Pick of the Day rules

A pick day is a kickoff day in US Eastern time, midnight to midnight. The archive, the dated permalinks and the email all key off that date. The selector starts at 5 AM Eastern and rescans every fifteen minutes through the day. A day carries up to six picks, one per game. When no game qualifies there is no pick, and a skipped day never enters the record.

Each pick has its own release time: one hour before its own game, inside an 11:00 to 23:00 UTC window. The page fills the slot at the drop, and the times on the page show in your device's own timezone.

Every pick starts with wallets graded S, A or B on their settled Polymarket record. The grade comes from resolved positions and realized profit. Follower counts and self-reported results play no part.

For each game, the selector reads how much of that money sits on each side and backs the side that holds more of it. Recent buying on that side counts as agreement, and a game whose graded flow runs against the side is dropped. Every holding behind a pick is confirmed against Polymarket before the price freezes.

A pick can also carry a specialist: a graded wallet with a measured, positive record in that sport across enough independent games. No sport requires one. A pick with a specialist outranks a pick without one, and the specialist's position is captured at selection and checked again at release.

How a wallet earns its grade: Smart Score.

A game clears a fixed set of checks, in a fixed order, on every scan. The checks are the same for every sport: basketball, football, baseball, hockey, soccer, tennis, cricket, MMA, boxing and esports, on moneylines, spreads, totals and props. A sport without a measured cohort of graded wallets is held out until it has one.

The market must be a real game with a known start time. It must carry enough liquidity to fill, at a price clear of both a coin flip and a near-certainty. Enough profitable wallets must hold one side, and the split between the sides cannot be near-even. A side the market already prices in line with those wallets is skipped. A market the record carries from the last few days waits out a cooling period.

The floors behind each check, the weights, and the tie-break order are the selector's and are not published. The output is: the side, the frozen price, the wallets on it, and the result.

Until release, the selector can swap a standing pick for a clearly stronger one. A locked slot on the page shows its release time and kickoff only; the market, the side and the price are not sent to the page until the drop.

At release the side and the price go to the page, the API, the Pro email and the bell at the same instant. The price on the card is the market price at that moment. It is written once and never rewritten. The market keeps moving after that, so your own entry can differ from the price on the card, and your fill and Polymarket's taker fee come out of any edge.

The receipt trail sits outside our servers. The Pro email goes out at release with the matchup in its subject, so its timestamp is on a mail server we do not run. Our Discord posts the date and the slot number at the same time; the side stays on the page.

Where the capture succeeds, the archive records the market's last valid price before kickoff, or before Polymarket's live feed first reported the game in progress, beside the frozen price. Closing line value is the move between the two. A positive CLV means our side's price rose between publish and that close. It is the standard test of whether a call had edge before the result was known.

Each measured pick shows its CLV and the two prices behind it. A pick whose close could not be captured shows CLV as unavailable, and a pick published after kickoff is marked not applicable rather than scored. The stats page shows the average and how many settled picks it covers.

The market decides. We read the outcome Polymarket resolved the market to and grade the pick against the side we backed. A cancelled market is a void. A postponed game stays pending until Polymarket resolves the market, however long that takes. We never guess a result.

A job re-reads every published pick every thirty minutes and posts the result. A pick whose market was corrected to a different outcome follows that correction, so a pick can never disagree with the market's settled result.

Win rate is wins divided by decided picks, where decided means wins plus losses. Voids and picks still pending count in neither half, so an open pick does not move the rate before it settles.

Profit, ROI and units assume a flat $100 stake on every settled pick that carries a frozen price, filled at that price, before Polymarket's fees. Profit is the sum of the returns minus what was staked. ROI is profit divided by the total risked, and the page shows the total risked beside it. One unit is one $100 stake.

Today's games, their sports and each release time are public. Pick 1 is full for any signed-in account at release: the side, the frozen price and the wallets on it, the same card Pro sees. The day's other picks at release, the email and the bell, and the API and MCP reads need Pro. Every settled pick, with the side we backed and the result, is public.

• A published pick is never deleted, re-pointed at another market, or re-priced.
• Every loss and every void stays in the archive, in the same table as the wins.
• A skipped day never enters the record, and neither does a pick that never published.
• The result of every pick is read from the market's own resolution by a scheduled job, never entered by hand.
• No pick is sold on its own and no pick carries a fee on a win. Pick of the Day is one part of a Pro subscription, cancelled at any time.
• Every day with a pick has its own page at /pick-of-the-day/YYYY-MM-DD, with the frozen price and the result.

The selector's liquidity floor, its price band, the wallet counts and share floors a side must clear, the margin a rescan needs to swap a pick, the weights, and the rank order. Those are not on this page, and this page will not change that. What is published in full is the output: every pick, its frozen price, the wallets on it, its close where captured, and its result.

Pick of the Day shows where profitable wallets are positioned. A position is not a guarantee. Prediction markets carry risk. Never stake more than you can afford to lose.