time (weight: 20%) ranks the rookie's Weeks 1–8 snap share and early routes (WR/TE) or early snap rate (RB) against their positional cohort. A binary 'earned by Week 4' flag adds a bonus for players who secured an immediate starter role in the first month. Higher = the player was on the field consistently from the start of the season, which correlates with sustained volume and reduced depth-chart risk.

Why it matters

Rookies who earn early playing time tend to hold roles longer and have a clearer path to full-season opportunity. A low Early score signals the player spent Weeks 1–8 on the bench or in a rotational role — which can recover, but warrants scrutiny.

How to use

Use Early alongside Role Expansion: a low Early but high Role player likely emerged in the back half of the season (a late-season riser). A high Early + high Role = consistent starter all year = strongest signal profile.

Watch out: High Early score can reflect an injury-created opportunity rather than earned role. Check the 'Injury-Driven Role' badge — if present, the Early component may be less durable than the score implies.

e.g. Early 82 — starter from Week 1; Early 28 — depth piece through the first two months

used in:Rookie Breakout Model

src: v_rookie_breakout_inputs: snap_pct_early (Weeks 1–8) + earned_by_wk4 flag. Faithful: snap%, earnedByWk4. Proxied: routes_early as route-volume proxy for WR/TE.

Eff (Breakout Component)0–100 cohort percentile

Intermediatebreakout_efficiency

Measures how productively a rookie converted their opportunities into output.

Efficiency (weight: 20%) scores how well the player produced given their usage. For WR/TE: yards-per-route-run (40%), EPA-per-target (35%), and receptions-per-route (25%). For RB: EPA-per-rush (50%), rushing yards over expected (30%), and EPA-per-target (20%). For QB: EPA-per-dropback (50%) and CPOE (50%). Cohort-percentile-ranked within position. When league-wide percentile columns are available (pre-computed in the database), they serve as a fallback when the rookie cohort is below the minimum size.

Why it matters

A rookie who converts limited opportunities efficiently is demonstrating skill despite usage constraints. Efficiency + Involvement together identify the most actionable profiles: a player with high efficiency on real volume is the most credible signal.

How to use

High Eff + low Invol = efficient but underused — watch for a role expansion. High Invol + low Eff = usage without production, possible scheme mismatch or poor play quality. High both = the clearest signal profile.

Watch out: Efficiency on small samples is noisy. The confidence modifier in the composite score accounts for this — a player with Eff 85 on 12 routes will have a lower adjusted score than one with the same Eff on 150 routes. Check the confidence band before acting on a high Efficiency score.

e.g. Eff 88 — top-10% efficiency in the rookie class; Eff 42 — below-cohort conversion rate

used in:Rookie Breakout Model

src: v_rookie_breakout_inputs: yprr, epa_per_target, rec_per_route (WR/TE); epa_per_rush, ryoe (RB); epa_per_dropback, cpoe (QB). Faithful: yprr, EPA metrics. Proxied: ryoe sourced from model estimates.

Hand Sizeinches

hand_size

Hand circumference (inches) measured at the NFL Combine.

Measured tip-to-tip across the spread hand. Widely used as a ball-security proxy for RBs and QBs, and as a catch-radius indicator for WRs and TEs.

Why it matters

Small hands (<9.0" for RBs, <9.25" for QBs) correlate with fumble risk and cold-weather performance degradation. Large hands (>10") at WR suggest catch-radius and boundary upside.

How to use

Flag QBs and RBs below position thresholds. Do not over-index for WRs — route separation matters more than hand size at that position.

Watch out: Hand size is seeded pre-combine for many 2026 players. Values marked as seeded should be treated as estimates, not official measurements.

e.g. 9.75" — above average for an RB, strong ball-security signal

src: rookie_measurables (hand_in); seeded where official value unavailable

Heat-Tint Scalepercentile (0–1)

heat_tint

Cell background color encoding a player's class-wide percentile rank.

Each measurable column is shaded by the player's class-wide percentile. Elite (≥90th pct) = green; good (75–90th) = light green; neutral (25–75th) = no tint; below average (<25th) = red. The darker the tint, the more extreme the value.

Why it matters

Scan a row for green or red cells to quickly identify a player's standout traits and weaknesses without reading every number.

How to use

Use tints for pattern recognition at a glance. Hover any tinted cell to see the raw value. For headline metrics (40-yd, Model Score), the positional rank #N is also shown.

Watch out: Tints are class-wide (all skill positions combined), not position-specific. A red tint for height at WR may not be a red flag if most short WRs still beat their overall class percentile on other metrics.

e.g. Green cell on 40-yd = top-25% sprinter in this draft class

src: v_rookie_measurables (pct columns); computed via percent_rank() OVER the full class

Hi-Val (Breakout Component)0–100 cohort percentile

Intermediatebreakout_high_value

Measures the share of a rookie's usage in the highest-value scoring situations.

High-Value Usage (weight: 20%) captures whether the player is trusted in the highest-scoring situations. For WR/TE: red-zone opportunity share (40%), air yards share (35%), and two-minute drill share (25%). For RB: red-zone opportunity share (45%), goal-line share (35%), and two-minute share (20%). For QB: air yards per game (50%) and two-minute share (50%). All situation-share inputs are 0–100 values (percentage of team plays in each situation going to this player). Cohort-percentile-ranked within position.

Why it matters

A rookie trusted in the red zone and two-minute drill has earned the coaching staff's confidence in the moments that drive fantasy scoring. Even a player with moderate overall volume can have significant touchdown upside if their Hi-Val score is high — they are getting the ball when it matters most.

How to use

The 'High-Value Usage' badge fires when this component is ≥75. Look for rookies with Hi-Val 70+ who have not yet converted that usage into TDs — they carry realistic TD upside heading into next season. The 'Red-Zone Role' badge is the narrower version: fires when rz_opp_share ≥ 30% or when Hi-Val is strong and goal-line share is meaningful.

Watch out: High Hi-Val can reflect an injury-driven vacancy in the red zone. If the player played 8 or fewer games due to injury and accumulated RZ looks from backup opportunities, the signal may not repeat when the starter returns.

e.g. Hi-Val 85 — dominant RZ/2-min presence among rookies; Hi-Val 22 — fringe situational usage

used in:Rookie Breakout Model

src: v_rookie_breakout_inputs: rz_opp_share (v_redzone_player), air_yards_share, two_min_share, gl_share. Faithful: rz_opp_share, two_min_share, gl_share. Proxied: air_yards_share sourced from seasonal aggregates (may include garbage-time plays).

Invol (Breakout Component)0–100 cohort percentile

Intermediatebreakout_involvement

Measures season-long volume of touches and opportunity within the offense.

Involvement (weight: 25%) is the highest-weighted component. For WR/TE: a blend of target share (40%), targets-per-route-run (35%), and weighted opportunity (25%). For RB: touches per game (50%), weighted opportunity (30%), and target share (20%). For QB: dropbacks and dropbacks per game. All sub-metrics are cohort-percentile-ranked within position — a player with 25% target share is scored against other rookies, not veterans.

Why it matters

Involvement captures whether a rookie is embedded in the team's offensive structure as a true participant rather than a depth piece. High involvement with early role = a significant usage profile. This is the strongest leading indicator of sustained production among the five components.

How to use

Prioritize players with Invol ≥ 65 as meaningful contributors to the offense. Invol 40–64 suggests a rotational or red-zone-specific role. Below 40 = fringe usage; score driven by other components (check which ones).

