Welcome to the third edition of my EPL Predictions newsletter, where each week, I'll post detailed forecasts for each game.
As a general rule of thumb, each issue will contain an assessment of the previous week's forecasts, plus naive and biased predictions for upcoming games.
(you can learn more about the general assumptions of the model here)
Let's get to it.
According to the latest simulation of the SOC model, Manchester City are now in a great position to win the league with a 66% chance of lifting the title – vs. the 58% predicted in the previous week.
Arsenal are another side that have improved their chances of a top-four finish, with the North London club having a 41% chance of finishing in a Champions League spot (vs. 33% in the previous week).
In the relegation battle, Forest have reduced their chances of relegation to 60% (vs. 72% at the start of GW2), whereas Everton went from 18% to 21% after their defeat at Villa Park.
The naive predictions are computed using historical data only, and it's the forecast system that I use to simulate the end-of-season probabilities for the Premier League.
The following table showcases each game's forecasts under this model.
In this section, I present my biased forecasts, which, are based on both my hunch and historical data. For each match, I'll present a detailed viz with probabilities for different events within the match and a small write-up.
The following table shows the summary of my biased forecasts.
AFC Bournemouth vs. Arsenal
Arsenal have really hit the ground running, and despite having some iffy moments during their first two fixtures, I expect them to get a comfortable win on this one.
I haven't watched a Bournemouth game yet, so I can't say much about them, but considering what I've seen of Arsenal, I wouldn't be surprised if Bournemouth manage to get their name on the scoresheet.
It's funny that the model constantly points to the 1 - 1 scoreline as the most likely outcome, however, I'm not sure the reasons why.
I believe that a 1 - 2 will be the result for this one.
Crystal Palace vs. Aston Villa
Palace, once more, proved why they're such a difficult team for the Premier League's Big Six.
I'm a big fan of what Patrick Vieira is doing at the London club and I think they'll finish in the top half of the table.
My model expects this to be a pretty even affair between both sides and there are likely to be some goals in there. If I had to choose a result, I'd say a 2 - 1 victory for the hosts.
Everton vs. Nottingham Forest
Kudos to Nottingham for getting such a great win against the Hammers on their first home game of the season. I sadly believe that they'll end up relegated, but, to be honest, I'll be happy if they don't.
Lampard's men are once again just struggling too much to get results and there must be a ton of pressure in the dressing room to avoid another relegation battle.
Everton can't afford to lose and Forest should be coming in with a big boost, so I think a 1 - 1 draw is a decent prediction for this fixture.
Fulham vs. Brentford
Man, what an amazing result that was for Brentford.
This will be a great fixture to watch in my opinion, since I'm just such a huge fan of Ivan Toney and Mbeumo.
I'm also hoping that the Bees manage to get the win, as this is one of the fixtures in which my model has a big difference versus the forecasts made by FiveThirtyEight – who are forecasting just a 40% chance of an away win.
Gun to my head prediction: 1 - 2.
Leicester City vs. Southampton
Funny that this is a fixture where my model isn't predicting a 1 - 1 outcome as the most likely, however, with both sides letting in 6 goals in their first two fixtures, it's no surprise that my model is expecting a goal-fest on this one – with a 65% chance of an over 2.5.
I think that the hosts will win comfortably as I just can't find anything positive to say about Southampton at the moment. For this match, I'm going with a 2 - 0 win for Leicester as my prediction.
Tottenham Hotspur vs. Wolverhampton
Despite being clear favorites, my model's predictions for a Tottenham win are still markedly below FiveThirtyEight's prediction of 65% for a Spurs win.
Conte's men really got away with it on Sunday, and with Cristian Romero's injury, it'll be interesting how Davinson Sanchez covers for the Argentinian. I don't expect an easy win here for the hosts, but they should be able to get a result.
Exact score prediction: 2 - 0.
Leeds United vs. Chelsea
Chelsea were extremely unlucky on the weekend, and I think that they're surprising a lot of people (myself included) who were skeptical of them being guaranteed a top-four spot before the season started.
I don't think they'll challenge for the title, but they should keep dominating these types of games and I expect them to keep a clean sheet on Sunday; I'd say that a 0 - 3 win is not completely out of the question.
Newcastle United vs. Manchester City
I read on Sky Sports that Eddie Howe is the only manager to have faced Pep Guardiola over ten times, and lost every single one of his fixtures. That's a pretty big stat.
Considering that the Magpies weren't able to put Brighton into any real danger on the weekend, I'm finding it hard to think of ways that they could pose some attacking threat to City on Sunday.
Full-time prediction: 0 - 2.
West Ham United vs. Brighton & Hove Albion
I felt a little bit sorry for West Ham after the match against Forest, despite strongly considering that the hosts fully deserved the win.
Brighton have been historically better away from home, and with great back-to-back performances in their first two fixtures, I think they should keep making life difficult for Moye's and his boys in the first few games of this Premier League season.
Funny enough, this fixture is the most likely to end up in a draw from GW3's fixtures, so I'm backing a 1 - 1 stalemate at the London Stadium.
Manchester United vs. Liverpool
Don't beat me up on this one, but with Darwin, Firminio and Jota most likely unable to play against United, I think that the hosts have a good chance of picking up some points against Klopp's men.
Either that or they get thrashed once more (lol).
If I had to pick a result, I'd say a 0 - 0 draw at Old Trafford is what we'll see on Monday.
How did we do last week?
Once again, we were beaten by 538's model in both the naive and biased predictions for the gameweek.
The model is still performing well, and I'm pretty happy with how the predictions are performing. Barring the West Ham game, I think we would have been able to get a much better score for GW2.
Hope you enjoyed today's issue! Let's see how we fare in this season's third gameweek, I'll post the model's error in next week's issue.
If you like my work, please help me by subscribing to my website and sharing my posts.
Catch you later! 👋