THE VIEW FROM HERE
The Shock of the Large

   Many of this issue’s articles deal with the challenge of working with large datasets. I find the idea of “large” quite intriguing. I always watch with rapt attention when profiles of families of six or more children appear on TV. (I have but one brother.) I’m a bit nervous in big cities. (My hometown had only 25,000 people and even Boston, the center of its universe, is on the small side of U.S. cities.) I enjoy exploring the mountains of the West, such as the Rockies. (I grew up climbing the 3,000 and 4,000 footers of New Hampshire.) These big concepts are intriguing to me because they are outside my experience. With some research, some preparation, and some trial and error, these and other “large” challenges can be managed and even grow even larger, in some cases, with even greater returns.

   This idea was driven home last month. I participated in something big, something outside my experience, the Pan-Massachusetts Challenge (PMC). This is the 25th year of the cross-Massachusetts bike ride to raise money for cancer research and treatment at Dana-Farber Cancer Institute through its Jimmy Fund. The PMC is large. This year’s event hosted 4,000 riders who were supported by 2,000 volunteers, of which I was one. The ride crosses 46 towns, with 6 different one- and two-day routes. The only thing that’s not large about the PMC is its paid staff of just eight full and part-time people.

   Consider for a moment the geographical extent of this event. Each resident along the route is sent a personal letter explaining what to expect during the weekend of the ride. And, either because the route changes, or because new houses and businesses are built, the notification list must be updated each year. Then there’s the challenge of signage. The route is marked with arrows the night before the ride. And, despite the letters and the high profile of the event, some residents along the route, for whatever reason, simply remove them.

   Mapping the route is not trivial, either. Each turn is noted and numbered, creating a list of segments. Each ten-mile or so section is mapped on 8.5 x 11 paper, paired with its corresponding directions, and copied into notebooks carried by each support van. (The software of choice? Microsoft’s MapPoint.)

   Let’s get back to the riders, since their sheer numbers are an intriguing management challenge. Consider something simple: serving lunch for 4,000. The good news is that riders don’t all hit the lunch stop at the same time. The first rider appeared at 9:30 a.m. and the last around 3:00 p.m. In between, all 4,000 stopped by for food, Gatorade, a massage, sunscreen, and a healthy dose of encouragement before heading for the day’s final destination on Cape Cod.

   To provide for that many hungry, tired riders the lunch crew had to assemble enough bike racks for safe parking; stock a half-dozen very long buffet tables with thousands of hand-made sandwiches, fruit, cookies, granola bars, and other goodies; unpack, chill, and manage case upon case of water, juice, soda, and yogurt smoothies. And then there was trash. No sooner were bags removed, than they were full again. But, despite my amazement at the number of bikes on the racks by 10:30 a.m., the big group never seemed unmanageable.

   Why? The large group was expected, and plans were in place months before the first rider appeared. The group’s needs were defined, broken down into chunks, and each one was assigned to a different team. And, of course, there was good management.

   While managing people and bicycles is quite different from managing megabytes and gigabytes of data, there are clearly themes from one that can illuminate solutions in the other. While large datasets may be intriguing, the goal of this issue is to keep them from becoming overwhelming.

Adena Schutzberg, Editor

Back