Dynasty rankings are a forecast, not a ledger. The right way to read ours is to look at three things together: where a player ranks, what tier they sit in, and the trajectory next to their name. This page explains how each of those gets computed, where the data comes from, and the cases where we override the numbers.
The fields, in order
Rank
Two numbers per player, both at the overall and positional level. For NFL, the primary is Superflex (SF), which assumes you can start two quarterbacks, with a secondary 1QB column. For MLB the primary is the standard format rank. Position pages always show the positional rank as the headline and the overall rank as a small secondary number, so you can read either view at a glance.
Tier
Tiers compress small ranking gaps into the groups that actually matter at the draft table. Tier 1 is the franchise core. Tier 2 is reliable starters with upside. Tier 3 is the bulk of your starting lineup. Tier 4 is high-variance flex types and younger players climbing. Tier 5 and 6 are deep roster shots and stash candidates.
Tier breaks are placed where the underlying value gap is large enough that trading across them should require a meaningful sweetener. If two players are in the same tier, the order matters less than the cluster they belong to.
Trajectory
The arrow next to a player captures where their stock is heading over the next 12 to 18 months. Up means we expect to be moving them higher, flat means we expect stability, and down means we expect to be moving them lower as the situation, usage, or age curve plays out. Trajectory is forward-looking on purpose: dynasty decisions are made on tomorrow, not last week.
Trade window
A short label that tells you how long this player can carry your roster at their current level. A two-year window means you should treat them like a depreciating asset and consider cashing in before next offseason. A five plus year window means you can build around them. Trade window is the most opinionated field on the site and the one we revisit most often.
One-liner
A single sentence under each player that captures the core thesis. It is not a stat dump. It is the reason we have them where we have them, written so you can read it during a draft without losing your place.
How the rank is calculated
Each ranking pass blends four inputs, weighted by position:
- Long-window production. Three-year rolling production at the position, with the most recent season weighted slightly higher. We use opportunity stats (targets, carries, snaps, pitches thrown) alongside output (yards, touchdowns, wOBA, strikeouts), because opportunity is stickier year over year.
- Situation. Team, role on the depth chart, surrounding talent, quarterback (or pitcher) play, coaching tendencies, and contract length. A situation shock (trade, free agent signing, coaching change) moves a player inside the next update cycle, not at the end of the season.
- Age and position curve. Each position has its own dynasty age curve: receivers age more gracefully than running backs, starting pitchers age differently than relievers, tight ends peak later than the other skill spots. The curve does not pick the rank by itself, but it sets the ceiling on how high a player can climb at a given age.
- Consensus. A blended consensus from a small number of trusted dynasty sources. We do not chase the consensus and we do not contrarian for the sake of it. The consensus check is a sanity rail; if we are more than a tier apart from the field we ask why, and the answer either holds up or the rank moves.
When we override the model
The blended rank is the starting point, not the finish. We override it for any of the following:
- Injury news that the model cannot reflect quickly enough, especially soft-tissue and ligament injuries that have known dynasty implications.
- Coaching or scheme changes that fundamentally alter usage. A play-caller swap that turns a 60 percent route-runner into a slot-only role gets reflected the week it is announced.
- Off-field situations that have a credible chance of affecting availability, handled with judgment and only when reporting is confirmed.
- Rookie news in the months between the combine and the draft, where the model has the least signal and the situation has the most.
The Superflex premium
A quarterback is worth meaningfully more in Superflex leagues than in 1QB, because you can start two of them. We treat 1QB as the canonical ranking and derive Superflex value at render time by applying a multiplier to each quarterback based on their position-only rank, then re-sorting the player pool. Non-quarterbacks have a 1.0x multiplier; their Superflex rank only shifts because quarterbacks get inserted around them.
The multiplier table, keyed off QB positional rank:
- QB1 through QB4 (elite): 2.0x
- QB5 through QB9 (high-end starter): 1.8x
- QB10 through QB15 (mid starter): 1.5x
- QB16 through QB24 (back end starter): 1.3x
- QB25 and below (streamer or backup): 1.1x
The curve was calibrated against the SF ranks we set by hand for the current quarterback room, so it reproduces our manual ordering within a few percent. Quarterback profiles surface this multiplier as a Format Premium stat. When our gut and the formula disagree (a young rookie before Week 1, for example), we set a manual override on that single player and leave the rest of the model alone.
