INCURABLE DATA GEEK

I recently completed a 5-Day Cycling Stage Race (or event), Haute Route Pyrenees. Each day is about ~100 miles (~160km) with ~10,000ft+ (~3000m+) of vertical climbing, which translates to about 5 hours of racing per day. For rider safety, not all high mountain downhills were timed, so the official riding times are a bit less.

In the end, I placed 3rd in the General Classification (GC) among the ~250 finishers after five days, so I thought to share a few performance data points needed for that level or results at a well-attended Haute Route cycling event. Of course, rankings completely depend on the competition. 

My cycling “numbers” going into the event around July 2022:

Rider weight: 79kg (174lbs)

FTP at sea level elev: ~405 Watts or 5.13 W/KG

20min power at sea level elev: ~425 Watts or 5.38W/KG

FTP at 9000ft elev: ~340 Watts or 4.30 W/KG

20min power at 9000ft elev: ~370 Watts or 4.68W/KG

Relative performance in the group

Given these events are mostly individual events (although many top-riders have a person or two helping as ‘team members’) and the riding profile has several daily 30-to-60minute sustained climbs with 10,000ft+ of daily vertical gain, these are quite clearly light climbers events.

That said, I noticed that given my high’ish FTP (400W+) and reasonably good relative power of 5.1W/KG (at FTP), I was..

Very competitive from flat to up to ~5% climbing grade

Within top ~five at steep long climbs and about 5-7% behind the fastest. In other words I’d need ~5.5W/KG (instead of 5.1W/KG) FTP to be competing for the win in the steep climbs in this particular group of riders (I’d guess all the top riders were -20% lighter than me or about ~60kg vs my ~80kg)

Very competitive in sustained efforts after the first 3 hours. Majority of riders’ performance dropped substantially at the last climbs, which maybe hydration/nutrition, bad pacing or simply lack of training. I usually picked up competitors in the second half of each stage.

Real-world performance at Haute Route

Looking at Strava results for the finishing climbs after 2-3 hours of riding each day, I was typically among the Top 3 fastest in the last long (1h+) climb. For example, these finishing climbs at Stages #1, #4 and #5:

Col du Pourtalet (1st fastest): 73minutes, climbing to ~7000ft

Col du Cap-de-Long (2nd fastest): 69minutes, climbing to ~7000ft

Col du Portet (4th fastest): 66minutes, climbing to ~7200ft

Notably, Col du Pourtalet was the least steep (4.7% average).

I could ride all these finishing climbs at about 340W (4.3W/KG) given the heat (typically 80F+), elevation (5000-7000ft) and end-of-day effort. That would be about 92% to 95% of my FTP at that elevation or about 85% of sea level FTP in optimal conditions.

Training and Nutrition

Nutrition: I took a pretty simple approach to nutrition: as always, I train and live with little to no sugar or obvious carbohydrates. It works perfectly fine. Then in the race, I took about 300kcal/hour of carbs from liquids (gel or drink mix). At the end of each stage I immediately ate about 800kcal of fast carbs to fill glycogen stores for the following day. Dinner and other meals were healthy whole foods with little to no carbs. I never bonked and had strength to ride at my threshold even at the end of each 4-5 hour stage, so I think my nutrition worked just fine. For hydration, I think I replaced about 50-75% of the sweat losses in each stage (I sweat 2.1L/hour) and that didn’t seem to negatively affect performance.

Training: While minimalist training can work fine (in my experience) for a single-day event like an Ironman triathlon, I assumed that it’ll be impossible to recover and feel fresh unless I can repeat the performance multiple days, back to back. Consequently, I increased cycling training with the following two:

Several weekends with back-to-back 4+ hour rides

One 4-day “camp” where we rode 5+ hours each day about 3 weeks out of the event

Secondly, I did a fair amount of event-specific training with both threshold (FTP and above) and “sweet spot” training (~90% of FTP). For example, I did a 20 minute hard lactate threshold (FTP+) effort in the beginning or 2*10min the same, then follow that with 2-3 x 30minutes at sweet spot for a total of 3-4 hour (weekend) ride. Sometimes I flipped that around, finishing with a 10-20min hard above threshold climb. I tried to keep these high intensity workouts to max 2, maybe 3 days a week. This seemed to work well for me.

How to go even faster at Haute Route events?

There are obvious ways to ride even faster based on my experience...

Ride your own, climbing-optimized bicycle. I rented a bike the day before start of the race. Pinarello Dogma F10 isn’t bad, but it didn’t exactly fit me and wasn’t optimally light for climbing.

Have a “team member” to help pull on the flats and generally help when something goes wrong.

Have support to not stop at aid-stations. The top 20+ riders never stop at aid-stations and get hydration/nutrition from a supporter. I lost a group couple of times and lost bunch of time because I didn’t have that support.

Rest when timing is stopped. Appropriately so, organizers stop timing on dangerous descends. I usually just kept going, but it would be possible to take an extended recovery break without losing any time. Some riders took advantage of this.

Improve relative threshold (or FTP) i.e. higher W/KG. These are climbers’ events.

The obvious: don’t fly across 9 time zones to arrive 19 hours before start of the first stage. This clearly wasn’t optimal for recovery and readiness...

Should you do a Haute Route cycling event?

