Bankroll Management Isn't Sexy But It's Why You're Losing
Every punter I've ever met wants to talk about two things: their winners and their "system." Nobody wants to talk about bankroll management. It's boring. It's maths. It involves discipline, which is the thing most gamblers have the least of.
It's also the single biggest determinant of whether you go bust or stay in the game. I've seen smart bettors with genuine edges blow themselves up because they staked too aggressively. And I've seen moderate bettors with small edges grind out consistent profits because they understood position sizing.
This article is going to be dry. Sorry about that. But if you take one thing from this entire site, make it this.
What's a Bankroll?
Your bankroll is money dedicated to gambling that you can afford to lose entirely. Not your rent money. Not your savings. Not "spare" cash from your current account. A separate, defined sum that you've allocated specifically for betting.
If you don't have a defined bankroll, you're not managing money — you're just spending it. And you probably don't track your results either, which means you almost certainly think you're doing better than you are. Every problem gambler I've ever spoken to has the same trait: they remember the wins in vivid detail and the losses are a vague blur.
Step one, before anything else: define your bankroll. Write the number down. Put it in a spreadsheet. It's real now.
Stake Sizing: The Bit That Actually Matters
How much of your bankroll should you put on a single bet? This is the key question and there are basically three schools of thought.
Flat staking. The simplest approach: every bet is the same fixed percentage of your bankroll. Most commonly 1-2%. So if your bankroll is £1,000, every bet is £10-20 regardless of how confident you are.
Flat staking is boring and it works. It protects you from yourself during losing runs (your stakes shrink as your bankroll shrinks, slowing the bleed) and it prevents you from getting over-excited about "certainties" and over-staking. I'd recommend 1-2% for anyone who isn't running a quantitative model.
Kelly Criterion. The mathematically optimal approach, and the one used by most serious bettors and traders. The Kelly formula tells you exactly what percentage of your bankroll to stake based on your edge and the odds:
Kelly % = (bp - q) / b
Where b = decimal odds minus 1, p = your estimated probability of winning, q = 1 - p.
Example: you think a bet has a 55% chance of winning (p = 0.55) and the odds are 2.10 (b = 1.10). Kelly % = (1.10 × 0.55 - 0.45) / 1.10 = (0.605 - 0.45) / 1.10 = 0.141 = 14.1%.
Full Kelly suggests staking 14.1% of your bankroll. Which is insane. Nobody should use full Kelly because it assumes your probability estimates are perfect, which they never are. If you overestimate your edge by even a small amount, full Kelly will eventually bankrupt you.
The standard practice is to use fractional Kelly — usually quarter or half Kelly. So in the above example, quarter Kelly would suggest about 3.5% stake. Much more reasonable. Still aggressive compared to flat staking, but it optimises your bankroll growth if your edge estimates are roughly accurate.
I used quarter Kelly when I was value betting seriously back in 2015-2016. It worked well, but it requires you to have genuine probability estimates for every bet — not just "I reckon this looks good" but actual numbers. If your estimates are wrong, your sizing will be wrong. Garbage in, garbage out.
Variable staking based on confidence. The "I'll put more on when I'm really sure" approach. Three units on strong fancies, one unit on speculative bets, that sort of thing. This is what most recreational bettors do and it's the worst of all worlds. Your "confidence" is just emotion dressed up as analysis. The bets you're most confident about are often the ones where you're most anchored to a narrative rather than genuine probability assessment.
If you're going to use variable staking, at least use Kelly or a model to determine the sizes rather than your gut. Your gut is wrong more often than you think.
The Practical Rules
Here's what I actually recommend to people who ask (and a few who didn't):
Never stake more than 5% of your bankroll on a single bet. Not ever. Not on a "dead cert." Not on a 1.05 favourite. The number of times I've seen supposedly guaranteed outcomes lose is too many to count. Leicester won the Premier League at 5000/1. Stuff happens.
Track everything. Every bet, recorded with: date, event, selection, odds, stake, result, profit/loss. Spreadsheet, betting tracker app, notebook — doesn't matter. What matters is that you have a complete record. After 100 bets, look at the data. After 500, look harder. The numbers don't lie even when your memory does.
Set a loss limit per day/week/month. When you hit it, stop. No "one more to get it back." No "the next one will turn it around." Stop. Go do something else. Come back when the period resets.
Never chase losses. This is the single behaviour that ruins more gamblers than any other. You lose a bet, so you increase your stake on the next one to "win it back." This is a mathematically proven path to ruin — it's essentially a martingale strategy, and martingale strategies always blow up given enough time.
I chased losses once. Early days, 2014, stuck £200 on a League Two match to recover a bad day. Lost that too. Then put £300 on an in-play market. Lost. Ended the day down £700 instead of the £180 I'd have been down if I'd just walked away. Never again. That £700 was about three weeks' disposable income at the time and it took months to rebuild both the bankroll and my confidence.
Separate your bankroll from your daily finances. Different account, different mental category. When you look at your betting bankroll, you should see a tool for gambling, not rent money. This psychological separation matters enormously. I use a separate e-wallet (Skrill) exclusively for gambling transactions. Money goes in monthly from my current account. That's my bankroll. Everything else is off-limits.
How This Connects to Everything Else
Bankroll management is the foundation that makes every other strategy viable. Value betting only works over large samples, which means you need to survive long enough for the maths to play out. If you're staking 10% per bet and hit a bad run (which you will — even a genuine 5% edge will produce runs of 15+ losses), you'll be wiped out before the edge manifests.
Same applies to casino play. If you enjoy slots or table games, set a session bankroll (a portion of your total gambling bankroll), use stakes appropriate to that amount, and accept that losing the session bankroll is a possible — likely, even — outcome. Casino games have a house edge. You're paying for entertainment. Size your stakes so the entertainment lasts and doesn't wreck you financially.
I covered the bonus angle in my piece about casino bonuses — bonus structures often push you to bet larger than your bankroll management would suggest. That's by design.
The Uncomfortable Truth
Most people who gamble will lose money over time. The house edge in casinos and the bookmaker's margin in sports betting ensure this. Bankroll management won't change that fundamental reality for most players.
What it will do is ensure that your losses are controlled, budgeted, and acceptable. You'll lose less than you otherwise would. You'll never lose more than you can afford. And if you do have a genuine edge — through value betting, arbitrage, or card counting — proper bankroll management is what turns that edge into actual, realised profit rather than a theoretical advantage that blows up in practice.
Nobody ever went bust betting 1% stakes. Think about that.