Verification of Sports Betting Tipsters & their TipIQs

A good tipster will never be afraid to show his betting record. When a tipster proofs his history of bets made, he provides a quality assurance to a prospective customer that his tips are reliable and verifiable. If they don't, a punter should always seek to find out why not if they are looking to follow their tips. Without a proper tipster-punter relationship of trust, the punter can never be sure that what they are getting (or buying) will be reliable.

Sports-Tipsters has proofed over 250 separate online sports advisory services and currently monitors around 35 or so of them. The tipsters league page summarises the outright records together in the form of a league table (there are other leagues for 3-month, 6-month and 12-month performance), whilst analysis of specific betting records and information about each tipster can be accessed by following the tipster links to the left. The remaining material on this page provides further information on how Sports-Tipsters analyses each betting record.

There are a number of indicators which can be used to assess the success of a particular betting record. Traditionally, the size of the betting bank has been viewed as the main factor which indicates the worth of a betting system. The value of the betting bank alone, however, provides little information on how many bets have to be placed and how much money has to be staked in order to achieve the profits shown, and how long such profit growth will take.

A significant piece of information that a punter should have when considering whether to follow a particular betting advisory servive is the profit over turnover (or yield). Profit over turnover or yield is simply the ratio of profit made to total stakes wagered. Of course a successful betting system allows one to recycle money already won into further bets so that little new money is actually risked. Nevertheless, the betting yield is considered to be a more reliable indicator of success, since any value of the percentage increase in betting bank will be dependent upon the initial bank, and the risk level of the betting system and the size of stakes placed as a percentage of betting bank.

Yield on its own, however, is not the whole story. How, for example, do we compare a record of 10 bets with a 50% yield to a record of 10,000 bets with a 5% yield? Clearly, the longer a record remains in profit the better it is. A simple way of comparing two records like this is to multiply the number of bets by the yield, to get a kind of profitability rating. Such a rating provides a measure of how much profit a tipster has been able to deliver in absolute terms, although it is not the same as actual profit or an actual bankroll size, since staking strategies are different and people's bankrolls are variable. It does, however, provide a reasonable relative assessment of one record against another. Until the end of 2013 it was Sports-Tipsters' chosen rmethod of rating and ranking verified tipsters. Unfortunately, there is an inherent weakness in rating the quality of a tipster in this way. Most importantly, it doesn't take into account the role of chance, that is whether a tipster has generated a profit purely by being lucky or by some other means (hopefully skill). For this reason, beginning 2014, Sports-Tipsters has introduced its new rating scheme: the TipIQ

How can a punter be sure that a tipster's betting record has not arisen through luck? It's a bit like asking how many times we must flip a coin we are told might be weighted before we can be sure that the coin really is weighted. If we flip it 3 times and it land heads each time we aren't going to learn a great deal. If we flip it 300 times and return 200 heads we might start getting a bit more interested. Similarly, a betting yield of 50% from just 10 tips is more likely to be the result of chance; a 5% yield from 10,000 picks, on the other hand, more probably reflects the influence of skill, simply because the record is so much longer.

Of course, one can never be completely (100%) certain that chance plays no part at all, but there are statistical techniques available to test a tipster's skill at making a profit. My book How to Find a Black Cat in a Coal Cellar: The Truth about Sports Tipsters sets out in detail how to analyse a betting record and how to test for the influeces of chance versus skill. Sports-Tipsters has now introduced this statistical technique to estimate the potential skill levels of each tipster. To be clear, really what this technique is doing is estimating the probability that a profitable betting history could have occurred just by random chance only in the absence of any other cause, it actuallty say nothing about the skill level of the tipster. Nevertheless, it is still reasonable to conclude that the smaller the probability a profit has arisen only by chance, the greater the likelihood that the tipster's skill is the causal factor.

So what is the TipIQ? Well, a TipIQ of 10, for example, says that there's a 10/1 chance (about 9%) that such a profitable record (or better) could arise purely by chance in the absese of any other cause (like skill). For a TipIQ of 100, the odds are 100/1 (about 1% probability). The lowest possible TipIQ a profitable tipster can have is 1, in other words an even-money (1/1) probability that his profit happened by chance. For simplicity, where a tipster's history is showing losses, his TipIQ has been assigned a value of 0. Implicit within the calculation of the TipIQ, it is assumed that anyone with no skill at all at forecasting can on average (that is to say over the long term) break even by using an odds comparison to find 100% payout value.

What, then, is a good TipIQ? Well, here a subjective judgment has to be made, but that's no different to using any other measure to determine whether we want to follow a tipster. Traditionally, statisticians don't really start ruling out the influence of random chance until there's only a 5% or even a 1% probability that something being measured could have been caused by it. This equates to a TipIQ of about 20 and 100 respectively. Sports-Tipsters considers a TipIQ of 100 to be a benchmark figure. That's not to say that a lower figure is evidence of little or no skill, and certainly over shorter time frames where there have been fewer bets advised (for example 3 or 6 months) it will naturally be less likely to see higher TipIQs. Similarly, values over 100 are equally not conclusive evidence of skill at work. Remember, all the TipIQ is telling us is how likely something will have happened by chance. People who win the lottery, for example, do so at odds of about 14 million to 1, and there's nothing but chance going on there. What should be clear from the Tipstrs league tables, however, is that there aren't very many tipsters, in Sports-Tipsters' opininon at least, who can rightfully claim to have demonstrated a significant measure of ability beyond what could reasonably have been expected to happen by chance alone, even over the long term.

The TipIQ is a function not just of the number of bets and the betting yield as one would expect (the longer a profitable history is the more confident we can be that it hasn't just happened by chance - recall the coin flipping example), it is also a function of a tipster's typical betting odds. The beauty of the TipIQ is that it allows us to compare similarly profitable tipping histories, with similar yields, but which have arisen by different betting odds. For example, two tipsters with the same yield from the same number of bets will not exhibit the same TipIQ if they are betting different odds. The one with longer odds will potentially be less skilled than the one with the shorter prices, because of the greater inherent randomness in betting lower-probability (higher odds) propositions. Conversely, when we have two equally skilled tipsters at work, we would typically expect the one betting longer odds to have a better yield. This is simply the consequence of taking on more risk to deliver a bigger reward.

Undeniably there will be those who disagree with the new method of tipster rating and ranking, in particular those tipsters who have long but very weakly profitbale records. Yes, it might be argued that such tipsters have nevertheless delivered more profit. However, their TipIQ makes it much clearer that this profit was more probably the result of good fortune, and will less likely prove to be sustainable over the even longer term. With that in mind, however, users of Sports-Tipsters should appreciate that the reliabiliy of the TipIQ is lower for smaller betting histories, and in particular samples smaller than about 30 or so tips.

Punters who are interested in finding out more about how to assess the difference between chance and skill, and get behind the vacuous proclaimations some tipsters make about their winning runs, are recommended to read the Truth About Sports Tipsters.