It’s time to take a deep dive on another famous gambling strategy: The Kelly Criterion.
We’ve investigated plenty of gambling strategies named after their inventors—like Martingale and Fibonacci—and shown that they don’t really work the way you might think. But the Kelly Criterion, known as Kelly Betting, is a bit different than most. Developed by Bell Labs researcher John Larry Kelly Jr. in 1956, it’s built on solid, repeatable math. A mathematician wrote a Harvard paper on it, and it’s gone on to become popular not only with bettors, who use it both to manage their bankrolls and figure out wager amounts, but also with many financial investors, who use Kelly’s formula to try to encourage steady, reliable growth in their portfolios.
So the Kelly Criterion is a legitimate method of working with odds. But it does have its fair share of risks, relying on players to do some very real math and analysis. And even then, there’s no guarantee of success.
What is the Kelly Criterion?
There’s no way to sugarcoat it: The Kelly Criterion is math.
More accurately, it’s a formula. Plug in a few numbers based on a specific bet, and it will tell you what percentage of your bankroll you should spend (if any at all) in order to win more than you lose over time.
At its simplest, the Kelly Criterion formula looks like this:
((Odds – 1) x Probability of winning) – Probability of losing = Suggested wager
Odds – 1
The “odds” value is the decimal odds for a given wager, supplied by the sportsbook or casino. For example, an NHL game between the Oilers and the Flames might have odds of 3.0 that the Oilers will win, meaning a winning $1 bet would be multiplied by 3.0 and return $3.
Next you need to figure out the probability of winning the bet. And for this, you’ll likely need to do your own analysis to identify bets where you may have what Kelly advocates call an “edge.”
An example: From what you’ve researched on the Oilers, you have a hunch that the sportsbook is undervaluing it. Based on its odds, the sportsbook thinks the Oilers have a 33% chance to win, but you’ve noticed the Oilers have won half their games against the Flames this season and think it’s closer to 50%. That translates to a probability of winning of 0.5. That’s the number you’ll use for the formula. It also means the probability of losing is 50%, or 0.5.
We now have all the numbers we need. Time to plug them into the formula and do the math.
((3.0 – 1) x 0.5) – 0.5 = Suggested wager
3.0 – 1
((2.0) x 0.5) – 0.5 = Suggested wager
2.0
1.0 – 0.5 = Suggested wager
2.0
0.5 = Suggested wager
2.0
0.25 = Suggested wager
Your suggested wager for this particular game is 0.25 (or 25%) of your bankroll. If your sports betting bankroll for the month is $100, that means your optimal bet is $25. According to the Kelly Criterion, betting more would be too risky, betting less would mean you’re not taking advantage of your so-called edge. If the suggested wager turns out to be a negative number, Kelly would counsel avoiding the bet altogether.
If you follow this method for all of your wagers, the Kelly Criterion says that, over the long run, you’ll win more than you lose and end up in the black.
Does the Kelly Criterion work?
Mathematicians have proven that the Kelly Criterion is sound in theory. No one will argue against that.
But here’s the thing: It only works if your numbers are accurate and your environment is controlled.
Given enough time, it will always work in a controlled setting in which the probability of winning is fixed and provable. But probabilities in gambling are, at best, educated guesses. Highly unlikely events occur all the time in gaming. That’s why there’s no such thing as a “sure bet” in sports, or casino games.
What’s more, the Kelly Criterion tends toward aggressive betting. In our example, your bankroll could disappear quickly if you lose a few times in a row, leaving you on the sidelines. There’s no fun in that.
Some bettors use variations on the Kelly Criterion that involve reducing the suggested wager by half or even a quarter. But this isn’t really a solution. It may make your bankroll last a bit longer, but winning over time still depends heavily on the accuracy of your probability numbers.
Should I use the Kelly Criterion?
Each suggested bet recommended by the Kelly Criterion serves as a percentage of your bankroll. Once your wagers add up to 100% of your bankroll, you’re finished. Working within a set budget and sticking to it are keys to playing smart, so there’s no harm using Kelly Betting like this.
That said, there’s also no real benefit. It won’t increase your odds of winning, and there are easier ways to work within a budget. Plan a set amount you can afford to spend on bets in a given week or month, and call it quits if you hit it. If you need more structure, PlaySmart tools can help. Deposit Limit, for example, is an easy way to create, monitor, and stay within your budget.
In the end, it comes down to maximizing your fun. If you enjoy math and numbers, the Kelly Criterion could be a satisfying way to calculate and manage spending within your budget. If, on the other hand, you’d rather not bust out an equation each time you make a wager, you might have more fun choosing your bets the way most people have since the dawn of gambling: by choosing what feels right to you and enjoying the excitement of watching the results unfold.
