The dirtiest secret in sports betting is this: most losing bettors are not actually bad at picking winners. They are bad at sizing bets. The math behind their picks would be marginally profitable in the long run if executed correctly, but the execution gets them killed before the long run arrives. They blow up during a normal losing streak that any disciplined bettor would have survived, decide they were “doing it wrong,” change their entire approach, and start the cycle over with a fresh deposit. This pattern repeats until the deposits stop coming.
The fix is bankroll management. It is not glamorous, it is not interesting to write about, and it is the single most important skill in the entire discipline. A bettor with a small edge and good bankroll management will outperform a bettor with a much larger edge and bad bankroll management every time. The math on this is unambiguous, and it has been unambiguous since Edward Thorp wrote about it in the 1960s. The only thing that has changed is that the tools to apply it are now free.
Why Most Bettors Go Broke
The mechanics of going broke are almost always the same. The bettor starts with a bankroll – call it $1,000. The bettor decides to bet “a unit” per game, where a unit is some percentage of the bankroll. The percentage is usually too high. Five percent is common. Ten percent is not unusual. Some bettors size by hunches and end up with 20% or 25% of the bankroll on their “lock of the week.”
The bettor wins some, loses some, and starts feeling either confident or nervous depending on how the first few weeks go. Then the inevitable losing streak hits. Eight in a row, ten in a row, twelve in a row – it does not matter, because every winning bettor experiences losing streaks of this length over a long enough sample, and most bettors do not understand how routine they are.
By the time the streak ends, the bankroll has been cut in half or worse. The bettor, looking at the damage, decides one of two things. Either “the system is broken” and changes everything, or “I need to win it back” and starts betting larger amounts to make up the gap. Both responses lead to the same outcome: the bankroll goes to zero, the bettor concludes that sports betting does not work, and the cycle ends.
None of this needed to happen. The losing streak was statistically expected. The bankroll could have absorbed it if the bet sizing had been smaller. The system was not broken; the execution was broken. This is what bankroll management exists to prevent, and the prevention works if you do it.
The Math of Ruin
Risk of ruin is the probability that a bettor with a given edge, bet size, and bankroll will eventually go broke. The formula is straightforward but the results are counterintuitive. A bettor with a 4% edge – which is a serious edge by any standard – betting 5% of their bankroll on each play has a risk of ruin of roughly 13%. Same edge, betting 10% per play: risk of ruin jumps to over 60%. Same edge, betting 2% per play: risk of ruin drops to about 2%.
The lesson is brutal. Cutting your bet size in half does not cut your risk of ruin in half. It cuts it by an order of magnitude. The relationship between bet size and survival is nonlinear, and most bettors are on the wrong side of the curve.
Before placing any bet, use a betting unit size calculator to determine your optimal stake based on bankroll, win rate, and risk tolerance. Plugging in real numbers – your actual bankroll, your honest estimated win rate, the odds you typically bet at – and seeing the recommended stake is usually the moment a bettor realizes they have been betting two to three times too much for years. The number the calculator returns is almost always smaller than the bettor’s gut would have chosen, and the gap between the two is the gap between long-term survival and inevitable blowup.
The Kelly Criterion, in Plain English
The most-cited bankroll management framework is the Kelly criterion, developed by John Kelly at Bell Labs in 1956. The full math is more involved, but the practical version is simple: bet a fraction of your bankroll equal to your edge divided by the odds. If your edge on a +100 (even money) bet is 4%, full Kelly says to bet 4% of your bankroll. If your edge on a +200 bet is 4%, full Kelly says to bet 2% of your bankroll, because the higher payout means a smaller stake achieves the same long-run growth.
Full Kelly maximizes the long-run growth rate of the bankroll. It also produces wild swings in the short run, which is why almost every serious bettor uses fractional Kelly – usually half Kelly or quarter Kelly – in practice. Half Kelly captures roughly 75% of the long-run growth rate of full Kelly with dramatically lower variance. Quarter Kelly captures roughly 44% of the growth rate with much lower variance still. The trade-off between aggression and stability is a personal choice, but the choice should be made consciously, not by accident.
The two ways Kelly can hurt you are also worth understanding. First, Kelly assumes your edge estimate is accurate. If you think your edge is 4% and it is actually 1%, Kelly will tell you to overbet by a factor of four, which can blow up the bankroll just as efficiently as flat betting too much. The defense is to be conservative about your edge estimates and to use fractional Kelly as a margin of safety. Second, Kelly assumes you have a real edge in the first place. A bettor with no edge running Kelly is just betting smaller amounts on a losing proposition, which loses money slower but still loses money. Kelly is a sizing tool, not an edge tool.
