I never thought I’d be asking this question. But the coronavirus pandemic seems to be making all of us think a bit differently at present.
A friend of mine — statistics professor at another university (and a Man Utd fan, but still we’re friends!) — said this to me a couple of days ago:
This seems an ideal use of your ranking method.
That was immediately after this news from Italy: Coronavirus: Serie A season may not be concluded says Italian football federation. Since then, there has been discussion of whether to cancel football matches in other countries too.
My first reaction to that suggestion from my friend was to dismiss it as fantasy: surely there must be a better way to decide who wins the league, who gets relegated, etc., in the event that the full season’s fixture list cannot be completed. Surely even matches played behind closed doors would be better than just stopping the season at some arbitrary point and using the league table — whether it’s the official league table or the (better!) alt-3 table — to determine the season’s ultimate outcomes.
But now it’s starting to look as though even the possibility of finishing the season without spectators could be in doubt.
So I got to thinking a bit about the idea of using the current league table to decide things. In some sense, that would be fair. But some clubs might reasonably complain, perhaps, that current points totals disadvantage them – either because they would have faced a relatively easy run-in (having played all the strongest teams in the league already, perhaps), or because they have played fewer matches than other teams.
That’s where the alt-3 method comes into its own. It provides a mathematically principled way to order the teams at any point during the season, in such a way that:
- strength of opponents already faced, home advantage, and games in hand, are coherently accounted for;
- there is perfect agreement with the official league table whenever there is no imbalance between teams in their fixture schedules — as is always the case at the end of a full season, after every team has played every other team at home and away.
So indeed (as my friend suggested) the alt-3 table could reasonably be used as a fairer basis than the official league table, in the event that final standings need to be decided before all matches have been played.
Still there would be worries, I reckon, about using (fairly advanced) mathematics to determine league outcomes that affect many people and a whole lot of commercial interests. Until such a time (which does exist in my dreams!) as alt-3 becomes embodied in football league rules as the accepted way to settle an unfinished fixture list — in much the same way that the Duckworth-Lewis method became part of cricket’s rules — I would be nervous about the use of alt-3 for that purpose.
But here’s the thing. It just so happens that right now the official standings and the alt-3 standings, at the top and bottom of all eight leagues currently covered by alt-3.uk, are in perfect agreement. That applies to the Premier League and Championship in England, to Serie A in Italy, to the German Bundesliga, Spanish La Liga, French Ligue 1, Dutch Eredivisie and Portugal’s Primeira Liga.
That is actually quite remarkable, to find that there is no schedule imbalance of any consequence, at the top or bottom of any of the major leagues across Europe just now! It is most unexpected. For example, even as recently as a couple of weeks ago there was substantial imbalance around the relegation zone in the Premier League. And the last time I made a blog post about Serie A, it was to say that Juventus were really top, despite being placed 2nd at that time in the official table. But all such imbalance has evaporated, for the time being — and this applies to all 8 of the listed leagues!
The conclusion? As things stand right now, it could reasonably be regarded as fair to use the official league table standings to determine who wins each league, and who gets relegated.
Direct links to the up-to-date alt-3 league tables for 8 leagues are here: