Stephen Jackson

Likelihoods for this weekend's Premier League matches from the goalprojection model

This is the impact on Premier League likelihoods of a (for example) 60-point deduction for Manchester City. Looks good for Arsenal.

These are the current goalprojection Premier League likelihoods assuming no Manchester City points deduction.

I've used the goalprojection model to look at the impact of potential points deductions on Manchester City's Premier League season. 60 points would make things interesting.

Likelihoods for first weekend of Premier League fixtures from goalprojection model

Premier League projection from my model. Can't split Arsenal and Man City. Has the promoted teams in relegation places. But (as we saw with Sunderland) it's difficult to account for the high player turnover among promoted teams. Still gives ITFC 53% chance of staying up!

After completing my Championship projection, I realised I needed to recode the model for the EFL's new bottomless playoff format. Here are the resulting promotion chances. West Ham quite likely to be promoted.

Pre-season Championship projection from the goalprojection shot-based model. Allows for Southampton's points deduction.

George Hirst is incredibly underrated. Main forward in 2 Champ promotion campaigns. Role is about much more than goal scoring, yet scored 11 times last season. Delivers in big matches (QPR, Norwich, and the iconic goal at Barnsley).

Updated ITFC recruitment strategy Venn. To be fair, they seem to be casting the net a bit wider this time, although I think there's been a Tom Fellows link over the summer.

The third place match usually delivers goals. Both not this many. The problem is that it will skew the data sample forever.

Now for the real prize. Likelihoods for the 3rd place fixture. Don't expect a seven-man defence in this one. I had to look it up, and it seems scarcely believable, but they would make the players endure extra time if it's level after 90 minutes.

Only two more matches before I mothball the model until 2030. Perhaps I should try to anticipate FIFA and adapt it for a 64, 96 or 128 team format in the meantime. Anyway, here are the likelihoods for the final.

Likelihoods for the momentous England v Argentina semi final on Wednesday. The model (like the market) has England as favourites. I can't say I'm as confident.

The two best teams in the tournament meet in the first semi on Tuesday, and it's a coin toss.

So the top 4 teams from the goalprojection model (each with >30% pre tournament likelihood of reaching the semi-finals) reach the semi-finals. Here are the updated likelihoods. England most likely team to reach final because two stongest teams play in other semi

Argentina v Switzerland likelihoods

Norway v England likelihoods

Spain v Belgium likelihoods

France v Morocco likelihoods