Contract values for salary cap leagues
For salary cap dynasty leagues, every player at the top of the board carries a recommended contract: an AAV expressed as a share of cap, a length of one to five years, and an overpay ceiling and bargain floor that bracket the recommendation. We store everything as a share of cap so the same number renders correctly in a $200 Sleeper league, a $500 league with finer granularity, and an MFL league that scales with the actual NFL cap.
Share of cap
We use share of discretionary cap, defined as the league cap minus the minimum salary times the roster size. A 25-man, $1 minimum, $200 league has $175 of discretionary cap. The recommended AAV is a fraction of that number, derived from a player’s Superflex rank:
share = 0.15 / (1 + rank_sf × 0.08)
At the top of the board this produces roughly 14% for a true SF1, 10% around SF5 to SF6, 8% around SF11, 5% around SF25, and 1.7% around SF100. The overpay ceiling is 1.35 times the recommendation and the bargain floor is 0.75 times. The bounds matter more than the midpoint: under the floor is a steal, over the ceiling is regret.
Years
Contract length runs from one to five years and is driven by the share itself rather than the overall ranking tier, because elite quarterbacks have a low overall rank in 1QB but a high share in SF. The base length comes from the share bucket, then we adjust for age and trajectory and clamp to one through five.
- Share at or above 10%: 5 year base
- Share 7% to 9.9%: 4 year base
- Share 4% to 6.9%: 3 year base
- Share 2% to 3.9%: 2 year base
- Share below 2%: 1 year base
Age adjustment: 22 or younger adds a year, 23 to 28 is neutral, 29 to 30 subtracts a year, 31 and older subtracts two. Trajectory adjustment: rising adds a year, stable is neutral, declining subtracts a year. Final length is clamped to one through five.
Public ranges versus your exact league value
A player’s actual contract value swings hard based on your league’s scoring. Half PPR versus full PPR moves wide receivers and pass-catching running backs by 10 to 15 percent. Six-point passing touchdowns instead of four move quarterbacks by 20 to 25 percent. A TE premium league bumps tight ends 30 to 40 percent. IDP changes the entire cap allocation.
Showing one precise number on a public player page would mislead anyone whose league does not match the assumed scoring. So on public pages we show a share-of-cap range (typical floor to ceiling) and a contract length range. The range is generous on purpose; your league’s exact value lives inside it.
Signed-in users get the exact value. Paste your Sleeper or MFL league ID once and Cap Penalties detects your scoring config (PPR, pass TD value, premium, IDP, roster size, cap amount) and recalculates every recommendation against your settings. The same calculator powers the league import tool and will drive every contract surface on the site once auth ships.
What this model does not do yet
- No support for franchise tags, restricted free agents, or escalator clauses. The model assumes flat AAV with no in-contract step-ups.
- No position-aware age cliff. A 29 year old wide receiver and a 29 year old running back get the same age penalty even though running back longevity is shorter. We handle that case by overriding individual players where it matters.
- No dead cap modeling. The contract is the contract; cutting it next season is a league-specific question we will tackle when the trade tool ships.
The rookie board
The rookie board is a separate model. It ranks the incoming class, weighting college production by age, athletic testing by position, and projected draft capital. Once the draft happens, draft capital becomes the dominant input. Tiers on the rookie board are tuned for startup superflex and 12 team formats by default.
Rookies appear on the main rankings only after they sign with a team and enter their first preseason, with starting tiers shaped by their draft slot and landing spot.
Update cadence
NFL rankings update weekly during the regular season and offseason news cycles, and at major milestones (combine, free agency, draft, training camp opens, week 1 inactives, trade deadline). MLB rankings update weekly during the season and monthly in the offseason. The last updated date on each rankings page is the source of truth; the timestamp in the database is what the page reads.
What we do not do
- We do not rank by name recognition. A 30-year-old running back coming off a highlight season is still a 30-year-old running back.
- We do not run two separate rankings boards (one cautious, one bold). The board you see is the board we would draft from.
- We do not bump players for storytelling reasons. If a player has had a good week and the situation has not changed, the rank does not change.
Corrections
If you see a player on the wrong team, the wrong age, or the wrong tier given the analysis, write us at hello@cappenalties.com. We log corrections and push them out with the next update cycle. Larger methodology changes are documented on this page and announced in the newsletter before they take effect.