What would we all do without the ritual humiliation that is prediction-making before the start of a new Premier League season?
Will Manchester United win the league? Will Chelsea? Will Swansea City make Europe? Will Everton be closer to being relegated or Liverpool? These are the kind of questions that can make anyone look like a fraud before September has even started.
We all know the feeling of the slow, steady, shameful comedown as what seemed like such sensible ideas in August wilt alongside the autumn leaves. We’ve all seen grown men crying over their Premier League predictions. If not directly then I’m sure it was at least something to do with it.
There is, then, only one thing for it. The only way out this time is to get radical with your predictions—and by that I mean get completely and utterly conservative. When you’re asked to give your Premier League predictions this week, whatever you do, don’t try and calculate exactly who’ll finish where based on the various factors involved in deciding it— the transfers, the confidence, the managers, the players, the fixture-lists. That’ll get you nowhere. It’s simply too complex. There are too many unknowns to ever create a worthwhile formula. Tottenham’s Champions League bid could be derailed by a poison lasagne at any moment and you couldn’t possibly predict it, unless you’re the chef.
Premier League on Sportsnet: Watch matches from the Barclays Premier League live on Sportsnet every week. Consult the broadcast schedule || Sportsnet NOW
Instead, this is what you do: you look backwards to last season and you steal all of its ideas. We can’t possibly calculate how the new campaign will play out, but here is the closest thing you can possibly get to a model of what will happen. Think about it. Most of the teams are the same, most of the managers are the same, most of the players are the same. Yes, of course there are loads of crucial variations, but as a starting point nothing else even comes close. The bulk of the prediction is in place, fully formed, with far more factors taken into account than you sitting secretly drawing up a table at your computer could ever get at.
Here’s how this system runs. Chelsea finished top last season by eight points. That gap makes it huge favourites to do the same again, and that assumption would require an extremely compelling reason to overthrow it. Manchester City on the other hand finished second by four points. That gap means that a slightly less compelling reason would be required to overthrow the assumption of City staying in the same place, particularly as in the second half of last season (the most recent sample on offer) it appeared to be going backwards. And you go on like that.
Practically, this season, this system would have the top four remain in place, but with Arsenal overtaking Manchester City to move into second. The compelling reason for the overtake would be this: Arsenal was the in-form team in the division in the second half of last season (again, the most recent sample of competitive league football), while City had begun to drift by mid-January when it was beaten comfortably by Wenger’s team, and neither side’s transfer activity suggests a drastic rearrangement in that form, relative to each other.
City’s real battle will be with Manchester United if you make the equation “last season plus whatever variations you can see to put in.” The club that finished nine points behind City last season and may well be about to lose David De Gea has added more substantially to its squad than Manuel Pellegrini has and also had more forward momentum at the end of last season. Those factors, combined, could make for a close competition between the two clubs, with United maybe even slight favourite.
Of course using the season before as a starting point for predictions in this way might not always work (outliers such as Liverpool’s title run two seasons ago would always have been difficult to predict), but the method would have been regularly effective in the past. Arsenal’s great start to the season before last would have been seen coming using its great ending to the campaign before that, for instance. Or Chelsea’s title win last season became quite predictable once its steady progress to third place in 2013-14 was added to the “compelling new evidence” of Cesc Fabregas and Diego Costa to suggest it could overtake City and Liverpool in one go.
The essential idea behind it all is that the summer isn’t some restart button, even if each team starts on zero points in August. Really, the summer is more of a pause button, more than ever in a short summer like this one, with no World Cup as a major disruptive influence. It takes a lot to overtake another team in the Premier League—especially further up the table where quality is stacked up—and last season doesn’t just disappear once the trophies are handed out. A huge percentage of factors from the old model remain in place and certainly infinitely more than you could ever draw up on your own.
So with the prediction-making underway in the final week before the new Premier League season, remember, this is what you do: you look at what happened last season and go from there. It’s the closest model we have. It might not even work, but you’ll be able to point to your reasons and that should, at the very least, enable you to hold off those fraud accusations for one more year. It’s better than nothing.
Ethan Dean-Richards is a London-based writer. Follow him on Twitter