How to Use the Kelly Criterion in Betting

Ever spotted odds that looked way off, like the bookie must’ve missed something, and wondered how much you should stake? Maybe you’ve been sure a fighter was undervalued or a football team had a better chance than the market was giving them… but then froze at the crucial moment. Too little feels like wasting the opportunity. Too much feels like gambling the whole roll.

Or maybe you’re the type who bets consistently but has no system for deciding how much. $20 on one game, $80 on another—no real reason, just a gut feeling. And when things go sideways, you end up chasing or second-guessing yourself.

That’s the problem the Kelly Criterion solves. It’s a formula—not for picking winners, but for deciding how much to stake when you think you’ve found an edge. It’s designed for people who take betting seriously and want to grow their bankroll over time without blowing it all on one bad call.

We won’t lie. To a novice, this formula and way of betting can be mega confusing, but when used correctly, it can remove the stress of having to manage your bankroll by taking emotional decisions out of your hands.

What is the Kelly Criterion?

The Kelly Criterion is a formula used to calculate the ideal percentage of your bankroll to stake on a bet when you believe you have an advantage. It’s designed to grow your bankroll efficiently over time by adjusting your bet size to reflect how strong that advantage is.

If you bet too little, you limit your profits. If you bet too much, you risk wiping out your bankroll during a losing streak. The Kelly Criterion aims to strike the right balance—big enough to benefit from a real edge, but conservative enough to avoid unnecessary risk.

The formula was introduced in 1956 by John Kelly Jr., a researcher at Bell Labs. Although it was created for communication theory, it was later adapted by gamblers and investors. Ed Thorp famously used it to size blackjack bets when the deck favored the player, and similar logic is now used by professional bettors and traders who want to manage their capital more precisely.

Why use the Kelly Criterion?

People make mistakes when betting on sports. Most people bet too much when they’re confident, too little when they’re unsure, and often adjust stakes based on emotion or how the last bet went. Over time, that leads to inconsistent results, unnecessary losses, or missed opportunities. The Kelly Criterion cuts through all of that by giving you a structured way to bet smarter.

• It grows your bankroll faster than flat betting—when used correctly.
• It protects you from overbetting and going bust.
• It scales automatically with your wins and losses.
• It forces you to bet based on actual edges, not emotions.
• It works across live betting, sports, casino games (with an edge), and trading.

What is the Kelly Criterion formula?

Ok, let’s look at the formula behind this highly technical model. Stick with us for this bit, we’ll get to some examples that make this much more understandable soon.

For a simple win-or-lose bet, the formula for the optimal fraction of your bankroll to wager is:

f*=p-qb

• f* is the fraction of your current bankroll to bet (the Kelly fraction).
• p is the probability of winning (your estimated chance of success for the bet).
• q is the probability of losing, which is simply 1 – p.
• b is the net odds multiple – how much you profit relative to your bet. Specifically, it’s the net profit you earn for every dollar you stake.

For example:

• If a bet pays even money (1-to-1 odds), you double your money on a win, so b = 1.
• If a bet pays 2-to-1 (you win $2 profit for every $1 bet), then b = 2.
• In decimal odds, b = decimal odds – 1.

This formula assumes that if you lose, you lose your entire stake, and if you win, you gain b times the stake. The result, f*, is the proportion of your bankroll to wager on that bet.

For example:

• A positive f* of 0.20 means you should bet 20% of your bankroll.
• An f* of 0 (or a negative number) means you should not bet.

In fact, a negative Kelly result theoretically means the other side of the bet has the edge, so if it were possible, you’d bet against your initial pick.

Example: Sports bet with an edge

Suppose you’re betting on a football game and you think Team A has a 45% chance of winning. The sportsbook is offering odds of 2.50, which implies just a 40% chance of winning (because 1 divided by 2.50 = 0.40). That means you believe you’ve found a small edge—your estimate is higher than the bookie’s.

Now let’s work out how much to bet using the Kelly formula.

• First, calculate the net odds multiple. With decimal odds of 2.50, you’d make $1.50 profit for every $1 you bet. So b = 1.5.
• You’ve estimated p = 0.45, which makes q = 0.55 (because 1 – 0.45 = 0.55).

Here’s the formula:

f* = p – (q / b)

f* = 0.45 – (0.55 / 1.5)

f* = 0.45 – 0.3667

f* = 0.0833

That means you should bet about 8.33% of your bankroll.

So, if your total bankroll is $1,000, Kelly suggests betting around $83 on this game.

Even though your edge is small, it’s still worth a calculated bet. Kelly keeps the stake modest, because there’s still a decent chance you’re wrong. And if your edge was bigger, the formula would recommend a larger bet. If there was no edge—or if the sportsbook’s odds were better than your estimate—Kelly would tell you not to bet at all.

How to use the Kelly Criterion formula for betting (step-by-step)

You’re probably still a bit bamboozled by all of these formulas, numbers and examples. That’s understandable. Most people don’t grasp Kelly Criterion straight away. So, let’s break it down step-by-step, so you can easily apply it to your bankroll and any of your bets.

Step 1: Estimate your chance of winning

Use whatever research, stats, or gut feeling you trust to work out how likely your bet is to win. Let’s say you think your team has a 60% chance of winning—that means p = 0.60.

Then work out the chance of losing: q = 1 – p. In this case, q = 0.40.

