Saved by Google Earth: A Last-Minute Check Barely Averts Disaster. Or, Why We Measure Twice
It had been a long day, starting when I was awakened by a call from Jim Gerweck,
who had heard that the Cherry Blossom 10 Mile had been diverted due to a police
emergency. At Jim's suggestion I called race director Phil Stewart, who confirmed
the story and then described exactly how and at what point the runners were
diverted from the course. Since course measurer John Sissala was out of town, the
next question was, when could I measure to find the actual distance run? I told Phil I
would do it that day.
With beautiful weather and cherry blossoms at their peak, this was not really the
greatest day to measure downtown. At least I knew there would be no place to park,
so I made the trip (15 miles there and back) by bike. (Frustrated driver to me on
bike: "Do you know where I can park?" Me to driver, shaking my head: "Why do you
think I'm on a bike?")
The bad news is some of the streets were chock full of cars; the good news is, they
weren't moving, so you could squeeze alongside as long as you kept a sharp lookout
for doors popping open (folks jumping out to get to a bathroom, to take a picture, or
to actually get somewhere other than the traffic jam).
Idle question: you are stuck in a traffic jam and cannot get to the actual cherry
blossoms. If you jump out and take a picture, does that count as being there? Or
something like that?
After a couple of slow-moving hours, I had my results: the course was short by
0.46357 miles, or about 816 yards. I reported all this to Phil Stewart, and I worked
out 10-mile equivalent times for winners and other key finishers. Phil told me they
were going to wait a day or so before announcing this result. It turned out that one
of those converted times would have "broken" an American record (the event offers
a bonus up to $10,000 for breaking Greg Meyer's 46:13 set in 1983, and/or Janet
Bawcom's single-sex American record of 52:12 set in 2014). The race committee
needed to decide how they were going to handle that before word got out. The
delay would turn out to be very lucky for me.
Fairly late on Sunday evening, I decided to write up the measurement results while
everything was still fresh. To further illustrate what had happened I thought I would
create a Google Earth measuring path for both versions of the course: the correct
course, and the path the runners were forced to take that morning. I remember
thinking, "These paths are really good-this result is going to be within a few meters
of what I found out there on the course." Then I clicked on the two distances,
subtracted, and found to my horror that the difference was something like 1078
yards, nowhere near the 816 yards I thought I had measured!
I look again at my notes. Here's the point where the emergency course and the real
course come together. I write "CP (for common point) 1st pole on Ohio after curve".
Oops. There is no number written there. The line below has a number, but beside it,
the note "S (south) edge, 14th St Bridge". Well that point is beyond the common
point. Right away I get it: looking at these notes earlier, I was thinking "I had to have
written a number down at the common point-it must be this one on the line
below." Wrong-the truth is I just didn't write down a number at the common point.
I just wasn't willing to face the truth at the moment (and I sure didn't want to go
back through the traffic jam and measure again!).
I write a quick note to Phil ("Hold everything . . . I messed up bigtime . . . will
remeasure . . ."). It was after midnight but I knew I wouldn't sleep well if I left this
for the next day. I went down, calibrated, rode both ways, recalibrated, and basically
confirmed the GE measurement. The course was 0.61347 miles short, about 1080
yards-the actual distance run was 9.38653. The conversion factor is (time run, in
seconds) x 1.06536 = (projected 10 mile time in seconds). Significantly, this means
that the best American male time of 43:38 converts to 46:29 (>46:13).
I texted all this to Phil Stewart, and then drove home. I slept well, knowing I had
narrowly escaped extreme embarrassment-thanks to the Google Earth check and
to the race committee's delay in making an announcement. And I had relearned
some lessons I had allowed to rust a bit:
- Always write down a number and what the number means! (A good habit is read the counter number, write it down, read the counter again, look again at what you wrote. If only I'd done half of that!)
- Measure twice. It helps avoid getting seriously embarrassed!
- Check your work.
- Measuring at night can work really well. My midnight measurement took 30 minutes, including both pre- and post-calibration-compared to about 2 hours during the daytime (not counting commuting time). Of course you need lots of lights, flashing and other, when you are doing this. I use a bright helmet-mounted light that works better than constantly fishing out a flashlight.
- Sometimes a little delay can be a good thing. It was this time.