Lockdown wrap: Is the weighted points-per-game system fair?

According to a report on The Athletic website, the Football League are edging towards a resolution about how to conclude the League 1 and League 2 season in the light of the coronavirus outbreak. It reports that all promotion and relegation will be decided by applying a ‘weighted points per game formula’. 

If applied, then Oxford will be promoted to the Championship. Understandably, there’s been plenty of handwringing and arguing about the merits of the system. Suddenly the game has been changed, though isn’t that the whole world right now? 

The oversimplified argument is that every team and fan will back the system which benefits their own team. Let’s be honest, there’s no glory by being promoted like this, nobody in their right mind will sink to their knees, purple faced on their living room floor at the epic conclusion of an Excel spreadsheet calculation. No Oxford fan should gloat at the misfortune of others if it pans out like this.

My preference is still to complete the season, even if it means adjusting future seasons, we’re nine games behind, cancelling the EFL Trophy and replacing it with a league fixture for a couple of years, or forever, can shorten the seasons so we’ll quickly catch up the lost time.

But if a re-start isn’t possible, then is there a way of testing the accuracy of the options available? If you stopped the last 10 seasons on the same day as this season – 14 March – and applied the options, comparing them to what actually happened, it’s possible to assess their accuracy.

I looked at the League 1 top 10 on the 14th March for the last 10 seasons. A nice round 100 teams. There were two reasons to look at only the top 10 – for one, I couldn’t be bothered to look at more and two, the further down the table you go the more impact points will have on your league position. A win has a greater impact when everyone else is losing. So, we’re looking at how accurately any system would predict the most successful teams in League 1.  

One option is to simply freeze the table and award promotion to the top three. Historically, that would give you the right result 76% of the time. The leaders on the 14 March have never failed to be promoted, but the third-place team have only blown it once more than the second-place team. It’s simple, it’s quite accurate, but is there a better option?

There are two weighted options – a simple points per game system and one that is weighted according to home and away form. The weighted formula is straight forward – the average number of home points achieved to date applied to all the remaining home games, and a similar calculation for remaining away fixtures.

The argument for weighting the points is a strong one; teams consistently accumulate a different number of points home and away, and usually higher at home. Plus, not everyone has the same number of home and away games left. 

A weighted system is more accurate than stopping the season. Both a simple PPG and a weighted PPG are 80% accurate in predicting the top 3. Across the 100 clubs – a bigger sample – the simple and weighted PPG systems accurately predict the final position 82% of the time. On average, for both systems, they’re out by less than one place (about 0.6 of a place). Sometimes teams have a late season meltdown or surge meaning these systems can out by as much as 7 places, but it’s very rare that the error rate is more than one place, and it’s only impacted the top three once.

On the 18 occasions the two systems disagree with each other, 60% of the time the weighted PPG was more accurate. Put another way, of the 100 clubs, the weighted PPG was either the same or better than a simple PPG at predicting clubs’ final positions 95% of the time. The weighted system is just about more accurate than a simple PPG.

But is it fair? It’s certainly fairer than Peterborough chairman Darragh MacAnthony’s approach on Twitter which arbitrarily made Posh obvious promotion candidates based solely on a series of Peterborough-friendly variables. It reminded me of my favourite story about former Scotland manager Ally Macleod. When he was manager at Ayr United he took them to the Scottish Premier League. He was asked how they’d do in their opening fixtures – Aberdeen? Win. Motherwell? Win. Hearts? Win. The first six games, all wins. Celtic? Win, he said. When asked why he was so confident of beating the Scottish giants he said ‘Well, after that winning streak, they’ll be terrified playing us.’

These systems aren’t unfair in that they’re an objective way of concluding an unknown. Is it fairer than, say, the play-off system? The third best team in the division have only been promoted twice in the last 10 years; meaning they’re only 20% accurate in identifying the third best team in the division. The play-offs are an arbitrary system created to generate revenue and excitement. This is an arbitrary system created to conclude a season impacted by a pandemic. I mean, I’m all for the play-offs, but when they’re so inaccurate are they fair? They appear less fair than either PPG system. 

Abandoning the season altogether is another option; I’ve often felt like giving up during this pandemic, so it’s very tempting. But I would always prefer a conclusion based on what’s been achieved, rather than scrubbing seven and a half months of hard work from the records. James Constable missed out on the club’s goalscoring record because Chester failed to complete their fixtures in 2009. We lost the three points gained that night, whereas I’d have preferred the result to have stood and every remaining team in Chester’s fixture list awarded a default 3-0 win. It feels fairer to me to recognise what’s been achieved than to pretend it didn’t happen.

Perhaps an abandonment is the best option; but look at the top 10s over the last 10 years, you can see how pivotal a good League 1 season can be. Of the forty-nine teams featured, six are in the Premier League; in addition two more have been in it recently, three are in the Conference with another eight experiencing it in the last decade. Financially, the difference between Championship and Premier League football and the rest is massive. Sometimes a club might get just one shot; this summer we are likely to lose the core of our successful team, we could struggle to replace it, next year could be a relegation scrap, all this year’s investment and work just goes up in smoke. Is that fair? Coventry are in an even worse position, without promotion, which they thoroughly deserve, they could easily plummet. Promotion in this way is not about glory, it’s about survival. At the other end, do Southend genuinely deserve another season in League 1?

If I were a club chairman, I’d vote to continue the season when possible, but it means people making compromises about safety and contracts. There are plenty of industries pivoting in the face of the crisis and football won’t be unique in having to balance safety with sustainability. But, if a system is needed to bring the season to a close and give everyone a breather, then the weighted points per game system does seem the best.

Published by


Oxblogger is a blog about Oxford United.

One thought on “Lockdown wrap: Is the weighted points-per-game system fair?

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s