Get access to everything we publish when you sign up for Outside+.
In today’s Power Analysis, we’re going to do something different, and look at the similarities and differences between virtual and in real life (IRL) racing, and see how the numbers of indoor specialists compare to real-life pros.
This past weekend was the Virtual Tour of the Gila on RGT, the second installment of the Echelon Racing League on RGT.
I raced it and placed second overall, so I can offer a little personal perspective in addition to my usual presentation and interpretations of the numbers.
The RGT platform is similar to — but different from — Zwift. RGT has unique algorithms that replicate many of the physics of riding outside, like braking force in the corners, for example, causing the riders to slow down, and making it harder (and much more life-like) to get back on top of the pedals and regain speed coming out of a tight bend.
These nuances were felt the strongest during Stage 1 of the Virtual Tour of the Gila on RGT, a 34km (21mi) criterium.
For each Virtual Gila stage, Project Echelon uploaded GPX files from the actual Tour of the Gila courses, which RGT then converted into “Magic Roads” with matching elevation and course profiles to the real race.
Stage 2 of the Virtual Gila finished atop the infamous Mogollon climb, while Stage 3 traversed the Gila Monster finale, an epic, stair-stepping climb with pitches of over 16 percent.
On Sunday, the pro men’s field tackled the Gila Monster course on RGT, with Tom Thrall (EVOQ.Bike) coming out on top in the final sprint ahead of me (Project Echelon Racing) and Hayden Warner (Above and Beyond Cancer Cycling).
Back in 2019, James Piccoli (Elevate–KHS Pro Cycling) secured his biggest win to date by taking the overall win at the UCI 2.2 Tour of the Gila after finishing sixth on the final Gila Monster stage.
Before we dig into the numbers comparing Piccoli’s numbers to those of Thrall’s and mine, let’s agree that two major differences between virtual and IRL racing at the Tour of the Gila affect power: altitude and distance.
First, the real Gila is based in Silver City, New Mexico, and the stages are held between 1,800m (5,900ft) and 2,300m (7,500ft). More altitude means less oxygen which means lower power, in acclimated and unacclimated riders. In virtual racing, there is no ‘altitude adjustment’ setting (yet); the oxygen is as rich as the air that surrounds the trainer, no matter how ‘high’ the climb.
Second, the five-stage race has substantially more distance than the three-day RGT event, which means more fatigue. Take the Mogollon stage, for example. In real life, the peloton arrives at the base of the Mogollon having already covered 144km of rolling desert road and nervy crosswind sections, which also require energy. In contrast, Virtual Gila riders began the final climb up the Mogollon after just 44km – not at altitude, not in the desert heat, and with the only crosswinds coming from the industrial-size fan blowing in their face.
2020 Virtual Gila on RGT vs. 2019 Tour of the Gila
Stage 3 of the Virtual Tour of the Gila began in the replicated ghost town of San Lorenzo, New Mexico. The peloton stayed mostly intact for the first 45 kilometers, with only a handful of riders falling off the back as the race began in earnest.
Thrall and his EVOQ.Bike teammates patrolled the front of the race, helping to keep breakaways in check and defend Thrall’s GC lead before the final climb of the Gila Monster.
Thrall – first 45km of Virtual Gila Monster stage:
Average Power: 321w (4.4w/kg)
Average Speed: 45.7kph (28.4mph)
In the 2019 Tour of the Gila, Piccoli and the rest of the pro men’s field had already covered 35km and climbed over 600 meters by the time they reached San Lorenzo. The early breakaway had been established over these first few climbs, and Elevate-KHS Pro Cycling began to control the pace in order to protect Piccoli’s GC lead.
Thanks to a combination of altitude, heat, and fatigue, the real-life race went across the valley significantly slower than the virtual peloton — covering this 45km stretch a full ten minutes slower — but also making a much easier ride for Piccoli who was saving his matches for the few final climbs.
Piccoli – first 45km of IRL Gila Monster Stage:
Average Power: 207w (3.1w/kg)
Average Speed: 39.6kph (24.6mph)
After 45km, the Virtual Gila continued straight on to the final climb of the Gila Monster, while the real-life race turned right and headed up and over the climb to the Cliff Dwellings. The climb on the way out is 11.2km (7mi) long at an average of 4.3 percent, but it is the return that truly does the damage – 7.6km (4.7mi) at an average of 7.1 percent, and pitches over 12 percent, shredded the Gila field down to less than 30 riders before the Gila Monster has even begun.
On the climb out, Piccoli averaged 4.8w/kg for 26 minutes while his teammates rode the front and began clawing back time on the breakaway. After a lightning-fast descent where riders reached speeds in excess of 90kph, the peloton did a U-turn and headed straight back up the harder side of the climb. After pushing over 500w at the bottom, Piccoli settled into a steady rhythm, and crested the return leg well within the front group, with the Gila Monster looming at the bottom of the descent.
