Glossary

What the numbers on the board mean, and roughly how each one is built. Organized the way the app is.

Rankings

The board everything else is built from.

Ozzie Score: A single 0–100 grade for how good a player is, all things considered. It starts from VOR, rescales it 0–100 across the whole pool, then bends the scale so the gap between good players shows up more than the gap between the scrubs at the bottom.
VOR — Value Over Replacement: Points above the player who'd be available for free at that position if this one were gone — the actual replacement level for your league's size and scoring, not a number from a chart. Everything downstream (Ozzie Score, trade value, waiver priority) is built on this rather than on raw projected points, because a running back's 200th point is not worth what a kicker's 200th point is worth.
Tier: A group of players with no real gap between them — "any of these four" rather than a strict rank. A new tier starts wherever the drop to the next player is unusually large compared to the drops around it.
Floor / Ceiling: A plausible bad week and a plausible great one — not the literal worst or best case. Computed from the player's own projection and volatility, so a boom-or-bust receiver shows a wider gap than a steady running back with the same average.
ADP: Average Draft Position — where the market actually drafts a player, shown for reference. It isn't a Stiff Arm number; it's what makes a ranking that disagrees with it worth noticing.

Research

The inputs behind a recommendation, not the recommendation itself.

Implied total: How many points Vegas's own game lines expect an offense to score. It has nothing to do with fantasy scoring directly — it's a proxy for how many opportunities that offense will hand out, which is why a game with two high implied totals is where the fantasy points live.
Lean bar: Who is favored in a game and by how much, built from the gap between the two teams' implied totals. The bar leans toward the favorite; it says nothing on its own about the total.
Defense rating vs. [position]: 1.00 is league average. A defense at 1.20 gives up 20% more than average to that position; a defense at 0.85 gives up 15% less. It is a multiplier on production, not a raw ranking.
Boom rate / Bust rate: The chance a player clears a big spike-week outcome (boom) or falls below a floor bad enough to cost you the week (bust), given that player's own projected mean and volatility for the week.
Confidence: How sure the projection itself is — driven by role stability and how much usage history there is to project from. It is not a chance of winning, and a low-confidence player can still have a high projection.
Teams / Players tabs: Two questions on one page. Teams is where the points are this week: game environments and the defenses at each position. Players is what each player does with them: his projected stat line, the chance it clears a line, his previous weeks and Ozzie's read.
Projected stat line: The same line the engine scores for fantasy points, shown stat by stat: carries and rushing yards for a back, targets, catches and receiving yards for a receiver, passing yards and touchdowns for a quarterback. It is the player's season pace multiplied by the defense rating, the game environment, the expected script and his usage — the drivers row on the card lists each one.
Pace: What the player has been doing per game this season, before this week's matchup is applied. The small number beside each projected stat. When the projection sits above pace, the matchup and the game environment are lifting him; below it, they are holding him back.
Reference line: The number a stat is measured against, set to the player's own season pace rounded down to the half — 60.5 rushing yards for a back averaging 61 — or "at least one" for touchdowns. It is not a market line and nothing about it comes from a sportsbook. Tap − / + to ask about a different number; yards move by five, counts by one.
Chance to beat: The probability the stat lands above the reference line, from a spread around the projected value. Yards use a bell curve whose width depends on the position and the player's role stability; catches and carries the same with whole-number rounding, so beating 4.5 means five or more; touchdowns use a Poisson rate, so 0.7 expected touchdowns is about a 50% chance of at least one. Tapping the line recomputes it with the same function the server used.
Game script: Whether the offense is expected to run or throw more than usual, from the gap between the two implied totals: a team favored by a lot runs out the clock, a team behind throws to catch up. Run-leaning lifts a back's carries; pass-leaning lifts targets.
Role stability: How settled a player's share of the work is — a bell cow near 100%, a committee back or a rotating receiver much lower. It widens or narrows the spread behind Chance to beat and feeds Confidence, and it says nothing about how good the player is.
Previous weeks: The fantasy points this league recorded for the player each finished week, dimmed for weeks he sat on a bench. That is all the weekly capture stores — no yardage, no targets — so this is a row of scores rather than a box score, and a player nobody rostered has no rows at all.

Trade

What a deal is actually worth, and to whom.