Our four person mixed (3 men and 1 (wonder)woman) team Fat Chance just completed the 2018 edition of the 3,070 mile cycling race Race Across America (RAAM). Although our priorities were safety first, then personal relationships, and lastly speed, we finished reasonably fast in 6d 9h 33mins i.e. averaging 20 mph (or 19.99 mph as the race organizers liked to report it!) and won the 4 person mixed team category.

I personally ended up riding exactly 42 hours and 934 miles (averaging 22.2 mph) during those six plus days. Before the event, I wondered what might be the optimal effort and pacing for an event that is essentially an individual cycling time trial for up to seven hours per day for a week or more. I got the exact answer - or one answer - during the race, which I’ll detail below. I’ll also share a few other lessons that may be helpful for others preparing for a future RAAM adventure (which I’d highly recommend as a life experience and sports event).

How hard can you time trial 42 hours in six days?

My initial hypothesis was that the optimal effort is somewhere between

Well below threshold / FTP or Zone4/5 (380-400 watts for me) effort even if each continuous riding effort between our four riders was limited to just 20 minutes at a time. Riding six hours (or even two!) at threshold effort per day seemed impossible even with optimal nutrition and sleep.

Definitely above pure aerobic endurance effort that is almost exclusively fat burning, Zone 1-2 riding (or 220-240 watts for me). I’ve done a few bike touring trips, riding 5-6 hours a day for a week at this effort.

So my best pre-RAAM guess was that with optimal rest, nutrition and riding schedule, I should be able to hold Zone 3 effort (280-300 watts) for these 42 hours, which is roughly my Ironman triathlon riding effort. My guess was pretty good.

What is the optimal riding schedule for four riders?

We debated the optimal 4-person riding schedule for weeks and ultimately settled on the first principle that sleep is the most important performance enhancer after the first two days. For this reason, we decided that we should:

Leave 12 hrs for full recovery, sleep and nutrition, each day

Keep circadian rhythm the same each day for each rider

Leave flexibility for each 2 person pair to optimize actual riding time

The net result was that one pair of two riders rode 12noon - 00AM and rested the second half. I rode the opposite schedule, starting at midnight each night.

The length of each pull was left for the pair and my pulls ended up being between 40 minutes and 2 hours, averaging just under 1 hour. The long 2 hour pulls were at night when it was raining hard and we wanted to minimize clothing change and getting cold. While those long pulls were mentally much harder, the effort that I was able to hold was about the same.

What actually happened during RAAM?

The first day started with 2.5hrs of riding in two pulls. I consciously rode hard at Zone 4 or 350+ watts just to drop other teams and to give my rested legs some real work. I knew I’d have almost 10 hours to recover before our regular 12 hour schedule would start. I think this was the right way to start, but definitely not the effort I could hold for 42 hours!

During my first real 12 hour shift, I continued with upper Zone 3 or 300-320W average efforts, but as we climbed to 6000-8000 feet of elevation, I was forced to bring it down 10-15% to 270-280 watts.

After two days of riding and coming back to sea level, I was able to average 300W+ again, but with heavy perceived effort. 280-300W felt much more sustainable.

Then I had my first totally failed night of sleep or about 3 hours only in our moving RV. I lost another 5-10% and could barely hold 260-280W for hours. We all realized that the decision to use a moving RV (is it even legal?) as our sleeping platform was a really bad decision. The 12 hour “recovery period” turned into something less restful for all of us.

After a bit more sleep (at most 5 hrs per night) the following nights, I was able to ride the second half (3 days) at around 280-290W average, while riding hills at 310-320W and several one hour pulls at 300W+ when another team was getting too close for comfort.

All and all, I ended up averaging about 290 Watts over those 42 hours, or almost exactly my Zone 3 Ironman triathlon efforts. With hindsight, I would say to further optimize pacing for maximum speed during RAAM, I would:

Optimize sleep: riders in 12h on - 12h off schedule should be shuttled to a hotel to sleep in a quality bed for 7-9hrs after each riding shift. Trying to sleep in a moving RV that is an errands-running vehicle is not going to work. I probably lost at least 5% if not more of my effort due to really bad sleep.

Optimize effort: when riding 24mph on a TT bike on flats, I’d drop effort even more (10-20%) to not waste energy on fighting the wind (wind resistance increases to the square of velocity). Same with any downhills: After 30mph there is no point in pedaling. Instead, add +10% on headwind and uphill sections.

Schedule: I would keep the same riding schedule: 12h on, 12h off and make most pulls about 40-60 minutes unless in very high altitude on very high heat, in which case I might shorten the pulls.

Summary of my RAAM efforts relative to my zones and watts:

Thoughts on optimal bikes

RAAM is definitely a two bike event: I had a road bike (Trek Domane with disc brakes, good climbing gear and clip on aero bars) and a TT bike (Cervelo P5 with a disc rear wheel, but no disc brakes) and I would go with the same setup again. California, half of Colorado, Kansas, Missouri, Illinois, Ohio, most of Indiana and some of West Virginia and Maryland are made for riding on a fast TT bike. 

