February 26, 2003
Endless hotcakes make for hot debate among hungry engineering students
The International House of Pancakes restaurant just west of the University of Washington campus in Seattle is not usually considered a hotbed for science.
But introduce a few hungry electrical engineering graduate students who’ve been working all day on problems involving probability to an all-you-can-eat offer, and soon the formulas — and the forks — begin to fly.
“We go to the IHOP after studying sometimes,” said Andrew Morabito, a first year EE grad student from New Jersey. On a recent outing, the group noted the restaurant’s latest promotional offer: bottomless pancakes for $3.99. Naturally, talk of a pancake-eating contest ensued, but not before the conversation took a scientific turn.
“We’d spent most of the day working on a problem set, which had all types of probability problems, so I guess we just had that on the brain,” Morabito said. “So one of the guys said that it wasn’t really official because the size of the pancakes wasn’t consistent. It just got more engineering-oriented from there.”
Fellow students Scott Philips and Jennifer Price began to devise a model for the size of a typical pancake. The variances introduced by such factors as an imprecise batter dispenser, how the batter splatters on the griddle, and the height from which it’s poured led the group to conclude that each serving could be modeled as “pancake plus noise” and that size must be a Gaussian random process, or a probability of pseudo-random numbers given to a specific source.
Richard Powers, another student in the group, summed up support for the Gaussian random-process view of pancake phenomena. “Everything in nature is Gaussian,” he said.
The discussion turned to application. “We talked about using a low-pass filter,” Morabito said, which in pancake terms would mean stamping all the pancakes with a cookie cutter smaller than the smallest pancake to get an even size. “Then we realized that there are variations in height, and after we stamped it we would have to feed it through something that takes the top off. Then, of course, you can still get into questions of uneven mass inside because you could have bubbles, so the question becomes ‘Is it standardized in volume or mass???'”
The students went ahead with the contest that night anyway ? Morabito won with 13 flapjacks to Philips’ 8 — but the debate continued.
Word got back to the Department of Electrical Engineering main office, where amused staff members forwarded the debate to the faculty. Soon professors weighed in with their views on flapjackology.
“I was thinking that, because of the insoluble globs of goo in the batter, the diameter would be roughly quantized to integer multiples of the mean glob size,” wrote Professor Bruce Darling in an e-mail. “That would make the diameter a Poisson distribution.”
Professor Rich Christie added that the group should take into account the unlikelihood of having a negative pancake.
“I would also point out that since it is not possible to have a pancake with a diameter less than zero, the distribution size is probably closer to log-normal than Gaussian,” he said.
The debate fueled a second contest at the IHOP, with nine student competitors downing a total of 100 pancakes. “We pretty much took over one whole side of the restaurant,” Morabito said.
The discussion continues, he added, but he doubts the group will be returning to the IHOP anytime soon.
“The special ended last Sunday,” he said, and the prospect of grant funding to continue the research doesn’t seem likely. “Besides, I think everyone”s a little tired of pancakes right now.”
###
For more information, contact Morabito at morabito@u.washington.edu. For the students’ account of the events, check the Web at:
http://students.washington.edu/rpowers/20030201_pancakes.html
