Table that calculates the probabilities

[TippingDutchmen]

In his book “Taking Chances”, John Haigh drafted a table that calculates the probabilities of a team scoring 0, 1, 2, 3 and 4 or more goals based on their average number of goals per game.




Does anyone know how to calculate these probabilities?
So how did he get to these %?
Maybe someone could explain it by showing how to calculate it for for example "average = 1.0", "average = 1.4" and "average = 1.8"

My math skills are not that good ;-)
Your help would be highly appreciated.

[dremeber]

Doesn't look like a poisson distribution cause, there also de defense strength of the other team is taking into account.
Sorry, can't help you with this one.

[leveller]

It's Poisson distribution, you get exactly these probablities when you run Poisson.

There are many Poisson calculators around, and I especially reccommend the one found at http://www.cgmbet.com/tools.html , it has exceptional capabilities.

If you want it exactly in this format like you've shown, here is a small Excel file in that format, you can download it from here (right click and save as): http://www.xenbet.com/files/Poisson-probabilities.xlsx

[leveller]

You can not use poisson distribution in soccer betting.

Care to elaborate further? Why not? You can use any probability estimation method you want.

There are countless poisson calculators around the web that try to predict sport events.

Will they yield the profit, though, that's different story.

It's well known that Poisson has it flaws when used for football betting, but it's also known there are methods to mitigate those flaws.

I'm not going into details, as I'm not an expert on that field, but I fail to see why couldn't use Poisson for football betting.

The original question, however, was how to reproduce probabilities in the given table, and there is no doubt they were calculated using Poisson distribution - calculate yourself and check.

[d3li6h7]

I think Poisson can make a profit!
I follow and develop at the same time a system based on the type of bet you attend "Team A" and "B" and use a different coefficient for each championship! (Currently it is only championship course). But I believe that put side by side, ie the statistical Poisson get a pretty good percentage of which can make a profit with a plan to bet very well done.
Attach the Premier League stage that is closed!

In the past I watched surprises odds over 4 on a file that has odds of opening and closing, and we all know that bookmakers can intervene and change the course of the game. Probably opinions divided about betting! But I think that the way you manage the bank's success in betting!

[Raabe]

No profit, it's just delusion that you found alchemic stone. Test it in long period and you will see.
Even with adjustments to poisson model you are NOT taking into consideration many factors such as injuries, weather, motivation in championship, etc.

[dremeber]

No profit, it's just delusion that you found alchemic stone. Test it in long period and you will see.
Even with adjustments to poisson model you are NOT taking into consideration many factors such as injuries, weather, motivation in championship, etc.

I cannot disagree with the arguments you are making here. But in my case one of the positive side of a Poisson model is that it also doesn't include emotions, which can go both side.
At the moment I'm running a poisson test for the Belgian Jupiler Pro League and I do have the feeling that it could work. I would have delivered a profit for the regular season but that was in a test phase.
One of the problems I do see is that a Poisson model has it much harder when all teams are more or less of equal strength.
Also Poisson is according to me better suited for the over/under and BTTS market.