I spent more than half of my time on the TT bike and would have ridden even more if I had disc brakes in it for rainy weather. I felt it was too dangerous to ride my TT bike on hilly and windy roads when it was raining. On the other hand, road bike was a must especially in West Virginia with constant 8-12 degree uphill grades with little shoulder to play with. And climbing those 8000ft+ and 11,000ft+ peaks in Colorado was also optimal for a non-TT climbing bike.

Thoughts on nutrition

I followed pretty closely our Fat Chance Row nutrition plan (i.e., real food and little to no carbohydrates), except that I consumed ~30 grams of carbohydrates per riding hour and some right after each of our 12 hour riding shifts. The only area that didn’t work well was that I assumed I’d be able to eat lots of solid food (e.g. macadamia nuts, nut butter) during the 12h riding shift in between my pulls. I couldn’t. Next time, I’d focus on hydration and liquid food only during the riding shift and and have the real meal(s) during off hours.

If I’m having fun and making money, why should I care about the real purpose of my company?

I ask all my new hires “you’re super smart, you work harder than I do — there must be so many opportunities for you out there — why did you join us?”. The equally interesting part of the answer is typically why people considered leaving their previous job. These are three actual, recent quotes from my team members:

“Our CEO [of a prominent unicorn tech company] stood in front of everyone and said ‘our purpose is growth at any cost’. I just couldn’t see being part of that anymore.”

“I was sick to my stomach going from one meeting to another trying to figure out how to sell more opioids to people who didn’t even need them.” (used to work for a top pharma co)

“Our purpose was making rich people even richer. It wasn’t very inspiring.” (used to work for a top investment bank)

Unsurprisingly, most people care about the “why”, the real purpose behind a company they work for. It’s obvious that if an organization’s purpose borderlines criminal, it’s unlikely that the most talented and thoughtful people will dedicate their careers to its cause.

But most companies are somewhere in between: some try to engineer more meaning and purpose by setting aside some percentage of profits to be given to charities. In other words, “let’s make money, so we can give some of it away for truly meaningful causes”. And others wonder whether a generation of talent is actually wasting their time optimizing ad click-through rates for living rather than doing something more impactful.

“Why?” and “What’s the purpose?” that resonates on a personal level matters to me, and I believe to most of us. I’ve asked this question myself many times. Not being able to answer the question well, was a key reason I decided to move on from the company I co-founded more than a decade ago and very much enjoyed building.

NYTimes columnist David Brooks recently argued that decision-making in life is not a calculated and rational chess game: “…but when it comes to the really major things we mostly follow our noses. What seems interesting, beautiful, curious and addicting?”

I agree. And “Our purpose is growth at any cost”, is likely not “interesting, beautiful, curious and addicting” to most people, or at least to the person, who wanted to leave the unicorn company she was working for.

My career advise to self (and to founders contemplating starting a company) is to always ask two questions:

Do I understand the true purpose of the company I’m starting / joining?

Does that purpose resonate at some deeper level than just making a buck or only provide momentary excitement?

If the answer is an obvious Yes, magic can happen. I call this alignment Profits Fuel Purpose and I believe it’s a perfect foundation for a long-term success as a for-profit company .

There’s very little published material -- and no peer-reviewed scientific publications that I’m aware of -- about multi-day stage race performance under nutritional ketosis, which I thought could be a performance maximizing tool when rapid recovery and ability to keep going are essential.

I raced an 8-day Mountain Biking stage race, riding a total of more than 40 hours, almost 400 miles and climbing more than 50,000 vertical feet in those eight consecutive days. Here’s what I learned through this N=1 person and N=8 day experience.

Goals

Going into the stage race, I was simply trying to optimize the following:

Rapid recovery given the back to back racing days

Overall race performance each day

The challenges that I knew I’d have to overcome:

Sleep (=performance enhancer #1) would be difficult in tent camps

Potential GI distress with 6+ hours of racing per day in African heat, so I’d want to avoid eating large quantities while racing

Despite super long days, mountain biking requires frequent, all-out an-aerobic efforts, where one can’t be grinding in “fat burning zone”

Approach

I had already trained my body to utilize fat as the primary energy source even at higher intensities and I knew from our Pacific rowing adventure that it’d be possible to keep exercising 10+ hours a day while limiting inflammation, recovering fast and getting quality sleep even in questionable conditions. So I hypothesized I would maximize my performance if I could stay in nutritional ketosis, while loading up my glycogen stores and doing race-time feeding with limited carbohydrates to be able to “attack” on all those technical hills where an-aerobic efforts would be impossible to avoid. In other words, my goal was to rely primarily on fat utilization for energy (fuel that is practically unlimited), yet be able to perform an-aerobic efforts that are almost exclusively fueled by glucose and/or stored glycogen.

Data

Below tables detail most of the key markers I tracked during each race day.

I measured my state of “nutritional ketosis” by doing a finger-prick blood test first thing every morning to get an accurate measure of BOHB concentrations (mmol/L). Concentrations above 0.5 mmol/L are generally considered to be in the range for “nutritional ketosis”.

The “Post Race (g)” carbohydrate intake includes all the carbohydrates between finishing the race and going to bed. However, I consumed about 90-95% of them immediately after each race day around 2PM.

