Predicting the 2017/18 Scottish Premiership

Predicting the 2017/18 Scottish Premiership

By Stefan Bienkowski

It’s two days before the new Scottish Premiership season, so naturally we at TheTwoPointOne have decided to put our neck on the line and make predictions for the coming league campaign up until the final day before the split.

 

Of course, we’ve tried to use a little bit of data to help us out. So let’s cut immediately to the chase, before we get to the theory.

 

As you’ll see below, we once again have Brendan Rodgers’ Celtic side running away with a seventh, consecutive league title, while the real intrigue comes in the race for the Premiership’s European spots. 

 

 

Rangers fans won’t relish the thought of them not only losing ground to Aberdeen, but also allowing Hibernian and St Johnstone to hassle them for third place. If you’ll allow us a moment, we’ll quite happily explain it all.

 

How it’s done

 

Plenty of websites already have their own formulas and theories for predicting league seasons, with varying degrees of complexity and accuracy. For example, our friend 11tegen11 studies every aspect of each team and runs 10,000s of simulations to see who finishes top of the English Premier League. You may have seen the Financial Times adopting his model for their own sports coverage. 

 

For our Scottish Premiership model we’ll borrow the concept of applying each team with two numbers or ratios – one for the amount of goals they score and concede at home and one for the same rate in away games. 

 

We’ve simply called this the Home G Ratio and Away G Ratio. Once we have that we can pick any particular fixture and decipher which team is most likely to score more goals. If Celtic have a home G Ratio of 0.85 and Ross County have an away G Ratio of 0.43 then Celtic are far more likely to win that particular game at Parkhead. Any fixture that has two figures that fall within 0.05 of each other is counted as a draw. 

 

 

These ratios are determined by last season’s results in the Scottish Premiership (Hibs took a little more work, but we’ll explain further below) and will be updated as the coming season progresses.

 

We plan to make this a rolling prediction table. So after four match days of the 2017/18 season the first four from the previous campaign drop out of the data set. As we move towards the pinnacle of the league season our predictions should also grow more and more accurate. 

 

Rather than tediously update this table each week, it’s probably best to update every six weeks. That means by matchday 30, we’ll be in a position to predict who should be on the right or the wrong side of the league split, and paint a fuller picture of how the Scottish Premiership is playing out. 

 

Why are Hibs so high up the table? 

 

Although the concept of a promoted team jumping straight in to a European spot may seem completely and utterly bizarre in most leagues around Europe, it’s actually quite common in Scotland. 

 

We saw Rangers do it last season and Hearts before them, and even before both of those giants returned to the Scottish top division we had Hamilton Accies get promoted and run to the top of the table before settling in to seventh spot. The general rule over the past five years has been that if you look good in the Championship you’ll usually do quite well in the Scottish Premiership too. 

 

 

But this hasn’t been an exercise in guesswork or to try and woo our friends in Leith. There’s some basic maths behind our predictions for the Easter Road side. 

 

Taking the goals scored and conceded home and away in the Scottish Championship from the last four teams to get promoted  – Rangers, Hearts, Dundee and Hamilton – we then tracked the goals they scored home and away in the corresponding Scottish Premiership season.

 

Assuming four teams of varying quality was a fair judge of Hibs’ own quality, we found that their predicted Home G Ratio was 0.49 and their Away G Ratio for the coming season was 0.53. 

 

Before Hearts fans throw their laptops out the window, it’s worth considering that Neil Lennon’s side have enjoyed a rather fruitful Scottish Championship campaign and like the Jambos did under Robbie Neilson, the maths predicts Hibs to take to the top flight like a duck to water. It just so happens to come right after a rather disappointing season for their cross-city rivals. 

 

Why are Rangers so far behind Aberdeen?

 

Although our prediction has Rangers picking up six more points than they did after 33 matches last season, it also suggests that Aberdeen are likely to dramatically increase their lead too.

 

There are a few reasons for this. The first one is that the model simply states that Aberdeen were better at home and on the road than the Ibrox outfit, which states that as far as the maths is concerned Derek McInnes’ side are expected to beat his old club at every encounter.

 

Although that’s unlikely to happen, it’s far more likely than for Rangers to upset the odds in all three games.

 

The second reason is that although Rangers may have bought well this summer, they still have to prove that they aren’t the side that picked up just eight wins on the road last season. Even if we were to discount the one-on-one clashes between both sides, Aberdeen have a far better record and therefore ratio in away games.

 

Ultimately, Rangers can’t expect to be the second best team in Scotland until they start playing like it.

Why do you hate Dundee and Motherwell? 

 

We don’t hate Dundee and Motherwell. Our maths hates Dundee and Motherwell. Kinda. 

 

Perhaps the most obvious reason is simply that Inverness CT dropped out of the top division and in their place are Hibs, who we expect to shoot in to the top half of the table and push everyone below them down one spot. 

 

Our predictions for both sides may seem harsh but bear in mind that Motherwell took six points off Inverness before match day 33 and Dundee claimed four. We don’t expect either team to beat Hibs home or away. 

 

Indeed, once you take Motherwell’s six points against Inverness out of their tally for last season they stand at 26. Just eight points above where we have them for this season. We’re sure both teams will probably do better than the tally we’ve gave them but for now that’s how they look.