Watch out: Weighted opportunity includes target-based inputs for non-RBs. Because WR/TE target data is available and route data is sometimes proxied (see Early component), involvement scores for WR/TE are more data-complete than Early scores on average.

e.g. Invol 78 — clear offensive role with real usage; Invol 31 — situational contributor

used in:Rookie Breakout Model

src: v_rookie_breakout_inputs: target_share, tprr (targets-per-route), weighted_opportunity (Faithful). Proxied: none for core sub-metrics; cohort fallback to league-wide percentiles when cohort < 6.

Role (Breakout Component)0–100 cohort percentile

Intermediatebreakout_role

Measures whether a rookie's role expanded or stabilized over the course of the season.

Role Expansion (weight: 15%) captures the trajectory of the player's role from early to late in the season. For WR/TE and RB: snap share delta (late-season snaps% minus early-season snaps%, cohort-ranked, weight 0.45/0.50) + opportunity delta (late opportunities per game minus early, weight 0.30/0.25) + stability (inverse of opportunity standard deviation — a consistent role scores higher, weight 0.25). For QB: snap share delta (0.60) + stability (0.40). A player whose role expanded meaningfully and then stabilized scores highest.

Why it matters

A rising role at the end of the year is a direct predictive signal for Year 2 usage. The NFL depth chart is path-dependent: a player who earned more snaps each month is more likely to be a starter going into training camp than one whose early role declined. Role Expansion is the forward-looking component of the model.

How to use

The 'Late-Season Riser' badge fires when Role Expansion is ≥70 and the snap delta is ≥10 percentage points. Those players are the clearest buy-ahead-of-market targets in dynasty and best-ball formats. Low Role Expansion despite high other components means the role plateaued — monitor for depth chart changes in the offseason.

Watch out: Stability (inverse of opportunity variance) penalizes players who had one or two outlier usage games. A player who was dominant in 10 games but missed 2 due to injury can see a lower stability score than their actual consistency warrants. Use the games-played count alongside Role to calibrate.

e.g. Role 91 — expanded snap share 22pp from early to late season; Role 38 — flat or declining role trajectory

used in:Rookie Breakout Model

src: v_rookie_breakout_inputs: snap_pct_late, snap_pct_early (delta), opp_pg_late, opp_pg_early (delta), opp_stddev (stability proxy). Faithful: all snap% and opp% inputs. Proxied: opp_pg uses combined carries + targets as touch proxy for RB; target-share delta folded into opp delta for WR/TE due to sparse weekly data.

RZ TD

rzTd

Touchdowns scored from inside the red zone.

Counts only TDs scored when the play started inside the opponent's 20-yard line.

Why it matters

Red zone volume is a leading indicator for touchdown upside.

How to use

Combine with red zone target share (receivers) or red zone carry share (backs) for true TD outlook.

Watch out: PPG can overrate unsustainable TD scoring. A player with 4 TDs on 4 red zone looks is less reliable than one with 8 red zone targets per game.

src: play-by-play aggregate

Signal Tags (REPS / HYPE / INJURY / DEPTH / ANALYST)categorical

signal_tags

Source-type labels on rookie highlight cards.

Each highlight card in the Rookie Watch stream carries one of five tags: REPS = on-field practice observation; HYPE = media or fan buzz with less verified basis; INJURY = health concern; DEPTH = depth-chart movement signal; ANALYST = named analyst take or projection.

Why it matters

Not all highlight types carry the same signal weight. REPS from a beat reporter at a practice rep is stronger evidence than HYPE from a national media take. Tags help you calibrate how much to act on each card.

How to use

Treat REPS and ANALYST tags as higher-confidence inputs. Weight HYPE lower until corroborated by REPS. INJURY tags should immediately prompt a check on current practice status.

Watch out: ANALYST tags do not imply accuracy — they simply mean a named analyst made the observation. Always cross-reference with REPS-based evidence before acting on analyst takes.

e.g. REPS tag: 'Love took all first-team RB reps in OTAs' — more actionable than a HYPE tag

src: rookie_highlights (tag column); tagged manually by the Blindside content team

Team Draft GradeA+ – D

team_draft_grade

Positional surplus value grade (A+ – D) for a team's full draft class.

Our grade aggregates the positional surplus value of each pick: expected fantasy/dynasty value at that slot minus the pick cost (Jimmy Johnson-style curve). Positive surplus = class performed above expected pick value.

Why it matters

Teams with consistently strong draft grades tend to build sustained depth. For fantasy, teams that nail skill-position drafts tend to have more depth-chart volatility — and more opportunity.

How to use

Use our grade as a context layer. A team with an A draft and a clear depth-chart vacancy is a prime target for rookies. Two external analyst grades (A1/A2) are also shown for comparison.

Watch out: Draft grades are inherently early-signal. Even the best classes take 2–3 years to validate. Do not use grade as a primary dynasty value driver in Year 1.

e.g. A- grade = above-average class that slightly exceeded expected pick value

src: team_draft_grades (analyst_a / analyst_b) + internal surplus model

Common Traps

Metrics frequently misapplied — understand these before making decisions.

PPG:Can overrate unsustainable TD scoring. Check if a player has high red zone looks or if TDs are luck-driven.

EPA:Noisy in small samples (< 50 snaps). Early-season EPA for low-usage players is often misleading.

YPRR:Needs route volume to stabilize. Players with fewer than 40 routes have volatile numbers that can overrate low-usage receivers.

Target Share:Must be paired with team pass volume. A 30% share on a team that throws 20 times ≠ 30% share on a team that throws 40 times.

ADP:Is market sentiment, not truth. ADP reflects consensus opinion – it can be systematically wrong in thin markets or early in the season.

Implied Probability:Raw implied probability includes vig (juice). Use the vig-removed probability for true edge calculation.

DvP:Can overrate matchups if the defense heavily suppresses a specific sub-role (e.g., tight ends). Always check position subtype.

Loading…

Core Fantasy

FPTSpoints

Basicfpts

Total season fantasy points.

Cumulative PPR fantasy points across the current season.

Why it matters

Useful for dynasty value and overall seasonal production context.

How to use

Use for season-long context; prefer PPG for week-to-week projections.

Watch out: Penalizes players who missed games – a player with 10 GP at 22 PPG beats 16 GP at 14 PPG.

e.g. 221 FPTS in 12 games

used in:Player ProfilesStats Leaders

src: player_stats_season

Basicgp

Games played in the season.

Number of games the player has appeared in this season – drives the PPG denominator.

Why it matters

Durability is a skill. Games played is a baseline availability signal.

How to use

Filter out low-GP samples before ranking by PPG.

used in:Player ProfilesStats Leaders

src: player_stats_season

PPGpoints / game

Basicppg

Fantasy points per game (PPR scoring).

Average fantasy points scored per game played. Uses PPR (1 point per reception).

formula:Total Fantasy Points ÷ Games Played

Why it matters

Better predictor of future output than season totals – normalizes for games played.

How to use

Rank players within their position group. Pair with GP to check consistency.

Watch out: Can overrate players on touchdown-dependent scoring runs. Check floor/ceiling spread.

e.g. 18.4 PPG in 12 games

used in:Player ProfilesStats LeadersDraft CenterWeekly Projections

src: player_stats_season

Usage

aDOTyards

Intermediateadot

Average depth of target.

Average air yards per target. Measures how far downfield a receiver is being targeted. High aDOT = downfield role; low aDOT = underneath/slot role.

formula:Total Air Yards ÷ Total Targets

Why it matters

Receivers with high aDOT carry more big-play upside but inconsistency. Low-aDOT players have high floors in PPR.

How to use

Combine with YPRR and target share. A high-aDOT receiver with declining target share is a concern.

