The most overlooked skill in sports betting
It's not bad picks - it's bad bet sizing. You can be a winning handicapper and still lose money if you size bets poorly. Conversely, proper sizing protects you during losing streaks and maximizes gains during hot runs.
A unit is your standard bet size, expressed as a percentage of your total bankroll. It provides a consistent framework for sizing all your bets.
If your bankroll is
and 1 unit = 2%
Then your unit size is
per standard bet
| Bankroll | 1% Unit | 2% Unit | 3% Unit |
|---|---|---|---|
| $500 | $5 | $10 | $15 |
| $1,000 | $10 | $20 | $30 |
| $2,500 | $25 | $50 | $75 |
| $5,000 | $50 | $100 | $150 |
| $10,000 | $100 | $200 | $300 |
Bet the same amount (1 unit) on every wager, regardless of confidence level.
Vary bet size based on your confidence level, using a 1-5 unit scale.
| Rating | Units | % of Bankroll | When to Use |
|---|---|---|---|
| 1 Star | 1 unit | 1-2% | Standard play, slight edge |
| 2 Star | 1.5 units | 2-3% | Above average confidence |
| 3 Star | 2 units | 3-4% | Strong edge, high confidence |
| 4 Star | 2.5 units | 4-5% | Rare - very strong edge |
| 5 Star | 3 units | 5% MAX | Extremely rare - max play |
A mathematical formula that calculates optimal bet size based on your edge and the odds offered.
f* = (bp - q) / b
You believe the Chiefs have a 58% chance to cover -3, and the odds are -110 (decimal 1.91):
Full Kelly suggests betting 11.8% of your bankroll - but that's aggressive!
Most bettors use a fraction of the Kelly recommendation to account for uncertainty in their probability estimates:
Maximum growth but extreme variance. Not recommended.
Recommended. ~75% of optimal growth, much lower variance.
Conservative. For uncertain edges or risk-averse bettors.
Many successful bettors recalculate unit size weekly or monthly. Don't adjust mid-week based on short-term results - stick to your schedule.
Risking 10%+ per bet means a short losing streak can wipe you out. Even 5 losses in a row at 10% each leaves you down 40%.
Doubling after losses to "get even" is a recipe for disaster. This is the fastest way to go broke.
Betting $50 one game and $500 the next based on "feel" removes the benefits of proper bankroll management.
If you don't know your total bankroll, you can't calculate proper unit sizes. Define it first.
Money you can afford to lose entirely
Conservative starting point
1 unit per bet until profitable