Molali is a prediction model for the Spanish La Liga. Based on the statistics from the last 5 seasons, it calculates the possibilities of the result for all La Liga games.

Created by: Cédric Schmitt

On Scorum there are some great prediction models for several leagues, but a big league was still missing: the Spanish La Liga. Not anymore, now there is a prediction model for the Spanish La Liga too, and of course I hope it will give some good results.

Predictions

I will post the predictions before every round. For every game I will post a table like the following one.

Example post

For each game the model calculates the probability for all outcomes between 0-0 and 5-5. Based on that, it calculates the probabilities for each team to win and for a draw.

To calculate the results the model takes the statistics for the last 5 seasons and the running season. The unweighted average shows the normal average of all these results. The weighted average gives the results more importance the newer they are. The stats of last season count more than the results of an older season. I did this trying to achieve better results, as the teams evolve and the statistics seem more relevant to me the newer they are.

Calculations

The model takes in account only the last 5 La Liga seasons and the current season. The calculations for teams that did not play the whole time in La Liga like Huesca are based on less games. Huesca is the only newcomer in La Liga, all other teams have played at least one full season in the last 5 years in the highest league.

Betting

I will place bets on the games I see some value compared to the odds of the bookies. I will always write the odds I play and the bookie where I place the bet.

Reporting

I will post 1 prediction post before every round and a review post after every round. If I see some potential I will try to adapt the model as needed and always inform about that. I am really excited about this project and hope to achieve a good performance. Feel free to leave your feedback in the comments at any time.