How Tiebreakers Are Ordered When Records Match

Two teams finishing with identical records is ordinary rather than rare. What separates them is a published sequence of tests, and the order of that sequence is a policy choice.

Head to head comes first for a reason

The first test in most sequences is the result between the tied teams themselves, because it is the most direct evidence available.

It is also the test both teams could control. A team that beat its rival in every meeting has a claim that no aggregate measure can undo.

Where more than two teams are tied, head-to-head becomes a mini-table among them, which frequently resolves nothing and pushes the decision down the list. Unbalanced schedules make that outcome common rather than exceptional.

Division and conference records come next

Subsequent tests usually compare records within the division, then within the conference, on the grounds that those opponents are shared.

This is the schedule-fairness argument made explicit. Comparing performance against a common set of opponents is more meaningful than comparing overall records.

It also rewards teams that performed in the matches the structure emphasises, which is consistent with why divisions exist at all.

Point differential is used cautiously

Aggregate scoring margin is an effective measure of team strength and is deliberately placed low in most sequences.

The concern is behavioural. A tiebreaker that rewards margin encourages teams to continue scoring in decided games, which spectators and opponents dislike.

Placing it late means it rarely decides anything while remaining available when everything above it has failed. A sequence still has to terminate, and a coin toss is the only alternative to a measurable test.

Multi-team ties follow separate procedures

Rules distinguish between two-team and multi-team ties, because applying two-team tests to three teams produces contradictions.

Standard practice breaks the multi-team tie first, establishing one team's position, then restarts the sequence from the top for those remaining.

Restarting matters, since a head-to-head test that was inconclusive among three teams may be decisive between the two that remain. Skipping the restart would resolve the remaining pair on a weaker measure than the rules intend.

The sequence must be published before the season

Teams manage rotations and rest against known incentives, so a tiebreaker discovered late would change decisions retrospectively.

Publishing also removes discretion. Once the order is fixed, no committee decides a playoff place, and the standings resolve themselves mechanically.

That is the underlying purpose. A tiebreaker sequence is less about fairness in any individual case than about ensuring no one has to make a judgement call.

For a long time the deep coverage was the closest thing basketball had to a safe answer. Keep the big near the rim, make the guard fight over the top, live with a mid range jumper. Then somebody put a screen on the retreating big himself, and safe stopped being safe.

The action is easy to describe. Standard ball screen on the wing. As the centre backs towards the paint, a third offensive player, usually whoever is guarded by the least reliable helper, sets a screen on him from behind. The big is now being asked to backpedal into a stationary body he cannot see. He fights through and arrives late, or he goes around and the roller has a straight line to the rim, or the defence switches and a guard is left trying to box out a centre.

What I like about it as a piece of design is that it attacks the coverage rather than the personnel. Most counters hunt the weak defender. This one punishes the strong one. The better your rim protector, the more committed he is to that retreat, and the more the screen behind him hurts.

Every defensive answer is uncomfortable. Pre switch the screener before contact, which needs two players to talk through crowd noise while both are watching the ball. Jump the big out to the level and refuse to be a stationary target, which gives up the pocket pass. Or tag the backscreener early with the low man, which by definition opens the corner you were trying to protect in the first place.

None of it is free, and that is the whole point. Good design does not manufacture an open shot. It manufactures a decision where every branch costs something, taken by a defender with a third of a second and eighteen thousand people shouting at him.

The version I would run more often is the fake. Show the screen, never set it, and let the big react to a threat that does not arrive. He hesitates for a beat. A beat is a layup.

Whoever first drew this up understood something coaches say far too rarely, which is that the most vulnerable man in any scheme is usually the one doing exactly what he was told to do.