Why “Units” Are a Worse System Than Kelly
Most amateur bettors size by units. A “unit” is some percentage of the bankroll, often 1% to 5%, and the bettor bets one unit on every play regardless of the odds, the perceived edge, or the variance of the bet. This is simpler than Kelly and worse than Kelly in almost every way.
The biggest problem with flat unit sizing is that it treats every bet identically when the bets are not identical. A 1-unit bet on a -110 favorite is a fundamentally different proposition from a 1-unit bet on a +400 underdog, even though the bettor is staking the same amount. The variance is different, the optimal Kelly fraction is different, and the impact of the bet on the long-run growth of the bankroll is different. Sizing them the same is leaving money on the table at best and inviting a blowup at worst.
The smaller problem is that unit sizing does not adjust as the bankroll changes. A bettor who starts with $1,000 and bets $50 units (5%) is overbetting at the start. If the bankroll grows to $5,000 and the bettor still bets $50 units (now 1%), they are underbetting at the new size. Either error costs growth. Kelly automatically scales with the bankroll. Units do not, unless the bettor recalculates manually after every win and loss, which almost no one does.
The Three Numbers Every Bettor Should Track
Bankroll management is impossible without recordkeeping, and most bettors do not keep records that are detailed enough to be useful. The three numbers that matter are your closing line value, your win rate, and your stake-adjusted ROI.
Closing line value is the gap between the odds you got and the odds the line closed at. If you bet a -110 favorite and the line closed at -130, you got positive CLV – you beat the closing line. CLV is the strongest predictor of long-term profitability that has been identified in sports betting. A bettor who consistently beats the closing line will eventually be profitable even if their short-term win rate looks ordinary. A bettor who consistently loses to the closing line will eventually lose money even if their short-term win rate looks great.
Win rate is obvious but commonly miscalculated. The honest version is wins divided by total resolved bets, including pushes counted as half-wins. Most amateur bettors quietly exclude their bad weeks from this calculation, which produces a flattering number that has no predictive value.
Stake-adjusted ROI is the percentage return on the total amount staked, not the percentage return on the bankroll. A bettor who stakes $10,000 over a year and ends $200 ahead has a 2% ROI even if their bankroll only grew from $1,000 to $1,200. ROI is the metric that actually measures betting skill, separate from bankroll management. A bettor with a 4% ROI has a serious edge. A bettor with a 1% ROI has a small edge that needs to be compounded patiently. A bettor with a negative ROI has no edge, no matter how their bankroll happens to be doing in the short run.
The Psychological Side
The hardest part of bankroll management is not the math. It is the psychology of sticking to it during streaks. After a long losing streak, the urge to “make it back” with bigger bets is overwhelming. After a long winning streak, the urge to “press the edge” with bigger bets is almost as bad. Both impulses are wrong. The correct response to a losing streak is to keep sizing exactly the same way – because the size was correct before the streak and is still correct now. The correct response to a winning streak is also to keep sizing exactly the same way, because the size was correct before the streak and is still correct now.
The bettors who survive long-term are the ones who can run the same staking plan through a bad month without flinching. The math has not changed. The plan has not changed. Only the recent results have changed, and recent results are noise. If you cannot tell the difference between noise and signal during a tough stretch, you will eventually act on the noise, and acting on noise is what kills bankrolls.
What a Working Plan Looks Like
A working bankroll management plan in 2026 has four components. First, a bankroll that is genuinely separate from the rest of your finances – money you can afford to lose entirely without changing how you live. Second, a sizing rule based on Kelly or fractional Kelly, with honest edge estimates and a margin of safety built in. Third, a recordkeeping system that tracks CLV, win rate, and stake-adjusted ROI bet by bet. Fourth, a written rule for when you stop and reassess – usually a drawdown threshold, like “if the bankroll drops 30%, stop betting for two weeks and review every recent decision before resuming.”
None of this is exotic. None of it requires special tools beyond a free calculator and a spreadsheet. The bettors who do it are not smarter than the bettors who do not. They are just more disciplined, and discipline in bankroll management is the entire game.
The Hard Truth
If you have been betting for a year or more and your bankroll keeps ending up in the same place, the problem is almost certainly not your picks. It is the size of your bets. Cut your bet size in half for the next month. Track everything. Calculate your real ROI honestly. If the picks have an edge, the bankroll will start to behave differently almost immediately. If they do not, the cut in bet size will at least make the eventual lesson cheaper. Either way, you will know more than you knew before, and knowing more is the only way out of the cycle.
