RICHARD WALKER / RWALKER@STARBULLETIN.COM
Parts of Kapiolani Park were pits of mud the day after the Honolulu Marathon. Heavy rain the day of the race made for muddy conditions at the park.
Marathon mired in mess
A technical glitch has jeopardized the results
STORY SUMMARY »
Inclement weather created massive problems with a relatively new electronic timing system in the 35th annual Honolulu Marathon on Sunday, leaving hundreds -- perhaps thousands -- of runners without proof that they completed the grueling 26.2 miles. Others did not receive accurate finishing times.
"We didn't have our equipment weatherproofed as well as we should have," said SAI Timing's David Simms, who has worked with the Honolulu Marathon for 21 years. Intense rain caused an electrical short just before the start of the race, leading to problems with the transponders that competitors wore on their shoes and the software that interacted with them.
Apologetic officials admitted that the effort to streamline the system and save money backfired, and promised to do whatever they could to obtain the information that competitors needed.
FULL STORY »
RICHARD WALKER / RWALKER@STARBULLETIN.COM
Participants in Sunday's Honolulu Marathon picked up their certificates of completion yesterday at Kapiolani Park.
Dave Stackhouse ran the Honolulu Marathon on Sunday in 3 hours, 8 minutes. But he can't prove it, and now worries that he won't be able to qualify for the Boston Marathon.
"I trained hard for this race, and to see my results not recognized is disappointing, to say the least," said the Honolulu firefighter. His half-marathon split registered, but his finishing time did not, making it look like he dropped out. "But I really feel bad for the people who paid a lot of money to come out here. It's not like you can pick up next week and try again. It takes a lot out of you."
Heavy rain created massive problems with a new electronic timing system in the 35th annual Honolulu Marathon, leaving hundreds -- perhaps thousands -- of runners without proof that they completed the 26.2-mile race.
"We didn't have our equipment weatherproofed as well as we should have," said SAI Timing's David Simms, who has worked with the Honolulu Marathon for 21 years. Intense rain caused an electrical short just before the start of the race. Some transponders on tags attached to competitors' shoes rebooted correctly; others did not. And some remained dead, leading to problems at every checkpoint.
Wilfredo Duran picked up his official finisher's certificate in Kapiolani Park yesterday, only to discover that an incorrect time was recorded on it. He stopped his own watch at 4:24. Instead, the certificate indicated that he completed the event in 4:39.
"It was my first marathon, so I was really hoping for something accurate," said the Big Island resident. "It's upsetting in that sense."
Dr. Jim Barahal, president of the Honolulu Marathon Association, took full responsibility. "We're very sorry it happened," he said. "It was a big mistake, and it was my decision to move to newer technology."
Barahal said they decided to switch because the system had been successful at other races with more than 10,000 competitors, including the Las Vegas Marathon. It is also more portable than the previous ChampionChip system, which involved the use of heavy mats, and would allow them to implement split times at every mile in the near future. In addition, it would have saved the marathon up to $300,000 in the next five years.
The waterproof ChampionChip system is used in major swimming, triathlon and running events, but the chip needs to be collected and reused. The demographics and sheer volume of the Honolulu Marathon make this almost impossible. Volunteers can remove chips from 1,700 Ironman Triathlon competitors, for instance, but not 23,000 runners.
The marathon has spent hundreds of thousands of dollars on lost chips.
Nobody is sure how many runners were affected, because nobody can figure out how many people started the race. Simms told officials that only a couple of hundred results are missing, but Barahal fears it could be as many as 3,500. Typically, registration numbers are "soft," said Barahal, because 15 percent fewer people actually show up at the starting line. About 90 percent of those who start end up crossing the finish line successfully. Mathematically, he estimates that 23,500 out of nearly 28,000 registered runners began the race.
"Did a lot of local people say, 'I don't need this'?" asked Barahal, referring to the rain. "I don't know. But based on previous statistics, we may have lost two or three thousand runners in the database." Complicating matters is the fact that some people might not have understood that they needed to remove the tag from their race number and place it on a shoe. Wearing it in the wrong place also would affect the system.
The Honolulu Marathon has always been on the cutting edge. It was one of the first major races in the country to introduce the ChampionChip system, well ahead of iconic races such as the Boston Marathon. But now people have come to expect the precision unavailable when volunteers managed those details manually.
Barahal thinks this incident could have major implications in terms of race sponsorship, though he does not believe people will sue or ask for refunds in light of the efforts marathon officials are willing to make.
"This is a fairly major screw-up," he said, "but we're going to do our best to get finisher data."
More than 95 percent of runners have accurate times, said an apologetic Simms, "but that's still a lot of people who don't. There's no excuse for what happened. Missing one person is too many when you've run 26.2 miles."
How to get your official finishing time
Honolulu Marathon runners who did not receive a finishing time (or did not get an accurate one) may e-mail marathon officials at email@example.com. List your name, race number and approximate finishing time. Everyone was recorded at the finish line on four video cameras used for MyMarathonDVD, with a precise running clock in each frame. Officials can review the tapes and record your time in the database.