Sa Calobra sub-30: same watts, 54 seconds faster

I’ve ridden Sa Calobra more times than I can count, and I’d been getting dangerously close to thirty minutes. You ride down to the port first, then race back up. On this trip I decided to stop being a nearly man. I thought it would be a sure thing: arrive on Monday, go under thirty on Tuesday morning, then spend the rest of the week on the limestone roads and on the beach with my family.

I underestimated it. It took three attempts and two failures.

Under thirty is considered a high benchmark, although Tom Pidcock’s knee-shattering KOM of 22:46 makes it look pretty mediocre. But I’m a parent with a full-time leadership role as head of content and creative director at Mindvalley. My training window is six to eight in the morning, mostly on the turbo, so everything I do is about intensity, and a thirty-minute effort suits that life exactly.

The goal also got me through the training. Halfway through my third ten-minute VO2 max interval one morning, the kind that sits right at your limit, every part of me wanted to stop pedalling. Picturing Sa Calobra kept me going.

Preparing to succeed

I worked with Team Sky and British Cycling from 2011 to 2019, Britain’s golden years, and I learned first-hand from Sir Dave Brailsford what preparing to succeed means: leave no stone unturned, and keep looking for the next marginal gain.

The bike came in sub-7kg, with a waxed chain, latex inner tubes, titanium bolts and no bottle cages. I wore a skinsuit and aero socks. I won’t rant about how aero socks look. They work.

I wore my grandad’s medal from a 25-mile time trial he won in 1949. I always wear it for luck, and a pendant is essential for any hill climb. I countered its minimal weight by leaving my heart rate monitor at home and riding to power.

Breakfast was two cups of my own coffee, made with the grinds and Aeropress I take everywhere (I’m a complete coffee snob), then a bar with over 40 grams of carbohydrate and an electrolyte drink. An hour before the climb I had some chews, and at the top of the descent a caffeine gel, 75mg, with another 22 grams of carbs.

Attempt one: the seatpost

My son was teething, so I’d barely slept, but I can survive one or two bad nights. The roads were empty, my nutrition and hydration were dialled, the bike felt smooth and light, and I felt confident. There was a headwind, but the mountain helped shelter me from it.

The first steep hairpin comes about six kilometres in. The racing line is the fastest way round, but it needs an extra kick, and I rarely get out of the saddle, so I sat back in my leather chair and put the power down. FUCK. My seatpost twisted to the right.

I’d torqued it to the recommended 6Nm when I rebuilt the bike after the flight, with a torque wrench and plenty of carbon grip paste. I knocked the saddle straight and kept going. Then the post dropped a few millimetres, and holding my target power got slightly easier. Eddy Merckx once told me my saddle looked a little too high. I should have listened. Then it dropped a few more millimetres, then a few centimetres, and holding my target power got twenty times harder.

There was only one sensible option. I sat up (metaphorically, since I couldn’t put any weight on the saddle), saved my legs for the next morning and rode to the top, where I’d hidden a musette bag in a bush. I fished out a multi-tool, sorted the seatpost and drank a recovery shake.

Attempt two: 30:47

The next morning I lowered the saddle 2mm and torqued the bolt to 6.2Nm. It held. I rode to power and cadence: 95rpm and 340 watts, pushing past 380 where the road steepened, and shifting down the cassette to get back to target whenever it flattened and the power dropped below 300.

As the clock approached thirty minutes, I knew I wasn’t going to make it. Where had the time gone? How had I fucked this up? I’d thought it was a sure thing. I went under the bridge and gave it everything I had left. My head unit chirped at the top: new PB, 30:47. I could taste metal, and my lungs were burning.

Back at the villa I uploaded the ride to Strava, and I was disappointed with the numbers. I’d averaged 325 watts. I know I can hold 340 for thirty minutes, and every time I’d glanced down I’d seen 340, give or take. So where had the fifteen watts gone?

Into gear changes. Chasing 95rpm, I was dropping power every time I changed gear. I’d also panicked and swapped the 53x39 chainset I’ve always ridden for a 52x36, to make that cadence easier to hold.

The fix was to ride a harder gear, like the 39 would have given me, and let my cadence vary: 75 to 85rpm on the steep sections and 95 to 105 as it flattened, instead of reaching for the shifter. On the shallower first six kilometres, I’d get lower and more aerodynamic.

Could I find 47 seconds? Could I give ten per cent more? No, I thought I’d already given everything. Five per cent? Maybe. Three per cent? Definitely. I had to try. I rode easy on Thursday and saved my legs for Friday.

Attempt three: 29:53

It was the worst night’s sleep of the trip. At half past three in the morning I was standing in the kitchen boiling the kettle, doubting I could do anything on so little sleep. I said to myself, “Fuck the data. Fuck self-doubt,” and decided to do it in spite of both. I made the strongest coffee I could, then another, until the sun came up over the mountains at quarter past six, and I packed an extra caffeine gel.

This time I raced myself. I turned on Strava’s live segments on my head unit, so I could see in real time whether I was ahead of my previous attempt.

At six kilometres I was sixty seconds up. Then the gap started to close. I’d clearly gone hard on the steep final 3.5 kilometres last time, and holding this pace took everything I had.

So I thought about my son. He’d spent the week with bright red cheeks, four top teeth coming through at once, chewing ice-cold cucumber to soothe his gums. I want him to grow up seeing what it looks like to work hard for something. His pain was my strength. I attacked the final straight clinging to 400 watts, and the last ten seconds felt like ten minutes.

My head unit chirped: new PB, 29:53. I fucking did it, with seven seconds to spare.

Relief.

Where the 54 seconds came from

My second and third attempts averaged 325 and 326 watts. I’d promised myself three per cent more and found a third of one per cent. The splits show where the time went instead. At six kilometres, the end of the shallow section, I was a minute up on my previous attempt. On the steep final 3.5 kilometres I gave six seconds of it back.

On that shallow first section you’re going fastest, so the air costs you most, and I rode it lower and more aerodynamic, as I’d planned. I’d also gone in with a harder gear and a plan to let my cadence rise and fall with the road. I’d expected that to win back the fifteen missing watts, and it won back one, so whatever it was worth, it wasn’t power. And Strava’s live segment showed me, second by second, whether I was still ahead of my previous attempt, and when the gap started closing, I thought about my son.

I didn’t record the wind on either morning, so I can’t rule it out. But if you’re chasing a time on Sa Calobra, the shallow first six kilometres are where I’d look first.

The sleep is the part I’m not proud of. I rode a maximal effort on almost no sleep and a lot of extra caffeine, and I got away with it. I put that down to luck, and I wouldn’t tell anyone to copy it.

I didn’t win a stage of the Tour. But as a parent with a full-time job, this was my definition of success. I’m already planning the next trip to Mallorca. Sub-29 is on the cards, so let’s call it sub-28.

← All journal entries