Bankroll Management Fundamentals

By 4Football Staff · October 10, 2026 · 10 min read

Bankroll management is about how much to risk, not which team to back. Learn budgets, unit sizes, losing streaks and why stake size matters more than most people think.

Two friends make the same ten bets across a football weekend. Both win four and lose six. One of them finishes the weekend a little down and carries on as usual. The other finishes the weekend with an empty account and a text message to a relative asking for a loan. The picks were identical. What differed was the size of the stakes.

That difference is what people mean by bankroll management: the set of habits that decide how much money goes into betting, how much goes onto each bet, and what happens after losses. It is not about choosing winners. It is about making sure that a bad run, which is normal, does not turn into a disaster. This guide covers the basics with worked arithmetic. It is educational only, with no picks, and it is not financial advice. Betting is legal only in some places and for adults of legal age, and the safest amount for many people to wager is none.

What a bankroll is

A bankroll is the amount of money set aside for betting. Two conditions define a sensible one.

1. It is money you can afford to lose completely. Rent, food, bills, savings, tuition and debt payments are not part of it. If losing the whole amount would cause hardship, it is too much.
2. It is separate from everything else. Keeping it in its own account makes it easy to see how much has been spent. A bettor who dips into the checking account "just this once" has lost track of the boundary.

A helpful way to frame the bankroll is as an entertainment budget. A person who spends $40 a month on streaming services, or a few tickets to a game, does not expect a return on that spending. The same mindset fits betting. The money buys an experience, and a win is a bonus rather than income.

Setting the budget

Before any bet, decide three numbers in advance:

• The total for the month or season.
• The weekly limit, so a single bad weekend cannot use the whole amount.
• The stop point, the loss at which you stop for the day or week, whatever the schedule looks like.

Many licensed sportsbooks offer tools that help: deposit limits, loss limits, time-outs and self-exclusion. Using them is not an admission of a problem. It is the same logic as setting a budget before a shopping trip. Our guide to responsible gambling lists the tools and the warning signs that tell you it is time to step back.

Stake sizing: what a unit is

Bettors often describe stakes in units. A unit is a fixed fraction of the bankroll, chosen once, and every bet is described as a number of units. A common convention is to define one unit as around one percent of the bankroll, and sometimes as little as half a percent. The idea is not a rule from any authority; it is a way to keep each bet small enough that a losing run cannot do serious damage.

Here is the arithmetic for a hypothetical $1,000 bankroll.

The last row shows the danger. At 10 percent of the bankroll per bet, ten losses in a row ends the bankroll. Ten straight losses sounds unlikely, and it is for a single stretch, but bettors place hundreds of bets, and long runs of losses appear in any large enough sample. Big stakes turn a rough patch into a wipe-out.

Flat staking versus percentage staking

There are two ways to apply units.

• Flat staking. Every bet is the same dollar amount, based on the starting bankroll. It is simple and easy to track.
• Percentage staking. Every bet is a fixed percentage of the current bankroll. As the bankroll falls, the stake falls with it, which slows the decline. As it grows, the stake grows.

Take the 10 percent stake again. With flat staking at $100, ten straight losses remove $1,000, the whole bankroll. With percentage staking at 10 percent of the current balance, ten straight losses leave 0.9 to the tenth power of the bankroll, about 35 percent, so the bettor retains roughly $349 of the original $1,000. The bettor has lost about 65 percent but still has money to bet. Percentage staking never reaches zero in theory, though stakes can become too small to be worth placing.

Variance: the part that surprises people

Variance is the natural swing in results around the average. Even a bettor with a real edge experiences long losing runs, and a bettor with no edge experiences runs of both kinds.

Consider a simple case: bets with exactly a 50 percent chance of winning. The chance of losing five in a row is 0.5 to the fifth power, about 3 percent. That looks small, but over a hundred bets there are many chances for such a run to start. Streaks of five or six in a row, in either direction, are ordinary in a hundred coin flips. They are not evidence that anything has changed.

For a bettor winning 55 percent of bets at -110 (a very strong result, as break-even is about 52.4 percent), losses still come in clusters. This matters for two reasons. First, the stake must be small enough to survive the cluster. Second, a short run of results is a poor basis for judging skill; our look at sample size and small edges explains how many bets it takes before a result means much.

The opening story, with numbers

Return to the two friends from the start. Each places ten bets at -110 and wins four, losing six. Suppose each started with $500.

The two weekends are identical in shape, and the second friend lost ten times as much. If the losses came early, the second friend might not even have been able to place all ten $100 bets from a $500 balance, which is part of the point: the stake was too large for the bankroll to absorb an ordinary 4-6 weekend. A stake that is small relative to the bankroll leaves room to continue, and room to continue is what makes it possible to stop on purpose instead of by force.

The Kelly formula, and why it comes with warnings

Some bettors have heard of the Kelly criterion, a formula that gives the fraction of a bankroll to stake to maximize long-run growth if you know your true chance of winning. For a bet that pays b to 1 and wins with probability p, with q equal to 1 minus p, the Kelly fraction is (b × p − q) divided by b.

