Duane Gish was a biochemist with a doctorate from Berkeley who spent the 1970s and 1980s debating evolutionary biologists in public halls, and he almost always won.
He won because of a format problem. In a live debate with fixed speaking times, Gish would deliver a rapid sequence of claims — the fossil record contains no transitional forms, the second law of thermodynamics forbids increasing complexity, radiometric dating is unreliable, this particular fossil was a fraud, the eye could not have evolved incrementally, and thirty more. Each took ten seconds to state. Each required several minutes to answer properly, because a proper answer meant explaining what a transitional fossil is, what an open thermodynamic system is, how isochron dating works.
His opponent had the same number of minutes as he did. He could answer perhaps four of the thirty. The remaining twenty-six stood, and to the audience they looked like twenty-six unanswered points.
Eugenie Scott of the National Center for Science Education named the technique the Gish gallop, and the underlying asymmetry was later stated as Brandolini's law: the amount of energy needed to refute nonsense is an order of magnitude larger than the energy needed to produce it.
Two cognitive facts turn that asymmetry into a reliable win.
The first is that audiences under time pressure use a quantity heuristic. Petty and Cacioppo's work established that when elaboration is low, the number of arguments influences persuasion more than their quality — people count rather than weigh. A list of thirty is more persuasive than a list of three even if all thirty are wrong, provided nobody has the time to check.
The second is that silence is read as concession. In an adversarial format, a claim that goes unanswered is scored as standing. The audience does not distinguish between not answered because it is unanswerable and not answered because there were twenty-six of them.
The technique is not confined to creationism, and it is not always deliberate. It appears wherever the format allows assertion at low cost and requires refutation at high cost: televised political debate, hostile parliamentary questioning, cross-examination, and above all online argument, where the cost of producing claims has fallen to nearly zero and generative tools have taken it further.
What does not work as a response is attempting to answer everything, which loses on arithmetic. What has some traction is refusing the format's terms explicitly. Name the technique to the audience, since a named technique is much less effective than an unnamed one. Select two or three claims, preferably the ones the other side considers strongest, and demolish them in full — the argument being that a source that is wrong on its best material is not worth checking item by item. State plainly that the remainder are unanswered because of time rather than merit, and where the exchange is not live, publish the full answers separately.
And decline live formats that structurally reward the technique, which is a strategic decision rather than a rhetorical one, and often the correct one.
Cross-ref: firehose of falsehood chapter