What is the break-even point at -110 odds?
At -110 odds, the break-even win rate is 52.38%. Win less than that over a large enough sample and the math is bleeding, even if the picks sound sharp.
That is why a 50% true coin flip at -110 is negative expected value. The book did not build marble counters for charity.
How much edge clears standard juice?
In a -110/-110 two-way market, each side implies about 52.4%, while the no-vig fair probability is 50%. The per-side margin is roughly 2.4 percentage points.
Your estimate has to beat the no-vig fair probability by more than that friction to be meaningfully +expected value at the posted price.
Do lower-juice markets change the edge needed?
Yes. Reduced-vig markets lower the break-even bar because the sportsbook is taking less margin from the price.
That does not make every reduced-juice line good. It just means your model has less tax to overcome before the bet can show positive expected value.
How should edge connect to bet size?
Once the edge clears the vig, sizing still matters. The Kelly Criterion converts edge and odds into a bankroll fraction, while fractional Kelly can tone down volatility.
No edge, no bet. Thin edge, small bet. Big claimed edge with no evidence, sunglasses stay on and the stake stays small.
How large must your edge be before the vig stops eating the bet?
The edge has to exceed the sportsbook's margin before a bet is positive expected value. At the familiar -110/-110 spread price, each side has a posted break-even win rate of about 52.38%. A bettor who wins exactly 50% at -110 loses money because the payout is smaller than the risk. That extra 2.38 percentage points is the visible cost of standard juice on one side of a balanced two-way market. No-vig math makes the hurdle clearer. In a -110/-110 market, both sides imply roughly 52.4%, and the market totals about 104.8%. After normalizing to 100%, each side is 50%. If a model estimates 51%, it is above the fair market probability but still a very thin advantage. If the estimate is 53% or 54%, the edge is more meaningful, assuming the model is calibrated and the line is still available. Lower-juice markets reduce the hurdle. A market priced closer to -105/-105 has less hold, so a smaller advantage can become worthwhile. Higher-vig markets, including many props, demand a larger edge because more of the price is tax. This is why devigging matters before sizing. Without it, a bettor may confuse the posted break-even rate with the fair probability or miss how much hold is embedded in the market. Bet sizing should respond to both edge and uncertainty. Kelly-style math can translate an edge into a bankroll fraction, but full Kelly assumes the probability estimate is correct. In real betting, estimates are noisy, so fractional Kelly is usually more defensible. A two-point edge with wide uncertainty deserves a smaller stake than a five-point edge supported by strong model history and positive CLV. Beating the vig is not about finding bets that can win. It is about finding prices where the expected return remains positive after the market's margin and model error are accounted for.
Which tools and guides support this answer?
Which free desk tools are referenced?
Which guides expand this answer?
What else should bettors know?
Is 52.38% always the break-even win rate?
No. 52.38% is the break-even win rate for -110 odds. Different odds have different break-even probabilities.
Can a 51% model edge be profitable?
It depends on the price. At -110, 51% is not enough. At plus money or lower juice, the same probability may be closer to playable.
Why compare against no-vig fair probability?
No-vig fair probability removes the sportsbook margin, so it shows the market baseline your estimate must beat before staking.