Introduction
Bankroll management is the difference between a hobby and an investment. If you treat sports betting like a business, you won’t let short losing runs wipe you out. This guide explains how much to bet, why the commonly recommended 1–2% rule works for most bettors, how the Kelly Criterion informs optimal sizing, and a simple, defensible plan you can follow today.
Why bankroll management matters more than any single pick
You can be right about edges and still lose money if your bet sizing is reckless. Risk management protects the edge: it keeps you in the game long enough for positive expectation to compound. Two practical consequences:
- Volatility control: smaller, consistent stakes reduce drawdowns and mental pressure.
- Survival: good sizing prevents ruin during inevitable losing streaks.
A real-world way to see this is to compare two bettors with the same edge but different stake sizes: over time, the more conservative bankrolled bettor will compound steadily while the overleveraged bettor swings wildly and may hit zero.
Rule-of-thumb: 1%–2% units (conservative, effective)
Across industry guides and betting educators, a conservative starter unit is 1%–2% of your bankroll per bet (one “unit”). This range appears repeatedly in practitioner resources and bookmaker advice (see Sports Insights and Sports Betting Dime for practical guidance recommending 1%–3% and noting 1%–2% as conservative).1,2 For most bettors who are still learning their edge and model error, 1% is a sensible default, 2% is appropriate if you have documented edge and low variance markets.
Example: with a $1,000 bankroll, 1 unit = $10 at 1%, 2 units = $20 at 2%. With a modest +2% ROI on stakes, a 1% unit system grows capital slowly and survives losing runs.
Break-even math: why the market’s vigorish matters
Understanding the implied hurdle rate helps size bets rationally. At a common market price of -110 (risk $110 to win $100), the break-even win rate is:
Break-even = 110 / (110 + 100) = 110 / 210 ≈ 52.38%
That means if you place many -110 bets, you need to win roughly 52.4% to break even before considering edge sizing and variance. That simple arithmetic anchors expectations and explains why staking conservatively is prudent until you’re confident in sustainable edge.
The Kelly Criterion: the math for long-run growth (and why professionals use fractional Kelly)
The Kelly Criterion gives the fraction of bankroll to wager that maximizes long-term growth given you know your edge and odds. The full Kelly fraction f* for a binary bet can be written as:
f* = (bp - q) / b
where b = decimal odds - 1, p = probability of winning, q = 1 - p.
Kelly is attractive because it optimizes geometric growth, but it has two practical problems for real bettors: model error and high volatility. Because probability estimates are imperfect, full Kelly often produces bet sizes that are too large for practical use.
Practitioners routinely use fractional Kelly: commonly half Kelly: to reduce variance and the chance of ruin. Wikipedia and practitioner sources note that fractional Kelly (e.g., 50% Kelly) sacrifices some long-term growth in exchange for dramatically lower drawdowns and more tolerable swings.3,4
Example: if full Kelly recommends 8% of bankroll, half Kelly would stake 4%. The math preserves directionality (bet more when edge is large) while preventing destructive volatility if your p estimate is optimistic.
How to choose between units and Kelly in practice
Which system should you use? Simple guidance:
- New bettors / unclear edge: fixed-percentage units (1% rule): predictable, easy, low emotional cost.
- Growing confidence / quant edge: fractional Kelly (e.g., 0.25–0.5 Kelly) using conservative edge estimates.
- Professional-grade bettors with robust models: dynamic sizing tied to Kelly with strong variance controls and stress-tested simulations.
Units are blunt but effective. Kelly is precise but fragile. That’s why most successful bettors combine the two: estimate a Kelly fraction, then cap it inside a unit-based framework (never more than X% of bankroll, where X is your chosen safety cap).
A concrete, step-by-step plan (recommended)
Follow this playbook for consistent, low-friction bankroll management:
Establish your “true bankroll.” Include only money you can afford to lose.
Start with a conservative unit size: 1% of your bankroll.
- Example: bankroll = $5,000 → unit = $50.
Log every bet. Record date, market, stake (units), price, result, and closing price. Discipline on data collection is non-negotiable.
