After a bit of a down week, the prediction model from Charlie and Mitchel righted itself during Matchday 6, getting 6/10 matches correct. For once, Villarreal actually pulled it’s weight! Barcelona having a 70% chance to win against Cadiz, (66% according to the official betting sites) and yet drawing was by far the most hilarious thing that happened this past week.
For the first time this season, the algorithm picks Villarreal to lose. Frankly, I can’t blame it. Their data set now involves matches where we were without Parejo and Gerard, and we won’t have anything close to a full strength attack going against the top team in La Liga. They are vulnerable and will concede some goals if we can be adventurous on the break, but it’s going to be a really tough go on the road in Madrid.
Levante is only being given a 13% chance against Barcelona, but frankly, I think they have better odds than that given Levante’s performance against Madrid and Barca’s general disarray. Elche has only a 3% shot against La Real and that, actually, seems about right to me. As glad as I am to have our first win of the season, I’m not going to act like they are any good. Betis vs Getafe should be absolute chaos, with the home side seemingly always both scoring and conceding and the away side an increasingly desperate shambles.
Overall, the predictive algorithm we are using sits at 28/58 matches correctly predicted. Given that the model almost exclusively predicts a win one way or another for the most likely result (this week’s Valencia match against Athletic Club being a rare exception) there have been a lot of instances where the second most likely result- often close in likelihood to the first- is what actually happened. A solid week this week should get us back to 50% overall, which is good in a sport where there are three possible results.