Why a single reason persuades better than five

A proposal is backed by one compelling argument. A colleague suggests adding four supporting points for completeness. The revised version persuades fewer people. Nothing was removed and nothing was contradicted, but the strongest argument now shares a page with four weaker ones and is being read as part of a set.

What everyone sees

More support is assumed to be more convincing, on the model of weight accumulating on a scale. That model predicts additivity, and evaluation behaves closer to averaging. Under averaging, the marginal item helps only if it is stronger than the current mean, which most supporting points are not.

What is actually happening

Nisbett, Zukier and Lemley documented the dilution effect: adding non-diagnostic information weakened the inferences people drew from genuinely diagnostic information, even though the addition contained no counter-evidence. Weaver, Garcia and Schwarz identified the presenter’s paradox from the other side, showing that presenters expect additive evaluation while audiences average, so a package including a mildly attractive item is valued below the same package with that item removed. The two findings converge on a counter-intuitive operating rule. Removing your fourth-best argument can raise the persuasiveness of the whole, and the person assembling the case is systematically the least able to see that, because they are adding while their audience is averaging.

Why it stays hidden

The asymmetry hides because the presenter and the audience are running different arithmetic and both feel obvious from the inside. Anyone who has spent a week on a proposal knows what each point cost to produce, and cost is naturally read as weight. The audience never sees the effort, only the mean.