Dynasty Owl Subscribe

Inside the Owlgorithm

The Owlgorithm is Dynasty Owl's dynasty value model. It combines what players are doing on the field with what the trade market will pay for them, then adjusts for age, role, and your team's window. Here's how it thinks.

How a ranking gets built

Step by step

Every player and pick goes through the same seven steps, every week.

  • On the fieldPoints scored, and the points his usage should produce
  • RoleSnap share, target share, and air yards
  • AgeWhere he sits on his position's curve
  • Trade marketWhat managers will actually pay
OwlgorithmSeven steps, every week

One player, three values

  • ContenderThis season
  • BalancedAny team
  • RebuilderFive seasons out
  1. Gather the data

    Weekly stats and play-by-play from this season and last season, plus a fresh snapshot of the dynasty trade market. Last season counts, but this season counts more.

  2. Project production

    Estimate half-PPR points per game, leaning mostly on expected points (what a player's usage should produce) and less on the actual box score, so luck gets filtered out.

  3. Compare to replacement

    Subtract what a replacement-level starter at the same position scores. What's left is value over replacement (VORP): the edge a player gives your lineup each week.

  4. Project the next four seasons

    Project each of the next four seasons on its own: hot and cold streaks fade toward expected points, each player ages on his position's curve (adjusted for his type of player), and the odds he's still playing and his injury history trim the later years.

  5. Adjust for context

    Nudge the value for durability (games played), QB efficiency, role and target share, draft capital and contract, and the quality of the player's offense.

  6. Blend with the market

    Combine the model's view with the trade market. The more games a player has played, the more the stats count, and the stats count a little more when the market price is shaky (swinging week to week, or for a player few leagues roster). Rookies and players with small samples lean on the market.

  7. Fit it to your league

    Re-run everything for your team's window (contender, balanced, rebuilder), your scoring (PPR and TE premium), and, with a linked league, its size and starting lineup.

What the Owlgorithm looks at

The inputs, in depth

Twelve inputs, with what each one measures, why it matters, and an example of each. The examples are a snapshot from Week 6 of 2026 (October 21); every player's current numbers are on his player page, updated each week.

Expected fantasy points

01 · Opportunity

What a player's targets, carries, and pass attempts should score, based on where on the field and how far downfield each one happens.

Usage is more stable than results. A player getting great opportunities will usually score, even if he hasn't yet.

Right nowJosh Allen (21.9 per game), Dak Prescott (21.3) and Christian McCaffrey (21.1) lead the league in expected half-PPR points.

Touchdowns over expected

02 · Regression

The gap between a player's actual touchdowns and what his usage would normally produce.

Touchdowns swing on luck. Players far above expected tend to cool off; players far below tend to bounce back. The projection leans toward expected, so hot streaks don't inflate values.

Right nowRunning hot: Dallas Goedert (+0.37 TDs per game over expected). Due for more: Dak Prescott (−0.64), Jacoby Brissett (−0.56), Daniel Jones (−0.45).

Value over replacement

03 · VORP

Projected points per game minus a replacement-level starter at the same position. In a 12-team superflex league, replacement is roughly QB24, RB30, WR42, and TE14.

Points only matter relative to who else you could start. A 15-point TE is worth more than a 15-point WR because good TEs are scarce.

Right nowJahmyr Gibbs (+11.1 points per game over replacement), Christian McCaffrey (+10.9) and Puka Nacua (+9.3) lead.

QB scarcity

04 · Superflex

Replacement level is set for leagues that start two quarterbacks, so the 24th-best QB is the baseline.

With only 32 starting QBs in the NFL, even a mid-tier starter is a valuable asset in superflex.

Right nowSeven of the top 25 Owlgorithm assets are quarterbacks.

Position-specific aging

05 · Age curve

Value is projected over the next four seasons. Running backs start declining after about 26, wide receivers after 28, tight ends after 29, and quarterbacks after 33. Player type moves that: rushing quarterbacks, small backs and deep-threat receivers start a year earlier; pass-catching backs and possession receivers hold on longer. Older players also carry the odds of retiring before each coming season. Young players get a small bump for development.

A dynasty roster is a multi-year investment. Two players with the same stats can be worth very different amounts. See when players peak.

Right nowChristian McCaffrey, at 30, is #6 for contenders but #132 for rebuilders.

Games played

06 · Durability

Running backs, receivers, and tight ends get a small adjustment for how often they've been on the field: the share of their team's games they played over the last three seasons and this one, with recent seasons counting more. A player who misses a lot of time loses up to 10% of his value; an iron man gains up to 2%. Quarterbacks are left out, because their games played mostly reflect who started.

