The how good was my prediction post.

With most of the European top leagues reaching the half way mark I thought it was time to start looking at my league projections again.

The model is based on machine learning and is described here and here. And before I start presenting new projections. Lets look at how well the model performed with 8 rounds to go in the top two Swedish divisions last autumn.

The predictions can be found here. Now, the expected accuracy would be at around 30% for exact league positions predicted correctly and at about 70% for ±1

The biggest miss, obviously would be that IFK Norrköping was predicted do finish 4th. They won. I wrote..

IFK Norrköping may have to hope that all three teams ahead of them under perform which seems unlikely

…and that was more or less what happened. But that one still goes in to the error-percentage. The percentages where the model was right however were:

Spot on: 7 of 32 = 22%
Accuracy of ±1: 25/32 =  78%

That would more or less be as expected. The top 3 positions in Superettan was projected spot on and in Allsvenskan 14 of 16 projected league positions were within ±1 position.

And with no further due I now present my first prediction that’s not made for Swedish football. And since the French Ligue 1 restarts this weekend I thought I’d start of with that one. So here is my projection with 19 rounds to go of Ligue 1:

Ligue 1 projection at round 19

The expected accuracy at 19 rounds to go is:

**************ACCURACY: 0.226210350584 *****************
**************ACCURACY +-1: 0.495826377295 *****************

And I guess PSG would be within the 23% that will be spot on 🙂

The possibility of PSG taking 100 points or more is well within 50%! The team predicted to make the biggest climb is Marseille. From 10th to 4th. This will be interesting to follow!

UPDATED (2016-01-10). Unfortunately I totally missed out oncontributing with my own model to Constantinos Chappas‘ excellent compilation of PL predictions at round 20.
I now have run my on model and the result is a bit interesting since it is one of few that doesn’t rank Arsenal as number one. Also, Leicester are ranked quite high in my model. They are at at least a 75% chance of finishing top 4. At the bottom end Aston Villa is at a 6% chance of avoiding relegation – at most.

Published by

Ola Lidmark Eriksson

Football analyst/programmer

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s