Watch out: aDOT alone can overrate downfield receivers who rarely convert. Pair with catch rate and CPOE.

e.g. 12.4 aDOT (deep receiver) vs. 4.1 aDOT (slot receiver)

used in:Player ProfilesTape RoomStats Leaders

src: player_season_metrics

Air Yardsyards

IntermediateairYards

Total yards in the air on pass attempts to this player.

Air yards = yards the ball traveled in the air from the line of scrimmage to the intended target. Measures downfield targeting, not YAC or actual receiving yards.

Why it matters

Air yards predict target quality. A 6-catch, 80-yard game means different things at 2 avg air yards vs. 14 avg air yards (aDOT).

How to use

Use aDOT for per-target efficiency. Total air yards for volume.

e.g. 280 air yards in 4 games

used in:Player ProfilesTape Room

src: player_season_metrics

ATT

pAtt

Pass attempts (for QBs).

Total pass attempts. Includes incompletions and sacks taken in some sources.

Why it matters

Volume for QBs. High pass-attempt teams boost all receivers.

How to use

Use team-level ATT to contextualize receiver target share.

src: player_stats_season

AY%%

airYardsShare

Share of team air yards.

Sum of intended yards on a player's targets divided by team total air yards. Captures downfield workload.

Why it matters

Downfield air yards convert to big plays and TDs. Share above 30% is true X-receiver volume.

How to use

Use to identify deep threats whose ceiling exceeds catch-volume scoring.

src: player_season_metrics

CAR

carries

Rushing attempts.

Count of rushing attempts (designed runs + scrambles for QBs).

Why it matters

Volume is the floor for RBs. More carries = more scoring chances.

How to use

Combine with snap share to see if a player gets their carries in critical game situations.

src: player_stats_season

I5 (Inside-5)count

rz_i5

Opportunities from inside the opponent's 5-yard line — the highest-value red zone zone.

Why it matters

A player who leads their team in I5 looks is the closest thing to a guaranteed TD machine in fantasy. Even a player with mediocre overall RZ volume can be a top-10 scorer if they hoard I5 looks.

How to use

Sort by I5 absolute count first, then RZ share second. A player with 8+ I5 looks in 10 games has a clear path to 8+ TDs. Use as a tiebreaker in DFS for stacked game environments.

Watch out: I5 is volatile week-to-week (a player can go 3 games without a goal-line look). It is a season-level signal, not a weekly edge.

e.g. 12 I5 opps → 4 TDs scored (expected: ~5.5 for a QB, ~4.8 for a RB)

src: v_redzone_player: count of plays where field_position ≥ 95 (5 yds to goal)

Leverage Score Basis (Within-Pos vs. Overall)0–100 percentile (two separate windows)

leverage_basis

Whether Leverage Score is percentile-ranked against positional peers or all offensive skill players.

Why it matters

How to use

Default to within-pos for all single-position decisions. Switch to overall when comparing across positions for the same flex/UTIL slot. The basis toggle is in the page header.

Watch out: The overall basis can make slot receivers look weak (they have fewer GL looks structurally) and goal-line RBs look dominant. Within-pos is almost always the more actionable view.

src: v_leverage_score: two separate percent_rank() windows

Leverage Weightsweights: GL 30%, RZ 25%, C/L 20%, 2M 15%, LD 10%

leverage_weights

The five situational usage categories and their weights in the Leverage Score model.

Why it matters

How to use

Watch out: Leverage Score is a ROLE metric, not a raw-production metric. A team with a low-scoring offense can produce a BELL-COW player whose ceiling is capped by the team's scoring volume.

src: v_leverage_score: weighted composite, /methodology#leverage

Route %%

IntermediaterouteShare

Share of routes run out of total team pass routes.

How often a receiver runs a route when the team passes. Routes run per game divided by team passing plays. Separates receivers who are truly in the passing game from those who block on pass plays.

Why it matters

Filters out 'snap share illusion' – a receiver in the backfield or blocking on pass plays won't generate YPRR opportunities.

How to use

Pair with YPRR. Route Share × YPRR = opportunity-weighted efficiency.

Watch out: Low-volume receivers can post elite YPRR rates on 20 routes/game. Route share exposes the ceiling.

e.g. 78% route share

used in:Player ProfilesTape Room

src: player_season_metrics

RZ Opportunitycount

Intermediaterz_opportunity

Total red zone looks (targets + carries) inside the opponent's 20-yard line.

Why it matters

Red zone opportunity is the primary driver of touchdown probability. Players who receive consistent RZ looks have a repeatable path to 6-point scoring plays regardless of game script.

How to use

Compare RZ opportunity share — not raw counts — to correct for team offense volume. Pair with rzOppRank (%ile within position) to see whether volume is elite for the position.

Watch out: High raw RZ opportunity counts can come from low-efficiency offenses that need many chances to score. Always check conversion rate (TD%) alongside volume.

e.g. 35 RZ opps in 14 games = 2.5/game — top-tier RB volume

used in:ModelsPlayer Profiles

src: v_redzone_player: sum of rz_targets + rz_carries where field_position ≥ 80

Snap%%

BasicsnapShare

Percentage of offensive snaps the player was on the field.

Snap share = player snaps / total team offensive snaps. A proxy for role stability and usage ceiling.

formula:Player Snaps ÷ Total Team Offensive Snaps × 100

Why it matters

Players who are on the field can only score if they're on the field. Snap share is the on/off switch.

How to use

Look for players trending up in snap share – usage often leads points. Trend direction matters more than a single week.

Watch out: High snap share doesn't mean high target share. A blocker can have 90% snaps and 0 targets.

e.g. 85% snap share

used in:Player ProfilesOffensive LineTape Room

src: player_season_metrics

TGT

Basictargets

Times targeted by a pass attempt.

Count of pass attempts intended for this player. Strong leading indicator of receiver value.

Why it matters

Targets are the raw opportunity signal. High-target players have more paths to big games.

How to use

Pair with target share – raw targets vary by team pass volume.

Watch out: Target share must be paired with team pass volume. A 30% share on a team that throws 20 times is not equivalent to 30% on a team that throws 40 times.

e.g. 9.2 targets / game

used in:Player ProfilesStats LeadersTape Room

src: player_stats_season

TGT%%

BasictargetShare

Share of team pass targets.

Targets divided by team pass attempts. Captures opportunity independent of team volume.

formula:Player Targets ÷ Team Pass Attempts × 100

Why it matters

Target share is the cleanest pass-game opportunity stat. 25%+ is elite WR1/TE1 territory.

How to use

Pair with team pass volume – share of a high-volume offense beats share of a low-volume one.

e.g. 29% – clear WR1 target

used in:Player ProfilesStats LeadersTape Room

src: player_season_metrics

TPRRratio

tprr

Targets per route run.

Targets divided by route participations. Pure per-route efficiency.

Why it matters

Filters out players who only got targets because they ran more routes. 0.25+ is elite per-route conversion.

How to use

Use to find efficient pass-catchers buried in low-route roles.

src: player_season_metrics

Verdict: BELL-COWverdict label (categorical)

verdict_bell_cow

Elite positional leverage — dominates the highest-value usage situations within their position group.

Why it matters

How to use

Watch out: Bell-cow status can change overnight with an injury, trade, or depth chart addition. Monitor weekly snap share and carry share to confirm the role persists.

src: Derived verdict in leverage-score ModelLeaderboard config: leverageScore ≥ 90

WOPR0–1+

Intermediatewopr

Weighted Opportunity Rating – 1.5×TGT% + 0.7×AY%.

Combines target share and air-yards share into one opportunity score (Josh Hermsmeyer formula). Standardized so 0.6+ is elite.

formula:1.5 × TGT% + 0.7 × AY%

Why it matters

Best single number to measure receiver opportunity – ahead of target share alone.

