If you look closely you’ll observe that the sports betting is actually a very uncertain business by nature. It’s entirely up to the sports bettors if they can handle all the randomness that’s involved in the process. Let’s look at how you can best tackle the unavoidable randomness in the field of sports betting and score consistent profits for yourself.
When it comes to sports betting, the perception that outcomes can sometimes be a result of the application of knowledge and information related to a player or a team, can result in an over-exaggerated sense of self-belief in one’s prediction abilities. As some of the experts have suggested, little information is equal to dangerous information actually!
In addition, as we’re all marred by a self-serving attribution bias, it makes sure that we most likely associate our prediction successes with some internal attributes, believing that we are highly skilled predictors and our skills are solely responsible for making correct calls, while associating failures with nothing but external attributes such as lack of luck.
Regardless of the fact that it doesn’t cater well to our cravings for control, to tell you the truth, just like stock markets and weather forecasting, the business of sports betting is also inherently very uncertain in nature, wherein games evolve in a very chaotic, complex and sometimes even in a completely nondeterministic way if we accept the fact that whatever takes place is largely impacted by quantum world.
Obviously, a large majority of sports bettors are appreciative of this fact, albeit on a bet to bet basis, that luck, whether bad or good, does play a huge part in their losses or gains. However, to what extent do you think chance or probability has an influence over sports bets over longer time periods?
In order to make sure that we don’t get fooled by randomness easily, we can practice a useful exercise involving analysis of the extent of random variability existing in the sports outcomes.
A good way of doing that is by analysing some hypothetical betting returns obtained from fair odds, in order to figure the extent of their variance over certain timescales. In our opinion, betting odds are merely a representation of the probability estimates, as per our expectations.
The wisdom of the betting crowds makes sure that these odds, on average, serve as a highly reliable indicators of the actual probabilities. However, the involvement of randomness results in frequent deviation of the outcomes from the idea of market expectations. Let’s find out how.
Looking at a time series plot of the yield or ROI of moving average level-bets of 10 football matches, for the bets placed on all the away, draw and home outcomes for 10 straight seasons of the English Premier League football matches played between 2005 – 06 and 2014 – 15, we can infer that the time series is actually quite staggered! The betting odds are derived from the actual market’s average match betting odds, after removing the profit margin of the bookmakers such as Bet365 etc.
It is found that unsurprisingly a large majority of punters recognise and accept the fact that over small samples of around 30 bets, any unexpected results will lead to major deviations from 0% expectation.
For instance, any lucky underdog winners are going to push the line well over zero, whereas any excesses in winning favourites delivering relatively smaller returns are going to push the plot below zero mark.
However, some people may nevertheless look surprised when they see the extent of these fluctuations over a range as high as 50%. If you get the returns of 50% post 30 bets, would you attribute that achievement to your skill, or simply shrug it aside as a matter of good luck?
If you were to go below -30% on the other hand, would you be confident enough to attribute it to lack of luck, and expect your performance to slowly regress towards the mean?
A second time series was studied showing the moving average returns of level-bets over 100 football matches. One could see a considerable amount of inherent random variability over 20% range from break-even. 23.5% was the largest deviation observed from 300 bets.
What will be your take after posting such impressive results? Would you consider yourself plain lucky, or go about telling everyone that you finally unearthed the holy grail of making correct sports predictions?!
Thereafter, a third time series plot was studied showing the moving average returns of level bets over 1000 football matches. The extent of inherent random variability continued to be high. Please keep in mind, 1000 football games convert into 3000 bets!
This is pretty large compared to what a large majority of sports bettors may bet over several seasons. Still, even in case of samples as big as this, we witnessed sizeable deviations over a considerable time period.
Looking at all these analyses you may argue that it all sounds academic to you, and no one will place such blanket-bets on all away, home and drawn football games. Yes, you’re right in a way. However, you’re going to witness the same kind of noisy random variability when you pick out a sample from even your own betting series.
What we can infer from this study is that simply because you’ve gotten 20% ROI from 300 bets, or a 4% ROI from 3000 bets, it is no indicator that it was your skill at play. You are most likely to get deceived by your own attribution biases unless you learn about the randomness involved in sports across various timescales. The sharpest sports bettors are able to recognise this and are thus able to separate themselves from the competition. Apart from that, majority of whatever happens in the field of sports betting can be attributed to chance.