Glossary

Every abbreviation and tier label we print, in plain English. If a term on the site is underlined with dots, you can tap it to see the short version without leaving the page.

Reading a number

pp: Percentage points — the plain gap between two percentages, not a percent change.If our model says 56% and the market says 40%, the gap is 16 percentage points (16pp). It is NOT '16% more likely' — it is 16 points of probability. We write pp instead of % for gaps so a gap can never be mistaken for a probability.
edge: Our win probability minus the market's, in percentage points. Positive means we rate the side higher than the book does.Edge = LINEHAWK win probability − market-implied win probability, expressed in percentage points (pp). A +4pp edge means we make the side 4 points more likely than the price implies. Edge is a measure of disagreement with the market, not a measure of how likely you are to win. Small edges (roughly under 3pp) sit inside our own error bars and should not be treated as signal.
Conf: Confidence — the LINEHAWK model's win probability for the side we picked.Short for confidence. It is the same number we elsewhere label 'LINEHAWK win probability'. 67% confidence means the model expects that side to win 67 times out of 100 — it does not mean we are 67% sure the pick is good.

Pick tiers

LOCK: Our highest-conviction tier: both model brains agree, the edge is large, and the price still holds value.LOCK is a tier label for conviction, never a guarantee. It means both brains landed on the same side, the edge cleared our top threshold, and market movement has not yet erased the value. Locks lose. Past tier performance is on the Track Record page.
TARGET: A solid edge worth watching — the brains lean one way, but it is not a slam dunk.
FADE: The model rates this side LOWER than the market does. We are calling it overpriced, not recommending it.A FADE is a warning label, not a pick to bet the other way. It means the price is worse than our estimate of the true chance.
No play: We ran the analysis and it did not clear our publish bar. There is no bet here.Every game gets an analysis and a model call. 'No play' means that call did not clear the threshold we require before we are willing to describe it as a bet. It is an honest abstention, not a missing pick.
CHALK: The heavy favorite — the side the public and the price both expect to win.
OVERPRICED: The market's price implies a higher chance of winning than our model gives it.

Market signals

MARKET AHEAD: The book has already moved past our number. Any value we saw is gone, so we are not calling it a bet.MARKET AHEAD means the market-implied probability now exceeds our own. We still show the analysis for transparency, but a MARKET AHEAD card is explicitly not a recommendation.
MARKET ECHO: Our model landed on essentially the same number as the market — we are agreeing, not finding an edge.
Line +5.2 24h: How far the price has moved in the last 24 hours, in percentage points of implied probability.'Line +5.2 24h' means the market-implied probability for that side rose 5.2 percentage points over the past 24 hours. Rising = money coming in on that side.
Buzz: Public-chatter volume for this matchup: Quiet, Normal, or Loud. It measures attention, not quality.Buzz reads volume of public discussion. 'Buzz Quiet' means almost nobody is talking about this game — which is neither good nor bad on its own, but it does mean less public money shaping the price.
Sharp edge severity: How far Pinnacle's price sits from the US books, in percentage points: MEDIUM 5–8pp, HIGH 8–15pp, CRITICAL 15pp+.Pinnacle is the reference sharp book. When it disagrees with DraftKings/FanDuel/BetMGM/Caesars, we grade the size of the disagreement in percentage points of implied probability — MEDIUM 5–8pp, HIGH 8–15pp, CRITICAL 15pp and up. These are probability points, not points on a spread.
Weather: Forecast conditions at the venue at kickoff. Hidden entirely when we have no forecast — we never show a placeholder.

Bet types

ML: Moneyline — a straight bet on who wins, with no points involved.
Spread: A handicap in points. The favorite must win by more than the number; the underdog can lose by less than it.
RL: Run line — baseball's version of a spread, almost always set at 1.5 runs.
O/U: Over/Under — a bet on the TOTAL, not on who wins. The unit depends on the sport.O/U is the combined total for the event and the unit changes by sport: points in basketball and football, runs in baseball, goals in soccer, and ROUNDS on a UFC or boxing card. 'O/U 1.5' on a fight means rounds; the same number on a soccer match means goals. We label the unit on every card for this reason.
K-of-N ladder: A round-robin: N legs split into every combination of K, so hitting only some of them still pays.A 2-of-4 ladder takes 4 picks and builds all 6 possible 2-leg parlays from them. You do not need all four to win — each pair that lands pays on its own. It trades top-end payout for a much higher chance of returning something.

Track record

n=: Sample size — how many graded picks a number is based on.n=545 means the figure next to it was computed from 545 settled picks. A big n is more trustworthy than a small one. Anything under about 30 graded picks is labelled PROVISIONAL SAMPLE and should be read as a hint, not a record.
u (unit): One unit = one flat, equal-sized stake. We stake exactly 1 unit per pick, so results are size-independent.Every result on the site assumes a flat 1 unit on every published pick, win or lose. '+26.00u' means the strategy is 26 flat stakes ahead of even. Your unit is whatever you choose it to be — that is the point of measuring in units rather than dollars.
CLV: Closing line value — whether the price improved or worsened between when we published and when the market closed.CLV compares our published price to the final closing price, in percentage points. Positive CLV means the market moved toward our side after we called it, which is the standard evidence that a pick was priced well. Negative CLV means the market moved against us. It is measured on price movement, not on whether the bet won.
ROI: Return on investment — units won or lost divided by units staked, at the prices we published.ROI is the honest scoreboard: a win rate can look good and still lose money if the winners are all heavy favourites. We settle every priced pick at the price we published for the side we picked, vig included, staking a flat 1 unit each time. Picks we never recorded a price for are counted in the sample but excluded from the money figures — they are never estimated.
BEAT THE CLOSE: The share of our picks that got a better price than the market's closing price.
PROVISIONAL SAMPLE: Too few graded picks to be a record yet. Read it as an early indication, not a track record.We label any figure computed from a small number of settled picks as PROVISIONAL rather than hiding it. Small samples swing wildly; one week does not make a record.

Missing a term? Email support@linehawk.bet and we will add it. LINEHAWK is analysis only — not betting advice. 21+.