Since I didn’t have a power-meter in my mountain-bike, the only objective performance measure was heart rate, perceived effort/feel and how I did against the rest of the field. I was racing with a partner, who had to drop out after day three. After that I was able to race consistently at my own pace, averaging around 130bpm for each of the days, which corresponds to 300 Watts on a power-metered bike for me (160bpm is about 400 Watts). I finished every stage very strong, except the last day. That was due to a pretty bad stomach bug that I got after the second to last day, after which I spent the entire night in bathroom. (Sidenote: I attribute the stomach issues to the 1,200+ racers living, eating and washing in a tent camp setting for 8 days -- not to my own decisions or dietary choices)

Observations

These are my N=1 observations based on very limited data and specific to my own biology, so take them with a grain of salt:

Performance: I slept well, recovered well and felt fantastic every day, except the very last day after my stomach bug. I could visibly pick up riders in the last quarter of each day’s stage. I didn’t observe any “lack of top end” and was able to attack on all the steep hills. Based on my average HR, my effort corresponded to about 300 Watts average over the 4-6hour stage each day.

Would it optimize performance to eat nothing during race: I was primarily utilizing fat for fuel, so I could finish the 7-hour stage #2 with zero food other than water. But it was (anecdotally) clear that by feeding just ~100kcal/carbs per racing hour was much better than nothing. My perceived effort was lower and I was able to ride a bit harder.

How many carbs could I eat to stay in ketosis: My BOHB values were all above 1.0mmol/L (generally above 0.5mmol/L is considered “nutritional ketosis”) with up to 200grams of carbohydrates ingested in one sitting right after the stage. Adding up the nutrition during the race, I consumed 300grams of more of carbohydrates per day, yet was in ketosis each morning when waking up. This is obviously highly dependent on the amount of exercise and carb oxidization during the race. Yet, it was surprising that I was able to fill up the muscle glycogen tanks after each race day, yet keep burning fat for primary fuel. 

Are those 100-150grams of carbs needed right after each stage: I don’t have a definite answer how else to restore glycogen levels for a back to back 2-6hour all out effort the next day. That amount seemed to be enough for me and I felt ready to go as hard as needed the following day. Based on this experience, I’d say it seems to help.

What would I change: The test worked out very well. I was especially impressed by how quickly I recovered between stages, considering that my training only included up to two 3+ hour rides per week. It’d be interesting to test what happens to the following day’s performance if one completely eliminates the post-stage carbohydrates. 

What exactly did I eat?

My daily routine was almost exactly the same every day:

Breakfast: Scrambled eggs, sausage, coffee with coconut cream and generously butter and lots of table salt (actual picture above).

Race feeding: UCAN Superstarch and food from aid stations included cashew nuts and honey-nut bars. (actual carbohydrate amounts in the table above). Sodium I carried separately.

Post race: Immediately after finishing each stage, I ate real food and about ~140 grams of carbohydrates from cooked rice or banana.

Pre-dinner snacks: Unsweetened beef jerky and macadamia nuts.

Dinner: Massive green leafy salad with fatty protein source (chicken / beef / fish) and very generously butter, oily dressing and table salt.

I did not use any vitamin/nutrition supplements or pain killers, except additional magnesium tablets after each day’s race.

What happened with the Fat Chance Row, our unsupported 2750 mile row from California to Hawaii, to raise awareness around the dangers of sugar?

Ever since I experienced the first miserable bonks on the bike (i.e., ran out of fuel and could barely move my legs to get home) and also saw my first unflattering test results from the Stanford Human Performance Lab in 2009, I’ve tried to figure out how to make myself bonk proof. 

What if you could move your “hitting the wall” point from 5 hours to 80 hours of running or riding a bike – wouldn't that be amazing!? The advice I got was: you need to do plenty of training in your fat burning zone to become an efficient fat burning machine for long endurance events. I disagreed. I had done plenty of that (relatively speaking) and was left with NO fat burning zone anywhere.

Endurance athletes have several performance or effort limiters (lactate, fuel, heat, muscular endurance, etc.) and some scientists still argue whether eventual slow down is caused by our brain (i.e., central governor) or some peripheral  limitation, such as running out of fuel in the working muscles. Either way, for practical purposes, I've personally found that the following measurable limiters determine most of your performance in a race:

< 2h race (Half-marathon, Sprint & Olympic distance Triathlon): Lactate/Anaerobic threshold i.e. who can go hardest until your muscles burn and breathing becomes tough

2h+ race (Marathon, Half-Ironman, Ironman): Fuel limitations i.e. who bonks last 

The reason for the fuel problem is quite simple: Our “carb tank” carries up to 2 hours worth of heavy exercise fuel, but even the leanest healthy athlete carries more than 24 hours worth of fuel in his “fat tank”. So the question is: how do you become a fat burner to be able to go longer and faster in a race lasting more than 2 hours? (Of course, in addition, you want to be energy efficient in general, and have a large cardiovascular engine to burn any fuel that might be available)

THREE TESTS: From a Sugar burner to a Fat Burner

What did Stanford human performance lab found out about me in 2009? I was tested using my own time-trial bike on a Computrainer with a "gas exchange" tube in my mouth after an over-night fast without breakfast. I had been following what I thought was the "healthiest diet" i.e. super low-fat and consequently super high carb diet for more than ten years and I already had 5 years of triathlon training under my belt leading into this test:

I couldn’t burn more than 200Cal/hr from fat after a few minutes

At the “comfortable” efforts of 250-300W I was burning 900-1000Cal of carbs per hour, which would mean bonking in 2-3hrs (assuming a 2000Cal glycogen fuel tank) even if I managed to eat some. 

