Every year before the first ball is kicked, and even before the transfer window closes, pundits the world over make gut level predictions of who will win the Premier League. More than anything these predictions only show a pundit’s biases. I don’t mean biases in the harshest way, we all have biases, I mean that instead of doing an objective analysis of each team’s strengths and weaknesses, player transfers and managerial effect, they simply “guess” as to who will win the League. In theory, their guesses are “better” than you and I because they have been around the block a lot longer than us, but in reality they aren’t much better than we are at guessing. After all, anyone can guess that the big clubs will finish near the top and the little clubs will struggle at the bottom.
Every once in a hundred years, however, the big clubs, with the big-mouthed managers, finish 10th and the team that was supposed to only get 40 points wins the League. Does this disprove anything? No. No one, apart from the Leicester supporters who put a hopeful punt on the club, predicted that Leicester would take the League by storm, that Mourinho would destroy a team that should have made a run on the title, that Louis van Gaal would try to play some of the worst football ever seen in Manchester, and that Man City would fire their manager half way through the season and keep him on as a caretaker like some kind of dead man walking. Even the analytics crowd had a hard time predicting the league last year: many, using week-to-week data, had Arsenal winning almost all the way to the end.
The reason why gut level predictions work is because they are usually based on some kind of previous data. The Arsenal injury record is a great example — has there been a season where Arsenal haven’t been rocked by injuries? Similarly, if you were a pundit during the late Ferguson era, predicting Man U to win the title was an easy choice.
There are a lot of ways to do an “objective” prediction. For example, you could use team value as a predictor. Using just the simple team value as assessed by Transfermarkt.com I get a .70 R-Squared correlation between team value and points earned in a season. That’s a pretty incredible correlation. But it’s exactly what you and I think
Using just the simple team value as assessed by Transfermarkt.com I get a .70 R-Squared correlation between team value and points earned in a season. That’s a pretty incredible correlation. But it’s exactly what you and I think of when we think of success in football: the more valuable the team, the more likely they will succeed.
Of course, this is correlation and not causation and every season we see teams which are costly but underperform (Chelsea and Liverpool are the big culprits over the last 5 years) and teams which are cheap but overperform (Southampton is a great example… also Leicester!). That is the fun of the Premier League. Literally, watching Chelsea finish 10th was great fun! And for most people, who aren’t Arsenal supporters, watching Leicester win the League was great fun. Personally, the scenes of their fans ecstatic in the run up to lifting the trophy will stick among my greatest memories of the Premier League.
But last season was a massive aberration. How big? .17 R-Squared correlation between team value and points earned. It was such a weird season that it would be like me predicting Bournemouth to win the League this season.
What is interesting about using the transfermarkt valuations is that they use a market approach to player value. So, for example, Mahrez is now worth €30m and last summer he was valued at just €7m. Vardy’s value has gone up and so on. Leicester have also added Slimani and Musa and with their performances last season their overall team value has gone from €91m to €210m in a single season. That’s a jump from 19th most valuable club to 9th most valuable club.
Man U also jumped but significantly more than almost any other team: they were valued at €374m last season, to €534m this season. Chelsea had the biggest drop, €580m last season and just €515m this season. And Arsenal, owing mostly to their purchases this summer, went from €408m to €469m.
If we just use team valuations for our prediction we get the following premier league table:
- Man U
- Man City
- Chelsea
- Arsenal
- Liverpool
- Spurs
- West Ham
- Everton
- Leicester
- Southampton
- Stoke
- Palace
- Watford
- Bournemouth
- West Brom
- Swansea
- Middlesbrough
- Sunderland
- Hull City
- Burnley
That actually looks like a pretty accurate table. There are always a few surprises and based on player purchases (types of players and whether they were recycled from other English clubs) my adjustments to the table are as follows: Stoke and Boro change places, West Brom relegated, Burnley in 16th, Sunderland 20th, West Ham drop to 9th, and Leicester up to 5th. But as far as Arsenal winning the League, we have to admit that Man U are planning on not playing any of their first team players in Europe, a total disgrace as a club but what Mourinho has already admitted. And we know that Chelsea have no European distractions, so anything below third place would be a massive failure for them as a club.
Therefore, my prediction are as follows:
- Man U
- Man City
- Chelsea
- Arsenal
- Leicester
- Liverpool
- Everton
- Spurs
- West Ham
- Southampton
- Middlesbrough
- Palace
- Watford
- Bournemouth
- Swansea
- Burnley
- Stoke
- West Brom
- Hull City
- Sunderland
What do you think will be the final table, and why?
You must be logged in to post a comment.