If you have a model that estimates p, calculate full Kelly then use fractional Kelly (0.25–0.5) to set a target stake. Do not exceed 3% of bankroll on a single wager without institutional justification.
Recalculate unit size monthly (or after a meaningful capital change): rebase unit = round(bankroll * 0.01). Avoid changing unit after each small win or loss: rebasing too often creates noise.
Cap single-bet exposure (hard limit): never risk more than 5% of bankroll on any single bet, for most bettors the practical upper cap should be 2.5% unless you have high conviction and model validation.
During long losing streaks, pause and review. If your results diverge from expected variance, investigate biases rather than simply increasing stakes to chase losses.
Use bet sizing rules as the tie-breaker between two close ideas: prefer the one that fits your staking plan and keeps your risk limits intact.
Examples: unit system vs fractional Kelly (numbers you can run yourself)
Assume bankroll = $2,000 and a pick at -110 (implied probability 52.38%). Two scenarios:
A) Unit system, 1% unit: stake = $20 per bet. B) Fractional Kelly: model estimates p = 0.56 (56% chance). For -110, b = 100/110 ≈ 0.909. Full Kelly f* = (bp - q) / b = ((0.909*0.56) - 0.44) / 0.909 ≈ 0.044 (4.4%). Half Kelly → 2.2% stake = $44.
Result: fractional Kelly suggests more aggressive sizing when model edge is clear. Unit sizing is steadier and easier to follow emotionally. Both can be correct: the difference is how much variance you can handle.
Handling losing streaks and drawdown discipline
Losing runs are inevitable. You should plan for them:
- Expect sequences of 6–12 losses even with a positive edge, depending on stake size and variance.
- If drawdown exceeds a pre-defined threshold (for example, 25% of peak bankroll), stop and audit your system: are you mis-estimating probability? Has the market shifted?
A simple rule: if bankroll drops by more than 20% from peak, reduce unit size by 50% until you regain ground and validate your edge. Preventing ruin is the most valuable benefit of staking rules.
Recordkeeping: the professional advantage
Accurate records let you measure ROI by market, by bet type, and by timeframe. Track at minimum: date, sport, bet type, stake, odds, closing odds, result, and notes about why you placed the bet.
With five months of cleaned data you can answer crucial questions: which markets produce the best ROI, which sportsbooks consistently shade lines against you, and whether your model is overfitting. Good recordkeeping transforms anecdotes into decisions.
Practical FAQs
Q: "Can I bump to 3–5% units if I’m confident?" A: Only with documented edge and stress-tested simulations. 3–5% increases volatility substantially, for most bettors that level is unsustainable over time unless backed by institutional risk controls.
Q: "Should I withdraw earnings or reinvest everything?" A: Treat the bankroll like capital and determine a withdrawal policy up front. A common approach: withdraw a fixed percentage of profits quarterly (e.g., 25% of new profits) to lock gains and avoid overconfidence.
Q: "How often should I rebalance unit size?" A: Monthly or after a 10%+ change in bankroll. Keep rebasing rules simple to avoid trading noise.
Quick checklist before placing a bet
- Is the stake ≤ my single-bet cap? (default: 2.5% absolute max)
- Does the bet align with unit or fractional Kelly target?
- Have I logged the bet in the tracker?
- Do I understand the downside and potential closing-line risk?
If you answer yes to each, proceed. If not, either reduce the stake or skip the bet.
Closing: risk controls beat hot streaks
Bankroll management is boring on purpose. It’s not about finding the flashiest pick, it’s about staying in the game long enough for a real edge to matter. Start with 1% units, keep disciplined records, use fractional Kelly as your sizing compass when you have a validated edge, and set hard caps to avoid ruin. This combination of conservative defaults and data-driven flexibility is how the serious, lasting winners manage risk.
If you want, we can provide a spreadsheet template with the unit tracker, Kelly calculator, and monthly rebasing formula: drop a note and we’ll generate it.
References:
- Sports Insights / Sportsbook strategies recommending 1%–3% unit sizing (conservative bettors often use 1%–2%).
- Sports Betting Dime guidance on conservative bankrolling (1%–2% recommended for most bettors).
- Kelly Criterion overview: Wikipedia (fractional Kelly as practical mitigation).