In short, I was a highly efficient sugar burning machine, who could ride a bike at 300 Watts without much effort, but burning all sugar, and at least theoretically hitting a wall half-way an Ironman bike ride at that effort.

For those who aren’t familiar with bike “watts”, here are two reference points:

The very best professional triathletes typically average their Ironman races at 270-290 watts (or 3.7 – 3.9Watts per body kg for most, but the less hills and corners, the less the power-to-weight matters)

At 300 watts, a cyclists with decent aerodynamics moves at ~25mph on a flat course (calculator) 

My second, comparable test (same bike, Computrainer, testing equipment and test protocol) is after a three month “pretty high fat”, “moderate carbohydrate” experiment, during which I replaced almost all sugar and added large quantities of nuts, oil and avocado. There was no meaningful difference in my training; and certainly no increase in easy workouts in "fat burning zone".

Now my fat burning had more than doubled to more than 400Cal/hr. The only visible change in training or lifestyle that I could point to, was my diet.

This spring, a week after winning the Wildflower Long Course triathlon amateur race, I completed my third substrate utilization test. I went into the Wildflower race 20 pounds overweight, with very limited triathlon training and also decided to cut my typical race time eating (= all carbohydrate calories) by a massive -60% -- which I knew was a huge risk. Yet, I won the race and my time was as close to the best professional athletes as ever (~8% in overall time). I also felt that I could have kept running at the end. I knew that something had changed for the better to be able to produce that performance.

My third test was done with a KORR device, again using Computrainer and my own time-trial bike. I also used Powertap to double check Computrainer's power measures.

  The results were pretty unexpected. Now my fat burning peaked at close to 750Cal/hr and fat utilization was still contributing 50%+ of the energy at 300W. I had more than tripled my fat utilization from the bottom values. I would guess that this change is part of the reason why I was able to maintain my strong effort at Wildflower triathlon, even though I cut my race-time eating by 60% and came into the race overweight and undertrained.

BURNING FAT:  So What and How?

So what’s the significance of becoming a fat burner in endurance sports? Like with most things in life, if you can go longer and harder, it's usually better. My progress to a better fat burner is clear in the chart below:

I had pushed my "hitting the wall" moment* from 5.6hrs to almost 90hrs at a comfortable 200 watts. And now I could race a full Ironman (8+ hrs) at almost 300 Watts, where as earlier I would have hit the wall just after 2 hours at that effort. To me, this is as close to bonk-proof as it gets. 

*) For these calculations I assumed that the total glycogen stores are 2000Cal to begin with and could be supplemented by 250Cal/hr by eating. You can eat more, but it is questionable whether your body oxidizes much more than 250Cal of digested carbs. Maybe ~100Cal more but it wouldn't change the calculations much either.

What were the exact changes in making this transformation (as a N=1 experiment) in the third test:

Exercise: No significant change to my knowledge. Definitely no increased volume. If something, more shorter and high-intensity workouts.

Diet: No sugar, no processed carbohydrates. Roughly ~15% of total daily calories from carbohydrates in the 5-6 months leading up to the last test, most of which came from vegetables and nuts.

Supplements/sports products: none during training, unless it was a more than a 3-4hrs workout (=which for me is less than once a month) and even then I tried to stick to real food, such as bananas, cashew nuts, etc. I completed all <3hour workouts with only water, without compromising my hard interval workout performance at all.

Although my N=1 experiment makes no science, I would recommend that if you want to become a bonk-proof endurance athlete, you might want to consider diet first, then adjustments to your training regime. Doubling or tripling fat burning abilities sounds like something that would take 10 years of endurance training, but I did it in months by changing diet.

A couple of weeks ago I became the overall amateur champion of the Wildflower Long Course (=Half Ironman distance) triathlon, now second time. I was 20 pounds over weight (at 201 pounds and barely 6ft tall) and just a month before the race my running speed was almost a minute per mile slower than before. At that point I assumed it would be impossible to be competitive, especially on the run.

A 201 pound triathlete who doesn't look like he should be running.

I would agree with most (all?) of the running and endurance sports coaches in that the most likely way to achieving maximum running performance in the long run, is a long-term commitment to consistency and gradually developing all areas critical for running: cardiovascular capacity, muscular endurance, strength, speed and technique. But I also think that you can hack your way to running fitness quickly.  Admittedly, my recipe is based on only two experiments, one of which was the four weeks leading into Wildflower triathlon.

This is what I was facing before the Wildflower race:

Overweight: race week I weighed at 201 pounds (about +20 pounds overweight from my average triathlon race weight in the last 5 years) due some rowing training and the majority of that additional weight was either non-functional (=fat) or in the upper body that is completely counterproductive for running (or cycling).

Slow running: four weeks before the race I was running almost a minute per mile slower in my test workouts. I expected this would mean up to ~10minutes slower running time in the race.

No running base: Since January 1, 2014, I had done exactly 2 runs per week on average, of any length. Most of these runs were easy 1 hour jogs and my leg muscles were sore for two days after each run (try running with a 20 pound backpack!).

