Expected Goals implied odds

[Stevie D]

Hello Folks

I trade the overs/unders football markets and I have a model which is essentially a variation on expected goals. This gives me a rough idea of whether there is value in the market pricing but my question is is there a method for turning my expected goals into implied odds so I can see the exact implied value, this can be mathematical, empirical or rule of thumb, preferably the latter as I am certainly not a mathematician.

Appreciate any help.

Cheers

[jamesg46]

The baseline is 2.5 so its 50/50 from there (odds 2.0) its either going to be 2 or 3 goals. So the next step is what percentage over or under is your expected goals from the baseline of 2.5


2.2 expected goals (for example) 2.2/2.5 x 100 = 88% so your missing 0.3 is worth 12%

12 over 100 is 0.12 so 2.0 - 0.12 = 1.88

I have no idea if that's right (it's been a long night & math isn't my strong point either) but that's my process of thought... I'm sure someone will comment to verify or correct

[Stevie D]

Thanks James. I was thinking it was more complicated but if you apply your method across the probability curve for the various goal outcomes I think its probably good enough. Cheers !

[Andriy]

GExp Unders
2.0 1.48
2.1 1.54
2.2 1.61
2.3 1.68
2.4 1.76
2.5 1.84
2.6 1.93
2.7 2.03
2.8 2.13
2.9 2.24
3.0 2.36
3.1 2.49
3.2 2.63
3.3 2.78
3.4 2.94
3.5 3.12

[spreadbetting]

A simple method usually involves tweaked versions of poisson to get your overall probabilities of events occuring throughout the match. Basically poisson gives you a probability for each scoreline and you'd add up all the home side winning probabilities to get the home team odds etc, add all the under 2.5 scores to get probabilities of under 2.5 and so on. There are flaws with poisson and you can find work arounds all over the net, the calculaton itself is simple and excel has a function for it. Power of goals site is a decent place to start but really depends how far you want to delve into goal prediction as to what you'll get out. Remember there are data teams and each bookmaker has more resources than you'll have at their disposal so sometime best to use their hard work rather than putting in your own.

http://thepowerofgoals.blogspot.com/201 ... t-ten.html

[Stevie D]

Thanks, ill take a look

[Stevie D]

Thanks Andriy, is this for a particular league ?

[Andriy]

Thanks Andriy, is this for a particular league ?

Valid for all leagues. It's just Poisson mentioned in a previous post; whilst Poisson need to be adapted to make correct scores modelling accurate due to draw inflation, for total goals scores it's very accurate.

[StellaBot]

Interesting reading
Is there an expected over goals expectency?
GExp Unders
2.0 1.48
2.1 1.54
2.2 1.61
2.3 1.68
2.4 1.76
2.5 1.84
2.6 1.93
2.7 2.03
2.8 2.13
2.9 2.24
3.0 2.36
3.1 2.49
3.2 2.63
3.3 2.78
3.4 2.94
3.5 3.12

[Andriy]

O2.5=U2.5/(U2.5-1)

eg if Unders=2.5, Overs=2.5/1.5=1.67

[Stevie D]

Thanks Andriy, is this for a particular league ?

Valid for all leagues. It's just Poisson mentioned in a previous post; whilst Poisson need to be adapted to make correct scores modelling accurate due to draw inflation, for total goals scores it's very accurate.

Thanks VM!

[StellaBot]

I cannot understand that expected goal stats as a proper guide
It seems one should consider other factors

[Kafkaesque]

I cannot understand that expected goal stats as a proper guide
It seems one should consider other factors

The operative word in your own comment - and the gist of the comments before - is guide.

It can be used as a guide or starting point, if you know what's behind the numbers, and then factor in anything else you fancy.

And then sometimes data even has to be disregarded, or reset rather than take the season thus far. As an example anything pre-Pearson for Watford should probably be in the bin.

[StellaBot]

My understanding of football is very specific
and way out of above figures
Im talking about profitable stratedgies
However I make profit with clowns eg above saying and doing different

[StellaBot]

As Morse said

I want to speak to the head knob
not the Knob head