TNF 50 Mile Race Report: Reflections on pacing, nutrition, training, and a lot more

At the finish with old friend Jeremy Humphrey and his wife Brandi. Jeremy finished a few minutes and places ahead of me.

At the finish with old friend Jeremy Humphrey and his wife Brandi. Jeremy finished a few minutes and places ahead of me.

It almost certainly will not be another four years until I race again. But overtime my recollection of The North Face 50 Mile Championship will change: I’ll forget the finer details, and my reflections will be influenced by future developments and conversations. So my race report is meant to be as instructive to me as informative to you. Here it is, fresh:


If my pace analysis of seven runners in the 2013 race was indicative of how this year’s race would unfold, I knew that:

  1. All runners would slow down gradually throughout the race, and
  2. Runners who would not finish at or near the top would slow down disproportionately more than the winners.

If for no other reason, I expected my average pace to slow at least due to the course: the first 11 miles were among the fastest of the day, and the biggest climbs were later the race: miles 18-23, 33-36, 41-44. But otherwise I was hoping to minimize my “pace creep” by starting conservatively and finishing strong.

2014 course profile. Due to a washed out bridge we followed a modified route. The small increase in length was offset by a small reduction in vertical and technical terrain.

2014 course profile. Due to a washed out bridge we followed a modified route. The small increase in length was offset by a small reduction in vertical and technical terrain.

While I did run a more even race than others to whom I finished near, I still failed to remain completely disciplined in the race’s earliest miles. After a two-week taper, maintaining for the first 15 miles a 7:35 min/mile pace on terrain averaging 200 vertical feet of change (up or down) per mile seemed EASY. I also felt socially pressured to run uphills that 20, 30, or 40 miles later would force me — and everyone around me — into hike mode. By burning up my quads on some of those early steep pitches to save face and irrelevant seconds, I cost myself minutes later in the race because I wasn’t able to crank on low-angle grades like Coyote Ridge (mile 41-44) and Marincello (mile 45-48).

Between Tennessee Valley and Muir Beach (miles 14-18) I intentionally slowed to a more sustainable pace. Thereafter, my average pace crept upwards only 54 seconds, from 7:58 to 8:52. In comparison, my friends Jeremy Humphrey and Brett Rivers — who finished a few minutes ahead in 34th and 35th, respectively — slowed by 70 and 83 seconds, respectively. Still, I would have liked to see a more even effort. Sage Canaday, the winner, only slowed by 23 seconds.

The big spike at Mi 35 is due to a slow 2.8-mile climb that averages 525 vertical feet per mile. Notice my strong finish!

The big spike at Mi 35 is due to a slow 2.8-mile climb that averages 525 vertical feet per mile. Notice my strong finish!

I’ll write another post with a broader analysis of times and paces. In the meantime, view raw data in Strava.


For dinner the night before, I had rotisserie chicken with salad and baked potato fries. About 90 minutes before the 5 AM race start, I had two Clif Builder bars.

From my training sessions and my recent Metabolic Efficiency Test, I knew that at race intensity my body relies more on its fat stores than its glycogen/carbohydrate reserves for energy. Specifically, at my average heart rate of 142 bpm during the race, two-thirds of my energy came from fat. Hence, once the race started I would not need to consume the conventional 250 calories per hour in order to maintain my energy level and avoid bonking/hitting the wall.

Instead, I consumed only 100-150 calories per hour, with my first snack about one hour into the race. I ate Clif Shot Bloks exclusively, which the aid stations had in abundance and which I prefer anyway over gels, which are messy.

I never felt low on energy and I never had any GI issues. I think I probably could have eaten even less but I haven’t raced enough to be confident yet in the lower bounds of my caloric needs.

Results from my Metabolic Efficiency Test at Fuel4mance in Denver. During the race, my average heart rate was 142 bpm. In the rest, my body was relying on its fat stores for two-thirds of its energy, meaning that I don't have to consume lots of food in order to resupply my depleted glycogen stores.