How to use

Rank pass-catchers by WOPR within position; gaps reveal opportunity tiers.

used in:Player ProfilesStats Leaders

src: player_season_metrics

Efficiency

%ile (Within-Pos)0–100 percentile

pctile_within_pos

A player's percentile rank among players at the same position.

Why it matters

How to use

Use the within-pos toggle as the default for all research. Switch to overall only when comparing players across positions (e.g., evaluating a TE vs. WR for the same roster slot).

Watch out: A 90th-percentile RB is elite among RBs — but may trail many WRs on the same raw metric. Always anchor to the same basis when comparing across positions.

e.g. TD Dependency %ile = 85 means more TD-dependent than 85% of positional peers

src: Computed via percent_rank() OVER (PARTITION BY position) in DB views

CATCH%%

BasiccatchPct

Catch rate — receptions divided by targets.

formula:Receptions ÷ Targets × 100

Why it matters

How to use

Watch out: Raw catch rate penalizes deep routes by design. Use CPOE-equivalent receiver metrics when available to separate scheme-driven rates from genuine hands/separation quality.

e.g. 76% catch rate (deep WR, efficient) vs. 61% (slot, likely drops issue)

used in:Player ProfilesStats Leaders

src: player_stats_season: receptions / targets

CMP%%

BasiccompPct

Completion percentage — completions divided by pass attempts.

formula:Completions ÷ Pass Attempts × 100

Why it matters

How to use

Pair with CPOE (Completion % Over Expected) to separate scheme-driven high rates from genuine accuracy. A QB checking down in a quick-game system can post 70%+ CMP% with no actual accuracy edge.

Watch out: CMP% rewards short throws and penalizes aggressive downfield passing. A QB with 58% CMP% and a 14 Y/A is often more valuable than one with 72% CMP% and a 6 Y/A.

e.g. 68.4% CMP% — above average; 55% — low, check throw depth

used in:Player ProfilesStats Leaders

src: player_stats_season: completions / pass_attempts

CPOEpercentage points

cpoe

Completion % Over Expected.

Difference between a QB's actual completion rate and the rate expected given throw depth, location, pressure, etc. Above zero = beating model expectation.

Why it matters

CPOE isolates QB decision quality from receiver influence and scheme. A consistent +CPOE QB sustains receiver production.

How to use

Identify QBs who consistently beat expectation – their receivers benefit. Negative CPOE teams suppress receiver value.

Watch out: CPOE is not purely a receiver metric. Offensive line quality and scheme influence it heavily.

e.g. +4.1 CPOE

src: player_season_metrics

EPAexpected points / play

Advancedepa

Expected Points Added per play.

Average change in expected scoring outcome attributable to this player on each play. Positive = better than baseline. For QBs: EPA/dropback; RBs: EPA/rush; pass-catchers: EPA/target.

Why it matters

EPA is the gold-standard efficiency stat – it accounts for situation (down/distance/field position) that raw stats ignore.

How to use

Use for ranking within position group. Combine with volume metrics. An EPA leader with low volume may be scheme-dependent.

Watch out: EPA can be noisy in small samples. Requires 50+ snaps to stabilize. Early-season EPA for low-usage players is often misleading.

e.g. 0.22 EPA/target

used in:Player ProfilesTape RoomModels

src: player_season_metrics

RACR

racr

Receiver Air Conversion Ratio.

Receiving yards divided by air yards. RACR > 1 means the player turns intended yards into more than expected via YAC and contested catches.

Why it matters

Identifies YAC monsters and contested-catch specialists who produce beyond the throw.

How to use

Use to separate air-yards-dependent receivers from self-creators. RACR > 1.2 consistently = excellent run-after-catch skill.

src: player_season_metrics

Success %%

successRate

Share of plays that gain ≥ 50% / 70% / 100% of needed yards.

Percent of plays that succeed per the EPA framework: 50% of needed yards on 1st down, 70% on 2nd, or 100% on 3rd/4th. Position-agnostic efficiency measure.

Why it matters

Consistency metric – separates players with sustainable production from boom/bust players.

How to use

High success rate + high EPA = elite, consistent player. High EPA + low success rate = boom/bust.

src: play_participation + EPA framework

TD Share%

Intermediatetd_share

Percentage of a player's fantasy points sourced from touchdowns.

TD Share = (TD points) / (total fantasy points) × 100. A high TD share means the player relies heavily on scoring touchdowns to generate their fantasy output, which is volatile week-to-week.

formula:TD Points ÷ Total Fantasy Points × 100

Why it matters

Touchdown production is the least stable component of fantasy scoring. High TD share players carry regression risk; low TD share players have a more durable, volume-backed floor.

How to use

Compare within position group. A WR with 40%+ TD share is an over-performer relative to their target volume — fade in DFS. A WR with sub-15% TD share posting strong PPG has a sustainable floor.

Watch out: Don't conflate low TD share with low upside. Some players have low shares because they are volume monsters — TDs are a small fraction of a very large pie.

e.g. 38% TD share — fragile, regress-candidate; 12% TD share — volume-backed floor

used in:ModelsPlayer Profiles

src: v_td_dependency: td_pts / total_pts × 100

TD± (vs Expected)touchdowns (signed)

rz_td_diff

Actual red zone TDs minus the I5-weighted expected TDs from the opportunity model.

Why it matters

Raw TD counts are noisy. TD± normalizes for opportunity quality, isolating conversion luck from volume. This is the single best indicator of whether a player's TD pace will revert.

How to use

Watch out: Small-sample TD± swings wildly. Require 8+ RZ opps before treating the signal as meaningful. A player with 2 RZ looks can be +2.0 on a single fluky score.

e.g. TD± = -2.1 → scored 2.1 fewer TDs than expected given opportunity → BUY-LOW

src: v_redzone_player: rz_td − expected_td (I5-weighted model per /methodology#redzone-expected)

True YPRRyards / route (adj.)

trueYprr

YPRR adjusted for target depth and team pass volume.

YPRR normalized for average depth of target (aDOT) and the team's pass volume per game. Removes the advantage of running deep routes in a high-volume passing offense.

Why it matters

Compares receivers across different schemes and route trees on a level playing field.

How to use

Use when comparing a deep WR1 in a high-volume offense to a slot receiver in a run-heavy scheme.

src: internal model from player_season_metrics

TTTseconds

ttt

Avg time to throw (seconds).

Average seconds from snap to release for a QB. Lower TTT = quick-game scheme.

Why it matters

Quick-release offenses benefit slot receivers and underneath targets. Deep offenses boost high-aDOT WRs.

How to use

Pair with aDOT to understand scheme context. Low TTT + low aDOT = volume-first slot play.

e.g. 2.4s (quick-game) vs. 2.9s (timing offense)

src: nextgen_player_week.avg_time_to_throw

Y/Ayards / attempt

Basicypa

Yards per pass attempt — total passing yards divided by attempts.

Y/A = passing yards / pass attempts. Measures the average yardage gained on each throw, including incompletions. The simplest single-number measure of passing efficiency and downfield aggression.

formula:Passing Yards ÷ Pass Attempts

Why it matters

How to use

Watch out: Y/A can spike in garbage time when a trailing team throws downfield repeatedly. Filter to competitive-script situations before drawing conclusions about scheme.

e.g. 8.2 Y/A — aggressive downfield attack; 5.8 Y/A — short-to-intermediate scheme

used in:Player ProfilesStats Leaders

src: player_stats_season: passing_yards / pass_attempts

Y/Cyards / carry

Basicypc

Yards per carry — rushing yards divided by carries.

