Good analysis can still fail without bankroll management. Even a bettor whose probability estimates are sound will endure losing streaks, and how much you stake on each wager determines whether a streak is a nuisance or a disaster. This article explains bankroll sizing, betting units, the Kelly criterion, and risk of ruin using simple, checkable numbers.
What Your Bankroll Is (and Is Not)
A bankroll is money set aside specifically for wagering, separate from rent, bills, savings and emergency funds. It should be an amount you can afford to lose entirely without financial stress. Defining it in advance matters because it turns staking into a rule-based decision instead of an emotional one, especially after a loss.
Units: A Consistent Yardstick
A unit is a fixed percentage of your bankroll, commonly 1% to 2%. Using units lets you compare results regardless of bankroll size. With a $2,000 bankroll and a 1% unit, one unit is $20. A common guideline is to keep typical stakes between 0.5 and 2 units, though nothing here is a rule; it is a way to keep the size of each bet small relative to the total.
Variance and Losing Streaks
Variance is the natural scatter of outcomes around the average. Even at a true 50% win rate, eight straight losses happen with probability 0.58 = 0.39% for any specific set of eight bets, yet across hundreds of bets such runs are not unusual. Your staking plan has to survive them. For more on how prices relate to probabilities, see implied probability and expected value.
Worked Example: Same Streak, Two Stake Sizes
Take a $2,000 bankroll and an eight-bet losing streak, with each bet sized as a percentage of the current bankroll.
| Stake per bet | Calculation | Bankroll after 8 losses | Drawdown |
|---|---|---|---|
| 1% | $2,000 × 0.998 | $1,845 | 7.7% |
| 5% | $2,000 × 0.958 | $1,327 | 33.7% |
The same bad luck costs about $155 at 1% and about $673 at 5%. Recovery is also asymmetric: after a 33.7% loss you need a gain of 1 ÷ 0.6634 – 1, or roughly 50.7%, to get back to even. Smaller stakes preserve the ability to keep playing while your edge, if any, has time to show up.
Flat Staking Versus Kelly
Flat staking risks the same amount each time. It is simple and limits damage from estimation errors. The Kelly criterion instead sizes each bet by your estimated edge: f = (b × p – q) ÷ b, where b is the net decimal odds (decimal minus 1), p is your win probability and q = 1 – p.
Suppose a bet pays +110 (decimal 2.10, so b = 1.10) and you estimate a 50% chance. The expected value is 0.5 × 2.10 – 1 = +5%. Kelly gives (1.10 × 0.5 – 0.5) ÷ 1.10 = 0.05 ÷ 1.10 = 4.55% of bankroll. On $2,000, that is about $91.
The catch is that Kelly assumes your probability is exactly right. If you overestimate, full Kelly overbets. Many practitioners use half or quarter Kelly (2.27% or 1.14% in this case) to allow for that uncertainty. If your edge is zero or negative, Kelly says to stake nothing.
Risk of Ruin in Plain Terms
Risk of ruin is the probability of losing your whole bankroll, or falling below a level where you must stop. It rises with larger stakes, higher variance (long-shot bets and parlays), and a smaller or nonexistent edge. If there is no edge, a bettor facing a bookmaker’s margin faces eventual ruin no matter how carefully stakes are sized; staking only changes how quickly. Sizing cannot manufacture an edge. To gauge how large the margin is on a given market, see how to calculate no-vig odds and fair probability.
Practical Habits
- Write down your bankroll, unit size and maximum stake before you bet.
- Recalculate units periodically, not after every result.
- Keep a log so that you can compare outcomes with expectations.
- Stop or reduce activity if you find yourself chasing losses.
Choosing a Level of Risk
Your tolerance matters as much as the math. A useful exercise is to decide the largest drawdown you could accept, perhaps 20% of the bankroll, and work backward. If a plausible bad run is twenty losing bets, a 1% stake shrinks the bankroll to 0.9920 = 81.8% of its start, an 18.2% drawdown. At 2% the same run leaves 0.9820 = 66.8%, a 33.2% drawdown. Seeing these figures in advance makes it easier to keep stakes modest.
Finally, keep bankroll and spending separate. If you deposit more money after a bad run, the original plan no longer applies, and your true risk is higher than your unit sizes suggest. Setting deposit limits in advance is a sensible safeguard.
Frequently Asked Questions
How big should a unit be?
Many educational sources suggest 1% to 2% of bankroll, but there is no universal answer. Smaller is more conservative and more tolerant of error.
Does bankroll management guarantee profit?
No. It manages risk and survival, not results. Without a genuine edge over the odds, expected value stays negative.
Should I stake more after a loss to win it back?
No. Each bet is independent of past outcomes, and increasing stakes after losses accelerates risk of ruin.
Conclusion
Bankroll management is about limiting the damage of variance. Set a defined bankroll, use small consistent units, treat Kelly as a ceiling rather than a target, and remember that no staking plan removes the bookmaker’s margin. Pair these habits with careful price comparison, covered in comparing odds across sportsbooks.
Educational content only, not betting advice. Please gamble responsibly; 21+ where applicable.
