Here's an article I found in today's GIS magazine.
http://www.directionsmag.com/editorials.php?article_id=882 Enjoy. -- Justin
Man and Machine: Human Errors Lengthen the Lakeshore Marathon By: Adena Schutzberg
(Jun 10, 2005)
"We can put a man on the moon so why can't we cure the common cold?" I'd like to ask a geospatial question like that. "We can navigate cars via GPS, track palettes of goods with RFID chips and explore the world on Web portals, so why can't we get the length of a running race correct?"
A Run Too Long
I found (and was sent by several friends) a rather disturbing article about the Lakeshore Marathon held last on Memorial Day, in Chicago. The bottom line was that due to a number of issues the 26.2 mile course was extended a full mile, to 27.2 miles. (All marathons today are supposed to be 26.22 miles, to commemorate a run from Marathon to Athens, Greece, when Phidippides ran to announce the news of victory by the Athenian army over the Persians in the fifth century B.C. That first marathon in Athens was actually about 25 miles.) It was runners, some with just watches and others carrying GPSs, that suspected the error during the race, according to a Chicago Tribune article.
The founder and organizer of the race noted "last minute changes to the course" due to "a section that was no longer permitable" and "some last minute construction in the park." He wrote on a marathoning website those changes "caused us to miscalculate and we foolishly added an extra mileÉ" Apparently the addition of a new section around Navy Pier, announced on May 1, and some construction at Diversey Harbor caused those changes. He's looking for a new race director for next year.
Reaction
Runners, including one person who ran the race, felt further "corrective actions" should be taken. He reported his concern to the Chicago Area Runners Association and the Boston Athletic Association (BAA), which organizes the Boston Marathon. Some runners were attempting to use the Lakeshore event to achieve times that would qualify them for next year's Boston Marathon, one of the most prestigious (and one of the few with qualifying standards) in the country. The BAA listens to "extenuating circumstances" and sometimes grants waivers to runners who don't quite make the qualifying time.
How are Courses Measured?
Now, before I go any further, I've never been a race director and I've never run a marathon. But I do know organizing a race, even a smaller, shorter one, is tough. There are lots of things to worry about including creating a safe course, providing medical support, getting sponsors and volunteers, accurately timing, and basically worrying about every runner, be there 50 or 20,000.
I went right to the source to look at how today's modern races are measured here in the United States. The body that oversees official competitive track and field is USA Track and Field (USATF). They do everything from laying out rules for certified courses, sanctioning races and selecting those who will run in the Olympics.
It turns out it's not too hard to measure a course, it's just time consuming. Those doing the measuring are either experienced measurers or just regular people. They are directed by the USATF to use the calibrated bicycle method, which requires a specific type of measurement device that attaches to a bicycle. The number of wheel rotations to cover the course distance is measured using the bike, then compared to the number required to cover a known distance (a standard calibration course, measured using steel tape). The gory details are here.
Further, since courses are often on wide roads, exactly where on the road the bike is ridden will affect the distance. There are guidelines for that, too, but there are many judgment calls to be made. The bottom line is that everything is done to ensure that the course is not short. If a new record is suspected on a course, and it's found to be short upon re-measurement, the record will be thrown out.
Once all the measuring is done, the measurer produces a map. It need not be to scale but must identify, based on permanent features, the start, finish, intermediate mile markers and other considerations. Here's the map for a little 5k run in my city last year.
Lakefront Marathon Certification
Armed with that information, I found the Lakeshore Marathon's course certification number on its website: IL02023JW. I plugged that into the course search tool on the USATF website and learned that number was replaced by IL03045JW. (IL is for Illinois and the JW are the initials of the person who approved the submitted application for course certification.. The 03 indicates that for the year 2003, 45 courses were certified in the state up to that point in the year.) The map lays out a Å marathon loop (a bit over six miles with each mile marked). It was dated June 8, 2003.
The only map I could find from the marathon organizers was from an "old" website. (That website appears to have been taken off line this week, perhaps due to the controversy.) It seems to lay out a single "big loop" with a start/finish at 18th Street. The current marathon website does not even include a link for a map, though the website notes the start/finish "on the Great Lawn on the East side of Soldier Field next to Burnham Harbor on Museum Campus Drive" and "just south of McCormick Place Lakeside Center, adjacent to the Fallen Firefighters and Paramedics Memorial."
Certification and Course Changes
I contacted my local expert on courses and measurements, Steve Vaitones, the USATF New England Managing Director. He's measured many a race course (including the map noted above) and watched the process evolve. He even wrote an article on the topic.
