What The Bookies Say – Gameweek 3

Does anybody know when the last price changes will occur before tomorrows deadline?

Jeeeeesus I wasn't expecting to find all that when I clicked on the first two pages of comments! 😀

Bookmaking for Dummies - Part III - Summary so Far

Before going on to the specifics of football betting, it should be helpful to draw out and emphasise some of the main points so far.

The absolute key sentence is in the Part II piece above:

"The 6/4 horse isn't actually the one they think has a 40% chance of winning. It is the one they think 40% (or 33% if you take the over-round into account) of the money bet will be placed on."

This is essential. Bookmakers' odds don't fundamentally express their idea of probabilities as many people imagine. "The bookies don't care if the 6/4 favourite is the right favourite. They only care that 33% of the money bet is placed on it. That's all that matters to them."

In order to get this right when setting their initial odds they have to have an idea of how probable the possible outcomes are but much more important is to have a good idea on where the weight of money (punter preferences) will be. Their sole interest is in producing a balanced book, whether the horse most backed deserves to be the most backed is immaterial to them.

Now, bookmakers' odds can be said to represent probabilities in a way but if they do they represent the punters' view of probabilities rather than the bookies'. The bookmakers, obviously, will be interested in being in the know - if there is a special reason why a horse (or option) is a more likely winner they can price it defensively (lower odds). But the main thing they want to be in the know about - and are expert in - is predicting where the weight of money will be.

If you convert bookmakers' odds to a probability of success all you are therefore doing is summarising what the majority of punters think (or will be expected to think). It seldom represents any pure knowledge or information beyond that at all although statistical factors are taken into account in a similar way that actuarial tables for life expectancy are taken into account in setting life insurance premiums.

That's it, that's the key point. It is the biggest problem in transferring bookmakers' odds across to use them as a probability guideline for something like FPL. But I will look at that in more detail in Part IV.

Thanks YMA and Ruth.

RMT - Sky Fantasy Football

Pantilimon
Ivanovic Prodl Fonte
Hazard Silva Ritchie Noble Dorrans
Aguero (C) Wilson

Trying to get rid of Dorrans and possibly Ivanovic, not exactly who for though? Still have captain change this week if I transfer in a Liverpool/Arsenal player.

Looking forward to parts IV - VIII

Bookmaking for Dummies - Part IV - Fixed Odds (Football) Betting

I am posting this as the last part of my summary, I want to get it done before taking a break from posting for a while.

Football betting is slightly divergent from horse-race betting in that it is a fixed-odds market - there is no on-course book with changing odds. Therefore the bookies will offer fixed odds on specific outcomes and usually they won't change them.

It might be assumed that they do this by having very accurate algorithms to predict the probability of the various outcomes and there is no doubt that they do have such software. But it is unreliable. What they do have brilliant knowledge of and ability to predict is where the weight of money will go. And as already explained, that's what matters to them.

A typical set of odds might see the home team at 4/6, the draw at 2/1 and the away win at 5/1. That's a 110% over-round and as explained before, as long as they take $600 on the home win to $333 on the draw to $167 on the away win (or money in those proportions) they are home and dry. Whatever the result they will pocket $100 in profit for every $1000 they pay out. And that's what they are after.

It may seem strange to think that the bookies can predict the weight of money so well but you have to remember that they have the data for every single bet made. Calculating the patterns is relatively easy and can be done to quite a fine margin. Most football punters are habitual or recreational gamblers. City fans (by the vast majority) back City. United fans back United. Of course there are exceptions - a fellow Chelsea fan I know always backed Chelsea's opponents to win. He figured that he'd then come away from Stamford Bridge happy whatever the result. 😕

There are more United fans than Norwich fans. As a result the "big" teams (and "big" players) in the PL are always under-priced, whether to win games, to score goals or to keep clean sheets. The bookies are pricing the weight of money, not the probability. They know very well where the money will go and they watch it like a hawk. If, for some reason, it skews too far away from what they expect they call it "suspicious betting patterns" and report it as a possible sign of match-fixing. They don't want anyone betting with inside knowledge. They want the predictable bets of predictable punters so they can take their predictable 10%.

Probabilities do feed into fixed odds betting. The most common results of PL games statistically are 1-0, 1-1, 2-0 and 2-1. That is an influence and they will almost always be the shortest priced outcomes. But taking that as any prediction at all is a fallacy. It simply isn't.

So, what can be discerned from the bookies' fixed odds? Not much that is of any use to a FPL manager really. Sites like FFFix will give better percentage chances of clean sheets (for example) than the bookies' odds will because FFFix are using a probability algorithm rather than a betting algorithm. Odds to score 2 or more goals? Well, Aguero has done that 22 times in 4 seasons, 14 at home and 8 away. On that basis the actual odds of it in any game are around 6/1 (or let's say between 5/1 and 7/1 depending on the opposition). The bookies usually offer 7/2 or 4/1. The gap is their profit.

When you look at fixed odds, you are not looking at "what the bookies say". You are looking at "what the not very informed, not very professional punter says". That's the long and the short of it I'm afraid.