A worked example with hypothetical inputs: at -110 the payout is 0.909 to 1. If the true win chance were 55 percent, the formula gives (0.909 × 0.55 − 0.45) divided by 0.909, which is about 0.055, or 5.5 percent of the bankroll. If the true chance were 52.4 percent, the break-even point at -110, the formula gives zero. If it were 50 percent, the formula gives a negative number, which means do not bet.

The warnings are important.

1. It requires knowing the true probability, which nobody does. The formula is only as good as the estimate you put in.
2. Errors cost more than they help. If you think your chance is 55 percent and it is really 51 percent, the stake is too large for a bet that loses money on average.
3. Full Kelly is volatile. Many people who use it, in sports or investing, stake only a fraction of the figure to reduce swings.

For most recreational bettors, a small fixed unit is simpler and safer than a formula that depends on an edge they cannot measure.

Longshots, parlays and bankroll size

Different bets create different swings. A straight bet at -110 wins about half the time. A parlay or a longshot future wins rarely and pays a lot, which means long droughts punctuated by occasional wins. A bankroll that comfortably supports flat spread bets can be drained by a series of small parlays, because the chance of any one ticket winning is low. Our guide to parlays shows how the margin grows when the odds are long, and the same is true of futures.

A practical consequence is that the unit size for long-odds bets should be smaller than the unit for even-money bets, since the likely run of misses is longer. A bettor who sets the unit at one percent for spreads and decides on a fraction of that for longshots is applying the idea.

Common mistakes

• Sizing stakes to the excitement of the game. A big primetime matchup does not deserve a bigger unit than a quiet Sunday afternoon game.
• Changing the unit after a win or a loss. Resizing the unit up after a hot streak or down after a cold one tends to follow emotion, not arithmetic. Reset it on a schedule, such as monthly.
• Counting wins as proof. A good month on a small sample fits luck as well as skill.
• Spreading across too many bets. Ten small bets are ten margins, and the total at risk on a Sunday can be larger than it feels.
• Ignoring the cost of promotions. Bonus offers often come with playthrough rules that require more wagering than expected.

Rules for handling losses

The way a bettor behaves after losing is the most important part of bankroll management, and the hardest. These habits are widely recommended.

1. Never chase. Raising stakes to recover a loss is the most common route from a manageable loss to a serious one. A bet should be judged on its own price, not on the need to get even.
2. Stop at the limit. If the weekly budget is gone, the week is over. Decide this beforehand, because the decision is hard to make during a game.
3. Do not borrow. Credit cards, loans and money from family are not bankroll. If the bankroll runs out, the answer is to stop, not to refill from somewhere else.
4. Treat winnings as separate. Some bettors withdraw part of any profit so it cannot be lost back on a bad weekend. The reasoning is that a bankroll that only grows through deposits has a clear ceiling on losses.
5. Skip bets when not at your best. Tiredness, alcohol and anger are poor companions to a fast-moving betting screen, particularly with live betting.

The "house money" trap

After a win, some bettors begin to treat the winnings as less real than their original deposit, a habit often called playing with house money. The money is not different. A dollar won is the same as a dollar deposited, and losing it costs the same. Treating winnings as free tends to lead to larger stakes and more losses, and it is a well-known pattern in research on how people treat money.

Keep a record

Few bettors track their results, and fewer track them honestly. A simple log with the date, the bet, the price, the stake and the result takes a minute per bet. It answers questions that memory distorts: how much has been spent, how much has been won, and whether a particular type of bet, such as parlays, is costing more than expected. Memory favors wins, so a bettor who feels roughly even is often down by more than they believe.

Winnings may also be taxable depending on where you live. The rules differ by country and state, and a tax professional can explain what applies. Keeping a record makes that conversation easier.

When the numbers are not the problem

All of the above assumes a bettor who wants to stay in control. Bankroll rules cannot fix a situation in which gambling has started to take over. Signs include betting more than planned, hiding it from others, borrowing to bet, being unable to stop, or feeling distress about results. Anyone who sees these in themselves or in someone close should reach out to the National Problem Gambling Helpline at 1-800-GAMBLER. Free, confidential help is available.

Fans who simply enjoy following the games can do that with no wagering at all: the scores page has live results, the schedule shows kickoff times, and the odds page is there for those who want to understand how the market sees a game.

The short version

• A bankroll is money set aside for betting that you can afford to lose entirely, kept separate from bills and savings.
• Set a monthly total, a weekly limit and a stop point in advance, and use the limit tools sportsbooks offer.
• Define a unit as a small fixed share of the bankroll, often around one percent, and keep each stake to a unit or less.
• Variance is normal: streaks of five or six losses happen in a hundred coin flips, and large stakes turn them into disasters.
• Kelly staking needs an accurate estimate of your edge, which almost no one has; errors make it risky.
• Never chase losses, never borrow to bet, and keep a written record.
• If betting stops being fun, stop, and call 1-800-GAMBLER for free, confidential help.