Welcome to the fourth 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.
With Liverpool underperforming, City just keep increasing their chances of another Premier League win according to my model. However, Spurs and Arsenal are also increasing their chances – albeit marginally.
On the other end of the table, Fulham keep pulling away from the relegation zone and I wonder which of Everton and Southampton will take their spot as one of the most likely to be relegated this season.
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.
On my exact scoreline predictions from last week, I got four correct results and would've gotten two exact scoreline predictions right of it weren't for Mateta's late goal against Villa.
Southampton vs. Manchester United
Yes, United were clinical against Liverpool at Old Trafford. However, there's still a cloud of doubt looming on their away record – with the Red Devils having lost their last seven fixtures on the road in the Premier League.
If you've followed along with my prediction, you'll know that I'm not a big fan of the Saints, so I'm thinking that maybe, just maybe, United's poor away run might come to an end this Saturday.
My exact score prediction for this one is a 1 - 2 scoreline in favor of Ten Hag's men, which according to my model has an 8.6% chance of happening.
Brentford vs. Everton
Everton's matches have been a hard watch for me lately, and I won't be putting much attention to this one.
Brentford deserved to lose against Fulham, but they're still a massive threat from an attacking perspective. The Bees won the series against the Liverpool side last season and I think they'll make it three on the run against the Tofees in the Premier League.
I'm leaning towards a 2 - 0 win for the home side, with a 9.8% chance of happening according to my model.
Brighton & Hove Albion vs. Leeds United
Brighton have been great lately, there's no getting around that. However, they tend to underperform at the Amex, and I wouldn't be so sure of the Seagull's winning this Saturday.
I think it'll be a good watch, albeit with no goals in my opinion. So a 0 - 0 draw is what I'm going for.
Chelsea vs. Leicester City
Apparently, Brendan Rodgers is now favored to be the first managerial sacking of the season according to the bookies, and I don't see him getting anything out of the trip to Stanford Bridge this Saturday.
With Tuchel away from the sidelines, I think that he wouldn't have favored a better match to miss than this one. The Foxes have only kept a single clean sheet in their last 27 games away from home, and the Blues should exploit that.
Let's go for a 3 - 1 win in this one.
Liverpool vs. AFC Bournemouth
This has to be Liverpool's first win of the season. It has to be.
If they don't get three points here I'm not really sure what we can expect from Klopp's team this season. According to my model, Liverpool are the most likely team to win this gameweek and I couldn't argue much about it.
A 3 - 0 win with an 11.3% chance of happening seems good to me.
Manchester City vs. Crystal Palace
Palace are a team that tends to upset top 6 teams and have lost only once in their last four visits to the Etihad, so a draw is not out of the question in my opinion.
The Eagles have had a great start to the season and barring the game against Newcastle, City haven't had much of a challenge just yet. I think another high-scoring draw is on the cards, and I'm backing a 2-2 draw at Manchester.
Arsenal vs. Fulham
Honestly, Arsenal fans are getting kind of annoying so I really do hope that they won't get a win here. However, it looks unlikely.
The Gunners are playing great football, and even though Fulham have been solid, I just can't let my emotions get in the way of calling out a 2 - 1 win for Arteta's side.
Furthermore, I read on Sky Sports that Arsenal have a 28-win record in their last 33 home games against newly-promoted teams. So, there you have it.
Aston Villa vs. West Ham United
Talk about a balanced game. Same win probability for both sides according to my model, and a 26.2% chance of a draw.
This is a tough one to call out, but with Moye's men at the bottom of the table and a Villa side that even struggled at times in their game against Bolton, I'm inclined to back the Hammers with a 0 - 1 win.
Wolverhampton vs. Newcastle United
It's a shame what's happened with the Wolves' front-line since the Jimenez injury. They're a good watch and I think that Bruno Lage is doing a fantastic job at Mollineux with the tools at his disposal.
Kudos to Newcastle for their result against City, they really should've gotten the three points in my opinion and Saint-Maximin looks better than ever.
This shouldn't be an easy task for the Magpies, but I think they'll be able to come out with three points this Sunday.
Exact scoreline prediction: 0 - 2.
Nottingham Forest vs. Tottenham Hotspur
Which Spurs side will turn up? I never know how to answer that question.
Conte's men were woeful against Wolves in the first half and were lucky to get a point at Chelsea. However, they were devastating against the Saints and I think they'll give a similar display against Forest.
Forest have been decent offensively and I expect them to get on the scoresheet, so a 1 -3 win for Tottenham is what I'm backing.
How did we do last week?
Gameweek three was as unexpected as it can get, and we're still lagging 538's model by 5% when it comes to the Brier Score metric (in the biased model).
However, my model has had a better performance in 13 out of the 30 games this season and I'm pretty proud of that, to be honest.
Hope you enjoyed today's issue! Let's see how we fare in this season's fourth 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! 👋