Several recent studies emphasize, correctly, that the biomass of prey captured by an orb web is likely more important than the number of prey in driving the evolution of web designs. Using equations that estimate prey mass from the lengths of captured prey, one study concluded that rare, long-bodied prey contribute the large majority of energy obtained by orb weavers in general, and thus that the designs of orb webs should generally reflect selection favoring the capture of larger insects, especially the ability to absorb high-energy impacts. I show here that the weights of long prey have sometimes been seriously overestimated by these equations. In addition, the longer prey captured by spiders probably represent highly biased samples, in terms of both low weight/body length and low momentum/body weight, of the longer prey available in the environment, leading to overestimates of the kinetic energy that the orb needed to absorb to stop them. Deductions concerning how selection acts on orb designs that have been based on the prey captured are also flawed, because additional data on prey availabilities and prey escapes are needed to evaluate the possible gains and losses from different orb designs. Still another complication is that data on prey abundances in natural rather than altered environments are needed to understand past selection pressures that produced present-day web forms. I conclude that a dominant importance for rare large prey in orb web evolution has not been conclusively demonstrated to be a general rule for orb weavers. A more inclusive approach regarding orb functions is prudent, especially because many traits that improve some functions have opposite effects on others.