The Maths That Keeps You in the Game
I blew through my first MMA betting bankroll in six weeks. It was 2014, I had just started taking fight analysis seriously, and I was convinced that my edge was large enough to justify aggressive staking. I bet 10% of my bankroll on a single fight, then 15% on the next because the first one won and I felt invincible. By the time a three-fight losing streak hit, I had burned through two-thirds of my capital and had to rebuild from scratch. The analysis was fine. The staking was disastrous.
That experience taught me a lesson I have seen confirmed hundreds of times since: you can have a genuine edge in picking MMA fights and still lose money if your staking is wrong. Bankroll management is not the exciting part of betting. It does not generate podcast content or social media clips. But it is the structural foundation that determines whether your analytical edge translates into profit or evaporates through variance.
Ten percent of UK adults participate in online sports betting, and 15% of men place sports wagers — roughly four times the rate among women. Within that population, the vast majority have no formal staking plan. They bet what feels right, increase stakes after wins, chase losses after bad weekends, and wonder why their balance trends downward even when their picks are decent. This article is for anyone who recognises that pattern and wants to replace gut-feel staking with actual maths.
Why Bankroll Management Matters More in MMA Than Team Sports
A football bettor placing thirty wagers per weekend across multiple leagues can absorb a losing streak with relative ease. The high frequency of events means variance smooths out quickly. An MMA bettor placing five to ten wagers per card, with events spaced one to three weeks apart, faces a very different mathematical reality. Losing streaks last longer in calendar time, drawdowns feel more painful, and the temptation to deviate from a plan has weeks to fester between cards.
MMA also has structural volatility that team sports do not. Roughly 44% of UFC fights end by judges’ decision, but the remaining 56% end by stoppage — knockout, TKO, or submission. A stoppage can happen at any second. A fighter can be winning every metric on the scorecards and then get caught by a single punch that ends the fight instantly. Sportsbooks themselves acknowledge the challenge — the betstamp editorial team described MMA as notoriously difficult to correctly handicap, which is precisely why the odds carry wider margins and why disciplined staking matters more here than in sports where the bookmaker’s pricing is tighter.
This volatility makes bankroll management more critical in MMA than in any other sport I bet. If you are staking 5% of your bankroll per bet and you hit a run of eight consecutive losses — which is entirely possible even with a 55% win rate — you have lost 34% of your bankroll before variance turns in your favour. At 10% per bet, the same losing streak wipes out 57% of your capital. The maths is unforgiving, and no amount of analytical brilliance compensates for staking that exposes you to ruin.
There is also a scheduling factor unique to MMA. A Premier League bettor has matches every week for nine months. An MMA bettor might have one UFC event per week during busy stretches, but Fight Night cards often feature matchups where the pricing does not offer value. There will be weeks where you have zero actionable bets. If your bankroll is too small or your staking too aggressive, you run out of runway before the next opportunity arrives. The intermittent schedule of UFC events demands a staking approach that can absorb dry spells without forcing you into marginal bets just to stay active.
The goal of every staking system I describe below is the same: keep you in the game long enough for your edge to manifest. An edge that plays out over five hundred bets is worthless if your bankroll cannot survive the first hundred.
Fixed-Percentage Staking: The Simplest System
If you are going to use only one staking system for the rest of your MMA betting career, make it this one. Fixed-percentage staking means you wager the same proportion of your current bankroll on every bet, regardless of how confident you feel or how attractive the odds look.
The standard recommendation is between 1% and 3% per bet. I use 2% as my default. On a 1,000-pound bankroll, that means every bet is 20 pounds. If I win and my bankroll grows to 1,040 pounds, my next bet is 20.80. If I lose and my bankroll drops to 980, my next bet is 19.60. The stakes adjust automatically with the size of the bankroll, which means you bet more when you are winning and less when you are losing — without any emotional decision-making.
The beauty of this system is its simplicity. There are no formulas to calculate before each bet, no confidence ratings to assign, no probability estimates to plug in. You decide on a percentage, apply it to your current balance, and move on. For someone who has never used a staking system before, fixed-percentage is the fastest path from chaos to structure.
The trade-off is that it treats every bet as equally valuable, which is rarely true. A bet where you estimate a 10% edge should logically carry a larger stake than one where you estimate a 2% edge. Fixed-percentage staking ignores that distinction entirely. Over a large sample, you will leave money on the table by understaking your strongest plays and overstaking your weakest ones. But “leaving money on the table” is a luxury problem. The system protects you from the catastrophic downside of overstaking, which is the risk that actually ends betting careers.
