Lessons from My Almost-DNF Marathon

Have you ever had a marathon that did not go well?

How about one that made you anxious toward the end?

A few days ago, on January 15, 2017, I ran the 2017 Chevron Houston Marathon (“Chevron”).

Here are my final chip-times, splits at 13.1 miles, and Houston weather for my previous four Chevrons. The weather data are from Weather Underground. (No, they are not the leftist terrorists!)

If you knew nothing else, I should have gotten a similar 5:30-ish chip-time in 2017, right?

Not so fast. Literally.

Two days before 2017 Chevron, I got an email message with this subject-line: “HOUSTON MARATHON ORGANIZERS URGE RUNNERS TO SLOW DOWN AND ADJUST PACE FOR SUNDAY’S RACE”

“Slow down, slow down, slow down,” said Dr. John Cianca, medical director of the Houston Marathon Committee. “It’s going to be hot and humid – not ideal weather – so run responsibly and adjust your pace to reduce the risk of overheating. This is not the year to set a personal best. It will be more difficult than you anticipated, so make adjustments.”

Sure enough,…

• I saw many more medics on bicycles during 2017 Chevron than in earlier Chevrons.
• I saw many more medic tents along the route than in previous years.
• We started with yellow weather-alert flags (“MODERATE: less than ideal conditions”).
• We finished with red weather-alert flags (“HIGH: potentially dangerous conditions”).

Adding weather for 2017 Chevron to the table, we get…

I’m not trying to be boastful, but the forecasted and actual weather did not make me anxious.

1. I train in Houston, so the humidity did not bother me too much. 2017 Chevron had overcast skies and some nice gusts of wind along the route, which cooled us a bit. Visibility was down to 6 miles due to fog and rain, with tops of buildings covered in fog. And, I enjoyed a few minutes of cooling rain in the 25K-30K segment. Unfortunately, we then had clearing skies, with the sun steaming up the pavement.
2. I ran the Rock ‘N’ Roll San Antonio Marathon on November 15, 2009. Weather Underground says that the mean temperature was 71° F and that the mean humidity was 81%. Those are numbers from San Antonio International Airport. It felt closer to 85° F in downtown San Antonio on that day. My 2009 San Antonio chip-time ended up being 5:50:13.

No, the anxiety-inducing part for me was not the weather. I knew that it would slow me down. I should have expected something similar for 2017 Chevron as for 2009 San Antonio.

Now what would you guess my times for 2017 Chevron were?

If you guessed slower than in previous Chevrons, then you were correct!

Here is how those chip-times look when attached to the mean temperatures:

Between 2016 Chevron and 2017 Chevron, I made two positive changes:

1. I spent a full year running on the balls of my feet, using the Pose Method. Before I learned the method in late 2015, I was a heel-striker.
2. I dropped my body weight by 15% — to my ideal weight — in the second half of 2016 to de-load my lower joints. I did this by taking 10,000 steps a day and tracking consumed Calories with MyFitnessPal.

I hoped as 2017 Chevron started that these might compensate for crummier weather.

Wow, was I wrong!

Beyond the crummier weather, what could slow me down by almost a minute/mile between 2016 and 2017?

• I did not train as much. Bursitis in a hip joint in early November, 2016, led to four chiropractor visits and lots of home therapy. I also missed several Saturday-morning long runs in November and December.

The crummier weather plus the incomplete training showed up at 13.1 miles in the 2017 race. I started well in 2017 Chevron. But, my 2017 13.1-mile split (2:38:39) was eleven minutes and forty-nine seconds slower than my 2016 split (2:26:50).

Look at my segmented paces in 2016 vs. 2017. A “segmented pace” is the pace within that segment only, such as between the 5K flag and the 10K flag.

2017 Chevron was faster than 2016 Chevron for me in only two segments of the race.

• 0-5K: I started near the back of corral D in 2016; I started at the front of corral D in 2017.
• 25K-30K: I took a toilet break in 2016; cooling rain fell in 2017.

I trained with the 3:1 run:walk ratio for most of 2016. I even had a 10:27/mile mean pace across a 9-mile 3:1 training run in mid-October, 2016. This made 3:1 my ratio of choice for 2017 Chevron. I persisted with the 3:1 method for the first twenty miles. I covered the next two or so miles with more walking than running. I covered the next two or so miles with an even mixture of running and walking. And, I returned to 3:1 method in the final two or so miles.