Y/C = rushing yards / carries. The fundamental per-attempt efficiency metric for rushers. Captures average yardage gained on each designed run or quarterback scramble, before accounting for volume.

formula:Rushing Yards ÷ Carries

Why it matters

How to use

Watch out: Y/C can be inflated by a small number of long breakaway runs, which are not predictable. Median carry (not average) is a better stability metric, though it is rarely surfaced publicly.

e.g. 4.8 Y/C — solid efficiency; 3.6 Y/C — below league average; 6.2 Y/C — elite

used in:Player ProfilesStats Leaders

src: player_stats_season: rushing_yards / carries

Y/Ryards / reception

Basicypr

Yards per reception — receiving yards divided by receptions.

formula:Receiving Yards ÷ Receptions

Why it matters

How to use

e.g. 17.2 Y/R — downfield role with big-play upside; 8.4 Y/R — volume slot target

used in:Player ProfilesStats Leaders

src: player_stats_season: receiving_yards / receptions

YPRRyards / route

Intermediateyprr

Yards per route run.

Receiving yards divided by routes run. Best single-number receiver efficiency stat – neutralizes opportunity differences and surfaces per-route production.

formula:Receiving Yards ÷ Routes Run

Why it matters

YPRR reveals efficiency regardless of scheme or volume. The best way to identify elite receivers in pass-light offenses.

How to use

Pair with route share to get the full picture. Elite YPRR with high route share = top-tier WR.

Watch out: YPRR needs route volume to stabilize. Players with fewer than 40 routes have volatile numbers. Can overrate low-volume players.

e.g. 2.4 YPRR (elite); 1.0 YPRR (below avg)

used in:Player ProfilesStats LeadersTape Room

src: yprr_player_season + player_season_metrics fallback

Projections

ADPpick #

Basicadp

Average Draft Position across redraft leagues.

Average Draft Position is the mean pick number where a player is selected across many drafts. Lower = drafted earlier.

Why it matters

ADP is the public's expected price. Players whose projections clearly beat their ADP are draft-day value.

How to use

Pair ADP with projected rank. Sort by `proj rank – ADP` to find value (and reaches).

e.g. ADP 14.2 = late 1st round, 12-team

used in:Draft CenterADP Tracker

src: consensus aggregator (multiple sources)

Earned vs. Sell-Highbinary classification based on pos_week_rate ≥ 0.55

gap_earned_vs_sellhigh

Whether a positive xFP gap reflects real efficiency or unsustainable variance.

Why it matters

How to use

When a player has a large positive gap, check pos_week_rate. ≥0.55 = EARNED (hold or buy more). <0.55 = SELL-HIGH (fade or trade). The threshold is conservative by design.

Watch out: Early in the season, pos_week_rate has limited sample. A 4-for-6 rate (67%) with only 6 games is encouraging but not definitive. Weight the signal more after week 10.

src: v_xfp_gap.pos_week_rate: fraction of weeks player's actual > xFP

ECRrank

ecr

Expert Consensus Ranking – the position rank.

Expert Consensus Ranking aggregates dozens of analyst rankings into a single positional rank (e.g. WR7, RB12). Lower is better. Updated weekly.

Why it matters

ECR is the market price for a player – use it to find places your view diverges from consensus.

How to use

Compare to your own projection rank. Big gaps either way flag a real disagreement worth investigating.

Watch out: ECR moves slowly; injury news and depth-chart changes lag the market by a few hours.

e.g. WR7 – top-tier WR1

src: consensus aggregator

ECR Δranks

ecrDelta

Weekly change in consensus ranking.

Difference between this week's expert consensus rank and last week's. Negative = ranking improved.

Why it matters

Quickly spots risers and fallers in analyst sentiment.

How to use

Use as a sentiment indicator – not a direct value signal.

src: consensus aggregator

Floor / Ceilingpoints

floorCeiling

10th and 90th percentile fantasy outcomes.

P10 (floor) and P90 (ceiling) projected outcomes from the player's distribution. Width = volatility; midpoint anchors to base projection.

Why it matters

Shows the range of realistic outcomes, not just the midpoint prediction.

How to use

High floor = safe start. Wide range = upside play for showdowns/tournaments.

src: internal projections

PROJ (wk)points

projWeek

This week's projected fantasy points.

Model projection for the upcoming game in PPR scoring. Built from base ROS projection, matchup adjustment, weather, snap-share trend, and recent EPA. Implied, not guaranteed.

Why it matters

Starting point for weekly lineup decisions. Always pair with floor/ceiling.

How to use

Sort by projection then filter by floor to find safe floors. Use ceiling for upside plays.

e.g. 18.2 projected, 10/28 floor/ceiling

src: internal projections

Rk Spreadranks

rkSpread

Disagreement across ranking sources.

Standard deviation of position rank across analyst sources. High spread = volatile or divisive player.

Why it matters

High spread means market uncertainty – which can be exploited on either side.

How to use

Wide spread + your model has a firm view = a real edge. Agree = press it.

src: consensus aggregator

Tier1–5

tier

Cluster grouping of comparable players.

Tiers group players whose expected output is statistically indistinguishable. Within a tier, draft for need; across tiers, draft for value.

Why it matters

Tier breaks – not individual ranks – are where the real positional cliffs live.

How to use

When the last player in a tier is about to be drafted, prioritize that position.

src: internal cluster model

Verdict: BUY-LOWverdict label (categorical)

verdict_buy_low

The opportunity is real but the production hasn't arrived yet — acquire now.

Why it matters

Buy-lows are the core mechanism of fantasy value extraction. If you identify players being suppressed by small-sample variance before the market reprices them, you win trades and draft picks.

How to use

src: Derived verdict in ModelLeaderboard configs (xfp-gap, red-zone)

Verdict: EARNEDverdict label (categorical)

verdict_earned

The player consistently beats their expected output — this is real efficiency, not luck.

Why it matters

How to use

Watch out: EARNED has a sample requirement. 6–8 games minimum before trusting the pos_week_rate. Early-season efficiency that looks like EARNED is often just a hot streak.

src: Derived verdict in xfp-gap ModelLeaderboard config: positive gap + posWeekRate ≥ 0.55

Verdict: REGRESSION RISKverdict label (categorical)

verdict_regression_risk

High TD dependency relative to positional peers — scoring is fragile and likely to regress.

Why it matters

How to use

src: Derived verdict in td-dependency ModelLeaderboard config: tdDepRank ≥ 80

Verdict: SELL-HIGHverdict label (categorical)

verdict_sell_high

The production is ahead of the opportunity — regression is statistically likely.

Why it matters

How to use

src: Derived verdict in ModelLeaderboard configs (xfp-gap, red-zone)

Verdict: SUSTAINABLEverdict label (categorical)

verdict_sustainable

Volume-backed floor — TD dependency is low, meaning production is durable and not TD-reliant.

Why it matters

How to use

src: Derived verdict in td-dependency ModelLeaderboard config: tdDepRank ≤ 25

xFP (Expected Fantasy Points)fantasy points

Advancedxfp_expected

Model-derived expected fantasy points based on opportunity volume and field position.

Why it matters

How to use

e.g. xFP = 18.4, actual = 14.1 → gap = -4.3 (under-converted; buy-low signal)

used in:ModelsPlayer Profiles

src: ff_opportunity_player_week: opportunity model via volume + field_position + game_state

xFP Gapfantasy points

xfp_gap

Actual fantasy points minus expected fantasy points (xFP).

Why it matters

The gap direction and persistence together determine whether a player is a buy-low, sell-high, or earned outperformer. A negative gap with high opportunity is the cleanest fantasy buy-low signal.

