Quick one this week. Two new things on the board, and one honest result.

Skill and Cost

Every pick on the board now gets split into two numbers.

Skill: did the market move your way? We compare the fair price (vig removed) when the pick was posted to the fair price at close. Above zero means the line moved toward the pick.

Cost: the book's cut on the price you took. Usually around 2 points a pick.

To actually make money, skill has to beat cost. Most records only show you one number. That hides the reason good-looking bettors still lose.

Real example from this week: both outside builders took Montreal at +110 and +115. DraftKings closed Montreal at -110, so both beat the close. Montreal lost 5-4. Good read, bad result. That's skill on a losing night.

The honest result

We ran our own rules through it first.

Our best rules do read the line. On average, the market moves their way. But that skill covers about a third of the book's cut. We checked DraftKings, the better price of two US books, and Pinnacle. None of our rules clears once you pay for the bet.

We also ran the Tail Test: grading at the price you could actually get after seeing a pick, not the price it was posted at. Nobody passes yet. Not our rules, not the builders.

That's not the answer we wanted. It's the answer the data gave. The races keep running, and a rule only goes on sale if it clears the bar we published before the races started.

Cappers joined the board

The builders race now has two lanes. Model builders race their models next to our rules, and now our first capper, J Pet Bets, has joined too. Same grading for everyone, every pick logged before the game, losses included.

See the board: https://pixelforgeplays.github.io/pixel-forge/builders.html

NBA rules start racing Tuesday night too.

One question for you

Would you pay $6 a month for either of these?

A) every capper's full receipts, and what tailing them would have cost you

B) a price check before you bet, plus grading on your own bets

Both, one, or neither? Just hit reply. Every answer helps, especially "neither."

Nothing here is a pick. We don't know yet if we have an edge. If we find one, you'll hear it here first.

Casey