What Is a Points System? Turning Results Into a Ranking

A points system converts a sequence of individual results into a single ordering, and the conversion is not neutral, since the relative value assigned to each outcome determines what a team should rationally pursue. Where a win is worth twice a draw, a team facing a strong opponent gains as much from two draws as from one win and one loss, which makes the cautious approach equally rewarded. Raising the value of a win breaks that equivalence and makes attacking play more attractive in expectation, which is exactly why the change was made in football in the 1980s and 1990s, moving from two points for a win to three. Evidence on whether the change produced more attacking play is mixed, with some studies finding more goals and others finding the effect small, and the reasoning behind the change is sound regardless of how large the result was.

Sports have adopted arrangements suited to what they want to encourage:

• •Three for a win and one for a draw, which is the standard football arrangement
• •Bonus points in rugby for scoring a set number of tries and for losing narrowly, which rewards ambition and keeps losing teams competing
• •Points for a win in regulation and fewer for one in overtime, used in ice hockey, which produces the criticism that some games award three points and others two
• •Descending points down the finishing order in motor racing and cycling, which rewards consistency across a season
• •Match points in cricket competitions, awarded for results and for performance within drawn matches
• •Tiebreakers on goal difference, head to head or points scored, which are themselves a design choice with effects

Any system rewards some behaviours and the unintended ones are the interesting part. Awarding a point for a narrow defeat keeps a losing team attacking rather than defending a respectable margin, which is the intended effect and also means a team can be content with losing. Systems that cannot distinguish a heavy defeat from a narrow one remove any incentive to limit the damage, which tiebreakers on scoring difference restore. Awarding points for the manner of a result requires defining it, which introduces judgement. And any arrangement can produce a situation late in a competition where a specific result suits both teams, which is the structural problem behind the notorious matches where neither side attacked, and which is addressed by scheduling final rounds simultaneously rather than by changing the points.

What separates teams level on points is a design decision with as much effect as the points themselves. Goal difference rewards heavy victories and therefore encourages a leading team to continue attacking, which is why it was adopted widely, and it means a team can be eliminated on the margin of a match played weeks earlier. Head to head results between the tied teams are used in some competitions and produce different outcomes, occasionally reversing the order goal difference would give, which has decided titles. Goals scored rewards attacking regardless of defending. Away goals in two-legged ties were used for decades and were abolished in several competitions after argument that they had stopped serving their original purpose. Drawing lots is the last resort and has genuinely decided major tournament places. Each rule rewards something different, and choosing between them is choosing what the competition values.

Where competitors do not all play each other, a points system has to do more work and the methods diverge. Ranking points in tennis and golf accumulate over a rolling period and are weighted by the importance of the event, which means a ranking reflects a season rather than a set of head to head results and which produces arguments whenever a player skips events. Rating systems derived from chess estimate a strength for each competitor and update it by how each result compares with what the ratings predicted, which handles unequal schedules properly and is used in several sports and increasingly for team rankings. Committee-selected rankings, used in some collegiate systems, substitute human judgement and attract the criticisms that judgement always does. The underlying difficulty is that comparing competitors who have not met requires assumptions, and every system embeds a different one.

The relative value of each outcome determines what a team should rationally pursue, so two points for a win made caution equally rewarded and three broke that equivalence. Bonus points for ambition and for narrow defeat keep losing teams competing and let a team be content with losing. Where competitors do not all meet, rating systems that update on how results compare with predictions handle unequal schedules properly.