So, what was anxiety-inducing about 2017 Chevron?

The anxiety came toward the end of the marathon.

• I knew that I had to get a chip-time of six hours or faster to avoid a “DNF” (Did Not Finish) in the race.
• I never want a DNF. (Does anyone?)
• I especially did not want a DNF in this race. You must complete ten non-“DNF” Houston Marathons to become a “legacy” participant. From what I had read, legacy Chevron runners get some extra privileges. 2017 was to be my tenth Houston Marathon.
• Finishing in <6:00:00 requires a mean pace faster than 13:44/mile.
• My Garmin Forerunner 235 was reporting a mean pace of ~13:20/mile toward the end.
• I knew that this was wrong.
• I did not know HOW wrong it was until I finished the race.

Look at these two mean-pace values that I had at the end of 2017 Chevron:

• 13:18/mile was my final, Garmin mean pace.
• 13:33/mile was my final, actual mean pace.

Why the difference?!

• My actual pace comes from the timing mats measuring 5:55:08, divided by the official distance of 26.21875 total miles.
• My Garmin pace comes from its stopwatch measuring 5:55:07, divided by its GPS-driven estimate of 26.70 total miles.

My Garmin overestimated the 2017 Chevron course distance by almost 0.5 miles!

I wrote about this kind of GPS error many years ago in this article. Every GPS watch that I have worn or discussed with others can overestimate distance.

I started to see the overestimation from the first mile-marker flag. As I approached each flag, my Garmin beeped for that mile even farther away. And, after I entered the final stretch, my Garmin beeped for “26 Miles” even though I could not even see the “26” flag ahead!
[Tweet “When a GPS watch overestimates distance covered, you believe that you’re faster than you are!”]
I knew that my Garmin was lying to me about my actual pace, but I did not know exactly by how much.

Only after I finished the race — in 5:55:08 — did I appreciate how close I was to a 6:00:00+ DNF.

You see, I had not set up my Garmin to display elapsed (stopwatch) time. Instead, I kept trying to remember the clock time when I crossed the timing mat at the start. I wasn’t sure when I crossed that mat, so my anxiety rose in the final miles. I knew that I was close to the six-hour limit.

Each of the course’s timing clocks in the final miles showed a duration greater than 6:00:00. I knew that this could be okay, given that I crossed the start line’s timing mat long after the race began. What I did not know was exactly how much after the blast of the gun I had crossed the start mat. Yikes!

Lessons Learned

I never want to experience again this “ignorance anxiety” at the end of a marathon. (Who would?)

So, here are lessons that I learned from 2017 Chevron:

• The Garmin-watch “Average Pace” data-field is worthless for a marathon. It may be okay for training and for shorter-distance races. But, Garmin watches overestimate distance covered, and this becomes worse at higher distances.
• Garmin-watch marathon-completion predictors are useless for the same reason. The concept is admirable. For an example, read about the “Marathon Race Predictor” data-field.
• The Garmin-watch “Timer” (stopwatch) data-field is crucial during a marathon. We non-elite runners wait several minutes to cross the start line. The stopwatch clocks on a marathon course are for the elite runners. They are worthless to the rest of us unless (a) we can recall the start-line clock’s display and (b) we can do clock arithmetic throughout the marathon. Both (a) and (b) are difficult, if not impossible, especially deep into a marathon.
• A final-kilometers or -miles wristband may reduce anxiety, if there is any chance of finishing after a marathon’s DNF time. I may explain how to create one of these anxiety-blockers in my next book. To create one of these on your own, see Run:Walk Pace Tables in Minutes/Kilometer or its per-mile companion.

Don’t get me wrong. I have loved the Garmin watches that I have had over the years. They have several features that you cannot find in a non-GPS watch.

Still, I recall with fondness how my first marathon training-group taught me to use a Timex Ironman watch for its stopwatch and run:walk features.

Conclusion

GPS watches are great for training, but the “Essentialism” of a non-GPS watch is perfect for a marathon. The good news for owners of GPS watches is that they can configure them to display the essentials in a marathon.