Why Do Teams Finish Level and Still Not Be Level? Ranking by Rules

A league table converts a season of individual matches into a single ranking, and doing so requires decisions that are not obvious. How many points a win is worth relative to a draw determines how much risk a team should take, since three points for a win and one for a draw rewards attempting to win far more than two and one did. Whether all matches count equally determines whether a fixture list must be balanced. What happens when teams are level determines outcomes in a substantial fraction of seasons. And whether the table decides everything, or whether playoffs follow, changes what teams are competing for at the end. None of those is a neutral bookkeeping choice.

Different competitions separate level teams by quite different measures:

• •Goal difference, meaning goals scored less goals conceded, which rewards winning heavily
• •Goals scored alone, which rewards attacking regardless of defence
• •Results between the tied teams, which is used in most international tournaments
• •Away goals in the matches between them, which several competitions used and many have dropped
• •Wins, disciplinary records and drawing of lots as later criteria
• •A playoff match, which several leagues use for the title and for relegation

Tiebreakers are decisive often enough to have settled major competitions, and the arguments are about incentives rather than fairness in the abstract. Goal difference encourages a team already winning to keep attacking, which spectators prefer and which produces heavy defeats for weaker sides. Head-to-head results reward beating your direct rivals and can produce a situation where a team finishes above another it scored fewer goals than, which appears unjust to anybody expecting arithmetic to settle things. Away goals in two-legged ties rewarded the away team enough that several competitions abolished the rule on the grounds that it distorted how the first match was played. Every rule creates incentives, and the choice is about which distortion is preferred.

What sits at the bottom of a table matters as much as what sits at the top and the two arrangements produce entirely different sports. A system with promotion and relegation makes every position consequential, keeps late-season matches between poor teams meaningful and allows a club to rise from the bottom of the structure to the top, which has happened. A closed league with fixed membership removes that, which protects investment, makes revenue predictable and produces teams with nothing to play for by midseason, addressed by devices including reversing the draft order to favour the worst teams. Each arrangement follows from what the competition is for, and attempts to move a sport from one to the other generate enormous opposition, as a proposed European football competition found in 2021.

Ranking by aggregated scores recurs well beyond sport and carries the same difficulties. University and school tables combine measures with chosen weights, and changing the weights reorders the table substantially, which institutions know and respond to by optimising the measured quantities. Country rankings on development, corruption and competitiveness do the same. Any aggregation of unlike measures into one number requires weights that are choices rather than discoveries, and presenting the result as a fact obscures that. The same applies to tiebreakers, which is why a league table is a useful example, since everybody understands that finishing above somebody on goal difference is a consequence of a rule rather than a deeper truth about who is better.

How much a win is worth relative to a draw changes how much risk a team takes, which is why the shift to three points was made deliberately. Ties are broken by goal difference, by goals scored or by results between the teams, and each rewards different behaviour. Any ranking that aggregates unlike measures needs weights that are choices, which a table presents as though they were facts.