For anyone transitioning from no system to a system, start at 1% and hold it there for at least fifty bets. The point is not to maximise returns immediately. The point is to build the habit of disciplined staking, which is far harder than it sounds when you are staring at a fight you feel certain about and your stake feels embarrassingly small.
Unit-Based Staking: Confidence Tiers for UFC Bets
Unit-based staking is fixed-percentage’s older, slightly more sophisticated sibling. Instead of betting the same percentage on every fight, you define a “unit” — typically 1% to 2% of your bankroll — and then assign a confidence rating to each bet that determines how many units you stake.
A common framework uses three tiers. A one-unit bet is your standard play — you see value, but the edge is modest or you have lower conviction. A two-unit bet signals stronger conviction: the gap between your estimated probability and the bookmaker’s implied probability is wider, and your confidence in the underlying analysis is high. A three-unit bet is reserved for rare occasions where everything aligns — a significant edge, strong supporting data, and no red flags in your analysis.
The key discipline is in how often you use each tier. If you are assigning three units to half your bets, you do not have a tiered system — you have a flat system with extra steps and more risk. In my own betting, roughly 60% of my plays are one unit, 30% are two units, and fewer than 10% are three units. If I go an entire card without finding a three-unit play, that is a normal outcome, not a disappointment.
The advantage over flat-percentage staking is capital efficiency. You allocate more to your strongest edges and less to your marginal ones, which should — over a large sample — produce higher returns for the same level of risk. The disadvantage is that it introduces a subjective element. How do you decide whether a bet deserves one unit or two? If you are not careful, confidence ratings become a back door for emotional staking. “I really like this fighter” is not a valid reason for upgrading a one-unit bet to three units.
My criterion is quantitative, not qualitative. If my estimated edge over the bookmaker’s margin-adjusted probability is 3% to 5%, the bet is one unit. If the edge is 6% to 9%, two units. If it exceeds 10%, three units. That framework removes the emotional component and ties the stake directly to the analytical output. It is not perfect — my probability estimates are themselves uncertain — but it is far better than sizing bets by feel.
The Kelly Criterion Applied to MMA: Full Formula and Worked Example
The Kelly Criterion is the most mathematically rigorous staking system available, and it terrifies most bettors the first time they apply it. Developed by John Kelly at Bell Labs in 1956, it calculates the optimal stake to maximise the long-run growth rate of your bankroll, given your estimated edge and the odds offered.
The formula: f = (bp – q) / b
Where f is the fraction of your bankroll to stake, b is the decimal odds minus one (your net profit per pound), p is your estimated probability of winning, and q is the probability of losing (1 – p).
Worked example. You estimate a fighter has a 58% chance of winning, and the bookmaker offers decimal odds of 2.10. Here, b = 2.10 – 1 = 1.10, p = 0.58, and q = 0.42. Plugging in: f = (1.10 x 0.58 – 0.42) / 1.10 = (0.638 – 0.42) / 1.10 = 0.218 / 1.10 = 0.198. The Kelly Criterion says to stake 19.8% of your bankroll on this bet.
And there is the problem. Nearly 20% of your bankroll on a single MMA fight where roughly 44% of bouts go to decision and a flash knockout can end things at any moment. Full Kelly assumes your probability estimates are perfectly accurate, which they never are. It assumes you will place bets indefinitely into the future, which is unrealistic. And it produces a variance profile that most human beings cannot tolerate psychologically. Drawdowns of 50% or more are not just possible under full Kelly — they are expected over a long enough timeline.
Despite those limitations, the Kelly Criterion remains valuable as a theoretical benchmark. It tells you the ceiling of how much you should bet, and anything above that ceiling is provably suboptimal for long-run growth. Even if you never use full Kelly in practice, understanding the formula forces you to quantify two things that most bettors leave vague: your estimated edge and your honest assessment of uncertainty.
The value betting framework I use generates the probability estimates that feed directly into this formula. Without a reliable probability estimate, the Kelly Criterion is just a calculator waiting for inputs that never arrive.
Fractional Kelly: Reducing Variance Without Killing Returns
The solution most professional bettors adopt is fractional Kelly — staking a fixed fraction of what the full Kelly formula recommends. Half Kelly and quarter Kelly are the most common variations.
