What is Bankroll Management?
Bankroll management is the practice of treating your betting funds as a separate, finite pot of money — and applying disciplined rules around how much of that pot you risk on each individual bet. It is the single most important skill in long-term betting, and the single most overlooked.
Most losing punters don’t lose because their selections are bad. They lose because their staking is bad. They risk too much on individual bets, increase stakes after losses, decrease stakes after wins, and eventually a normal variance run takes their bankroll to zero before any analytical edge has had time to play out. The maths of betting is unforgiving in this respect: even a punter with genuine 5% edge can ruin their bankroll if they’re staking 20% per bet during a normal losing streak.
This guide covers the framework that protects you. It’s not glamorous, it doesn’t promise tips that win, and it requires the kind of discipline most people find hard to maintain — but it’s the bridge between having edge and capturing it. Apply the principles below consistently and your worst betting outcomes will get meaningfully less bad. That alone is usually the difference between sustainability and burnout.
- Bankroll = the total pot of money set aside specifically for betting
- Unit staking (1–2% per bet) is the foundation of every sensible approach
- Kelly Criterion gives the mathematically optimal stake size based on edge
- Variance means you can be a long-term winner while losing for weeks
- Detailed record-keeping is the only way to distinguish bad luck from a broken strategy
Use our free betting calculators to work out unit sizes, accumulator returns, and each-way settlements without doing the maths by hand.
Setting Up Your Bankroll
Before any staking framework can work, you need a clearly defined bankroll. This is the most fundamental discipline in betting, and the one most casual punters skip entirely.
Rule 1: Money You Can Afford to Lose Entirely
Your bankroll must be money that, if it disappeared tomorrow, would have zero impact on your essential outgoings, savings goals, or financial obligations. Not money you can “probably spare”. Not money you’d “rather not lose”. Money you have already accepted as gone before the first bet is placed. If you cannot identify such a pot in your finances, you do not have a bankroll — you have rent money you’re gambling with, and no staking system will save you from the consequences of that.
Rule 2: Separated From Your Everyday Finances
The bankroll lives in a separate account, e-wallet, or at minimum a clearly tracked sub-balance. The point isn’t just psychological — it’s practical. When your bankroll is mixed in with your current account, every bet feels like it costs nothing because the day-to-day balance noise dwarfs the wager. Separated, every loss is visible. Every win is visible. The relationship between staking and bankroll becomes real.
Rule 3: Sized Realistically for Your Edge and Volume
A bankroll needs enough size to absorb normal variance. The standard professional rule of thumb is at least 100x your typical unit stake, with 200–500x being the norm for serious value-seekers. So if you want to bet £10 per unit, your bankroll should be at least £1,000. Anything less and a normal losing run can wipe you out before your edge has a chance to compound.
| Typical Stake | Minimum Recommended Bankroll | Conservative Recommended Bankroll |
|---|---|---|
| £5 per bet | £500 | £1,000+ |
| £10 per bet | £1,000 | £2,000+ |
| £25 per bet | £2,500 | £5,000+ |
| £50 per bet | £5,000 | £10,000+ |
If you don’t have the bankroll to support your desired stake, the answer is to lower the stake — not to increase the risk by under-funding the bankroll.
Unit Staking: The Foundation
The simplest and most robust approach to staking is flat unit staking: every bet is the same size, expressed as a percentage of your starting bankroll. Most successful long-term bettors stake somewhere between 1% and 2% per bet.
Why 1–2% Is the Sweet Spot
The maths of bankroll ruin makes this clear. With 1% unit staking, you can survive a 50-bet losing streak with 60% of your bankroll intact (roughly) — comfortably enough to continue betting through normal variance. With 5% staking, the same losing streak wipes you out entirely. The sustainability curve is brutal: doubling your stake percentage doesn’t double your risk, it multiplies it.
Worked Example: A Year of Flat Staking
Suppose you have a £1,000 bankroll and stake 2% per bet (£20). Over a year you place 500 bets. Your results show:
- Pure flat staking, 5% edge, average odds 2.0: roughly +£500 profit, with worst-case drawdowns of around £200
- Pure flat staking, no edge: roughly £0 P&L over thousands of bets, with maximum drawdowns of around 20–30% of bankroll
- 5% per bet staking, 5% edge: theoretically higher profit but realistic worst-case drawdowns can reach 60–70% of bankroll mid-year
The pattern is consistent: conservative staking sacrifices some maximum upside in exchange for a far higher probability of staying in business long enough to capture your edge. Most failed bettors didn’t fail because their analysis was wrong — they failed because they staked too aggressively and ran out of bankroll during a routine cold patch.
Kelly Criterion: Mathematically Optimal Staking
For mathematically inclined bettors, the Kelly Criterion calculates the theoretically optimal stake size based on your assessed edge. Used correctly, it maximises long-term geometric growth of bankroll. Used incorrectly, it’s the fastest way to ruin yourself.