Trade value: A player's worth in a deal — VOR scaled so the best player in a trade counts disproportionately more than the sum of raw points would suggest, since a roster has one starting slot to fill with him, not several.
Redraft value vs. Dynasty value: Redraft is worth this season, full stop. Dynasty layers in age: a young player near his positional peak is worth more, an older one less, because a dynasty trade is buying seasons that haven't happened yet.
Market premium: A position-specific multiplier (quarterbacks worth more per point of VOR than the math alone would say, kickers and defenses worth less) that nudges pure value toward what leagues actually pay in real trades.
FAIR / A WINS / A WINS BIG / B WINS / B WINS BIG: The verdict on a proposed trade, from the percentage gap between what each side gives up: under roughly 7% is fair, 7–22% is a real but survivable edge, and over 22% is a blowout.
Rest-of-season impact vs. Lineup impact: Lineup impact is this week's starting-lineup change. Rest-of-season impact is the same thing averaged across every remaining week of the horizon below — the number to weigh a trade against if it's not about this Sunday.
Regular-season horizon: How far ahead a trade is judged: through the last week before your league's playoffs start, not the whole season. A trade is mutual benefit over the games both sides are guaranteed to play, so a week neither side will play for a spot doesn't belong in the math.
Review a trade: Uphold / Veto: A recommendation on a deal between any two teams, for a commissioner or a league that votes on trades. It upholds by default, because managers are allowed to make bad trades. It recommends a veto only when one side gives up far more value and gets neither of the usual reasons for taking less: a starting lineup that improves, or the best player in the deal in exchange for more bodies. Standings count too: a team out of the playoff race sending value to a contender is the pattern vetoes exist for, and a lineup that improves is no excuse for a team with nothing left to play for.
Waiting on the league: Trades an ESPN sync saw in the league's review period, listed at the top of Review a trade with a recommendation on each. In review means both managers accepted and the league has not decided; an offer has not been accepted yet. ESPN may only show trades your own account can see, and Sleeper and Yahoo do not publish other teams' pending trades, so those still go in by hand. You can be told when a new one arrives: Settings, Trades awaiting review.
Value gap: How far apart the two sides of a trade are in trade value, as a percentage of the average package. Under 25% reads as even. An unexplained gap past 60% leans veto, and past 100% is a clear veto. A team that also weakens its own lineup, or is out of the playoff race while sending value to a contender, reaches veto territory from 40%.
Clear call / Close call: How confident a trade review is. A close call is one where reasonable commissioners could land either way, such as a lopsided deal with no good-faith explanation that still falls short of a fleece. It judges the deal only; evidence that two managers are working together outweighs it.

Draft

Live, mock, or graded after the fact.

VONA — Value Over Next Available: Not "who is the best player left" — how much value you'd lose at a position if you wait one more turn to take it. A thin position with nobody good left past the next few picks can outrank a deeper one even with a lower raw projection.
Survival probability: The odds a specific player is still on the board when your next pick comes around, estimated from how far off his ADP the draft has already gone.
Need weight: How much the recommender leans toward filling an empty starting slot versus simply taking the best player left. Turned up, it drafts a mediocre starter over a great bench piece; turned down, the reverse.
Snake / Auction / Third-round reversal: Snake alternates the pick order every round; auction has no fixed order at all, just a shared budget bid on every player. Third-round reversal keeps round 3 from flipping, so nobody gets back-to-back picks across rounds 2 and 3 the way an ordinary snake would give the last pick of round 1.
Keeper: A player and price carried over from last season rather than drafted this one. Kept out of "best value" and "biggest reach" grading, since nobody actually paid this year's price for him.

Waivers

Who to add, who to drop, and what it costs.

FAAB: Free Agent Acquisition Budget — a season-long dollar pool you bid from on the waiver wire, separate from your roster.
Priority (0–100): How urgent a pickup is, in absolute terms rather than a rank against the other free agents — built from the lineup gain, the rest-of-season gain, and the player's Ozzie Score together.
MUST_ADD / PRIORITY / STREAMER / SPECULATIVE / DEEP_STASH: The categories a waiver target falls into, from "starts for you this week" down to "worth a bench spot on a hunch and nothing more."
Rolling priority vs. Reverse-standings priority: Rolling: a numbered queue where using your claim sends you to the back of it, so it's worth holding unless this claim beats whatever comes later. Reverse-standings: priority resets every week by the standings, so using it costs nothing next week.

Lineup & Start/Sit

Who plays this week.

Win probability: The modeled chance a player outscores every option he's being compared against — built from each player's whole range of outcomes, not just the average, so a boom-or-bust player and a steady one with the same projection can carry different odds.
Edge: The point gap at the actual cut line between starting a player and sitting him — how close the call really is, not just who is ranked higher.

Season & Odds

The whole year, not one week of it.

Playoff odds: The share of thousands of simulated season endings where a team makes the playoff cutoff, run against each team's actual remaining schedule. Withheld rather than guessed whenever the schedule, the cutoff, or the number of divisions isn't fully known — a blank here means the league hasn't synced enough to trust the number, not that the odds are zero.
First-place odds: The same simulation, counting only the endings where a team finishes in first rather than merely inside the cutoff.
Rest-of-season horizon vs. Regular-season horizon: Two different endpoints for two different questions. Anything about a player (rankings, the wire, this glossary's Trade section aside) runs through the league's actual last week, championship included. Anything about standing — the table, the odds, a trade — stops the week before the playoffs start, because that's the last week every team is guaranteed to play.

Ozzie

The assistant, when you ask it directly.

It's the same numbers: Ozzie Score, VOR, VONA, floor and ceiling, matchup ratings, boom and bust — Ozzie answers in plain sentences, but every number it gives you is one of the ones on this page, computed the same way. Nothing it says is a separate opinion.
Ozzie's read: The short write-up behind the Ask Ozzie button on a player card in Research. Ozzie is handed the projected line, the pace, each chance to beat, the drivers and the previous weeks already computed, and asked to explain which inputs are doing the work and what would make the line miss. He does not produce a number of his own. On a trade review it is a short post for the league chat that explains the uphold or veto from the same numbers, never changing the call. An Elite feature, like the chat.