How to use

Watch out: A large positive gap in 3 games is nearly meaningless — variance is high over small samples. Require 6+ games and check pos_week_rate before acting on any gap signal.

e.g. Gap/G = -3.2 → under-converting by 3.2 pts/game → BUY-LOW signal

src: v_xfp_gap: actual_fpts − xfp_total

Wagering

Alpha

alpha

Return above what market pricing implies.

Borrowed from finance: alpha is the excess value generated vs. the expectation embedded in the market price. In fantasy/DFS: how much a player beats their consensus ADP or projection.

Why it matters

Real edge is alpha – consistently finding value the market misses.

How to use

Track hit rate on alpha plays over time. Consistent alpha generation = a genuine edge.

src: internal model

Confidence0–100

confidence

Model's certainty in its own projection, from 0–100.

A composite of sample size, recent consistency, and input data quality. Low confidence means the model has high uncertainty – typically due to small samples, injury clouds, or missing inputs.

Why it matters

Helps you weight projections correctly. High-confidence calls are actionable; low-confidence ones should be tempered.

How to use

Filter for confidence > 60 before acting on any model output. Below 40 is a caution flag.

e.g. Confidence: 72 (solid)

src: internal model

DvPpoints / game allowed

dvp

Defense vs. Position – fantasy points allowed to each position.

How many fantasy points a defense allows to each position (QB, RB, WR, TE) per game, ranked across the league. High DvP = favorable matchup.

Why it matters

Matchup context is one of the strongest weekly adjustments to any projection.

How to use

Use as a tiebreaker between similar players. A player ranked 12th at their position against a top-5 DvP matchup is worth a look.

Watch out: DvP can overrate matchups if the opposing team heavily suppresses a specific sub-role (e.g., tight ends). Check position subtype.

src: player_stats_season + matchup tables

Edge%

Intermediateedge

Model probability minus implied market probability.

Edge = model win probability minus the implied probability derived from market odds. Positive edge means the model projects the outcome as more likely than the market prices it.

formula:Model Probability − Implied Market Probability

Why it matters

The core concept of quantitative sports research. A positive edge does not guarantee a win – it means the price is favorable relative to the estimated true probability.

How to use

Look for edges above +5% after accounting for the vig. Smaller edges are noise; larger ones warrant attention.

Watch out: Edge is only as good as the model generating it. A systematically biased model produces systematically wrong edges.

e.g. +8.4% edge on team total over

used in:Wagering Tools

src: internal model

Implied Prob.%

impliedProbability

The win probability embedded in the market odds.

Derived from American odds: -110 → 52.4% implied probability (includes vig). Removing the vig gives the fair implied probability.

Why it matters

The baseline you're comparing against when calculating edge.

How to use

Convert any line to implied probability, then compare to your model's probability.

Watch out: Raw implied probability includes vig (juice). Use the vig-removed probability for true edge calculation.

e.g. -150 → 60% implied prob (before vig removal)

src: market feed

Market Line

marketLine

The consensus odds price for a given outcome.

The line (spread, total, player prop) available in the sportsbook market. Used as the denominator for edge calculations.

Why it matters

The market aggregates thousands of inputs. Beating it consistently is hard – and valuable.

How to use

Compare model projection to the implied line. The gap = raw edge.

src: market feed

Neutral Script

neutralScript

Projected game state assuming a close, competitive game throughout.

Why it matters

Game script heavily influences player usage. A neutral-script projection is more predictive when spread uncertainty is high.

How to use

Use neutral-script numbers for close games (+/- 3 points). For large spreads, use the directional script.

src: internal model

Paceplays / game

pace

Plays per game or seconds per play for a team's offense.

Offensive pace = average number of plays per game (or seconds between snaps). High-pace offenses run more plays, creating more opportunities for fantasy production.

Why it matters

Pace multiplies opportunity. Two identical offenses with different pace have very different fantasy floors.

How to use

Filter for high-pace teams when stacking. Pace + high team total = volume upside.

e.g. 70 plays/game (fast) vs. 60 plays/game (slow)

src: play-by-play aggregate

PROEpercentage points above expected

proe

Pass Rate Over Expected.

How much more (or less) a team passes than expected given game situation (score, time, field position). High PROE teams systematically pass more than expected.

Why it matters

PROE is a stronger predictor of target volume than raw target counts alone. Teams with consistently high PROE support high receiver floors.

How to use

Look for receivers on high-PROE teams. Combine with target share for reliable volume plays.

e.g. +8 PROE (frequently passes more than situation calls for)

src: play-by-play aggregate

Signal

signal

Composite indicator of a predictive pattern in the data.

A signal is any data-driven observation that has historically been predictive of future performance. Blindside Alpha surfaces signals as tagged annotations on player cards and the Tape Room.

Why it matters

Signals separate actionable patterns from random noise.

How to use

Signals are starting points, not conclusions. Combine multiple signals to build conviction.

src: internal model

Team Totalpoints

teamTotal

Implied points a team is expected to score.

Derived from the game total and spread: team total = (total ÷ 2) + (spread ÷ 2). The consensus expectation for a team's scoring output.

Why it matters

Team total drives game-script projections. High team totals boost receivers; low totals boost backs in closer games.

How to use

Use as a multiplier on individual projections. A team total of 27+ supports WR upside. Under 20 = tighten receiver projections.

e.g. Team total: 26.5

src: market feed

VORPpoints above replacement

vorp

Value Over Replacement Player.

The fantasy points a player scores above the replacement-level player at the same position (typically the last starter in a standard league). Measures true scarcity premium.

Why it matters

The cleanest measure of a player's actual impact on your team. A WR2 with high VORP is worth more than a WR1 with low VORP.

How to use

Use for trade valuation and waiver priority. Sort by VORP within position – it already accounts for positional scarcity.

e.g. +42 VORP = 42 points above the last WR starter in a 12-team league

src: internal model

DFS

Boom%%

boom_pct

Percentage of weeks the player scored above their position's boom threshold.

Why it matters

DFS tournaments are won by hitting boom weeks. Boom% identifies players with the ceiling to deliver tournament-winning scores on any given week — independent of their average.

How to use

In GPP construction, target players with boom% ≥ 25%. Combine with bust% to find true darts (high boom AND high bust). Combine with low bust% to find cash/GPP hybrids.

Watch out: Boom% without context can overrate players on touchdown-driven blowout games. Check if boom weeks are touchdown-dependent (fragile) or volume-driven (sustainable).

e.g. Boom% = 35% → player exceeded boom threshold in more than 1/3 of games

src: v_boom_bust: count(weekly_ppr > boom_line[pos]) / games

Bring-Back

bringBack

An opposing-team player added to hedge a game stack.

After stacking a QB + pass-catcher from Team A, the bring-back is a player from Team B who benefits if Team B scores. Reduces the risk that only one team wins the scoring game.

Why it matters

Bring-backs lower the variance of a game stack while maintaining correlation upside.

How to use

Default to the #1 receiver or a high-usage RB from the opposing team of your QB stack.

src: DFS strategy

Bust%%

bust_pct

Percentage of weeks the player scored below their position's bust threshold.

Why it matters

Cash game survival is about avoiding busts. Bust% is the floor metric — a player with bust% < 10% is almost never a losing start in 50/50 or H2H.

How to use

Cash: eliminate any player with bust% > 25% unless they're the only viable option at their price point. In GPP, high bust% paired with high boom% defines the ideal tournament dart.

Watch out: Bust% can be inflated by injury-shortened games. A player with 3 bust weeks due to mid-game exits has a misleadingly high bust% for healthy projections.

e.g. Bust% = 8% → reliable floor; Bust% = 40% → high-risk tournament gamble only

src: v_boom_bust: count(weekly_ppr < bust_line[pos]) / games

Ceilingpoints

ceiling

P90 projected fantasy outcome – the upside scenario.