Within four weeks I turned around my run and got these results:

I won the overall amateur title.

Both my half marathon running time and total race time at the Wildflower Triathlon was only about +9% slower than the average time of top four professionals, which is as good as my best performance at Wildflower (I’ve raced the long course 5 times).

At the Wildflower finish, my legs and muscles felt completely fine and I could have kept going at the same pace.

In my race week run workout, I had taken 40seconds per mile off of my running speed in four weeks.

(Also, based on my speculation, if the Wildflower weather was a bit cooler than 90F+, I would have run significantly faster relative to competition; Heat at 90F+ becomes a major limiter when running at 200+ pounds and it was the factor limiting my running speed in the race. But that’s speculation.)

So what is the hack? It’s a simple four-step protocol for four weeks. My total running averaged 1h 55min per week:

Develop muscular endurance by running 10-15 minutes every day: This is important to be able to take 13.1 miles of pounding in the race, especially if you’re missing a solid running base, like I was. The conventional approach is to do a “long run” once a week and increase the distance each week. The downside of this is that it takes a long time to develop the muscular endurance with this one long run and each long run will knock you out from other hard training for 3-5 days each time. You can't afford to do that when you only have 4 weeks to prepare. I averaged 6 runs per week.

Build cardiovascular engine with an all-out 10*1min set once a week: I always do this workout on a treadmill, as it’s perfect for regulating effort. Between each interval I recover for 1 minute, so the total workout length is less than 35minutes with a warm-up included. There’s ample scientific evidence (non-scientific article here) that short high-intensity interval training develops the entire “aerobic engine” rapidly, which is conventionally done by slogging long, slow miles. You can’t afford to do that when you only have 4 weeks to prepare and expect to be fresh race day.

Plyometrics for quick improvements in running efficiency: Even highly trained endurance runners seem to improve their running times in a few weeks of plyometrics. I simply do the following ~5 minute routine three times a week: 3 * 12 * explosive box jumps 3 * 12 * jumps for max height (sometimes replace w/ 20X skipping into a hill)

No long runs, but one 1 hour strength run per week I continued to run an easy “normal distance” run once a week, typically 50-65minutes, but it was focused on strengthening running muscles by running 3-6minute hills a few times. I tried to run each hill strong, but not anaerobically.

This approach took about 40 seconds (or -10%) per mile off of my best running speed in four weeks leading up to Wildflower. Typically a 5% improvement per year is a huge jump. Also, simultaneously, I gained about a pound or two of body weight so my speed gain wasn't due to change in body weight (on purpose in preparation for this).

The other time I used a very similar approach was before Ironman Sweden, where I ran a 2h56min marathon - by far my fastest - after riding 112miles.

I’m curious --  do you agree or disagree with this approach?

What I hear your asking...

Do you really think your two experiments prove the case? No.

Will this work for everyone? I don’t know, but I would guess that for most.

Do you need a massive “base” before this works? I don’t know, but I would guess you don’t need a massive base, but you need to have basic running fitness and mobility, so that you don’t injure yourself with plyometrics and high-intensity intervals.

Would you achieve the fastest times by continuing this program indefinitely? However you train, if you continue the same plan forever, it won’t be optimal for fitness gains.

My wife wants to double down on marriage and the global fight against sugar and processed foods. The brilliant idea -- that no husband could say no to -- is to row across the Pacific Ocean from San Francisco to Hawaii completely unsupported, launching in June '14. So that's what we'll do: www.FatChanceRow.org.

Hi Lisa,

Good question. Since I'm no medical expert myself, I'd highly recommend someone who has researched the topic for 30+ years: Dr. Steve Phinney and his work http://www.artandscienceoflowcarb.com/

His books are excellent resources of scientific information, but written for an average reader.

Athletes have long been focused on their recovery state to optimize training and to arrive fit and fresh at the race start line. This certainly applies to myself as well.

Recovery/stress tracking is also increasingly applied in the corporate world (ex: Spire at LinkedIN, Fatigue Science case studies) and in recreational use, where sleep (or lack of it) typically gets most coverage.

With all this market demand for accurate stress/recovery tracking and availability of cheaper and more accurate sensors, entrepreneurs have sensed the opportunity; I’ve now seen about dozen devices or services focused on measuring one’s recovery/stress state, some of which are still in stealth mode. 

After beta-testing or simply using many of these services, I think there are three key design hurdles that these devices have to overcome before they can dream of a real success in the marketplace. Here are the killer app product specs based on my own experiences as both a business (=office productivity) and athlete (=race course performance) user:

Extremely easy measurement, preferably continuous and automatic. Except for the most motivated athletes (e.g. professionals or some larger than life goal), it is very unlikely that the user will continue to play around with wires, straps and tools every day, even if it only takes less than a minute or two. That may work for a life-saving device, but not for a recovery tracking tool that may not have anything new to tell you that day. Great example: Besides sleep, Beddit measures Heart Rate and Heart Rate Variability (HRV) every night while sleeping without the need to do anything before, during or after going to bed. Not even clicking a mouse. The sensor is installed once (under the sheet) and it keeps measuring non-stop each night while in bed. Suboptimal example: Testing the performance of peripheral nervous system via sensors and stickers delivering electrical currents. The test itself is less than 20 seconds, but the setup takes precision and a few minutes. The test itself may be one of a kind, but the testing itself is a high hurdle.

