Posted on Leave a comment

Barry-Roubaix 2023 – Race Report (and thoughts on Garmin Power Guide)

Roubaix in Barry

A week after Paris-Roubaix weekend, a bit less famous, yet way more crowded, race starts in West Michigan. Called Barry-Roubaix, and claiming to be a most popular gravel cycling event in the US, it attracts gravel cyclists from all of the Midwest region.

Held in March in previous years and infamous for its unpredictably bad weather (and gravel) conditions, it was moved to mid-April this year and attracted almost 4,000 cyclists to the town of Hastings (population 7,500), where they raced in 100, 62, 36 and 18 mile distances on the rolling hills and gravel of the Barry County.

Preparation (or Lack of It)

I was not training specifically for this race, but since other races scheduled for this year have hills involved, signing up to Barry-Roubaix with a good group of fellow athletes fit my plans well.

Of course, I adapted my triathlon training plan to include slightly more bike time. And since Chicago weather is not the best for winter and early spring riding outdoors, I did all of my training indoors on my Elite Direto XR. For hill-focused training, I used another Elite product – Rizer (pictures of the setup are in the gallery), that is used to simulate up to 10% downhill and 20% uphill by lowering/lifting the front fork. It is also supported by most popular virtual reality bike training apps, like Zwift or Fulgaz. My choice of app was Rouvy, primarily because it had the course of one of my upcoming race, and also a number of famous climbs featured in Tour de France and Giro d’Italia. And when not training for climbing, I spent my time on trainer watching GCN+ — the only TV subscription that I have 🙂

Spending around 5:30 hours per week in a saddle since December 2022, I felt quite prepared to have fun in my first ever gravel bike race (as when not doing triathlons, I’m more of a runner). Since I have only a TT bike, friends arranged me a a nice gravel bike, I did a couple (literally, two) bike rides outdoors, and felt like I’m ready to go.

Race Report (Short One)

Since the race descriptions are usually boring and interesting only to the athletes themselves (and maybe their very close friends), I’ll boil it down to the key points and switch to the more interesting topic – Garmin Power Guide numbers and if this tool is worth trying or not.


We left Chicago on the same day around 4:00 AM in the morning. Since I was not driving, took a couple of good naps, and around 8:30 AM (this includes a stop in Starbucks to meet other group of friends and +1h time zone change Illinois –> Michigan) we were at the ACE Hardware Store to pick up the bib numbers.

As it happens to me all the time before my first race in the season (technically, my first one was the Bank of America Shamrock Shuffle, but I was not specifically preparing for it), the stress and anxiety was going through the roof. And that is the reason why I always tell to myself and my athletes – you have to race as much as you can before your A race(s) so that stress won’t burn you out and you would come to your main race to do your thing and to celebrate the end of your grueling preparation cycle.

The Race

We started with the very first wave of 36-milers, yet not being the elite cyclists, we placed ourselves at the back of the starting group. Since the official race time was taken from the wave gun time, that most likely costed me “sub-2:00” finish which was kind of my goal (my official time is 2:00:03, while my Garmin start/stop was 1:58:02), but having in mind my experience, preparation, etc. I think it was a good decision.

During the race itself there were a few embarrassing moments when due to my lack of skill of riding in the sand (or avoiding it) I fell off the bike by potentially making some riders behind me to unclip as well (not sure about that, or at least no one yelled at me; maybe they were just nice people) and had to push my bike for a dozen of yards to clip back in.

There was also one sh-***-t moment, when going almost full speed downhill I saw the 90-degree left turn too late, mainly because other riders in front of me started braking, skidding and trying not to hit the gutter. I started braking as well with my cantilever brakes that were shrieking like banshee (I’m happy that I didn’t have disc brakes, that most likely would have been too efficient for my lack of bike-handling skill and I would have taken down riders that were behind me), yet with some help of God I managed to stay on the road, yelled ‘Sorry!’ just in case and moved on.

In general, course was somewhat challenging but hills were not that crazy, and with a bit of my aforementioned ‘hill training’ on the Rizer and watching my body weight they were completely doable. As mentioned before, I did not get under official sub-2:00, yet that didn’t screw my mood, as the actual goal was to get myself and the bike through the finish line in one piece.

