More Football Prediction with Math
I had an interesting talk with a friend over the weekend about the math behind the prediction. The outcome of which is that to come up with win, lose, draw predictions, all you need to do is add up the predictions for all point differences. Let me give you an example, if the point difference for the home team is 1 then they have one by one goal (1-0, 2-1, 3-2, …). If you take the draw then all draws are equal thus the goal difference of 0 is the chance of all draws (0-0,1-1, 2-2,…). If you take the lose position then the goal difference for the home time is -1 which covers (0-1, 1-2, 2-3,….). So, you do this up to a goal difference of 5. This makes it much easier to make a match prediction.
Let me leave you with a prediction. Today Liverpool is playing versus Aston Villa. Here is how it breaks down: Liverpool has a goals per game at home average of 1.71 and Aston Villa has a goals per game away average of 0.71 according to Soccer Stats. You can probably already predict the outcome just from that but here is the percentage breakdown:
- home win —> 60.92%
- draw —> 23.44%
- away win —> 15.15%
Celticist, Computer Scientist, Nerd, sometimes a poet…
What is the standard deviation? Is it even a normal distribution?
Is that a lifetime average? How does the average change if you only consider the current season or perhaps the last.
What is the history of these two teams when playing each other?
Are there any injuries?
Do the professional gambling sites agree with your odds?
You can find all the statistical information in my last post which links to the Wikipedia page for the distribution.
The average is for the season.
I have no clue of the history of the teams or injuries.
Betfair is giving these percentages as of 6:00pm GMT: Liverpool -> 62.11, Aston Villa -> 15.15, Draw -> 23.81. They are in the area of each other.