Using the previous example, full Kelly recommended 19.8% of the bankroll. Half Kelly cuts that to 9.9%. Quarter Kelly brings it down to 4.95%. The return on investment decreases, but the reduction in variance is dramatic. Half Kelly captures roughly 75% of the growth rate of full Kelly while cutting the worst-case drawdown nearly in half. Quarter Kelly captures about 50% of the growth rate but makes the ride dramatically smoother.
I run quarter Kelly for my own MMA betting. The reason is straightforward: my probability estimates are not precise enough to justify full or half Kelly. When I say a fighter has a 58% chance of winning, the honest confidence interval around that estimate might be 52% to 64%. That uncertainty compounds through the Kelly formula, and quarter Kelly provides a buffer against the inevitable occasions when my estimate is closer to 52% than to 58%.
There is a practical bonus to fractional Kelly that the textbooks rarely mention. It keeps your stakes small enough that sportsbooks are less likely to flag you as a sharp bettor. A bettor consistently placing 20% of their bankroll on single fights will attract attention regardless of the platform. A bettor placing 3% to 5% blends in with recreational punters, which extends the lifespan of your accounts — a genuinely material advantage in the UK market where operator tolerance for consistent winners varies widely.
Which Staking Model Fits Your MMA Betting Style?
The right staking model depends on two things: the reliability of your probability estimates and your psychological tolerance for drawdowns. If you cannot estimate fight probabilities with reasonable accuracy — and most bettors in their first year cannot — fixed-percentage staking is the correct choice. It demands nothing from your analytical process except the binary decision of whether to bet or not.
If you have been tracking your bets for at least two hundred wagers and your closing line value is consistently positive, you have evidence that your probability estimates carry information. At that point, unit-based staking lets you begin capitalising on the variation in your edge across different fights.
If you have a multi-year track record, a robust probability model, and the emotional discipline to ride out 30% drawdowns without flinching, fractional Kelly gives you the highest theoretical growth rate. But be honest with yourself about all three conditions. I have met exactly two MMA bettors in twelve years who run even half Kelly without flinching. Both have backgrounds in quantitative finance. For everyone else, quarter Kelly or a disciplined unit system produces better real-world results because it accounts for the gap between theoretical optimality and human psychology.
One final point. Whichever model you choose, commit to it for at least a hundred bets before evaluating. Switching systems after a bad week is the staking equivalent of chasing losses — it feels like action but produces nothing except noise. The system works over the long run or it does not. A sample of ten bets cannot tell you which.
Five Bankroll Mistakes MMA Bettors Repeat
The first mistake is the one I made myself: staking too much, too soon. A new bettor with a small bankroll and a large ego stakes 10% per fight because 1% feels meaningless. Three losses later, they have lost a quarter of their capital and the emotional damage outweighs the financial loss. The fix is simple arithmetic. If you cannot tolerate the idea of losing your stake on a single bet, the stake is too high. Reduce it until a loss feels like a transaction cost, not a crisis.
The second mistake is increasing stakes after a winning streak. This is not the same as the natural stake increase that comes from a growing bankroll under a percentage system. I am talking about the conscious decision to “press your advantage” by bumping from 2% to 5% because you have won four in a row. Winning streaks are not evidence that your edge has increased. They are evidence that variance is currently moving in your direction. The edge is constant — it is defined by your analytical process, not your recent results.
Third: treating the bankroll as fungible with everyday spending. Your betting bankroll is a dedicated pool of capital. If you withdraw money to cover dinner, a holiday, or a phone bill, you are not managing a bankroll — you are maintaining a checking account that occasionally touches a sportsbook. A bankroll only functions as a bankroll if its size is determined exclusively by betting performance.
Fourth: ignoring the bankroll during live betting. Seventy-six percent of young UK bettors aged 18 to 24 use their mobile phones for gambling, and the majority of live UFC bets are placed on mobile during fight broadcasts. The speed and excitement of live betting make it trivially easy to exceed your staking limits. I apply the same unit system to live bets that I use for pre-fight wagers, and I set a hard cap of two live bets per event. Without that cap, the adrenaline of a close fight can turn a disciplined bettor into a reckless one in seconds.
Fifth: never recalculating. Your bankroll changes after every bet, which means your stake should change too. Bettors who set a fixed pound amount at the start of the year and never adjust are effectively overexposing themselves during drawdowns and underexposing during upswings. Recalculate your unit size at least once per event — before the card starts, not between fights when emotions are running.