The whole race was extremely well organized, starting from the flawless and fast bib pickup, to a decent amount of port-a-potties (there was *no line* to hit one, even close to starting line), to a super fast tarmac finish where you could burn your last match, and a few hundred yard long ‘landing zone’ so that there were no congestion of riders stopping and celebrating their finish too early.


Things that worked

Things that I wanted to test/validate before the race, and they kind of went well (I think):

  • Taper, foam rolling/stretching and pre-race opener workout a day before
  • Nice amount of food (not too much, but not too little either) a day before, including a visit to my favorite Mexican place; even if it added more than a pound for the race morning, it made my pre-race day much happier
  • Pre-race meal plan in the morning (overnight oats 4h before and bagel-egg-dates-peanut butter sandwich 2h before the start)
  • Hitting a goal of consuming 100g carbs per hour during the race (no solid food, just drinks and gels on course); I was named “carb burning machine” during my VO2Max lab test, so it was really interesting how consuming a decent amount of carbs will help with the performance (and if I won’t have digestive issues), and it worked perfectly fine. I have tested the setup during the long indoor rides before, but testing it in racing conditions (with intensity and general stress) boosted my confidence even more.
  • Recovery snack immediately after the race homemade ugali and Rockin’ Refuel Muscle Builder from Shamrock Farms (read: overpriced top shelf chocolate milk with highest protein content)
  • 1x post-race IPA and 4 slices of pizza 🙂

Things to do differently

Not many of those, as gravel racing is not my target, however:

  • brushing gravel riding skills would help
  • more awareness to what’s happening around instead of focusing on speed/power
  • more careful shifting, as derailleurs may (and will) go out of service if you won’t

Thoughts on Garmin Power Guide

Since race organizers did a good job sharing race course GPX files well in advance, I decided to play around with the Garmin Connect feature called Power Guide, that is supposed to break the ride course into splits and suggest target power to each of them based on multiple parameters, including riders weight, FTP, bike weight, etc. Also, it is possible to tune the desired effort level according to the percentage of average FTP desired for the ride.

Here’s how my settings for the race looked like (the only thing that I had no clue is the ‘Bike & Gear Weight’, so the number my be random):

Since I always make pace, power and fueling plans for my races (yet never stick to it when racing and go by feel, with an exception of fueling plan), this time it was no exception. However, it makes it even interesting to analyze the actual efforts after the race and see if it matches to the plan, and what to learn from it if not.


Plan settings:

  • I set the Effort level to 82% of my FTP of 218W (FTP estimated by Garmin), so the average power for the race had to be 179W.
  • Sum of the predicted split times was 2:02:51 (honestly, I did not know that number until I started my post-race analysis)

Actual numbers:

  • Average power: 175W (Normalized Power [TrainingPeaks]: 197W). Difference from the estimate: 2.29%
  • Actual time: 1:58:09. Difference from the estimate: -3.83%. Important! this is not official race (gun) time, but the time from Garmin workout start when crossing start line to the finish line. The official race time is 2:00:03.

All in all, I would say numbers are chillingly accurate 🥶

So in my opinion, it is definitely worth playing around with the tool and finding out how can one leverage it.

Splits, Numbers and Conclusions

We can also dive into some boring numbers that are most likely interesting only to myself and maybe one or two other data geeks in this world, who would also like to try this feature and see if it’s worth it to stare at your Garmin screen and see that they have a good number of kilometers left in a split at the power that they cannot longer sustain 😉

Garmin broke the 36 mile (or 57.82 km, according to GPX) course into 30 splits. After loosely comparing the estimated and actual values I came to the following conclusions:

  • It seems that target for power output when descending is underestimated. I’m bad at descending, and even accounting that my power output at >1% descents in 6 of 7 segments was higher (sometimes up to four times) than Garmin suggests.
  • With climbing power output picture is more accurate, however Garmin thinks that you’re a decent climber. After all the training I’ve done I would consider myself to be in the middle of the climbing pack, but for inclines that are steeper than 2% I matched the suggested power just for 1 of 6 instances (I was not matching others by 10-20%)
  • Garmin does not account for final sprint 🙂
Leave a Reply

Your email address will not be published. Required fields are marked *