Obsessing Over That Half Marathon

Since my last actuarial exam 15 years ago, I hadn’t worked so hard toward a single event until signing up for the Bay Bridge Half Marathon. I’m still feeling the runner’s high a week later.

The Eve

Although Oakland was only 25 miles from home, the thought of having to cross a whole bay to get there in the morning was too stressful. The race didn’t start until 8:15, but traffic on the bridge could be unpredictable and parking in an unfamiliar neighborhood may be tricky. As such, I decided to stay at a hotel close to the start line to get that peace of mind.

Waterfront Hotel in Jack London square was the perfect choice. It was merely a mile away from the event, located in a beautiful waterfront commercial district, and was redeemable using Hyatt points. This hotel had a cute maritime theme and upgraded us to a junior suite.

In part (?) because our family would be in Oakland, Hong’s big sister at work decided to throw a party at her home in Piedmont. Aamra was famous for hosting some of the most awesome parties, and what we thought was a small gathering turned out to be a 25-plus-person event with their coworkers and extended family. It was such a fun crowd to be around, and I especially enjoyed hanging out with Aamra’s teenage sons and nephew. The main bummer was that, in order to prepare for the race, I had to approach the delicious food with moderation and leave for an early-ish bedtime. The big surprise of the evening was that everyone knew about my race and wished me luck. I did not expect that kind of attention and was totally flattered.

The Prep

On the morning of the race, I got an iced latte from the nearby Starbucks with an extra shot of espresso. Caffeine never used to have any noticeable effect on my body. However, consumption had become a daily ritual since last year’s Andalucia trip and now I had no idea how dependent on it I might be. Today wouldn’t be a fun day to test it out, so I erred on the side of overdosing.

Then I proceeded to eat 1.5 Snicker bars for breakfast. This teeth-rotting candy used to be among my favorite, but I had to stay away because of pre-diabetes. About two years ago, a friend told me that it was essential for marathon runners to replenish not only water but also high-energy foods during their race. Ever since, whenever I went on a long run, I fantasized about one day running a long race and indulging in Snickers for fuel. My moment of using that excuse to eat candy had finally arrived. Wow. And it was way too sweet.

I had thought that I’d jog a mile to warm up and do a deep stretch in the early morning, but when the time came I did not feel like it. I was content with leisurely walking that mile to the start line, with Hong, Xuan, and Ting by my side. I felt good. I had spent the past week stretching instead of running to keep my legs loose. Only my right calf remained a bit tight, which became uncomfortable about 3 miles into the race, but it was manageable.

Being a complete rookie, I spent a lot of time stressing about what gear to bring with me during the race. The rational side of my brain knew that I didn’t need anything, but my inner maximizer wanted to make sure I didn’t miss any incremental efficiency. I brought sunglasses just in case it was too bright, but opted to not wear them. I brought an extra hair tie just in case it was too windy and I needed to put my ponytail into a bun, but that got left behind as well. I brought the electrolyte tablets that the organizer gave me at check in, but decided that it was too silly. While I did decide to use my ear buds for music, it only worked for the first half of the run… sweat dripping down my long hair started messing with the touch control. I couldn’t fix the problem without slowing down, so they went into my pockets for the entire return trip. It was a familiar issue from my routine runs so there was no panic, though I’ll have to seriously rethink about music if I ever do this again. So much for compiling a killer playlist just for this race!

The Photos

Given the iconic race location, I contemplated taking a break on the bridge for a quick selfie. It likely wouldn’t have cost me more than 30 seconds, but at this point I was obsessed with the absolute best time I could manage. I was super glad then to realize that the event organizer had a photographer and posted these photos without making me pay extra money.

Look at me… I was having such a good time 🙂

The Analyses

With the race results data now in the public domain, I put my professional habits to work. Here are some visualizations of the data.

I knew that my 1:33 finish time was within the top 5%, and the median runner took just under 2:09. But what did the distribution of everyone’s finish time look like? Here it is:

The distribution of all runners can only go so far, of course, without some demographic insight. Despite some older people outpacing me and two ladies being in the top five spots, there was still strong evidence that physical ability differed by age and gender. On average, women took 15% longer than men, and 20-year-old dudes are 4% faster than my 40-year-old cohort.

I had wondered if I’d take more or less than half of the time to cover half of the distance. Seeing that the course was a perfect round-trip on the same route, I had assumed that the second half would be faster because it sloped downward. Although I did manage to make that happen, by spending 32 fewer seconds on the return trip, it sure didn’t feel like it while I began to drag my feet on the asphalt.

How did everyone else do? In the chart below, the gray dotted line represents people who spent the same amount of time during both halves of the run. Above that line were groups of runners who took more time during the second half, and vice versa. The top 10% of runners all maintained their speeds consistently, but slower participants ran out of steam visibly. Well, 13 miles certainly were a test of stamina!

Zooming in, some runners chose to conserve energy up front and use it for the home stretch.

Speaking of elevation gain, the course for this half marathon was quite flat. I wanted to see how exactly it compared to the streets of Millbrae, so I pulled out my records from the past three years. Dividing total elevation gain by total distance in each run, the char below illustrates the level of steepness where my workout runs took place. Across more than 1,400 total miles, I trained on roads that were six times more steep than the Bay Bridge.

According to the event organizer, 2023 was the third time this Bay Bridge Half Marathon happened. To see how each year’s group performed relative to each other, I plotted the finish time distribution below, by percentile. Since 2019, each year had improved over the previous one.

I’d love to fantasize this phenomenon as a result of everyone training more during the pandemic and becoming a better runner. Though the more likely explanation is that race conditions just happened to be better (e.g. less windy) in 2023 than in prior years. I consider myself lucky to have participated when it was easiest to go fast.

The Trackers

The event used an app called RaceJoy, which I didn’t realize until Ken showed me while waiting for the race to begin. It was a nifty tool that provided live updates to runners and spectators on the whereabouts of an individual and his estimated completion. If we had been more familiar with it, my family could have used it to monitor my progress while they waited.

But it was fine because Hong and I have location sharing with each other on Google Maps. As she got brunch with the kids, she was able to see how far I was from the finish line.

For detailed measurements, I counted on my Fitbit. Indeed, it indicated that I had run several of my fasted miles ever. Though there was a serious problem… it thought that I had only run 12.3 miles. The discrepancy bothered me for days, and might continue to in the future. How should I think of it? Was the event organizer incapable of setting up a 13.1-mile course? Or was Fitbit’s GPS tracking plus Google Maps inaccurate?

I had issues before with Fitbit under-counting distance when running on a trail with mileage markers, and giving me inconsistent reads when I ran on identical paths. This isn’t a problem I’d be able to regularly observe, but now I have to wonder if Fitbit underestimated all my runs by 5-10%. Perhaps this was a blessing in disguise… consistently tell me that I was 20 seconds per mile slower than I really was, so I could get a pleasant surprise during the actual race.

The Crash

After that half marathon, we had a lazy lunch with Ken’s family, went home, and melted into a puddle.

My last serious run before the race, according to Fitbit, was just 0.2 miles shorter. But I took 500 more steps and had a higher heart rate, so it was as comparable as it could be. It was tiring, too, but I continued on to a full work day as usual. Thus I was surprised to be so tired after the race.