Where does luck meet talent?
Jun 20th, 2011 - Posted by Andrew_Brocker in Betting Theory
bettingexpert blog editor. Always taking the alternative route to finding the value.
How do we know if a sports betting tipster is either lucky or has genuine ability? Andrew takes a look at some numbers as he attempts to discover at what point we can have confidence in a tipsters ability to find us long-term value.
Anyone who has bet long enough knows that losing streaks are all part of the bumpy terrain of sports betting. No matter how good you are, you're bound to be confronted by periods where nothing seems to fall your way. Even if my bets hit at a winning rate of 90% over thousands of tips (I wish), any one span of 10 tips, or 50 tips, or even 100 tips is vulnerable to a great degree of variance. And when we are talking about the more realistic goal of averaging a long-term return of between 3 and 6%, the subtle and very natural turbulence that will occur over shorter spans of time is something that even the most proficient sports bettors as susceptible to.
This subtle turbulence is something that anyone getting involved in the world of sports betting needs to have an appreciation of, whether you be a tipster or a follower of tips. In my days as a tipster on BettingExpert, like everyone else, I was not immune from the hostility of luck. I remember after having a poor weekend with my football tips, receiving a comment from a user telling me quite bluntly that I should quit tipping. Obviously this wasn't someone with an understanding of how over a short period, results can sway quite violently. And likely it was someone who would be vulnerable to those in the betting tips industry who prey on naive and wishful thinkers.
Every winning pick wins
If you've done any amount of exploration of the internet based betting and tip service industry, you've no doubt come across what I like to call 'Streak Cooks'. And if you haven't seen them, you'll no doubt know when you do. These are the sorts of services that promote 'winning' or 'hot' streaks, blaring figures like “13-3 winning streak” or “10-2 last 7 days”. For anyone who understands just how broad the distribution can be within such small samples for even the most prolific tipsters, these so-called winning streaks suggest nothing more than a fortunate run. And you have to ask yourself, if these services are so successful, why aren't they promoting their long-term results? Well, we could be cynical enough to offer two explanations for this. One, their long-term results are not worth promoting. Or two, steady long-term success with modest average returns just arent sexy. In other words, they don't sell.
Other services go to even more ludicrous extremes to promote their services. You'll often see tip salesmen promote their results given all sorts of combinations of conditions. They slice up their picks from all sorts of angles and then present the numbers as if to suggest............well to be honest, I'm not quite sure what it's supposed to suggest, success I guess. You'll see them promote nonsense like “Our Western Conference NBA picks are 65% when the line is between 1 and 4 points for the home team.” You think I'm exaggerating? I'm not. It's pure nonsense. They might as well be promoting their tips as “All our winning tips are 100% this season”. Congratulations Einstein.
Let's flip a coin
So how many tips do we need to see before we can make a fair assessment on a tipsters ability to find us profit? Without getting too bogged down in probability theory, let's look at the likelihood of different outcomes given a variety of sample sizes. To do this we will use the example of tossing a coin. We will assume that the probability of the coin landing head or tails is 50% either way.
Firstly we will look at the probability that you will achieve a particular outcome over 30 tosses. Here we can see that it is roughly a 42% chance that over 30 tosses the coin will land on either heads or tails 16 or more times. More to the point, we can see that it is around an 18% chance that you will hit at least 60% winners when tossing a coin (at least 18 out of 30). Or further, there is a 5% chance of hitting at least 67% winners (at least 20 out of 30). Keep this in mind when you see tip services promoting these sorts of winning streaks over such a small sample. Given enough tips, the chances are fairly good that sooner or later you'll hit one of these so called 'hot streaks' even if just making selections randomly.
The reality becomes a little more clear when we look at a sample size of 100 coin tosses. Here we can see that the chances of you hitting a minimum of 60% winners drops to around 3%. This however, is still quite high if you consider we are just picking heads or tails randomly. It means that for about every 30 '100 toss' sample groups, we can expect to hit a 60% streak once. So even if you see a tipping service promoting this sort of streak over a 100 or tips, approach them with a little scepticism.