The Kelly Formula
Kelly Stake % = (BP − Q) / B
Where:
- B = decimal odds − 1 (the “to one” portion)
- P = your assessed probability of winning
- Q = 1 − P (probability of losing)
A Worked Kelly Example
You’re considering a bet at 2.50 (3/2). You believe the true probability of winning is 50%.
- B = 2.50 − 1 = 1.50
- P = 0.50
- Q = 1 − 0.50 = 0.50
- Kelly = ((1.50 × 0.50) − 0.50) / 1.50 = (0.75 − 0.50) / 1.50 = 0.167
Full Kelly recommends staking 16.7% of your bankroll on this bet. That’s an enormous stake. If you have a £1,000 bankroll, Kelly suggests £167 on a single bet. This is why almost no real-world bettor uses Full Kelly.
Quarter Kelly: The Practical Choice
Most professionals use Quarter Kelly — staking 25% of the calculated optimal. Quarter Kelly captures most of the long-term growth while dramatically reducing the variance and drawdowns. In our example above, Quarter Kelly would mean staking 4.2% of bankroll (£42) instead of £167.
Why fractional Kelly? Because Full Kelly assumes you’re estimating probability with perfect accuracy. Nobody does. Even small overestimates of edge lead to dangerous overstaking under Full Kelly. Quarter Kelly builds in margin for error — your edge has to be off by 4x before you’re approaching ruin territory.
When Kelly Isn’t Right For You
Kelly only works if:
- You have genuine, measurable edge (positive Closing Line Value over hundreds of bets)
- You can quantify that edge as a probability assessment
- You’re emotionally able to handle the variance Kelly produces
If any of these are uncertain, flat staking at 1–2% is significantly safer. Kelly is for established, edge-confirmed bettors with disciplined records — not for those still learning their game. For more on identifying whether you actually have edge, see our value betting guide.
| Approach | Stake Size | Long-Term Growth | Variance / Drawdown | Best For |
|---|---|---|---|---|
| Flat 1% | Fixed 1% of bankroll | Slower | Very low | Beginners; preserving capital |
| Flat 2% | Fixed 2% of bankroll | Moderate | Low | Most retail bettors |
| Quarter Kelly | 25% of full Kelly recommendation | Strong | Moderate | Edge-confirmed bettors |
| Half Kelly | 50% of full Kelly recommendation | High | Significant | Pros with stable edge |
| Full Kelly | Full Kelly recommendation | Maximum theoretical | Brutal | Almost nobody, in practice |
Variance: Why Sound Strategy Sometimes Looks Broken
Variance is the single most psychologically destructive force in betting. It is also the most misunderstood. Even with genuine edge and perfect bankroll discipline, you will have losing weeks, losing months, and occasionally losing quarters. The maths is brutal — and most bettors abandon sound strategies during normal variance, mistaking bad luck for a broken process.
The Maths of Variance
Suppose you have a 5% edge — your bets are placed at prices that exceed true probability by 5% on average. Your long-term ROI is +5%. Sounds straightforward. The reality of how that plays out:
- After 100 bets: expected profit is 5 units (at 1-unit stakes), but the standard deviation is around ±15 units. You could realistically be anywhere from −10 to +20 units.
- After 500 bets: expected profit is 25 units, with standard deviation around ±35 units. The signal is starting to emerge.
- After 1,000 bets: expected profit is 50 units, with standard deviation around ±50 units. Edge is becoming clearly visible.
- After 10,000 bets: expected profit is 500 units, with standard deviation around ±150 units. Edge is unambiguous.
Most retail bettors never reach 1,000 bets at consistent stakes. Most who quit “because the system stopped working” simply hit a normal variance trough and assumed their edge had disappeared.
The hard truth: If your process is sound — value-based selection, disciplined staking, accurate record-keeping — you must persist through losing runs. The maths only emerges over hundreds or thousands of bets. Quitting at the bottom of a normal variance trough is how good bettors turn into losing bettors.
Drawdown Expectations
Even with a genuine edge, expect drawdowns of 20–30% of bankroll over a year. With higher variance approaches (Kelly, accumulators, longer-odds value bets), drawdowns of 40–50% are not unusual. Plan for these. Pre-decide what your “stop and review” point is — the bankroll level at which you pause to objectively assess whether the issue is variance or strategy. Setting that threshold in advance, while calm, is what stops you tilting at it.
Tracking Every Single Bet
If you’re not tracking, you don’t have a betting strategy — you have a hobby with delusions about itself. Detailed tracking is what separates winners who know they’re winning from losers who think they’re winning.
What to Track
The minimum useful record for every bet:
- Date and event — when and what
- Sport / market type — football match result, racing each-way, etc.