The 90th percentile outcome in the player's projection distribution. Used for DFS tournament selection and boom/bust analysis.

Why it matters

Tournament DFS requires ceiling, not floor. Knowing a player's upside shapes roster construction.

How to use

In cash games, weight floor. In GPPs, sort by ceiling. Best plays have both.

src: internal projections

Fantasy Sharperatio (higher = more consistent)

Advancedfantasy_sharpe

PPG divided by weekly standard deviation — risk-adjusted scoring consistency.

formula:Mean Weekly PPG ÷ Standard Deviation of Weekly PPG

Why it matters

How to use

Cash: sort by Sharpe descending. GPP: sort by Sharpe ascending AND boom% descending. The Sharpe + boom% combo identifies optimal tournament darts.

e.g. Sharpe 2.4 = elite consistency (floor player); Sharpe 0.6 = high-volatility dart

used in:ModelsBoom/Bust Analysis

src: v_boom_bust: mean(weekly_ppr) / stddev(weekly_ppr) across the season

Floorpoints

floor

P10 projected fantasy outcome – the downside scenario.

The 10th percentile outcome in the player's projection distribution. Used for cash game selection.

Why it matters

Cash games (50/50s, head-to-head) are won by not losing. Floor matters more than ceiling.

How to use

In cash games, sort by floor first. A player with a 10-point floor rarely busts you.

src: internal projections

Hit Rate%

hitRate

% of weeks a player scores above their projected point total.

The fraction of games in which a player exceeded their pre-game projection. High hit rate = consistent over-performer relative to market.

Why it matters

Hit rate is the most actionable measure of projection quality for a specific player. Chronic over-performers are buy-low targets.

How to use

Players with > 60% hit rate over 8+ games are systematically undervalued by the market.

Watch out: Small-sample hit rates are meaningless. Require 8+ games.

e.g. 65% hit rate over 13 games

src: internal model vs. historical projections

Iso-Sharpe Contourvisual reference (not a raw metric)

iso_sharpe

Lines of equal Fantasy Sharpe on the PPG vs. Volatility scatter plot.

Why it matters

Iso-Sharpe contours let you immediately see which format each player is built for without computing anything: upper-left = cash plays; lower-right = GPP darts.

How to use

Read the scatter by quadrant: top-left of a given contour = better risk-adjusted return. Players you'd start in cash sit above the S=1.5 line. Tournament darts sit below S=0.75 with notable boom%.

Watch out: Iso-Sharpe contours assume stationary distributions. A player with 3 early-season bust weeks (injury) and 10 boom weeks has misleadingly poor visual placement. Check games played.

src: Computed from v_boom_bust: sd=mean/S lines for display in model-scatter-hero.tsx

Leverage

leverage

Ownership-adjusted DFS value – how much you own vs. the field.

In DFS, leverage = the difference between your roster percentage and the field's roster percentage. Being the only person in a large contest on a player who overperforms is high leverage.

Why it matters

DFS is a relative game. Scoring 50 points with 80% of the field on your player advances nobody. Same 50 points at 5% ownership = a large move.

How to use

In GPPs, actively look for leverage by fading chalk and targeting low-owned value.

src: ownership data (DFS platforms)

Recurrence %%

recurrenceRate

How often a performance type repeats week over week.

The probability that a specific performance pattern (e.g., 8+ targets, TD + 80+ yards) repeats in the next game given it happened in the prior game. Autocorrelation signal.

Why it matters

Some players have highly correlated weekly outputs. Recurrence lets you ride streaks intelligently.

How to use

After a big week, check recurrence rate before fading the player.

src: internal model

Stack

stack

Rostering players from the same team or game to correlate scores.

A stack is rostering a QB + at least one pass-catcher from the same team. A game stack adds a player from the opposing team. Correlation amplifies upside when a game script plays out high-scoring.

Why it matters

In high-variance GPP formats, stacking is the primary mechanism to differentiate your lineup and achieve the score needed to win.

How to use

In large GPPs, always stack. Use bring-back to hedge the opposing team's scoring.

src: DFS strategy

Verdict: DARTverdict label (categorical)

verdict_dart

High-variance GPP play — boom or bust, but worth the ceiling in tournaments.

Why it matters

How to use

Watch out: Not all DARTs are good GPP plays — a player can be volatile without having a meaningful ceiling. Always check that boomPct ≥ 25% before calling a player a DART, not just bustPct.

src: Derived verdict in boom-bust ModelLeaderboard config: sharpeRank ≤ 30, boomPct ≥ 25, bustPct ≥ 30

Verdict: STEADYverdict label (categorical)

verdict_steady

Reliable cash game floor — high Sharpe, low bust risk.

Why it matters

How to use

src: Derived verdict in boom-bust ModelLeaderboard config: sharpeRank ≥ 70, bustPct ≤ 25

Volatilitystandard deviation (points)

volatility

Standard deviation of weekly fantasy output.

The standard deviation of a player's weekly fantasy scores over the trailing N games. High volatility = boom/bust. Low volatility = consistent floor contributor.

Why it matters

Tournament DFS favors high volatility; cash games punish it.

How to use

Sort by volatility for GPP roster construction. Combine with ceiling to find high-upside plays.

Watch out: Volatility without volume context is misleading. A player with 2 touches/game has inherently higher volatility.

e.g. σ = 9.4 (high volatility)

src: internal model

Other

10-Yd Splitseconds

split10

Time to reach 10 yards from the snap — the best single acceleration proxy.

The first 10 yards of the 40-yard dash measure pure short-area quickness and acceleration, which is more predictive of NFL separation and open-field running than the full 40 time.

Why it matters

Routes start with a release from the line. A WR with an elite 10-yd split can create separation in the first 3 steps regardless of their top-end speed. For RBs, it predicts cut-back burst.

How to use

Lower is better. Elite is sub-1.50s. Sort by 10-split before full 40 for route-running receivers. Compare within position group only.

e.g. 1.48s — elite acceleration, top-5% for WRs

src: rookie_measurables (split10); seeded for most 2026 players

20-Yd Shuttleseconds

shuttle

Short shuttle drill time — measures lateral quickness and COD speed.

The 20-yard short shuttle (5-10-5) tests lateral acceleration and change-of-direction quickness across a 10-yard span. Paired with 3-cone, it completes the agility profile.

Why it matters

For slot receivers, shuttle time predicts ability to gain separation on quick-game routes. For RBs, it correlates with lateral agility in tight spaces.

How to use

Lower is better. Compare within position. Use 3-cone as the primary agility metric; shuttle as a corroborating signal.

Watch out: Shuttle time is sometimes conflated with 3-cone. They measure different things: shuttle = lateral quickness, 3-cone = multi-directional change-of-direction over a longer course.

e.g. 4.15s — top-10 shuttle for a WR class

src: rookie_measurables (shuttle); seeded for most 2026 players

3-Coneseconds

cone3

Three-cone drill time — the gold-standard route-running agility test.

The L-shaped 3-cone drill measures multi-directional acceleration, deceleration, and change of direction. It correlates more strongly with NFL route-running efficiency than any other combine drill.

Why it matters

Receivers with elite 3-cone times (sub-6.50s for WRs) produce more separation on in-breaking routes. It is the single best combine predictor of YPRR for wide receivers.

How to use

Lower is better. Use for WR and slot receivers primarily. For RBs, it predicts break-tackle ability and open-field agility. Less predictive for TE due to blocking role.

Watch out: Many players skip the agility drills. Missing data (em-dash in the table) does not mean poor agility — it means no measurement. Do not treat null as a red flag.

e.g. 6.45s — elite for a WR, projects to consistent route separator

src: rookie_measurables (cone3); seeded for most 2026 players

Arm Lengthinches

arm_length

Arm length (inches) from shoulder to fingertip, measured at the NFL Combine.