So let's finally look at a sample size of 1000 coin tosses. Here we can see that there is around only a 3% chance of at least hitting merely a 53% strike rate. And as for hitting a 60% strike rate of 1000 coin tosses, forget about it. It just isn't going to happen.
So what's a good sample?
So in conclusion, let's look at the probability of hitting at least a modest 55% winners when tossing a coin over a variety of sample sizes. We will choose 55% as winning at this rate when betting at bookmaker 'even odds' of traditionally 1.92 for a 50-50 'coin toss' bet, would give us an average return of between 5 and 6%, a return that any talented tipster and sports bettor would hope to achieve in the long-term.
Here we can see that the bottom falls out at around 500 coin tosses with the likelihood of hitting 55% winners when randomly selecting which side a coin may fall becoming virtually impossible beyond 700 tosses.
We could then conclude that a sample size of between 500 and 700 tips would be the minimum number we should take into consideration before assessing whether or not a particular tipster or tipping service is blessed with either great luck or great talent as a sports betting tipster.
If you liked this article you might also like:
How to convert odds to their implied probability
How do you find betting value?
How the size of your bookmakers commission impacts your betting bank
What's the best way to manage your betting bankroll?
How does randomness impact upon our betting?
Related posts
-
May 18th, 2013Why Does Pinnacle Sports Welcome Winners?
Why does Pinnacle Sports have a policy of welcoming winning punters and how...
Full blog post -
May 16th, 2013Talking Betting With......Rory Jiwani
What's it like working behind the scenes at a popular bookmaker? How do boo...
Full blog post -
May 15th, 2013How To Avoid Betting Account Restrictions
What can you do if your betting account is restricted? How can you avoid ha...
Full blog post -
May 9th, 2013Bookmaker Terms & Conditions: As Hard To Read As Nietzsche
The average bettingexpert user is active at over 4 bookmakers. This means t...
Full blog post
You must be logged in to post a comment! Sign up + or log in in the top right corner.
Hmmm it might, but in general no. Given you've made 237 bets and have a very good return, it definately suggests that you have betting talent. The thrust of the article is more about reinforcing the idea that we need a large sample size before we can make any real assessment of whether someone is just lucky or not. But anyone who has made over 200 bets and is making a solid profit, luck or no luck, those numbers suggest there is some talent.
@ Andrew - thanks for the feedback, I always appreciate feedback from someone who's opinion I value alot. Would the fact that 90% of my picks are football and there are 3 possible outcomes unlike the tennis or basketball, would this change some of the figures provided above for the better........?
Thanks for your comments Game. Now as for whether you're lucky, let's have a look at your numbers. Ok, youre making an 8.8% return over 237 bets. If we were just betting randomly on 50-50 outcomes like a coin toss, and getting 'even money' odds of 1.92, you would have to win 56.5% of the time to achieve a return of around 8.8%. So that's 134 wins out of 237 bets. There would be roughly a 2.5% chance of achieving at least 134 wins out of 237 coin tosses. So what does this mean? Well it means you're either very very lucky or you have genuine betting skill. It will be interesting to see your numbers again when you have reached the 500 bet mark. Keep up the good work.
@ Andrew - brilliant again as per normal, another quality article to get us thinking. I am really enjoying your work and just wanted your opinion on my own personal record: Bets. 237 Stake. 1850 Average stake. 7.8 Average odds 2.05 Profit. +163.2 Yield. + 8.8% Would you put that down to luck or talent, please be honest even if you do think it's luck.........?
Thanks Buffalo. Look forward to seeing you back tipping in the new season. And yes, be sure to check out the blog during the off-season, there will be plenty of football betting analysis in the coming months.
awesome once again andrew, your efforts never cease to amaze me, im having a little break from the tipping till the new season european leagues commence, but i always check in to see what your up to, thanks for your great informative articles, they are always well thought out and a great benefit to us punters.
| 
|
Thank you sir.....