True skill in sports betting can be best measured by how closely your theoretical Expected Value correlates to your Return on Investment over a sufficient sample size. Your record over smaller samples or small data can be an indicator of strength of performance depending on how quantifiable the uncertainty of the bets in your portfolio were, but in general, the bigger the sample the better.
For a larger volume of bets, we might want to calculate a probability value (p-value) that tells us how likely a run of results was through chance alone. If we run our numbers through a p-value calculator by inputting our ROI%, average odds taken, and sample of bets (n), it will tell us the 1-in-x likelihood of our results being obtained by chance.
In statistics, a p-value of less than 0.05 is generally accepted as rejecting the null hypothesis, which in our case means it’s extremely unlikely that our sports betting results were due to chance.
Empirical evidence and intuitive data in betting
P-values don't measure the actual skill of a bettor, but what if certain sharp sports bettors can quantify uncertainty accurately? In my book, Hypnotised By Numbers, I talk about many skilled bettors being equipped with something called risk intelligence, and their brains working a certain way. These bettors can subconsciously run iterations in their head based on past empirical evidence accrued, and project or measure probabilities accurately for future events.
This is how the brains of several professional sports bettors are wired. Empirical evidence is information obtained through observation and the documentation of certain behaviours and patterns in data or through experiment. These expert sports bettors store it all up in their brains that work like a computer, processing all of the data. They will have built up countless hours of watching the sports they bet on and will generally specialise in certain sports and markets.
After all, aren’t stored up experiences a form of data? We all base decisions on experience built up over time, largely intuitively.
“Risk intelligence is the ability to measure probabilities accurately through intuition gathered from experience”
In betting, there are known knowns, known unknowns, and unknown unknowns, as Pinnacle’s Marco Blume once said. If a horse runs up the hill at Cheltenham and slows down significantly, or if a UFC fighter gasses a few times in round four of a fight, how big of a sample size is needed to know that this will keep happening? Should we include this variable in our estimation of the true (conditional) probability? In the case of the Cheltenham horse, we could argue that seeing it once is enough.
The logic behind the variables
There is high variance that needs to be ironed out with large samples in truly random events like a coin flip, but sports betting is not truly random and there are many instances where the uncertainty is highly quantifiable over smaller samples. This is what I often call the logic behind the variables.
The key to unlocking the value is watching sports to find angles that are not priced into markets. While watching the games play out, we should see a less volatile profit and loss graph, which stays closer to our theoretical EV line.
Handicapping is the best skill to have in sports betting
Several of the most gifted sports betting handicappers will be anonymous. The size of your stakes doesn’t necessarily reflect your ability as a sports bettor. In poker, many of the best players play low-to-mid-stakes games. True skill in sports betting is measured not by the figure of money you stake or make – it’s all about how much you can beat the odds by, vig included.
I’m not referring to closing line value, the actual true probabilities of all the bets you make over a big sample. Measuring this only comes through your long-term ROI%. The closer your predicted Expected Value (EV) correlates to your actual Return on Investment over the long run, and the higher your Return on Investment is, the bigger skill advantage you have over the bookmakers and the market.
Many of the top sports bettors will be handicappers, and usually the best sports handicappers will compile or estimate their own fair odds for the markets they are interested in. These odds will add up to a true 100% book rather than a false bookmaker percentage with a margin added.
The most highly skilled in the art of sports betting are people who study the sports and betting markets intrinsically and are then able to quantify their edge and spot any inefficiencies in the markets. The above is known as a bottom-up approach to sports betting.
What is Closing Line Value in sports betting?
Closing Line Value or CLV is another way to gauge performance and skill in sports betting, but is it efficient? The answer is that it depends on the sports and the market. We would prefer to have a Return on Investment that correlates to our projected or theoretical EV over a big sample of bets. We would rather achieve a 10% return on investment which equates to our theoretical EV, than a 5% ROI with 15% CLV.
Is Closing Line Value efficient? The answer is that it depends on the sports and the market.
To calculate Closing Line Value, we can use this simple formula from Hypnotised By Numbers:
((Decimal odds taken - closing odds)/closing odds) x 100
Closing Line Value has emerged as an alternative way to calculate and record aperceived longer-term edge, and it’s mainly used by “Steam Chasers” – also known as the top-down approach. Software that compares lines from different sportsbooks allows data-orientated bettors to benefit without necessarily knowing the nuances of the sports.
It’s not a new concept – it’s been going on for many years in European betting markets since the emergence of odds comparison grids/websites and Betfair. Closing Line Value has its merits as an alternative way to track performance, especially in the more popular and liquid markets.
You want to be on the right side of CLV on average over the long term in most markets, but CLV does not always correlate with EV. After building a large sample portfolio of bets, if your actual Return on Investment (ROI) equates to your predicted EV that’s the best benchmark of true skill in betting.
People who predominantly bet on props markets, derivatives or more obscure markets might not get too much Closing Line Value, or if they do, it might be on lines that were easily moved, sometimes due to their own action.
How can we learn to bet smarter?
As a rule, the more accurate you are at estimating true odds, the smaller the sample you will need to discover your true skill. This is true so long as the variables you are weighing and the parameters you are using remain consistent. “Uncertainty” can often be quantified to a degree.
If you are working with longer-odds markets, it will take more time to even out the variance, but there are several things we can do to put the odds in our favour as a sports bettor:
- Efficient risk management and staking strategy
- Work on our discipline
- Target lower odds to reduce variance
- Line shopping
- Networking and scaling up
- Game selection is huge in sports betting
- Document our bets and look for any strange patterns on the spreadsheet or in the gameplay variables themselves.
- Put in a high volume of value bets to lessen volatility and variance in the graph
Managing risk in sports betting
Having a consistent edge on the odds is only half the battle. Risk management is the other big factor to consider, and Kelly staking is key in this area. 30% Kelly is a strong starting point. What the Kelly formula does is mitigate risk in sports betting and it gives us a suggested stake size in proportion to our bankroll that is optimal to exploit the perceived value in the bet.
We can see the Kelly Formula below: (If using 30% Kelly, just adjust to 30% of Fafterthe calculation)
F* = (bp-q /b) where:
- F is the fraction of the current bank roll to bet
- b is the net odds received on the bet ("b” to 1 - if using decimals, 2.50 would be 1.5 to 1)
- p is the probability of winning
- q is the probability of losing, which is 1–p
Final thoughts on performance and true skill in sports betting
Return on Investment is the truest form of measuring skill in sports betting. If our ROI% matches up to our Expected Value over a decent sample of bets and there's a valid reason why we think the odds were wrong, then we can usually be sure that our performance is based on skill.
How much we beat the odds by is the best indicator of our ability as a sports bettor. Uncertainty in sports betting can usually be quantified to a degree, unlike casino games or other games of chance.
If you place positive expectation bets consistently, you'll win over the long term with an efficient staking plan and bankroll management strategy in place. Allow yourself room for error in your tissue pricing by setting a threshold regarding theoretical EV.
For example: if you make something a 2.000 fair price, you could look to take no less than 2.200 which would equate to a 10% EV threshold and allow you that margin of error.
Want to learn more about this and other betting strategies to get the edge in the markets? Head over to our Betting Resources hub. And follow Bryan Nicholson on Twitter for more information on his betting approach.