Why it matters

Short arms at WR (<31") can be mitigated by elite separation skill. Short arms at TE are a bigger concern because the position demands more physicality at the catch point.

How to use

Pair with hand size and contested-catch production. Use as a tiebreaker, not a primary driver. Arm length matters far more for OL/DE evaluation than for skill positions.

Watch out: Arm length in isolation explains relatively little variance in WR production. Use only as a supporting data point.

e.g. 33.5" — above-average arm length for a WR, supports outside-zone role

src: rookie_measurables (arm_in); seeded where official value unavailable

Block Grade0–100

blockGrade

Per-snap blocking grade.

Composite blocking grade derived from per-snap advanced charting (run-block, pass-block, hits/hurries/sacks allowed). Internal model – not a vendor surface.

Why it matters

Critical for evaluating RBs behind their blocking context and TE two-way utility.

How to use

Use to contextualize RB output – elite block grade + low carry share = a committee situation, not talent deficit.

src: internal model from play-by-play participation data

BMIratio

bmi

Body mass index — a proxy for play-strength relative to size.

BMI = weight (lb) × 703 ÷ height (in)². For skill positions it measures functional mass: an RB at 29 BMI is thicker and more likely to absorb contact than one at 26 BMI.

Why it matters

Extreme BMI values signal positional misfit: very low BMI at RB often means a gadget/returner role; very high BMI at WR signals contested-catch upside but route-tree limits.

How to use

Compare within position group only — raw BMI comparisons across QB/RB/WR/TE are meaningless. Look for outliers >1 SD above or below the positional mean.

Watch out: BMI does not capture muscle composition. A player with 33% lean muscle at 28 BMI is far more functional than one with 18% lean muscle at the same BMI.

e.g. WR at 28.3 BMI — dense for a receiver, projects to contested-catch role

src: college_profiles (weight_lb, height_in); seeded pre-combine

Comp FP

competitiveFp

Fantasy points in games decided by ≤ 7 points.

Filters out blowouts and garbage time. Surfaces who delivers when the game is competitive – more predictive of next-game value than raw totals.

Why it matters

Garbage time production is unreliable and one-game-script dependent. Comp FP removes it.

How to use

Use when evaluating players on teams with wide win/loss splits.

src: play-by-play aggregate

Early (Breakout Component)0–100 cohort percentile

Intermediatebreakout_early

Measures how quickly a rookie earned a significant snap role in Weeks 1–8.

Why it matters

How to use

e.g. Early 82 — starter from Week 1; Early 28 — depth piece through the first two months

used in:Rookie Breakout Model

src: v_rookie_breakout_inputs: snap_pct_early (Weeks 1–8) + earned_by_wk4 flag. Faithful: snap%, earnedByWk4. Proxied: routes_early as route-volume proxy for WR/TE.

Eff (Breakout Component)0–100 cohort percentile

Intermediatebreakout_efficiency

Measures how productively a rookie converted their opportunities into output.

Why it matters

How to use

e.g. Eff 88 — top-10% efficiency in the rookie class; Eff 42 — below-cohort conversion rate

used in:Rookie Breakout Model

Hand Sizeinches

hand_size

Hand circumference (inches) measured at the NFL Combine.

Measured tip-to-tip across the spread hand. Widely used as a ball-security proxy for RBs and QBs, and as a catch-radius indicator for WRs and TEs.

Why it matters

Small hands (<9.0" for RBs, <9.25" for QBs) correlate with fumble risk and cold-weather performance degradation. Large hands (>10") at WR suggest catch-radius and boundary upside.

How to use

Flag QBs and RBs below position thresholds. Do not over-index for WRs — route separation matters more than hand size at that position.

Watch out: Hand size is seeded pre-combine for many 2026 players. Values marked as seeded should be treated as estimates, not official measurements.

e.g. 9.75" — above average for an RB, strong ball-security signal

src: rookie_measurables (hand_in); seeded where official value unavailable

Heat-Tint Scalepercentile (0–1)

heat_tint

Cell background color encoding a player's class-wide percentile rank.

Why it matters

Scan a row for green or red cells to quickly identify a player's standout traits and weaknesses without reading every number.

How to use

Use tints for pattern recognition at a glance. Hover any tinted cell to see the raw value. For headline metrics (40-yd, Model Score), the positional rank #N is also shown.

e.g. Green cell on 40-yd = top-25% sprinter in this draft class

src: v_rookie_measurables (pct columns); computed via percent_rank() OVER the full class

Hi-Val (Breakout Component)0–100 cohort percentile

Intermediatebreakout_high_value

Measures the share of a rookie's usage in the highest-value scoring situations.

Why it matters

How to use

e.g. Hi-Val 85 — dominant RZ/2-min presence among rookies; Hi-Val 22 — fringe situational usage

used in:Rookie Breakout Model

Invol (Breakout Component)0–100 cohort percentile

Intermediatebreakout_involvement

Measures season-long volume of touches and opportunity within the offense.

Why it matters

How to use

e.g. Invol 78 — clear offensive role with real usage; Invol 31 — situational contributor

used in:Rookie Breakout Model

src: v_rookie_breakout_inputs: target_share, tprr (targets-per-route), weighted_opportunity (Faithful). Proxied: none for core sub-metrics; cohort fallback to league-wide percentiles when cohort < 6.

Role (Breakout Component)0–100 cohort percentile

Intermediatebreakout_role

Measures whether a rookie's role expanded or stabilized over the course of the season.

Why it matters

How to use

e.g. Role 91 — expanded snap share 22pp from early to late season; Role 38 — flat or declining role trajectory

used in:Rookie Breakout Model

RZ TD

rzTd

Touchdowns scored from inside the red zone.

Counts only TDs scored when the play started inside the opponent's 20-yard line.

Why it matters

Red zone volume is a leading indicator for touchdown upside.

How to use

Combine with red zone target share (receivers) or red zone carry share (backs) for true TD outlook.

Watch out: PPG can overrate unsustainable TD scoring. A player with 4 TDs on 4 red zone looks is less reliable than one with 8 red zone targets per game.

src: play-by-play aggregate

Signal Tags (REPS / HYPE / INJURY / DEPTH / ANALYST)categorical

signal_tags

Source-type labels on rookie highlight cards.

Why it matters

How to use

Treat REPS and ANALYST tags as higher-confidence inputs. Weight HYPE lower until corroborated by REPS. INJURY tags should immediately prompt a check on current practice status.

Watch out: ANALYST tags do not imply accuracy — they simply mean a named analyst made the observation. Always cross-reference with REPS-based evidence before acting on analyst takes.

e.g. REPS tag: 'Love took all first-team RB reps in OTAs' — more actionable than a HYPE tag

src: rookie_highlights (tag column); tagged manually by the Blindside content team

Team Draft GradeA+ – D

team_draft_grade

Positional surplus value grade (A+ – D) for a team's full draft class.

Why it matters

Teams with consistently strong draft grades tend to build sustained depth. For fantasy, teams that nail skill-position drafts tend to have more depth-chart volatility — and more opportunity.

How to use

Use our grade as a context layer. A team with an A draft and a clear depth-chart vacancy is a prime target for rookies. Two external analyst grades (A1/A2) are also shown for comparison.

Watch out: Draft grades are inherently early-signal. Even the best classes take 2–3 years to validate. Do not use grade as a primary dynasty value driver in Year 1.

e.g. A- grade = above-average class that slightly exceeded expected pick value

src: team_draft_grades (analyst_a / analyst_b) + internal surplus model