Last time out was not great for our prediction model, getting only 3/10 predictions right. We are now 66/118 on the year, just barely hanging on above 50%. The model bolding predicted Real Madrid losing to Granada and that just did not come off at all. Villarreal also injured the predictions by drawing Celta instead of losing to them.
Our prediction model predicts a lower number of draws than what actually end up happening, and that’s the largest part of the reason why we hover around 50% accuracy. We do have a predicted draw today, with Osasuna and Elche predicted to hold one another fast. We are MASSIVE favorites against FC Barcelona, which is super weird to me, and the model seems to have turned on Granada with ferocity, giving them only a 6% chance to win a week after they were favored to beat Real Madrid. Alaves vs Celta is on a razor’s edge. This is one of those matches that gets the system in trouble. The percentage odds basically say that it is more likely that one team or another win than it is that they draw, yet the difference in odds between the two of them is less than a single percentage point. It’s mathematical indecision that ultimately ends up being a random cast of the die. You end up seeing this play itself out in the ‘most likely scoreline’ which is a draw instead of either team winning or losing.
Tell us what you think about these predictions in the comments.