Vaitones noted that "There is hardly a 'major' race that isn't certified É Many smaller local races don't certify their courses; some don't see the need and feel a car or a bike computer is 'close enough.' Some see it as an unwarranted expense - but doesn't any race that charges a fee have a duty to give an honestly measured course to their runners? There are some [courses] that can be classified as uncertifiable - going over trails, through undefined parking lots and driveways without being coned, etc."
He explained that "The bigger the race, the more an event has to expect the unexpected." For "the Boston Marathon, we've measured out an alternative finish line on Commonwealth Ave. that could be used if something happened race morning or during the race that closed down Boylston St. Some other major events probably have back-up plans É I drive in to Boston on marathon morning from Waltham via the Newton hills just to be sure there are no surprises; one year there was a water leak in front of Brookline post office, but it was quickly remedied without excavation."
I asked if race directors are obligated to re-measure if changes occur to the course. "Well, if it is advertised as certified, yes." Vaitones went on to note that depending on when the change happens it may well cost some money (for a measurer on short notice or a police detail to accompany him or her) to re-measure.
He's quick to note that sometimes changes are required on race day. He points to a race my club hosts in Cambridge, MA. The day of the race there was a gas leak. The section of road where it occurred was used not once but twice in the race. The race director tasked Vaitones with coming up options as the starting time approached. Vaitones offered two and the local police selected the one to use.
But the course was no longer the promised 5 km (3.1 miles). How much longer was it? Well, there was no bike or wheel on site (it was only an hour or two before race time) so Vaitones resorted to pacing off the added distance and moving the mile markers accordingly. He estimated the distance at just over 3.4 miles instead of the promised 3.1. An announcement was made about the updated distance at the start, and Vaitones reported that his assignment during the race of calling out the distance to go near the finish did not go unnoticed. "É The front runners appreciated the notice of the length É and the back of the packers simply appreciated knowing how far they had to go." Had there been even a single day's notice, he offered, the course would have been re-measured and recertified.
More on Lakeshore Certification
Why, I asked, did I not seem to find a certification map for the marathon on the USATF website? Vaitones suggests there may be a good reason for that. First off, "the paperwork that is sent to the regional reviewer only needs to be postmarked before race day," meaning it may well have reached the USATF reviewer and not yet be posted due to the various people who need to look at it. He offers that there are two places in the process where the extra mile might have easily been found: at "1) the measurer and 2) the regional reviewer going over the paperwork. At least one of these should have spotted the extra mile. Part of most maps you will see on the site have a description of each mile mark location. A 27th mile could easily be spotted simply by listing the mile split locations."
As for the existing Å marathon map found under the marathon listing, Vaitones offers that due to human error the map's number might have been switched with another one. It may have occurred during updates to the website. "In 2003, the Road Running Technical Council was in the process of scanning every map from the thousands of certified courses, taking these from paper files." Jay Wright, whose initials JW are on the Å marathon course, notes on a discussion board about certification, "It certainly appears that the race director changed the course for the 2005 event but made no serious attempt to have the new course certified." He also notes that he has "no record of him [the race director] EVER having had the half marathon course certified."
Advice and the Future of Course Measurement Vaitones' advice when something like occurs? "É Rule number one is that race directors need to 'fess up right away." He also is sympathetic that "it's probably also more of a problem the longer the race gets, and certainly with a marathon where there is a cadre of people who are looking for meaningful times for long distances and in particular, qualifying times for the Boston Marathon." He learned that "the [Lakeshore] race director has provided the BAA a spreadsheet of runners with projected times for a mile shorter based on the course overdistance." He expects that the BAA will likely accept those runners whose recalculated times based on the distance run would meet qualifications into the Boston Marathon.
What about the current technology and future technology for measuring courses? "As a starting point for any measurement and particularly for course adjustments, a push-wheel produces a pretty good distance, but just to within a few yards. Right now, the only approved method for certification is the calibrated bicycle method, but the USATF road running technical council is doing a good bit of work on electronic and GPS systems as we speak. But even after all the measurement work is done and maps are produced, the course ultimately has to set up on race day as it was originaly measured to be sure the proper distance is covered."
Lessons Learned
What astounds me is that with all the technology we have, such grave distance and location errors can occur. And, that should not astound me. All the tools in the world, geographic or otherwise, need human involvement and oversight. Technology did not fail the runners, some (I like to think) well meaning people did. I give the organizer credit for stepping up and taking responsibility, and admitting his limitations as an organizer. This is a striking reminder that technology is not enough.