Step 2: Work out the profit you’ll get from the odds

If you’re looking at decimal odds, subtract 1 from the number to get your net profit per dollar.

So if the odds are 2.00, then b = 2.00 – 1 = 1.

If the odds are 3.50, then b = 2.50.

(For fractional or American odds, you’ll need to convert them to decimal first.)

Step 3: Plug everything into the formula

f* = 0.60 – (0.40 / 1)

f* = 0.60 – 0.40

f* = 0.20

That tells you to bet 20% of your bankroll on this bet.

Step 4: Work out your stake

Take the Kelly result and multiply it by your current bankroll.

If you’ve got $500 and f* = 0.20, then:

$500 × 0.20 = $100

So you’d stake $100 on this bet.

Step 5: Recalculate after every bet

Once the bet’s settled, update your bankroll and run the numbers again for your next bet. If your bankroll grows, your stakes get a bit bigger. If it drops, the formula tells you to scale down. That’s what makes Kelly so useful because it adjusts to how you’re doing and keeps you betting in proportion to your advantage.

Where the Kelly Criterion fits in sports betting

Esports, sports and live betting are a few of the few places where casual bettors can get an edge—if they’re switched on. Kelly helps you take that edge and turn it into structured, long-term bankroll growth. However, it’s not for everyone and not suitable for all sports betting situations.

When Kelly Criterion works well in sport betting:

• You’re betting on value, not vibes. You think the true odds are better than the bookie’s.
• You’ve done research, looked at stats, line history, and matchups, and have a reason for your estimate.
• You bet regularly and want your bankroll to grow over time, not just ride hot streaks.
• You want to bet more when your edge is bigger, and less when it’s slim.

When Kelly Criterion doesn’t work in sports betting:

• You’re betting for fun or emotionally (e.g. always backing your team).
• You’re making random or one-off bets without tracking long-term results.
• You’re copying someone else’s picks without understanding the reasoning.
• You don’t have a reliable way of estimating win probabilities.

Basically, Kelly is your friend when you treat betting like a skill game. If you’re just in it for the occasional Saturday flutter, you’ll probably want to keep things simpler.

Where the Kelly Criterion fits in online casino games

Most casino games aren’t suitable for Kelly betting because the house edge is fixed and you can’t gain a mathematical advantage.. But in specific edge-based strategies, it becomes a powerful bankroll tool.

When Kelly can be useful for online casino games:

• You’re counting cards in blackjack and the deck is in your favor.
• You’re playing video poker with a perfect strategy or a promo that boosts the RTP above 100%.
• You’ve found a genuine advantage play, like a loophole in a bonus or a short-term player edge.
• You want to scale your bets sensibly while playing with a small but real advantage.

When Kelly Criterion doesn’t help you out:

• You’re playing games with a fixed house edge, like roulette, slots, or baccarat.
• You’re trying to “out-stake” the odds in negative expectation games.
• You don’t actually know if you have an edge. You’re just guessing.
• You’re playing for entertainment but trying to apply investment logic.

Limitations of the Kelly Criterion

Kelly has a solid track record and strong logic behind it, but it’s not perfect. There are a few things to watch out for before you start betting like a mathematician.

You need to know your win probability (and get it right)

Kelly only works if your estimated chance of winning is accurate. But in the real world, we rarely know the exact odds of something happening. If you’re way off with your estimate, Kelly will have you betting too much—or too little. That’s where people get burned. It’s often said that Kelly is only as good as your guess. If your guess is bad, your bet size will be too.

It doesn’t care about your personal risk tolerance

Kelly is focused on maximizing long-term bankroll growth. It doesn’t care how stressful the short-term might feel. If the math says bet 30%, it’ll say bet 30%—even if that gives you heart palpitations. If you face a couple of losses early on, your bankroll will take a hit.

That’s why most bettors don’t use the full Kelly figure. They scale it down, using half-Kelly or even a quarter, to smooth out volatility and avoid those gut-punch losing streaks. If you’re the kind of bettor who values consistent returns and a manageable ride, full Kelly can feel like too much. And that’s fine. Betting should fit your risk tolerance, not the formula.

It’s built for the long game

Kelly is designed to work over hundreds or thousands of bets. If you’re only betting occasionally or just dipping in and out, the long-term benefits won’t mean much. In the short term, luck plays a bigger role, and small sample sizes can throw everything off. In that case, playing it safer might make more sense.

It gets tricky if you’ve got lots of bets at once

The standard Kelly formula assumes one bet at a time, with each result affecting your bankroll before the next one. But if you’ve got multiple bets running at once (like on a busy football weekend), it gets more complicated. You can’t just apply the formula to each bet individually—doing that can overexpose your bankroll without you realizing how much total risk you’ve taken on.

Going over Kelly is a bad idea

One final point—never bet more than what Kelly recommends. It’s not a target to exceed. Instead, it’s a ceiling. Betting beyond it will increase your risk without improving your returns, and in the long run, it can wreck your bankroll completely. Even pros who use leverage or big bankrolls have gone bust by ignoring this.

Put the Kelly Criterion into action with Cloudbet

Currently, we’ve got a sweet $2500 Welcome Package to boost your bankroll and make those Kelly Criterion calculations a bit sweeter. Sign up, place a bet in the sportsbook or casino, and you’ll start receiving 10% cashback on all bets and 30 days of daily welcome rewards—it’s that easy.