Consistent and reliable results, preferably scientifically proven. This sounds very obvious, but I’ve tested several devices while I’ve been truly sick and/or visibly tired, but the service tells me I’m perfectly recovered. You only need one or two of those experiences and you’ll completely lose your belief in the product. Great example: Restwise is a recovery tool that incorporates a number of both objective and subjective inputs to provide a recovery score. From my personal experience, this tool provides consistently the most reliable results of any service I’ve tested so far.  Suboptimal example:  Purely heart rate based solutions that use a short measurement period. Heart rate does react to major stressors, but usually with a 1-3 day delay and sometimes it’s off for other reasons (cup of coffee, excitement, room temperature, etc.). HRV is a step-up from HR as a recovery measurement, but that too has been found to be too unreliable and with a long delay.

Actionable recommendations, not just data. This requirement would apply to most of the first generation “quantified self” tools. It is quite exciting to read some data that is coming out of your body, but this excitement wanes quickly, unless it is very clear to the user what to do next based on the measured data. And if this recommendation isn’t insightful enough (e.g., your leg hurts -> don’t run should be obvious) the device isn’t worth the time and money.  Great example: Omegawave has quite a specific recommendation (muscular and cardiovascular readiness separated) based on the recovery status, and not just numbers. Suboptimal example: Unfortunately almost all the devices and services that I’ve tested fall into this category and the user is left asking, “so what?”.

So, does a killer recovery/stress measurement tool that has nailed #1, #2 and #3 already exist?

The one that I currently consider the best, is based on a sensitive neural network that is also self-learning. It is (1) completely ambient and automatic if properly trained, (2) the results are always consistent if properly trained and it has been scientifically proven, (3) the recommendations are almost always actionable.

That is, the human brain combined with a high level or self-awareness. I’ve written about this before. However, to the credit of the many services and devices I’ve used (especially Restwise), my own self-awareness and ability to detect my mood in a matter of seconds has improved as a result of using these services.

Because she’s a very fit and gifted runner and triathlete, unlike most others! Simple and true.

This could be your body at Hawaii Ironman triathlon.

However, there are a few fundamental factors that change the limits of endurance performance at the Hawaii Ironman triathlon significantly and thus make certain athletes much more likely high performers in such conditions. Having just raced at that event the 7th time and seen many write-ups about “optimal pacing in Hawaii” or “how much slower you should expect to go in Hawaii” type of articles, I thought it would be interesting to look at the very specific reasons why Hawaii is different and why race pace recommendations based on “averages” is quite useless.

Human endurance performance is limited by many factors, many of which are still debated and whether it is our brain (=the central governor theory) or our failing peripheral muscles (=the cardiovascular model) that slows us down to our maximum sustainable speed. However, it is widely agreed that extreme heat and humidity negatively affect performance at maximum efforts. Interestingly, heat and humidity doesn’t slow down much if at all in the swim and bike portion of triathlon, which I’ve personally experienced and race simulation research in heat shows that too. So something different happens during the run.

Our bodies and brain try to avoid the destiny of eggs on a hot frying pan and there are two key factors pushing us away and towards this homeostasis while running:

Heating power that is the result of our human inefficiency. For every 1000kcal we metabolize, we only use about 200kcal (or 20%) to the forward moving motion. The rest is wasted as heat. (Roughly the same applies to cycling too, a cyclist pushing a bicycle forward at 300 watts is actually generating more than 900W of heat in addition to that forward moving power) Interestingly the energy cost of running is independent of our running pace. It takes about 1kCal per body kg to run 1km, regardless of running pace or body size. The consequence of this is that the heating power rapidly increases as you run faster and as you add weight.

Wasted heat production (in watts) at certain marathon running paces for athletes of different sizes (weight, kg). Assumes that heat power equals 80% of the total energy metabolized for the movement. As an example, a 50kg person running a 2:50 marathon “only” needs to get rid of 750Watts of continuous heating. Meanwhile, a 65kg person can only run a 3:40 marathon at the same waste heat level. Fortunately, our bodies can get rid of some of the excess heat, so heat doesn't always become the performance limiter.

Cooling power that is primarily a function of convectional cooling and sweat evaporation (not just sweating) from skin. Unfortunately cooling power does not increase at the same rate with excess heat production, as cooling is more closely related to human skin surface area than weight. Secondly, cooling is also very sensitive to ambient temperature and humidity. So the bad news is that at some running speed, runners should theoretically hit the point at which heating power exceeds cooling power, (soon) after which our bodies either shut down or turn into eggs on a frying pan. This almost never happens in cycling - unless climbing a very steep hill at walking/running pace - because at 20-40mph speed the cooling effect easily exceeds the extra heat generated.

The best approximate for body cooling during running I could find was in Tim Noake’s Lore of Running, in the following chart, which seems to assume strong air flow at max running speeds (about 5min/mile) and large skin surface area (about 75kg athlete).

Considering the average weather in Kona, HI in October is 29.4C and 84% humidity and stays at a remarkable narrow band over the years, I estimated what happens if you run at 2:50 marathon pace in these conditions by using Noakes’ cooling model and adjusting it based on body size (skin area):

 It looks like a 55-60kg (120-130pd) runner is not going to be limited by heat for a 2:50 marathon, but anyone larger than that will suffer and overheat at that pace based on average athlete's heat loss (cooling ability).

