The Kelly Criterion is one of the most common concepts in sports betting. Someone calls it simply the optimal way to manage a bankroll, but there are also those who criticize this method – we will try to dot the i’s in this matter in our material.
What is the Kelly criterion?
American financial expert John Kelly developed his criterion back in 1956, working at the AT&T Bell Laboratory. Later, the Kelly criterion became more and more popular in the betting industry and in the financial sector, and now it is widely used by both bettors and traders, who often call it simply “Kelly”.
In general, the Kelly criterion is a formula that calculates the proportion of available funds that should be risked to maximize the potential return on a bet or investment. In other words, it takes into account how much money needs to be wagered, what is the likelihood that a bet will win and how likely it is that it will lose, and whether bets can be placed based on this.
While this is just one of many tried and tested betting methods, the Kelly Criterion is considered the best because it protects your bankroll and at the same time ensures that you bet in proportion to your positive expected value you have above the market.
How does the Kelly test work?
Betting strategies can be very different – from simple methods with flat or fixed rates (for example, bet the same amount each time) to sequential methods such as the Martingale method or catch-up (doubling the bet after a loss) and Fibonacci ( move up one step in the sequence of numbers after a loss and down two steps after a win).
However, the Kelly criterion differs from all the common betting methods listed above in that it is proportional. The amount you bet using the Kelly Criterion is always proportional to your bankroll in relation to the edge you own.
The main principles of the Kelly criterion are that if you lose your bankroll will never run out, and if you win, your funds will multiply exponentially. If you have a series of losses, the recommended bet amount will decrease to match your bankroll. Conversely, if your bets are profitable and lead to an increase in your bankroll, your further bets will also increase.
Calculation of the Kelly criterion
At first glance, the Kelly criterion formula may seem confusing to some, but if you break it down into its components, then it is very easy to understand and apply to your own rates.
F = (BP – Q) / B
F is the amount you need to wager (a fraction of your bankroll).
B – decimal odds for your implied bet – 1.
P is the probability of winning (calculated by you).
Q is the probability of losing (1 – P).
Next, let’s look at a practical example to clarify this formula. Imagine Roger Federer playing Rafael Nadal in the Wimbledon final. Federer has a winning odds of 1.598, while Nadal has 2.490. According to these odds, Nadal has a roughly 40% chance of winning, but you think his chance of winning is 48%.
Calculation of the Kelly criterion in this case looks like this:
B is the decimal odds for your implied bet (2.49) – 1 = 1.49.
P – probability of winning (calculated by you) = 0.48.
Q – probability of losing (1 – 0.48) = 0.52.
F = (1.49 × 0.48 – 0.52) / 1.49 = 0.13
As you can see, in the above example, the Kelly criterion suggests that you should bet 13% of your bankroll that Rafael Nadal will beat Roger Federer.
It is of course important to understand how to calculate the stake based on the Kelly Criterion formula, but you can use tools such as Excel, or the free Kelly Criterion Calculators, which are available on the Internet to automate this process.
Criticism of the Kelly criterion
The most common criticism of the Kelly criterion in the context of betting is that it does not account for market volatility and the effect that variance can have on results, which is a big disadvantage.
In practice, this means that you can build your bankroll with a few small bets at high odds when the edge you have seems small. However, if your model allows you to gain a large advantage in the market with small odds, your bankroll building efforts can be overwhelmed overnight if you lose the bet.
A lot of research has been done on this topic, and the solution seems to be quite simple – a fractional version of the Kelly criterion. Players can now apply a bankroll management strategy based on 1/2, 1/4 or 1/8 Kelly Criterion, using the same fraction sequentially throughout the method.
This in practice means the following: if the Kelly criterion indicates the need to put 10% of your bankroll, then using 1/2 Kelly criterion it will be 5%, 1/4 – 2.5%, and 1/8 – 1 , 25%, etc.
Another common complaint about the Kelly Criterion concerns how to manage multiple benefits while betting simultaneously. There is a potential scenario where a player finds an advantage over Team A versus Team B, while having an advantage over Team C versus Team D, and both events are taking place at the same time. In addition, using the money line market or the long list futures market as an example, it is possible that two, three, or even more outcomes of the combined bets will give the player an edge over the market
In each of the scenarios described, this popular claim to the Kelly criterion is challenged again. Depending on the number of concurrent events and the size of the perceived edge, using the Kelly criterion, a bet may be offered that will negate most of the bankroll. In some extreme cases, this can lead to an offer of a bet amount that even exceeds the current bankroll, and then all the meaning of the method is completely lost.
You also need to consider whether the Kelly Criterion is the appropriate method for calculating bets based on what kind of player you are. If you are disciplined, willing to expand your advantage and take the time it takes to build your bankroll, then the Kelly Criterion is probably your option. However, if you are placing your bets just for fun, or if your betting process is not carefully calculated, stick to relatively small bets out of your bankroll.
The main takeaway for players using the Kelly Criterion or its fractional version is that this method is based on your calculations of the probability of a particular outcome. Optimizing your bankroll management in relation to your edge is undoubtedly a good thing, but it also takes an effort to ensure that you truly have an edge in the betting market.
In any case, a lot of work needs to be done to formulate more accurate probability of outcomes than those available in the betting market, but you also need to devote time to perfecting your model and constant testing to eliminate the influence of the factor of luck and randomness on any positive results of work.