- Selection — what you backed
- Odds taken — the price you got
- Stake — exact pound figure
- Bookmaker — which book the bet was placed at
- Result — won, lost, void, half-won
- Profit/loss — net £ outcome
- Closing odds — the odds available at kick-off (the most important field)
- Notes — your reasoning, info changes, why you backed it
Closing Line Value: The Truth-Teller
The single most powerful metric is Closing Line Value (CLV) — the difference between the odds you took and the odds available at the moment the market closed. The closing line is the sharpest, most-informed estimate of true probability available, and consistent positive CLV is the strongest signal that you have genuine edge.
Suppose you back Liverpool at 2.10 four hours before kick-off. The closing price is 1.85. You took a price 13.5% better than the closing line — that’s a meaningful edge. Track this metric across hundreds of bets. Average positive CLV with negative P&L = bad variance, keep going. Average negative CLV with positive P&L = good variance, your process is broken.
Monthly Review
Review your tracking spreadsheet at least monthly. Calculate:
- Total bets, wins, losses, voids
- Total stake, total return, profit/loss
- ROI (profit ÷ total stake)
- Average CLV
- P&L by sport, market type, and bookmaker
- Bankroll growth/decline curve
This is what tells you whether your approach is working. A spreadsheet doesn’t lie, doesn’t tilt, and doesn’t tell you what you want to hear.
Common Bankroll Management Mistakes
| Mistake | What’s Going Wrong | Better Approach |
|---|---|---|
| Increasing stakes after losses | “Chasing” — accelerates ruin in normal variance | Fixed unit staking regardless of recent results |
| Decreasing stakes after wins | “Locking in profit” caps upside artificially | Stay at unit stakes; let bankroll grow naturally |
| Doubling up on “sure things” | No bet is safe enough to overstake | Same unit regardless of confidence |
| Chasing weekly or daily targets | “I want to be up £100 this month” — forces poor bets | Process goals, not outcome goals |
| Mixing bankroll with everyday money | Losses become invisible against day-to-day balance noise | Separate account or clearly tracked sub-balance |
| Not tracking every bet | Cannot distinguish edge from variance | Spreadsheet from day one — even £1 stakes |
| Tilting after a single bad result | Single bets shouldn’t drive emotional state | Pre-defined stop rules; walk away when triggered |
| Topping up after losing the bankroll | Re-betting losses with new money compounds the problem | Bankroll gone = stop and review, not refill |
Bankroll Discipline Across Bet Types
Different bet types have different variance profiles, and your staking should reflect that. The 1–2% unit principle applies as a baseline, but the practical application differs.
Single Bets
The cleanest application of unit staking. Each single bet at 1–2% of bankroll. Track CLV, review monthly, adjust unit size only when the bankroll reaches meaningful new highs (e.g. every 25% increase).
Accumulators
Higher variance by nature — most accas lose, the few that hit are larger. Stake accas at the same 1–2% unit as singles, treating each acca as a single unit-staked bet. Resist the temptation to stake higher because “the odds are better” — that’s the trap that compounds variance and destroys bankrolls. Our football accumulator strategy guide covers the optimal leg-count and EV maths in detail.
Each-Way Betting
Remember that an each-way bet is two bets — a £10 each-way bet costs £20 in total stake. Size accordingly: if your unit is £20, an each-way bet is £10 each-way (£20 total), not £20 each-way (£40 total). Most beginners get this wrong and double-stake themselves into unintended exposure.
Bet Builders
Treat bet builders the same as singles for staking purposes — one combined bet, one unit. The market complexity doesn’t change the bankroll maths. Higher correlation pricing means slightly higher variance than equivalent singles, but the staking principle is unchanged. See our bet builder walkthrough for more on the format.
System Bets (Lucky 15, Yankee, etc.)
System bets are multiple bets bundled into one — a Lucky 15 is 15 separate bets. Treat the total stake as your unit, not the per-bet stake. So a “£1 Lucky 15” is a £15 system bet, and that £15 should fit within your unit-staking framework. Our horse racing betting guide covers system bet structures in more detail.
Scaling Stakes With Bankroll Growth
If you’re winning, your bankroll grows — and your unit stake should grow with it. But the right cadence matters. Increase too aggressively and you erode your variance buffer; increase too slowly and you cap your growth unnecessarily.
The 25% Rule
Adjust your unit stake when your bankroll grows by 25%. Started at £1,000 with £20 unit (2%)? When your bankroll hits £1,250, increase the unit to £25. Hits £1,562? Move to £31. And so on. This compounding approach scales your stakes with your edge while preserving the same percentage risk per bet.
Decreasing Stakes When Bankroll Shrinks
The mirror image is just as important. If your bankroll drops by 25%, scale your unit stake down to maintain the same percentage risk. This is the discipline that prevents a small drawdown from becoming bankroll ruin. Most bettors find decreasing stakes psychologically harder than increasing them — but it’s the same principle, in reverse, and equally important.