So the obvious question is how fast can a runner of size X run in the Kona, HI conditions. Based on the above assumptions and also including a +8% heat acclimatization factor that has been show at least in some studies,  I came up with the following:

Estimated best possible marathon performance in Hawaii Ironman based on heat as the limiting factor for different size runners.

From the above chart, if you dream of running a sub 3hr marathon in Hawaii, you shouldn’t be larger than about 72kg (or 160 pounds). For reference, the 2:50 marathon women's record holder weighs about 52-53kg and many of the top men in the 60-65kg range for this event.

So what’s the take-away for someone racing in Hawaii Ironman and hoping to run really fast?

Unless you’re very small (under 55-60kg) heat is likely going to be a significant performance limiter on the marathon run (not bike or swim) assuming you’re well trained to run fast otherwise.

Acclimatize and cool as much as you can on the run.

Hope for a cool and windy day on the run.

Watch someone like Mirinda Carfrae (52kg) run really fast every time despite the weather conditions in Kona.

I’ve met several lucky tech entrepreneurs, who have built a profitable, self-sustaining business. From these lucky ones, I’ve often heard the question “we are wondering whether we should raise outside money, take a much bigger risk and aim for ‘world domination’ in our space -- or not”. 

I think it’s a good question to ask, but in my view there’s only one feasible answer: Yes. The reason is simple: the alternative under almost all scenarios is that you run out of business and become irrelevant over time. And why is that?

Contrary to what most people think, I believe there is only one long-term sustainable (if cultivated well) competitive advantage for a tech company: Continuous innovation. In fact, I think that’s the only defensible competitive advantage for any tech company and innovation is the most important output of any enterprise in the fast changing tech world.

In addition, scale in some important aspect of the business (customers, data, marketplace liquidity, network size etc.) can buy some extra time to produce more innovation. But scale as competitive advantage often masks the reality and fools company leaders (especially those focused on operational excellence vs. innovation) to think they have a sustainable competitive advantage other than continuous innovation. Two simple examples of this pitfall:

eBay: An auction platform and its demand-side increasing returns (the more people join, the more existing users benefit from it) is probably the most impressive example of competitive moat with a built-in network effect. But what happened to eBay in the last 10 years given all its struggles? (According to some ex-employees) the company was laser focused on website yield, profit maximization and  true innovation was non-existent. eBay’s still alive, but has struggled against other online merchants, including Amazon. It’s far from invincible anymore despite the untouchable moat it was supposed to have.

Nokia: Massive scale, strong brand, loyal users. Company logistics and operational excellence became the envy of everyone. Yet, without continuous output of true innovation, competitors forced Nokia to collapse in a matter of few years.

So what choice is a profitable and content tech company left with?

It has to build an innovation machine that attracts the best, brightest and most innovative people now and in the future, not just for the v1.0 or v2.0 product (unless the aim is to sell a company based on a great v1.0 launch, which is a short term, likely failing strategy)

It has to also quickly build scale that gives more time and resources to make mistakes and innovate more against smaller competitors.

In other words, you either grow and innovate more, or fall apart. That really leaves only one choice: shoot for the stars!

Related thoughts on innovation as the only competitive advantage:

Niche companies? In some rare occasion there might be a tiny niche in which a tech company can operate relatively safely without growth and radical innovation. However, even this “opportunity” is typically just a matter of time before it is gone. As an example, what might happen to vertical specific registration services (e.g. concerts, sports events, etc.) when a platform like Eventbrite really gets to scale.

A friend had to take a red-eye flight to NYC last night and it reminded me that I (used to) ONLY take red-eye flights in the name of efficiency. 

Optimally you'd sleep nicely during the flight and wake up 5 hours later fully refreshed, ready for the 8AM meeting. But that almost never happens - instead, the morning after I feel like crap. Over the years I perfected a hack that has worked like a charm and quite often I heard "really, did you take a red-eye .. you look so rested and fresh!?".

This is how I've killed the red-eye fatigue:

1. Choose early red-eye's that arrive ~5AM and immediately check in to your hotel, before coffee or breakfast.

2. At your room, drink an espresso or take 100-150mg of pure caffeine.

3. IMMEDIATELY after #2, jump into bed and sleep ~15minutes (MAX 20minutes) with blinds on, so the room is dark.

4. IMMEDIATELY after #3, go to hotel gym for 10-15mins. First 5-10min either run or pedal bike easy/medium effort. Then do twice 1 minute ALL-OUT sprinting.

5. Shower, change and get to your 7AM or 8AM meeting.

6. CRITICAL: Don't eat any big meals during the day and avoid almost all carbohydrates, especially sugar, sugary drinks, rice, pasta, etc. that might spike your blood sugar and send to a death spiral two hours later. That means focus on protein, fat and some vegetables when you eat.

This routine (vs. three cups of coffee and straight to meeting from airport) works for three reasons, (i) caffeine is a stimulant, (ii) even little sleep helps to refresh, (iii) high intensity effort forces your body to release hormones that keep you alert and up. Just don't screw it up with a huge burrito for lunch.