Pricing Loopholes

A couple of corrections to the FTL table from yesterday. Peter Nordsted* pointed out that I had recorded his Drawmaster* selection of Genoa* v Udinese* as a loss, and indeed I had, despite the game finishing 3:3, and despite it being one of my own XX Draws, which was a little embarrassing. Pete clearly fluked a winner, whereas my selection of the match as a draw was the result of deep data analysis.

The second correction was that a late winning selection from Jamie A was missed altogether. Jamie sends his selections in at different times, which not only complicates record keeping, but has also exposed a loophole that I will address later in this post. Here's the mid-section of the table after the corrections:

Steve M, of Daily25*, asked on Twitter*:

“Do we need a standard in place on how Tipsters record their odds? I think so, what do you think?

It's an important question, and one I spent a lot of time pondering over the summer. It's always a problem for a number of reasons, and unfortunately the inherent opaqueness provides a screen for the more dubious of characters to hind behind.

To me, there are two main issues. The first is with the recorded price and the second is with the availability of that price.

The price on a tip is not fixed, whether it is a share tip, an ante-post tip on the 2018 World Cup*, the next Presidential Election* or an upcoming sports event, so timing is key. After the tip is released, but before an event goes in-play, the price fluctuates as new information and money comes in, so at what point should the price be recorded?

At, or close to, the time the tip is made would be fair to the tipster, but what makes a claimed price valid? To my mind it needs to be commonly available, and for a period of time.

By commonly available, I mean a price that someone serious about betting can actually obtain. One trick some tipsters use is to send out tips with the top price they can find. When this price is ‘offered’ by a book known for restricting and closing accounts, e.g. BetVictor / Stan James, anyone serious about their investing is unlikely to have access to it. The exchanges arguably offer the truest price on liquid markets, but here the issue is often that the amount available is seriously limited.

The other logical time for recording the price is at kick-off or tip-off or SP, but this penalises any tipsters whose edge is time-sensitive, i.e. they have access to inside information which later becomes public knowledge. Not so much of an issue with football as with other sports perhaps, but certainly some tipsters would be impacted.

One might also expect any tipster worth his salt to state what they feel is the correct price on an outcome. Most don’t appear to do this. If any reasoning is offered at all, it tends to be subjective but if, over a period of time, backing the selections is profitable, they clearly have an edge. (Whether or not that edge will last for too long is another thing).

With prices likely to change after a tip is sent out, especially from some of the heavier hitting and successful services, it really is important to know at what price there is no value. It appears that some people plan their social lives around their chosen service’s emails, but if you are unable to act on the tip until several hours later, is the reduced price now available still value? Saying “I believe United will win” is one thing, saying “I believe United offers value at 2.5” is another, but saying “I have United at 2.2” is quite another.

For my Friendly Tipster League (FTL), everyone is measured against the price as recorded for Pinnacle in the Football Data web site*. It’s a standard, a benchmark, and for comparison purposes it does the job, but anyone tipping a team on a Wednesday* who might be at 2.0 at the time, is penalised if the price drops to 1.5 on the Friday* when the Pinnacle price is recorded. Even the 'boring' old draw price can often fluctuate within a broad range (e.g. AC Milan* v Atletico Madrid* 3.0 to 4.30 on Betfair* or Arsenal* v Bayern Munich* (3.30 to 4.40) .

Ultimately it is up to the individual which services they choose to follow and tables such as the FTL, imperfect as they are, do at least offer some clue as to who may or may not be finding value. The Pinnacle benchmark can often be beaten, so the returns should always be beatable. How much they are beatable by depends on the individual, their patience and their ability to get on.

Football Data’s collection of a weekend’s match prices on a Friday is also problematic, because it is at an arbitrary time, and the price at kick-off on a Saturday (one day later), a Sunday (two days later) or Monday night (three days later) may well be significantly different. Knowing the Friday price is the one being used for a Monday night game, there is a loophole that people could wait until the last minute and make their selections based on how the markets have moved in the intervening hours.

And in case anyone hadn’t thought of that before, which I believe is the case, I’ll close the loophole this week by requiring weekend selections (Friday to Monday) to be submitted by the first match of the Friday, and for midweek selections (Tuesday to Thursday) to be sent by the first match on Tuesday. Most people are already sending their selections in well ahead of time, and I don't believe anyone is currently exploiting this opportunity, but having exposed it, and to keep the contest as fair as possible, I feel this small adjustment should to be made.   

Incidentally, Pinnacle were chosen because they are generally perceived as being one of the better books out there – both in terms of price and of accessibility – although further research indicates that they are not as generous as generally believed*. They are generally willing to take your money though, with limits high enough to be all but irrelevant for the majority of us, except for big-hitter Steve of course, who triggered these thoughts.

* The views expressed in linked blogs, posts, articles or tweets are of course those of the authors of the blogs, posts, articles and tweets and do not necessarily represent the views, opinions, or positions, of myself and should not be shown to the police*.