Results from my Metabolic Efficiency Test at Fuel4mance in Denver. During the race, my average heart rate was 142 bpm. In the rest, my body was relying on its fat stores for two-thirds of its energy, meaning that I don’t have to consume lots of food in order to resupply my depleted glycogen stores.


At the NCAA Cross-Country Pre National Invitational in 2000, I learned the importance of staying hydrated on a travel day. I barely drank anything the day before, was dehydrated when the race started, and passed out at the finish line. My time was awful but I was too stubborn to drop out.

I was well hydrated on Friday night, but probably should have consumed more than just 24 oz of coffee and water in the morning. It’s hard to catch up on hydration once the race starts.

I drank water exclusively. I started with one full 10-oz bottle and refilled it at Mile 6 and 11, :44 and 1:22 into the race. I refilled both 10-oz bottles at the next eight aid stations, and consumed an extra 10 oz of fluids at Tennessee Valley at mile 45. I completely bypassed Alta at mile 48.

To replenish my electrolytes, I took one SaltStick capsule per hour.

I peed twice, at miles 23 and 34 (I think).

Conditions were comfortable all day: temperatures in the mid-50’s to mid-60’s, mostly cloudy or foggy, and high humidity.


Two weeks before the race, I quit drinking coffee — my only caffeine source — so that its effect would be magnified on race day. That morning, I drank 1.5 servings of Starbucks Via instant coffee (~200 mg caffeine). Out on the course, I consumed 100 mg of caffeine at miles 23 and 33, then 200 mg at mile 41.

Clothing, footwear, and equipment

Last week I posted my gear list for the race. My selections were flawless and I wouldn’t change anything for a race with similar conditions or aid station intervals.

Except one: I used a 100-lumen Rayovak headlamp that I bought last week at Home Depot for just $15 rather than my trusted Fenix LD02 flashlight. Since I didn’t have drop bags and or a dedicated crew, I had mediocre odds of ever seeing my light again if I abandoned it, and being out $15 is better than being out $40.

My clothing, footwear, and equipment for The North Face 50 Mile Championship

My clothing, footwear, and equipment for The North Face 50 Mile Championship


Starting in mid-October — after the backpacking season and an abbreviated elk hunt — I dramatically increased my training for this race. Over the next 42 days, I ran 525 miles with 167,000 vertical feet of change in a cumulative training time of 83+ hours. On average, that’s 12.5 miles with 4,000 vertical feet up or down in 2 hours everyday. (For more details, view my Strava profile.)

It’s impressive, but absolutely not enough. Decent genetic talent, a robust baseline fitness level, a solid 6-week training cycle, and a 2-week taper does not put me on equal footing with other elites who train year-round and who race much more regularly. Unless I dedicate myself fully to the sport, I’ll never reach my potential in it. And I’m probably okay with that: if I consider all the time, effort, and cost involved in training for and recovering from an ultra, generally I’d rather leverage it into a kick-ass backpacking trip or trips.

For many reasons, it’s likely that I will enter this race again. Next time, however, I’ll train with greater specificity to the course, which is predominantly run-able and non-technical. I should have included more higher-intensity interval training — notably, track intervals and tempo runs on moderate grades — instead of so many long runs on steep, technical trails at an “ultra shuffle” pace.

My body seemed shocked by the amount of running and by the running pace. It’s revealing that I cannot point to a single workout that resembled the first 15 miles, in which I averaged 7:35 minute/mile pace with 200 vertical feet of gain or loss per mile. It also explains why I did comparatively better once the course got hard — I’d trained for that. Consider, too, that I need to train even harder to be a runner because I spend so much of my year hiking.

Unfortunately, this “shock” manifested in two physical ailments that hindered my racing speed. Starting around mile 11, both of my hip flexors got tight. And then around mile 40, both of my calves began to cramp. Once these symptoms came on, I had to manage them carefully so that they wouldn’t blow up and force a DNF. I didn’t have any hip flexor or calf issues during training, so I suspect that they were caused by the race pace being so much faster than my normal long run training paces.