Production is measured per game, so without this a player who misses six games a year looks as good as one who never misses. Missed games are lost points, and they tend to repeat. Injury history also trims the seasons after this one: a long absence with a knee, Achilles or foot injury, or repeated hamstring, groin or calf injuries.

Right nowChristian Watson has played 70% of Green Bay's games over that stretch, which trims his value 6.5%.

EPA and CPOE

07 · QB play

Expected points added per dropback and completion percentage over expected, plus rushing production.

Efficient quarterbacks keep their jobs and lift their receivers. Rushing gives a QB a high fantasy floor.

Right nowBrock Purdy (+0.26 EPA per dropback), Jared Goff and Drake Maye (+0.22 each) lead qualified quarterbacks.

Target share, WOPR, and snaps

08 · Role

For receivers and tight ends: target share, air yards share, and WOPR, which combines the two. For running backs: share of offensive snaps.

Role is the best early signal of a breakout, and the most stable part of a player's profile.

Right nowJaxon Smith-Njigba's WOPR (0.89) is the best among receivers. Rookie Carnell Tate (0.75) already ranks third.

Draft capital and contract

09 · Insulation

Where a player was drafted (for his first three seasons) and how much his contract pays as a share of the salary cap.

Teams give high picks and big contracts longer leashes. That protects a player's role, and his dynasty value.

Right nowRebuilders weigh this most. It's why recent first-round picks hold up in the Rebuilder rankings.

Team offense

10 · Environment

The efficiency of the player's offense, measured by expected points added per play this season.

The same player scores more in a good offense, with more drives, red zone trips, and touchdowns.

Right nowIt's a small adjustment by design, because offenses change quickly.

How much to trust the market price

11 · Confidence

How much a player's market price has swung between weekly snapshots, and how many dynasty leagues roster him. For players few leagues roster, the Owl also checks where they go in dynasty startup drafts.

A price that jumps around, or comes from a player who rarely trades, says less than a steady one. When the market is shaky the stats get a little more say (up to 10 points of weight), and draft position fills in for missing trades.

Right nowStars everyone rosters barely move; the effect is mostly on fringe starters and rookies.

Trade market, momentum, and team window

12 · Market & window

The current dynasty trade market and how it's trending, then separate Contender and Rebuilder versions of every value.

A model shouldn't ignore what other managers will actually pay. Comparing the two is how the Owlgorithm finds undervalued and overvalued players. A player is flagged only when the two disagree on both price (6% or more) and ranking spots: 3 spots near the top of the board, rising to 15 deeper down, because at the very top a big price gap can be less than one spot. Overvalued stars are tagged "Sell high": move them only for a premium.

Right nowAbout one in five ranked players is flagged as undervalued and about one in seven as overvalued.

When players peak

Age curve

The share of fantasy starters whose best season, in PPR points per game, came at each age. Running backs peak first and fade fastest; quarterbacks are spread over the widest range. Every QB, RB, WR, and TE profile shows his position's curve with him on it.

  • QBmedian peak 27 · 60 players
  • RBmedian peak 25 · 143 players
  • WRmedian peak 25 · 182 players
  • TEmedian peak 25.5 · 92 players
PRIME WINDOW · 23–260%5%10%15%20%25%2123252729313335AGE AT THE START OF THE SEASON
Seasons 2012–2025 with 8+ games. Players with 3+ such seasons whose best year was a fantasy starter's (QB 14+, RB and WR 10+, TE 7+ PPR points a game). Lightly smoothed.

One player, three values

Windows

The same inputs, looked at over different time horizons.

Contender

Looks at this season. Current production counts most, rookies who aren't contributing and future picks count less, and proven veterans hold their value.

Balanced

The overall rankings: a multi-season view for any team, with tags comparing each player to the trade market.

Rebuilder

Looks five seasons out. Youth, development, draft capital, and picks count most, and aging veterans lose value faster.

PlayerAgeContenderBalancedRebuilder
Christian McCaffrey30#6#49#132
Derrick Henry32#20#104#155
Davante Adams33#35#101#224
Jeremiyah Love21#40#30#27

Ranks among players as of October 21, 2026, at 0.5 PPR.

Scoring matters too. On the Tiers page and in the Trade Analyzer, you can set points per reception (0 to 1) and TE premium (0 to 1.5). Each player's projection and his position's replacement level are recalculated from his projected receptions.

