The Edge Letter
How to Read the Desk
The desk uses a small set of terms with precise meanings. Here is each one in plain language — what it means, and why we bother. None of this is a wager or advice; it’s the vocabulary of holding analysis accountable.
The call
Our explicit probability on a market, plus which side we think has value. Every market a desk looks at gets one — most days the honest call is "the market's price is about right," and we say that too.
Why it matters
A view you can grade. Vague leans can't be scored; numbers can.
Falsifier
The result we state in advance that would prove our reasoning wrong — written down before the event, next to the call. If it happens, we don't get to explain it away; the model takes the hit.
Why it matters
Anyone can explain a miss afterward. Naming what would break your thesis beforehand is the difference between analysis and storytelling.
Brier score
The standard measure of how close a stated probability ran to what actually happened. Squared distance between the probability and the outcome — lower is better. A 36.5% call on something that doesn't happen scores 0.133; the market's 24.85¢ on the same event scores 0.062 and wins the round.
Why it matters
Win–loss records are mostly noise at small samples. Brier scores measure whether the numbers themselves are honest.
Devig (devigged price)
A sportsbook's posted odds include the book's built-in margin (the vig). Devigging strips that margin out to reveal the fair probability the book actually believes. A −155 price implies 60.8% — but devigged, the book's true number is closer to 57.8%.
Why it matters
You can't compare our number to a book's number until the margin is removed. And sometimes we beat the fair price while the vig still makes the bet a loser — see the value gate.
Closing-line value (CLV)
How the price you identified compares to the final price just before the event locked. Beating the close consistently is the strongest evidence of real skill in pricing — stronger than short-term win–loss.
Why it matters
Results are variance in the short run. The close is the market's best final answer, and beating it repeatedly isn't luck.
The value gate
A call only becomes a card when our probability beats the price you could actually get — vig, spread, and liquidity included — by a real margin. Our number can be sharper than the market's fair value and the bet can still be bad at the posted price.
Why it matters
The most instructive example on our record: a call we graded correct that we still refused to make a card, because the vig ate the edge. It hit. Passing was still right.
Conditional card
A call armed on a trigger we announce before the information exists — "if the lineups show X" or "if the price reaches Y, then Z is the play." If the trigger fires, it grades like any call. If it never fires, it's archived as a no-fire, not counted either way.
Why it matters
News creates most real edges. Logging the reaction before the news exists proves the reasoning wasn't hindsight.
Smart Flow
Our daily reconstruction of where Polymarket's most profitable wallets are actually positioned, built from public trade data. Read with one law: follow the flow, don't copy it — a wallet's overall rank says nothing about its skill at the specific bet you're watching.
Why it matters
We audited the "smart money" and found top-ranked wallets bleeding on the exact flows a copier would have tailed. Where the money is and whether it's right are different questions — answering the second is the desk's job.
Every term above shows up on The Record, where the calls are graded. 21+ only · Gambling problem? Call or text 1-800-GAMBLER · Analysis for informational and entertainment purposes only — we do not accept, place, or facilitate wagers, and no outcome is guaranteed.