Body weight

Extra body mass — that is, body fat — is a liability in most athletic competitions, but it’s especially damaging over an ultra mountain trail race. Gosh, if that 20-oz bottle feels like it is slowing you down, imagine the cost of those five pounds of belly fat.

I was able to maintain my weight between late-September — when I finished my last backpacking trips of the year — until race day. You’d think I would have lost additional weight by running 100 miles per week, but it seems that I lean out more when backpacking: I’m burning a higher proportion of fat and my caloric intake is limited by what’s in my pack.

If I disregard the pleadings of my wife and mother, I could probably shed another 5 pounds. But for now, I’ll enjoy being as trim as I’ve been in many years, and I’ll hold off on additional weight loss until I have something more specific on my calendar.


I’m definitely not Mike Wardian, who on Saturday beat me by 19 minutes and who on Sunday finished the California International Marathon in 2:33. But my recovery has been reasonable: I’m stiff, sore, and slow, but I made it through two airports yesterday and I can walk normally up and down stairs. After two rest days, I’ll probably run tomorrow.

Reputation in the ultra community

At the afterparty at the 2 AM Club, it became apparent that within the ultra community I have the narrow reputation as being “the guy who did the pace charts.” I would not expect more given my 4-year hiatus from racing and my general disengagement with the ultra social scene, but I’ll humbly submit that I’m much more than that.

Travel & lodging

Avoid San Francisco International Airport; instead, try Oakland, Sacramento, or even San Jose. The risk of a flight delay or cancellation due to fog is very high at this time of year. The rental car agencies are a 15-minute train ride away from the airport. There will almost always be heavy traffic when driving through San Francisco to Mill Valley. And the security lines for Southwest Airlines gates in Terminal 1 are among the least efficient I’ve seen anywhere.

Brett and Larissa Rivers are awesome. With them, I had a comfortable place to stay and I was immediately hooked in with the local scene since Brett’s store, San Francisco Running Company, is the region’s trail running hub.

Race management

It’s no coincidence that The North Face 50 Mile Championship is the only ultra race that I’ve run twice. Amanda and I also ran the marathon together in 2012. In short, I think it’s an excellent event. The entry fee is inexpensive. The location is superb. The prize money attracts the most talented field of any ultra race in North America if not the world. And its early-December timing is usually conflict-free.

But perhaps most important of all, lining up at a premier race next to some of the best ultra runners in the world is not a multi-year, low-odds effort that involves qualifying races and luck. I understand why other races have implemented such policies, but I view them as a major deterrent. I’m sorry, but I’m very reluctant to invest significant time and some money into a 100-mile qualifying race this year, destroying my body in the process, only to have extraordinarily low odds of admission into a race that I really want to do, e.g. 4.7 percent for Western States and 1.3 percent for Hardrock.

Next year I’m planning to run 2-3 ultras. TNF is almost a definite. For the aforementioned reasons, races like San Juan Solstice 50 Mile and Run Rabbit Run are appealing for the other slots.


  1. D Lonsdale on December 8, 2014 at 5:56 pm

    An excellent report and job well done!

  2. c l morris on December 8, 2014 at 6:36 pm

    Why the caffeine halway thru the run? Where the caffeine pills planned for that specific time or was that taken on how uou felt at the moment.

    • Andrew Skurka on December 8, 2014 at 9:30 pm

      The pace was fast enough at the start and I was still fresh, but later in the race they can be a nice boost.

  3. GZ on December 8, 2014 at 6:48 pm

    Excellent report hiker guy who does pace charts.

    No, seriously.

  4. Jeff Valliere on December 8, 2014 at 9:09 pm

    Great job Andrew and awesome report. Some of us give you flak for the charts, but it is friendly ribbing and really, I do find it interesting (and can relate, as I do similarly geeky things). Looking forward to getting out on the trails again with you soon.

