There are dozens of bankroll management systems floating around sports betting circles. Most of them share one flaw: they change your bet size based on confidence, streaks, or "hot hands" — which sounds logical but usually accelerates losses.
Flat unit betting does none of that. Same stake every single play. And it turns out to be one of the most defensible approaches you can use.
What a unit is
A unit is just your standard bet size — whatever amount represents one "normal" wager for you.
If your bankroll is $500, a common starting point is 1-2% per play, so $5–$10. If your bankroll is $5,000, a unit might be $50-$100. The number doesn't matter as much as the consistency.
One unit = the same dollar amount every time, regardless of how confident you feel.
Why flat betting works
1. It separates skill from variance
When you vary bet sizes, you introduce a second variable: how well did you size the bet, not just how well did you pick the game? Flat betting removes that variable entirely. The only question left is: are your picks positive expected value over time?
2. It protects you during losing streaks
Losing streaks happen even with a positive-edge model. The question is whether your bankroll survives until the edge shows up in the results.
If you bet 1u on every play and go 0-10 on your first 10 picks, you're down 10 units. That's painful but survivable. If you'd been betting 3u on your "high confidence" plays and happened to be wrong on those, you're in a much worse spot.
3. It makes your record auditable
When every bet is the same size, net units = wins minus losses adjusted for odds. Anyone can look at the history and verify the math. There's no "I bet bigger on that one" or "that was only a half-unit."
Transparency requires consistency.
How we track it
Every pick we post is tracked at flat 1u. Win on a +150 underdog: +1.50u. Lose on any pick: -1.00u. Win on a -130 favorite: +0.77u.
The net units number on the site reflects this exactly. Nothing is scaled up retroactively. Nothing is hidden.
What about variable staking?
Some systems recommend betting more on "high confidence" plays. The problem: confidence in a model is an estimate, not a fact. Over-betting high-confidence picks that turn out to be wrong is how people blow up bankrolls.
Unless you have a calibrated probability estimate (meaning you've verified that your 70% confidence plays win exactly 70% of the time across thousands of samples), variable staking adds more risk than it removes.
We don't have that sample size yet. So we use flat 1u.
Practical starting points
- Recreational bettor: 1-2% of your bankroll per play. Start at the lower end.
- Serious bankroll: Some people use a fixed dollar amount per unit that they don't adjust based on recent wins/losses.
- The number doesn't matter as much as keeping it the same every time.
The compounding effect of discipline — not losing track of your stake, not doubling down after losses, not getting overconfident after wins — is where most of the long-term edge in sports betting actually comes from.
For informational use only. Past results don't guarantee future performance. Bet responsibly. If gambling is affecting your life, call 1-800-GAMBLER.