Piccoli – Cliff Dwellings return climb of IRL Gila Monster stage:
Average Power: 347w (5.2w/kg)
Average Gradient: 7.1 percent
The Gila Monster finale is one of the toughest finishes in North America because it seems to go on forever. After an initial 5km climb that ascends over 400m, the road keeps on rolling upwards for another 11 kilometers, before finally relenting – but only for a moment. The final 4km to the finish are again uphill, with steep pitches of over 11 percent thrown into the mix, and a long drag to the line with pitches of 8-9 percent in the final few hundred meters.
In the Virtual Gila, the front group exploded on the lower slopes of the Gila Monster, leaving less than 15 riders left in contention after the initial 5km. I put in a few accelerations on the steep pitches, pushing over 7w/kg in hopes of dispatching as many riders as possible. By the top, the majority of the damage was done, and both EVOQ.Bike and Project Echelon Racing had three riders each in the front group, setting the stage for a challenging and tactical finale.
Nehr – first 5km of the Virtual Gila Monster climb:
Average Power: 387w (5.4w/kg)
Average Gradient: 6.6 percent
In the real-life Gila, a steeper gradient meant a slower time, but a similar power-based effort for Piccoli and the rest of the front group. Just 14 riders made the selection over the first 5km, with Piccoli now defending his lead GC lead against the likes of Aevolo, Team Medellín, and Floyd’s Pro Cycling.
Piccoli – first 5km of the IRL Gila Monster climb:
Average Power: 360w (5.4w/kg)
Average Gradient: 8.2 percent
For as many variables as there are in bike racing – as well as the potential differences between real-life and virtual racing – RGT might be as life-like as it gets. After the first 5km, the Gila Monster continues climbing for another 11km over rolling and twisting roads, and the efforts of Nehr and Piccoli were nearly identical over these stretches of virtual and real roads, respectively.
Nehr – Virtual Gila Monster rollers:
Average Power: 324w (4.5w/kg)
Average Gradient: 1.5 percent
Piccoli – IRL Gila Monster rollers:
Average Power: 302w (4.5w/kg)
Average Gradient: 1.2 percent
After a well-deserved rest down the descent from 13km to go to 4km to go, the final climb to the finish began. This final run-in is tricky, punchy, and never flat – it takes a special rider to win the final kick up to Pinos Altos at the end of the queen stage of the Tour of the Gila.
With a 46-second buffer over 2nd in GC, Piccoli was focused on sealing the GC instead of going for the stage win, but still put in a massive effort over the final few kilometers to finish 6th on the stage and win the Overall at the 2019 Tour of the Gila.
Piccoli – final 4km of the IRL 2019 Gila Monster:
Average Power: 347w (5.3w/kg)
Normalized Power: 390w (6w/kg)
Average Gradient: 3.8 percent
In the Virtual Gila, the front group of 12 stayed together for the first couple of kilometers on the road to Pinos Altos, but at 1.6km to go, a series of attacks blew the group to bits and left Patrick Walle (EVOQ.Bike) with a small gap off the front and just a few hundred meters to go. But bike racing is cruel, and just when Walle looked to have the win in the bag, Nehr attacked out of the front group and went across and then straight over the top of Walle with less than 300 meters to go. Unfortunately for Nehr, he had Thrall on his wheel, the strongest man in the race and the winner of the first two stages of the Virtual Gila – Thrall came flying over the top of Nehr, and powered away to the stage win and a dominant GC victory at the Virtual Tour of the Gila.
Nehr – final 4km of the Virtual Gila Monster:
Average Power: 389w (5.4w/kg)
Normalized Power: 413w (5.7w/kg)
Average Gradient: 3.3 percent
Final kick to the line: 0:40 at 639w (9w/kg)
Thrall – final 4km of the Virtual Gila Monster:
Average Power: 423w (5.8w/kg)
Normalized Power: 443w (6.1w/kg)
Average Gradient: 3.3 percent
Final kick to the line: 0:47 at 694w (9.5w/kg)
All in all, it is fascinating to see the similarity between virtual and real-life racing in terms of power output. Obviously, there are significant differences between each discipline – altitude, weather, positioning, gamification, overall fatigue, etc. – but in the end, the two disciplines have a lot more in common than many initially thought, and we wouldn’t be surprised to see more pro riders using virtual racing to simulate the physical effort of real-life racing.
Strava files for Thrall, Nehr, and Piccoli for the Gila Monster stage
Tom Thrall (EVOQ.Bike) – 2020 Virtual Tour of the Gila Stage 3
Zach Nehr (Project Echelon Racing) – 2020 Virtual Tour of the Gila Stage 3
Piccoli (Elevate–KHS Pro Cycling) – 2019 Tour of the Gila Stage 5
**Project Echelon and Echelon Racing Promotions helped organize and police the Virtual Tour of the Gila, using weigh-in videos, power calibration, and more to ensure that rider data and subsequent results are as accurate as possible. Special thanks for all of their hard work.