An odds convertor translates any betting price into its implied probability — the exact win rate you need to break even. American -150 implies 60%. Decimal 2.50 implies 40%. A spread at -110 implies 52.38% on either side. The format changes. The math does not.
I once bet a -54.5 college favorite at -110 because the converter said 52.38%. The team won by 53. Never again without checking dispersion (how much books disagree).
How does an odds convertor work?
An odds convertor turns any betting price — American, decimal, fractional, or a point spread with juice — into its implied probability, the break-even win rate baked into the price.
What the live market shows right now
On the August 22 slate, our single-book lines capture NFL home spreads from -1.5 (Commanders at Lions) to -5.5 (Chiefs at Buccaneers) . College football stretches wider: Hawaii at Stanford sits at -5.5 , New Mexico State at Florida State sits at -31 , and Arkansas-Pine Bluff at Missouri sits at -54.5 . Every line carries a price — usually -110 — and that price is what the convertor turns into a probability.
Why dispersion matters more than the line itself
A line is a single number. Dispersion tells you whether the market agrees on it. Every game in our data shows zero dispersion — only one book posted each line. Zero disagreement means the probability is clean. It also means you cannot line-shop (compare prices across books) for free EV (expected value = long-run profit). Our NFL picks feed and college football picks feed surface dispersion on every game so you know when the number is trustworthy.
The formula you can run in your head
You do not need a tool for the arithmetic. You need the tool for the live data.
- Negative American odds (-N): N / (N + 100). Example: -110 → 110 / 210 = 52.38%.
- Positive American odds (+N): 100 / (N + 100). Example: +150 → 100 / 250 = 40%.
- Decimal odds: 1 / decimal. Example: 1.91 → 1 / 1.91 = 52.36%.
The spread is a handicap. The juice is the probability. A -3.5 spread at -110 and a -3.5 spread at -120 are different bets — the first implies 52.38%, the second implies 54.55%. That 2.17 percentage point gap is the difference between a sharp price (pro book) and a recreational one (casual book). Line shopping is free EV. Our odds reading guide walks through the full conversion table.
No-vig fair probability: the number the market actually believes
Add the two implied probabilities from a -110/-110 market and you get 104.76%. The extra 4.76% is the vig (the house cut). Strip it out by dividing each side by the sum: 52.38 / 104.76 = 50% fair. That is the market's true estimate before the house tax. On a lopsided college game like Arkansas-Pine Bluff at Missouri (-54.5) , the spread juice is still -110, so the no-vig fair probability on the spread is 50/50. The convertor that skips this step overstates the favorite's chances by the full vig margin.
How to use this on a real slate
- Open the picks page, pick a game, and compare your win % to the fair probability shown there. If your number beats it by 2%, bet. If not, pass.
This workflow is exactly what our model builder in /tinker automates: you input your probability, it pulls the live consensus, computes closing-line value (CLV = did you beat the final line?), and tells you if the bet clears the vig. The convertor is step one. The builder is step two through infinity.
Where the convertor falls short
A standalone convertor cannot tell you:
- Whether the juice is -110 or -115 at the sharp book versus the recreational book you are logged into.
- Whether the line has moved since you opened the page (CLV requires a timestamped close).
- Whether the market type (spread, total, moneyline, prop) carries a different hold structure (how the house sets its cut per bet type) that changes the break-even threshold.
Our CLV explainer shows why the closing line is the only benchmark that matters. A convertor gives you the price at one moment. CLV tells you whether that price was better than the final consensus.
What would change our mind
If a single game shows dispersion above 3 points, I stop trusting any convertor's single number — I wait for the line to settle or I pass. Clean consensus is the only time the math lies flat.
Bet responsibly — set limits, never chase losses.
Expected bankroll growth at 55% edge
Expected geometric growth of a $100 bankroll under different Kelly multipliers across 1000 bets at p=0.55, decimal=2. Full Kelly maximises long-run growth but produces the deepest drawdowns; fractional Kelly trades growth for variance.
EV per $100 across win rate × odds grid
Expected value of a $100 stake at each combination of true win rate and market odds. Anywhere the cell is positive you have a long-run profitable bet; the magnitude shows how aggressive Kelly will size it.