Where the data comes from

Sources

  • nflverseOpen-source NFL data: weekly player stats, play-by-play with EPA and CPOE, snap counts, advanced stats, draft history, contracts, and player ages.
  • ffopportunityAn open expected-fantasy-points model built on NFL play-by-play. It's the source for expected points and touchdowns over expected.
  • KeepTradeCutCrowdsourced dynasty trade values (superflex, 0.5 PPR), used for the market side of the model and for momentum.
  • IDP (nflverse defense)Weekly tackles, assists, tackles for loss, sacks, QB hits, interceptions, passes defended, forced fumbles, and defensive snap share for every DL, LB, and DB. With no public IDP trade market, IDP values come from the model alone: points per game under your IDP scoring, above replacement at the position (set by your league's starting slots), adjusted for age (peaks end around 29 for DL and 28 for LB and DB), role, draft capital, and injury news. Because most leagues start only one IDP, that's the default, and IDP values scale with the number of IDP starters: in a one-IDP league only the very best defenders carry real value, while in a full IDP league (2 DL, 2 LB, 2 DB, and a flex) the top ones rank alongside solid offensive starters. Set your slots with the IDP control, or connect your Sleeper league and the Owl reads them for you. 2027 IDP rookies on the Rookies page and Rookie Draft board are weighted the same way.
  • Dynasty IDP, K and DST expert consensus (FantasyPros, via DynastyProcess)With no IDP trade market, the Buy/Sell labels on defenders compare the Owlgorithm's rank at each position (DL, LB, DB) with the weekly dynasty expert consensus rank there. Because the experts and a production model disagree far more than trade markets do, a defender is a Buy only when we rank him at least twice as high as the consensus (and at least 5 spots higher), and a Sell the other way around; defenders the consensus doesn't rank get no label. Kickers and team defenses get the same treatment against the dynasty K and DST consensus. The consensus is used for the labels only and doesn't change any value.
  • Trade WatchSome players would be worth more somewhere else. For every player with trade reporting in the last month, each story is weighed by how firm it is (a done deal, a report, or speculation) and how recent it is to estimate the chance he's traded before the deadline, blended with public prediction-market odds where a market lists him. Separately, the Owlgorithm estimates his upside in a better situation from the gap between his talent (draft capital, efficiency, his best season) and his production in his current offense, capped at +15%. His value includes chance x upside while the trade window is open; if he isn't traded by the deadline, the boost goes away on its own. Look for the ★ Trade upside note on Tiers, the Trade Watch panel on his player page, and the Trade Watch list on Trade Targets.
  • Kickers and team defenses (nflverse)Field goals by distance, extra points, and misses for every kicker; sacks, interceptions, fumble recoveries, defensive and return touchdowns, safeties, blocked kicks, and points allowed for every defense, scored with your league's settings. Turn them on with the K/DST switch, or connect a Sleeper league that starts them. These spots are drafted and streamed for this season's points, so their values are about 85 to 95% this season's production and only a small dynasty layer on top: a future projection pulled toward the average, a small age adjustment for kickers (who stay effective well into their 30s), and the age of a defense's core. Even the best K or D/ST lands around a mid-to-late 4th-round rookie pick. Weekly projections scale kickers with their offense's expected scoring and defenses with the opponent's expected scoring, sack rate, and giveaways.
  • Sleeper, MyFantasyLeague, Fleaflicker and ESPNPublic league data, read in your browser when you link a league: rosters, settings, drafts and trades. Sleeper also supplies player IDs for matching.
  • FantasyCalcDynasty values built from real trades, checked every three hours, and blended into the market side so values move between weekly refreshes. Also the source of the live ticker. FantasyCalc doesn't sponsor or endorse Dynasty Owl. The Trade Analyzer doesn't use it.

The rankings are refreshed weekly during the season with the latest stats and a new market snapshot. Between refreshes, the market side follows live FantasyCalc values and approved injury news.

What it can't see (yet)

Honest limits

  • Route data. Yards per route run and route participation need paid charting data, so target share and WOPR stand in for now.
  • College production. Rookies without NFL snaps lean almost entirely on the market until they play.
  • Breaking news. Injuries, trades, and coaching changes show up after the next weekly update.
  • A long track record. The model is new. Its rankings are saved every week, so you can see how each player's value has moved on his player page.
Market
Dynasty values since August 11, 2026Matthew GoldenWR▲ 78%Braelon AllenRB▲ 64%Isaiah LikelyTE▲ 56%Chuba HubbardRB▲ 53%Christian WatsonWR▲ 51%Jonah ColemanRB▲ 50%Omarion HamptonRB▼ 40%Justin HerbertQB▼ 34%Kyler MurrayQB▼ 34%TreVeyon HendersonRB▼ 33%Brian Thomas Jr.WR▼ 33%Daniel JonesQB▼ 32%2027 Early 1st▼ 6%2027 Mid 1st▲ 2%2028 Early 1st▲ 9%2027 Late 1st▲ 2%2028 Mid 1st▲ 6%2028 Late 1st▲ 13%QB median▼ 5%RB median0%TE median▲ 10%WR median▼ 2%
Values: FantasyCalc