Updated Sep 7, 2026 · Week 1 board.
Feeding one price at a time
Bovada has Falcons at Steelers, Sunday, Sep 13, 1:00 PM ET, at -185 on the Steelers and +160 on the Falcons. Same board, same kickoff, Jets at Titans: -125 on the Titans, +105 on the Jets. Hand any one of those four numbers to a converter and an implied chance comes back. It doesn't care which side you gave it, or whether you ever bothered with the other one.
Running the formula on each side
Negative price: divide its size by itself plus 100. The Steelers' -185 gives 185 / (185 + 100), which is 64.91%. Positive price: take 100 and divide by the price plus 100. The Falcons' +160 gives 100 / (160 + 100), or 38.46%.
Same two formulas on the other game. The Titans' -125 gives 125 / (125 + 100) = 55.56%. The Jets' +105 gives 100 / (105 + 100) = 48.78%.
Read -185 as an instruction about money instead of a chance and you land in the same place. Risk $185 on the Steelers to win $100. Risk $100 on the Falcons at +160 to win $160. The percentage and the stake are two descriptions of one price.
A converter that hands you the first one hasn't lied to you. It answered the question you asked it.
| Game | Side | Bovada price | Implied probability |
|---|---|---|---|
| Atlanta at Pittsburgh | Steelers | -185 | 64.91% |
| Atlanta at Pittsburgh | Falcons | +160 | 38.46% |
| New York Jets at Tennessee | Titans | -125 | 55.56% |
| New York Jets at Tennessee | Jets | +105 | 48.78% |
Add the two sides together
Now add them up. Steelers 64.91% plus Falcons 38.46% is 103.37%. Not 100%. Titans 55.56% plus Jets 48.78% is 104.34%.
Neither pair lands where two outcomes that cannot both happen are supposed to land, and that isn't rounding. It happens on every two-sided market on the board, because both raw numbers are still carrying whatever the book built into the price. The conversion was right. The output just isn't a chance yet, and a chance needs two outcomes that add to 100%.
That overflow is the real risk in the whole exercise. The arithmetic is exact. The input is a live market price. The output still isn't the fair chance of either team winning, not until both sides of that same market get converted and set against each other.
That second step is one I'm skipping here deliberately, so you can watch what the first step produces on its own.
It turns up on every two-sided moneyline on the Week 1 board, not only on these two games. A one-side conversion is exact arithmetic on a real price. The error is treating one side's raw output as though it had already answered the two-sided question. Nothing goes wrong inside the formula.
It goes wrong when you ask a converted number something its single input was never in a position to answer.
What this deliberately leaves out
The odds converter and our live board carry the same raw prices used here. What happens once you do run the both-sides step is in a separate explainer on vig and hold. And the full odds hub lists every Week 1 game, should you want to try this one-side arithmetic on a price of your own choosing.
This is arithmetic on two live Bovada prices and nothing more. Sunday hasn't happened yet, so nothing above has a real-world record standing behind it. The four numbers in the table are what Bovada's board shows and what the two formulas produce out of it. No second book went into any of this, no spread, no total.
And here is the habit I would build off it. Any time a page hands you a one-side percentage and calls it a chance, go find what the other side of that same market converts to. If nobody can tell you, then nobody has a chance to hand you, only a price with a percent sign on the end of it.
The Receipts Drawer
64.91% and 38.46%, adding to 103.37%, off two Bovada moneylines and nothing else. No college line, no MLB line, no prop, no PrizePicks number anywhere near it. That sum over 100% is the normal state of the world rather than a glitch, because the book's own take is living inside both halves of it. Let Bovada nudge either price and that sum lands somewhere else.
So next time you open that board before kickoff, run both sides through the same two formulas and see how far past 100% the pair lands that day.
