1. Spread will be calculated using an average of all snap polls from CNN, Ipsos, Morning Consult, YouGov, SRSS, Red Eagle Politics available 48 hours after the debate. The spread of each poll will be averaged. So if polls report A +3, B +10, B+9, then the final spread is ((-3 +10 +9) / 3) = B +5.2. I will not weight the polls by sample size or any other measure for simplicity. If outlets jointly sponsor a poll, such as CNN/SRSS, it will count only once.

2. The ranges widen to compensate predicting outliers. In the unlikely event the result lands exactly on a boundary the winner will be the category with the higher absolute magnitude. In other words, all terms are inclusive on the inside and exclusive on the outside. (For example, a result of exactly Vance +3 would resolve to Vance +3 to +10 instead of as a tie.)

2. Any outlet sampling fewer than 30 people will be excluded.

3. I will void the question if fewer than 3 separate outlets produce poll results on the outcome or if the debate is canceled.

Note: The above outlets include some with minor "leans" in either direction, but all of them produced debate snap polls that tracked the pack of polls for the Harris/Trump debate so seem reliable enough for inclusion and any biases should wash out.