PuckCast Edge
Every pick is free. The Edge is the rest.
Betting edges, stake sizing, props priced against the line, fantasy tiers and the graded ledger. $14.99 a month in season, or $99 for the year.
PuckCast Edge
Betting edges, stake sizing, props priced against the line, fantasy tiers and the graded ledger. $14.99 a month in season, or $99 for the year.
PuckCast Edge
Every Edge play this season, recorded before puck drop and graded at the price when captured. Updated Nov 3, 12:52 AM.
Running balance, units. Recommended stakes · flat 1 unit
123-111 over 234 plays. Recommended stakes: +4.23u on 18.70u risked (+22.6%), worst drawdown -1.78u. Flat 1 unit: -3.91u (-1.7%), worst drawdown -18.24u. An early sample. Past results do not guarantee future ones. Every play, with prices and timestamps.
| Month | Plays | W-L | Recommended stakes | Flat 1u |
|---|---|---|---|---|
| September 2026 | 7 | 4-3 | +1.81u (+67.1%) | +2.65u (+37.8%) |
| October 2026 | 218 | 112-106 | +2.26u (+14.2%) | -11.02u (-5.1%) |
| November 2026 | 9 | 7-2 | +0.16u on 0.10u: too little staked to rate | +4.46u (+49.6%) |
Common Questions
Every PuckCast Edge play this season. Each one is recorded from the board before puck drop, with the best price available at that moment across the major US sportsbooks, and graded at that price after the game. Nothing is repriced afterwards and nothing from backtests is mixed in.
The stake is half-Kelly on the gap between our win probability (the model blended with the market) and the price, sized for a 20-unit bankroll and then cut by confidence, quote freshness and market-efficiency discounts. 3 units for a strong bet, 2 for a bet and 1 for a lean are ceilings, not typical stakes: this season most stakes have been a fraction of one unit. A small gap gets a small stake, and a play whose price has gone stale is a no bet at 0 units.
Flat stakes bet 1 unit on every play, so they show the raw hit rate against the prices without our sizing. Recommended stakes show what following our sizing did. Both are always shown together.
The largest fall from a high point in the running balance: the worst losing stretch you would have sat through. It is shown for both stakes.
It priced the model's most confident straight-up picks at an assumed -110, because the archive it read had no odds. Those were mostly favourites, which cost more than -110, so it overstated the return, and they were not Edge plays. We retired it in October 2026 and replaced it with the live Edge record.