Why a 70% win rate can lose money
"70% winners!" sounds like a bettor who is making money. Maybe, maybe not. A win rate tells you how often bets won, not what they cost or what they paid. Without the prices, it says very little about profit.
The price sets the bar
Every price has a break-even rate, as Reading odds showed. A −300 favorite risks $300 to win $100, so it needs 300 ÷ 400 = 75% just to break even. Win 70% of those bets and you lose money: each loss costs as much as three wins, and 70% isn't enough wins to cover the losses.
Turn it around and a record that looks poor can still profit. At +150 the break-even is 100 ÷ 250 = 40%. Win 42% and you are ahead, even though you lost most of your bets.
Same 70%, different prices
A 70% win rate is a good record at −200, where the break-even is 200 ÷ 300 = 66.7%. Each $100 bet there wins $50, so the EV is 0.70 × $50 − 0.30 × $100 = +$5 per $100 (see Expected value). At −300, the same 70% is a losing record. The record didn't change; the price did.
Why short favorites tempt people
Favorites win often, and winning often feels good. That makes high win rates easy to sell. Someone who bets mostly short favorites can post a 70% record and still lose money, and the record alone won't show it. Two numbers tell the real story:
- ROI: profit divided by the money bet. It already accounts for the prices.
- Units won or lost, over enough bets to mean something.
Short favorites carry a second cost. At −300 one loss erases three wins, so an ordinary bad week does more damage than the win rate suggests.
What this means on Consensus Edge
Lists of the likeliest sides, our Best Bets included, show how often something should hit. Likely is not the same as good value. On Consensus Edge only a graded call, BET or LEAN, is a recommendation; a likely side without one is there for context, and a short favorite is the classic case of likely but overpriced.
When you judge any record, ours included, ask for the prices and the ROI, not just the win rate. And judge it over hundreds of bets, not a few dozen: a short run says very little either way (see variance).
Example. Two bettors each make 100 bets (illustrative, not live data). One bets favorites at −300, risking $300 to win $100, and wins 70. The other bets underdogs at +150, risking $100 to win $150, and wins 42.
| 70% at −300 | 42% at +150 | |
|---|---|---|
| Break-even | 75% | 40% |
| Wins | 70 × $100 = +$7,000 | 42 × $150 = +$6,300 |
| Losses | 30 × $300 = −$9,000 | 58 × $100 = −$5,800 |
| Result | −$2,000 | +$500 |
| Money bet | $30,000 | $10,000 |
| ROI | −$2,000 ÷ $30,000 = −6.7% | $500 ÷ $10,000 = +5.0% |
The 70% bettor won 28 more bets and lost $2,000. The 42% bettor lost most of the time and finished $500 ahead. The win rate pointed the wrong way both times; the price explained the result.
Terms in this lesson
Win RateBreak-evenFavoriteROIBest BetsGraded CallVarianceImplied Probability