Comfort Needs No Denominator Unbundling Emotional Value, Certainty and Guidance in the Market for Confidence

Abstract Success coaching, motivational content and much classics-for-life teaching are sold as one product containing three goods: comfort, which acts on the buyer's own state; certainty, which scales up the position a buyer takes without improving the buyer's estimate; and guidance, a claim that following the advice improves outcomes. This paper shows that the three differ sharply in who can read them and what evidence they need. For comfort the experience is the product; for guidance the effect is the product. The reading by which a buyer judges comfort meets all four conditions of reading ownership; the reading by which the market judges guidance fails three. By Wald's bound and the two-arm convention of [S1], about ten sessions alternating with and without a source of comfort let a buyer identify a comfort effect of one standard deviation; identifying the sign of a five-point change in a 20% success rate needs about 670 independent outcomes split across two arms — a denominator no single buyer possesses. Testimonial counts identify the product of audience size, success rate and reporting rate, not the success rate; under selective reporting the observed odds of success are the true odds times the ratio of reporting rates, a bias that does not shrink as reports accumulate. A sincere seller's certainty has two sources that must be kept apart: reading resonance as evidence about guidance makes the seller's subjective precision about guidance grow with the audience while the information about guidance stays at zero; reporting bias on the outcome channel survives every update rate. Humility therefore cannot be exhorted; it can only be engineered with a denominator. In a Kelly model, certainty that inflates the perceived edge or deflates the perceived variance by a factor m sets the relative position to m, and m ≥ 2 makes ruin certain; certainty can also flip the sign of the perceived edge. Demand for certainty is strongest among the buyers most exposed to personal ruin; but on favourable games with reinjection the tail index equals minus the survival number per unit of reserve, so the dose that does most harm to the collective tail is not the heavy one but the moderate one that pushes a near-Kelly buyer just past the ruin point. Bundling is selected by multitask incentives; unbundling keeps comfort and removes only borrowed credibility. But comfort's own side effects — closing the drive, rewriting set points, building dependence — are not readable by the buyer, and they need a denominator. Ten predictions are registered; executions on external data: zero.