The tiny Cuban tree frog can jump an impressive 1.7 m, but its relative, the larger and more muscular bullfrog, can rarely muster anything farther than 1 m in the lab. Initially, the bullfrog's dismal jumping performance was blamed on a trade-off caused by the need to jump and swim; however, Henry Astley, a PhD student from Brown University, USA, wasn't convinced: ‘Other papers suggested that they had a catapult mechanism like the tree frogs, where they stretch an elastic tendon and then quickly recoil, like a bow and arrow, and if so, they should be doing a lot better than they appear to be doing in lab.’ What's more, The Guinness Book of World Records documents the achievements of ‘Rosie the Ribeter’ a wild bullfrog who, in 1986, at the annual Calaveras County Jumping Frog Jubliee, jumped a colossal 2.2 m. Was Rosie's jump just a once in a million fluke or were bullfrogs indeed able to jump this far? If so, why weren't they performing to their maximum potential in the lab? Astley decided it was time to find out (p. 3947).

Along with other lab members, Astley made the trip out to Angels Camp, CA, USA, for the 84th session of the fair. Over the course of 4 days, the team filmed jumps from both ‘rentals’ (frogs rented from the fair organisers) and ‘professionals’ (frogs hand-selected from the wild by teams that competed annually). During the breaks, the team would lay down a measuring grid for calibration purposes, but on the whole the team were able to sit back and soak up the atmosphere. ‘It's a tremendously fun fair to go to and everyone was so enthusiastic about the frogs’, says Astley, jokingly adding, ‘Plus, how often do you have fieldwork were you eat funnel cake and hot dogs?’

Back in the lab, however, the work began in earnest, and several undergraduate students were recruited for the daunting task of digitising over 3000 jumps. Astley was then able to calculate take-off velocities and angles, and estimate the power used during the jumps. However, what really stood out was that most of the frogs outperformed frogs in the lab; rental and professional frogs' jumps averaged at 1.1 and 1.5 m, respectively. Even more impressive was just how close some frogs came to Rosie's world record – clearly bullfrogs are capable of enormous jumps. So, what was the trick? Astley explains that professional competitors are secretive about their ‘trade secrets’ of where to collect frogs and how to look after them and motivate them to jump on stage. However, overall they didn't look dramatically different to the rental frogs and the only change was that professional frogs were kept at warmer temperatures.

Astley wondered whether it was just a matter of probability, so he randomly took samples of frog jumps to find out what the chance of seeing a long jump was. He revealed a non-linear relationship between sample size and jump distance. For example, a sample size of 10 rental frogs (the equivalent of the lab frogs, which aren't selected by ‘professionals’ with decades of experience) gives you just a 14% chance of seeing a jump over 1.6 m, while increasing the sample size to 50 frogs gives a 56% chance. In conclusion, Astley says researchers would have to process a large number of frogs to stand a chance of seeing an impressive jump, but as he jokes: ‘We can order a dozen frogs from our supplier, but a hundred frogs? The animal care bills would bankrupt us!’ So, perhaps, fairs like the one at Calaveras represent an underused resource in the field of animal performance.

• © 2013. Published by The Company of Biologists Ltd