  5. Daily News, Tue, Dec 9 on December 9, 2014 at 5:17 am

    […] Skurka gives a very detailed analysis of his […]

  6. Joey Dawson on December 9, 2014 at 10:20 am

    I love every part of this, from the analytics to the breakdowns of each aspect of the race and training. I am a huge fan of your backpacking feats and am looking forward to rooting for you in the ultra scene.

    • Andrew Skurka on December 9, 2014 at 10:42 am

      Thanks, glad you enjoyed the read. Honestly, I wrote it more for me than anyone else so that I don’t forget what I learned during this experience, but obviously there’s a lot of information in there that could be relevant for others.

  7. james on December 9, 2014 at 10:48 am

    great report. i also suffered due to going out a bit too hot. I went 7:11 and know i could have went sub 7 if I would not got caught up in the early pace. I’ll likely be back in 2015 as well. I agree… it’s such a great race. Looking forward to hearing about more of your return to racing next year.

  8. Dave on December 9, 2014 at 6:24 pm

    Excellent report. I enjoyed the indepth analysis.

    Personally, I like reading about others’ successes and failures. No point for me to re-invent the wheel if others have already trodden the path ahead of me 😉

  9. Dan on December 9, 2014 at 9:03 pm

    Interesting to following your journey towards and through the run.

    I’ve been mulling over your pace creep musings, and after re-writing this post a few times this is what I got:

    Energy expenditure increases exponentially with running speed, so based solely on this a consistent pace would be the fastest – similar to how a car gets better average gas milage cruising at 50 MPH than it does alternating between 40 MPH and 60MPH. However, we are different from cars because our running efficiency declines over time due to physical, nutritional and physiological degradation. Thus the optimal pace also declines over time and some pace creep is ideal. Fitness determines how much degradation the body experiences, and thus fitness determines the ideal slope of the pace creep function. Optimal pace creep is going to be less for a fit person than a non-fit person.

    So what’s important is having an accurate idea of how fit you are, and then setting pace creep targets based on that. My guess is that someone with outstanding fitness might run optimally at 5% pace creep, while someone in moderate condition might have their best race at 20% creep.

    Fortunately it’s partially easy to solve race results to find these optimums. The race winner in a competitive field must have excellent fitness and very good pacing in order to win. Thus the average pace creep for a large sample of ultra run winners would tell you roughly the optimal pace creep for someone in near perfect fitness. One caveat being that some courses lead to greater degradation rates than others, so you’d want to compare similarly challenging courses.

    Finding the optimal pace creep for someone in sub-optimal fitness is trickier because if you look at the middle of the pack results you have two variables: condition and pacing. My guess is that you could pseudo-solve for fitness by taking the anti-derivative of pace creep function (accelerating pace creep indicates poor fitness and vice versa), but that’s a lot of calculus. It might be easier to look at winners of different age brackets. I.e. the average pace creep for winners of an over 55 group might be similar to someone who’s younger but in less than optimal fitness.

  10. […] his blog the other day and he has a graph where he has had his metabolic efficiency measured: TNF 50 Mile Race Report: Reflections on pacing, nutrition, training, and a lot more // Andrew Skurka Look at the 3rd graph under "Nutrition". Interesting to see the crossover point, i.e. […]

  11. Ryan on August 20, 2015 at 10:36 am

    How was the Rayovac light? Pacing for the Leadville 100 this weekend and need to pick up a light.

    • Andrew Skurka on August 20, 2015 at 2:18 pm

      It throws enough light, but the beam is too focused (i.e. a spot with little flood).

      The best light for the money that I’ve seen is the Coast HL27 Headlamp. Specs: 330 lumens, adjustable spot/flood, and nothing-to-full dimmer. Also just $60. This is the light I think I will be wearing during Run Rabbit Run 100 next month, which has about 12 hours of nighttime racing.

      Leadville and TNF50 have different lighting demands, especially if you are crew. At TNF, it’s dark for the first 90-120 minutes of the race, when the field is still packed up. At Leadville, you could be out there all night, with mostly just you and your racer.

Leave a Comment