Len Testa and the Math Behind Your Theme Park Vacation

Last month GeekMom Dak reviewed Touring Plans, a website and app that helps you plan your Disney vacation and knock hours off queuing times at the theme parks. Touring Plans‘ features included crowd calendars, wait times and customisable plans that allow you to select the attractions you are interested in seeing each day before the site gives you a detailed, unique itinerary. But where does the data for such a system come from and how do you go about creating a website that can instantly produce such a detailed plan for the millions of permutations each park offers on a single day? I spoke with Len Testa, the founder of Touring Plans and co-author of The Unofficial Guide to Walt Disney World, about the mathematical side of planning your dream Disney trip.

Probably the most straightforward example of the time-dependent travelling salesman problem is the kind of scheduling that a company like FedEx or UPS has to do for one of its drivers. The company’s goal is for the driver to deliver packages to customers in different locations while minimizing the overall cost, including labor and fuel. At any point in the day, the FedEx driver has to take into account not only the distance between his current location and the next customer, but how much traffic will delay him when he’s on the road to that next customer. For example, the driver may decide to take a 4-mile detour on a rural road to get to the next customer, rather than drive a 1-mile stretch of I-95 at 5 p.m. on a Friday. The I-95 segment may be shorter, but the rural road is faster because it has less traffic. The trade-off there is slightly higher fuel cost for much lower labor costs.

After I finished my undergrad degree (also in computer science), I visited Walt Disney World the summer before I started graduate school. One day during that trip I waited in line almost two hours for the Great Movie Ride. Sometime during that wait I thought that there should be an app for minimizing your waits in line at theme parks.

I went back to my thesis advisors and discussed the problem. They proposed a literature search, which showed it was a suitably difficult problem. Once they gave their approval, I contacted Bob to see whether he’d share his data from the book.

It turned out that he was using a different approach than I was envisioning, so we didn’t get to share data. But Bob was exceptionally generous with his time, explaining how his modelling worked and what to look out for when creating a schedule for theme parks. We stayed in touch through my graduation, and I started joining Bob’s team for in-park research in 2000. Because I was spending so much time in the parks for touring plan research, I started updating other sections of the book when it needed to be done. I became co-author of the Guide in 2007.

Two things make the Unofficial Guide book, the Touring Plans website and the Lines app different: First, our research is consumer-oriented. That means that we’ll tell you in plain language whether an attraction isn’t worth your time, or a restaurant isn’t worth your money. Second, we’re a data-driven organization. Our staff consists of scientists applying their knowledge to travel problems, which is unique in the travel publishing industry. This allows us to tackle things like touring plans, which are complex scheduling problems. It turns out that there are quite a few vacation questions that can be answered through science, math and operations research. Finding the least-expensive combination of Disney admission tickets, for example, is a bin-packing problem.

The other thing that makes our app different is that we’ll estimate how long you’ll actually wait in line at a given ride at a given time of day. Every other app just tells you Disney’s posted time, or (worse) tries to estimate Disney’s posted wait time because they don’t have people in the parks feeding them data. Any theme park veteran will tell you that the wait time posted outside an attraction isn’t how long you’ll really wait. Sometimes the posted waits are set artificially high on purpose, as a form of crowd control, to get people to get in line somewhere else. Sometimes the waits are set high at the end of the day to dissuade people from getting in line, so management can close the park on schedule and keep their labor costs low. And sometimes the posted waits are too low, because the kid staffing the sign got caught up doing something else.

A lot of it is the same for any organization, including Disney. We’re looking for bright, self-directed, team-oriented people. Because we’re both writers and scientists, we probably put more emphasis than other companies on the combination of fact-based decision making and strong oral and written communication.

I spent a long time doing architecture in American Express’ technologies group prior to joining the Guide. AmEx Technologies is an excellent place for computer scientists to learn how to run a company; their leadership team is level headed and fact-based. They make their tech teams responsible for rationalizing tech investment to the business group giving the funding. You learn how to verify that your idea makes business sense and how to communicate the investment to an audience whose skills are outside of technology.