Pre-Set Bankroll Triggers
Define in advance — when calm, before the bets are placed — the bankroll levels at which you’ll act. For example:
- Bankroll +25%: increase unit stake by 25%
- Bankroll −25%: decrease unit stake by 25%
- Bankroll −50%: stop, review CLV, examine for process drift before continuing
- Bankroll back to starting level after drawdown: re-enter at original unit; ignore “lost ground” thinking
Setting these triggers in advance is what stops you tilting at them. The decision is already made; you just execute.
Responsible Approach to Bankroll Management
Good bankroll management is the operational expression of treating betting as entertainment with strict financial limits. The disciplines in this guide — separated bankroll, unit staking, fixed limits, mandatory tracking — are the same disciplines that protect against problem gambling. They’re not separate frameworks; they’re the same framework, viewed from two angles.
If you find yourself increasing stakes during downturns, betting on selections you don’t have edge on, chasing losses, or repeatedly topping up a depleted bankroll, those are signs that the maths has stopped governing your decisions. Step back, set deposit limits at every bookmaker, and use the responsible gambling tools available. The deposit limits, time-out features, reality checks, and self-exclusion options at every UKGC-licensed bookmaker exist precisely for these moments.
Free, confidential support is available 24/7 from the National Gambling Helpline on 0808 8020 133, or via BeGambleAware.org. For self-exclusion via GamStop and the wider toolkit, see our guide to GamStop and self-exclusion.
Bankroll Management: Frequently Asked Questions
What’s the right starting bankroll for a beginner?
Whatever you can comfortably afford to lose entirely without affecting your essential outgoings or savings goals. There’s no “correct” number — £100, £500, £1,000 are all fine starting points. What matters is that the bankroll is genuinely separate from your everyday finances and that your unit stake is calibrated to the bankroll size (typically 1–2%, so £1–£2 stakes on a £100 bankroll, £10–£20 stakes on a £1,000 bankroll).
Should I use Kelly Criterion or flat staking?
For most retail bettors, flat staking at 1–2% is significantly better. Kelly Criterion only works if you have measurable, established edge and can accurately quantify probability — and even then, most professionals use Quarter Kelly to reduce variance. Until you have hundreds of bets tracked with positive Closing Line Value, stick with flat staking. It’s simpler, more robust, and harder to get badly wrong.
How often should I increase my unit stake?
The standard rule is to adjust when your bankroll grows by 25%. Started with £1,000 and £20 unit? When bankroll hits £1,250, increase unit to £25. Hits £1,562, move to £31. This compounding approach scales your stakes with your edge while preserving the same percentage risk per bet. Avoid increasing stakes more frequently than this — small bankroll moves can be normal variance, not real growth.
What should I do during a losing streak?
If your tracking shows positive average Closing Line Value, the answer is almost always: nothing. Continue with your unit staking framework and trust the process. Variance is brutal but mathematical — extended losing runs at positive CLV are normal and will reverse. If your CLV is negative, however, your process needs review. The temptation during losing streaks is to chase by increasing stakes; this is the single fastest path to bankroll ruin.
Can I top up my bankroll after losing it?
Topping up an emptied bankroll with new money is one of the most dangerous patterns in betting. It signals that the bankroll wasn’t genuinely separate from your everyday finances, and it removes the natural stop-loss that bankroll exhaustion provides. If your bankroll runs out, the right answer is to stop, review your records honestly, and only resume betting if and when you can fund a new bankroll from genuinely disposable money. Repeatedly topping up is one of the clearest warning signs of problem gambling.
How long does it take to know if I have edge?
Closing Line Value gives you a directional answer after 100–200 bets. Profit and loss alone is unreliable until 1,000+ bets due to variance. Most bettors confuse short-term variance with edge or its absence — which is why systematic tracking and CLV analysis matter more than any individual result. If your CLV is consistently positive, you have edge regardless of what your P&L looks like in any given month.
Should my bankroll be in a separate bank account?
It’s the gold standard, but not strictly required. What matters is clear separation — either a separate account, a dedicated e-wallet, or at minimum a clearly tracked sub-balance you mentally and recordwise treat as ringfenced. The point is that betting losses should be visible against the bankroll, not buried in everyday balance noise. If you’re using a single bank account, a meticulous tracking spreadsheet is essential.
What’s the difference between bankroll management and money management?
The terms are often used interchangeably. Strictly speaking, “bankroll management” is the broader framework — defining the pot, separating it from everyday finances, and applying staking rules. “Money management” or “staking” is specifically the rules around individual bet sizing within that framework. Most guides treat them as synonymous, and the practical advice overlaps almost entirely.
