The hot hand argument keeps changing shape to survive its critics

The claim and the test diverged immediately

Players and observers describe an experience in which shooting feels easier and the basket appears larger during certain stretches of a game. The statistical version of the question asks whether a make raises the probability of the next make, holding everything else constant. Those are not the same claim, because the experience concerns a state while the test concerns a sequence of outcomes.

A player could genuinely be in a better state and still show no sequential dependence if the state changes what they attempt. Much of the long-running disagreement comes from each side answering a different question and treating the other as refuted.

Shot selection moves during a streak

A player who has made several attempts will typically take harder shots afterwards, from further out and against closer defence. Defences also adjust, assigning more attention to a player perceived to be shooting well, which raises the difficulty of subsequent attempts. Both effects push the observed success rate down during exactly the period when the underlying state is supposed to be elevated.

A test that ignores difficulty therefore has the two effects working against each other and cannot separate them. Controlling for difficulty is possible with modern tracking and changes what the analysis is capable of detecting.

Counting rules do a great deal of work

How streaks are defined, how sequences are sampled and how attempts near the end of possessions are handled all affect the resulting estimate. Reasonable analysts make different choices and obtain different answers, which is a sign that the data underdetermines the conclusion. Subtle properties of how sequences are selected can bias estimates in directions that are not obvious without careful thought.

That such subtleties matter at all indicates the effect being sought is small relative to the noise in the measurement. A small effect is not the same as no effect, and it is also not the effect the original experience described.

Why belief survives regardless

The experience of shooting well is vivid and internally consistent, and no statistical result is likely to change how it feels. Human perception also reliably finds patterns in short random sequences, which means the belief would exist whether or not the effect did. Both of those can be true simultaneously, so the persistence of the belief is not evidence either way.

Coaches must act on something, and treating a player as being in rhythm changes deployment in ways that may be justified for other reasons. A decision can be reasonable without the belief that motivates it being literally correct.

Where the question actually stands

The honest summary is that the crude version of the claim is not well supported and that carefully controlled versions have produced mixed and contested results. The effect, if present, appears small enough that it would not justify the confidence with which the concept is invoked during broadcasts. It is equally unjustified to declare the matter closed, since the analyses depend on choices that remain under active dispute.

This is a genuinely unresolved question in which both confident positions have outrun the state of the argument. Sitting with that is harder than picking a side and is the only defensible option available. The dispute has run long enough that both camps have institutional commitments, which makes a clean resolution less likely rather than more.