Training

Power Analysis: Zwift team time trial vs Tirreno-Adriatico team time trial

Take a dive into the numbers from two different TTTs — one on Zwift and one from the 2018 Tirreno-Adriatico — and see just how realistic Zwift is.

Many of the complaints about Zwift stem from its lack of realism, especially during large group rides and races. From the power-ups to the blob effect, to the super high speeds and presence of weight dopers, Zwift is not the most realistic riding platform in the world; and nor does it pretend to be. The gamification of Zwift is part of its very fabric and is what helps separate it from other platforms like RGT Cycling, Sufferfest, and Rouvy.

But perhaps the most realistic aspect of Zwift is its team time trials (TTT) organized by WTRL. Their TTT series has been Zwift’s most popular race series for years, with thousands of riders from around the world participating in the weekly WTRL TTT series, which includes teams and categories of all ages and abilities.

The recent launch of Season 2 the Zwift Racing League (ZRL) saw a number of high-profile TTTs added to the calendar, including the most recent round of the ZRL on Zwift’s RGV course. The ZRL includes both invite-only Premier Divisions, as well as Community Divisions which are open to the public. Normally, the live-streamed Premier League is the star of the show, but Round 2 of Season 2 offered a rare opportunity for the Community Divisions to test themselves on the same course as the Premier League “pros” to see how they would stack up.

***

After the opening round in Richmond, ZRL Season 2 continued on France’s RGV course with a 25km TTT. The course was mainly flat, but not without its undulations – the Aqueduc KOM at kilometer 13 is just 900 meters at an average of 2.3 percent, but included painful pitches of 5 – 6 percent. And in the final 6 kilometers, there were three back-to-back rollers lasting 300 – 400 meters at an average of 2 – 4 percent. While that doesn’t sound like a lot, after 27 minutes of going full gas, and with your heart rate already at 190bpm, those rollers feel like mountains.

Competing in one of the top Zwift Community Leagues – Americas E Division 1 – was NeXT eSport, a newly formed Zwift racing team including some of the top racers on Zwift, as well as myself. My role was the same as my teammates’, and it was simple: once you hit the front, pull at 6.5 – 7w/kg for 45 seconds at a time. As soon as you finish your pull, sink to the back of the pace-line as quickly as possible, and accelerate hard to get back into the draft. Repeat for 30 minutes.

Just because it’s simple doesn’t mean it’s easy.

My teammates and I all understood our roles and the effort that would be required, and we slotted into our paceline as soon as we got out of the start gate. Also riding for NeXT eSport were: Brian Duffy Jr., Scott Catanzaro, Chris Uberti, Greg Grosicki, and former individual pursuit world record holder Ashton Lambie.

A few kilometers in, the effort began to bite. 300w is hard enough to hold on its own, but consider that being your “recovery pace” during a 30-minute over/under interval. We nailed our pulls during the first half of the race, with each rider holding 400 – 500w for each of their 45-second pulls. Looking at the power charts, we can see the brief mesa that forms during each 500w pull, followed by a steep drop (“sinking”), and then a sharp spike to help latch onto the draft at the back of the group.

Nehr (NeXT eSport) – opening kilometers of ZRL TTT
Time: 8:01
Average Power: 376w (5.2w/kg)
Normalized Power: 396w (5.4w/kg)
Average Power/pull: ~485w (6.7w/kg) for 45 seconds

Lambie (NeXT eSport) – opening kilometers of ZRL TTT
Time: 8:01
Average Power: 375w (5.1w/kg)
Normalized Power: 408w (5.6w/kg)
Average Power/pull: ~500w (6.9w/kg) for 45 seconds

After leading at the first intermediate split, it wasn’t until the final third of the race that NeXT eSport ran into some trouble. The 3 percent rollers took their toll, and with a few kilometers to go, the team had just four riders remaining. Since the team’s time was taken on the fourth rider across the line, they couldn’t afford to lose anyway else. More S-turns and undulations nearly blew the paceline apart, with each rider on their limit over the final few minutes.

Lambie (NeXT eSport) – final kilometers of ZRL TTT
Time: 3:54
Average Power: 451w (6.2w/kg)
Average Heart Rate: 194bpm
Max Heart Rate: 201bpm

As they rounded the final corner with 300m to go, the riders of NeXT eSport emptied their tanks and crossed the line in a time of 30:28 – fast enough to win Americas E Division 1, and also four seconds faster than the Premier League TTT winners, POAUTO-CeramicSpeed. Each rider hunched over their handlebars, struggling to breathe, and with a heart rate of close to 200bpm – it was an incredible victory for a team who had only started racing together a few weeks prior.

Lambie (NeXT eSport) – full ZRL TTT
Average Power: 386w (5.3w/kg)
Normalized Power: 407w (5.6w/kg)

Nehr (NeXT eSport) – full ZRL TTT
Average Power: 381w (5.3w/kg)
Normalized Power: 394w (5.4w/kg)

Duffy, Jr. (NeXT eSport) – full ZRL TTT
Average Power: 372w (5.5w/kg)
Normalized Power: 395w (5.9w/kg)

***

To compare this Zwift TTT to a real-life time trial, we can go back in time to the 2018 Tirreno-Adriatico which began with a 21.5km team time trial along the Mediterranean, in Lido di Camaiore. The pan-flat parcours and complete lack of turns – only the turnaround at Forte dei Marmi – make for one of the fastest time trial courses in professional cycling. Because the speeds are so high, riders have to push an incredibly high wattage in order to overcome the wind resistance and maintain 55kph+. That means shorter pulls for most riders, and a higher overall speed compared to other TTTs, and compared to ZRL’s TTT on RGV.

Team BMC lined up as favorites for the opening stage, with a line-up that included Rohan Dennis, Stefan Küng, Alberto Bettiol, and Greg Van Avermaet. As soon as the American outfit exited the start ramp, they accelerated down the course and settled into a rotation at nearly 60kph. Küng – the uber-strong time trial rider that he is – was one of the main workhorses for Team BMC, taking pulls at nearly 600w (7.2w/kg) for 30-35 seconds at a time.

Küng (Team BMC) – first half of 2018 Tirreno Adriatico TTT:
Time: 10:00
Average Power: 471w (5.8w/kg)
Normalized Power: 505w (6.2w/kg)
Average Power/pull: ~590w (7.1w/kg) for 32 seconds

Within Team BMC, we can see the discrepancy between the bigger time trial specialists and the lighter climbers. In a TTT like this where the speeds are high over a flat parcours, the power needed to maintain such an effort is 550-600w (given a world-class, low drag coefficient), which is a lot more “comfortable” for 80kg rouleurs than it is for 65kg climbers.

We can see this played out in the power numbers of Damiano Caruso, a 67kg climber on Team BMC who took just 10 to 15-second pulls for the duration of the TTT, but at a harder relative effort compared to his larger teammates.

Caruso (Team BMC) – first half of 2018 Tirreno Adriatico TTT:
Time: 10:00
Average Power: 395w (5.9w/kg)
Normalized Power: 505w (6.2w/kg)
Average Power/pull: ~565w (8.5w/kg) for 12 seconds

In the final few hundred meters, Caruso moved to the front and crossed the line first for Team BMC, leading the team to a four-second victory over Mitchelton-Scott, with an other-worldly average speed of 57.8kph (35.9mph).

Compared to Zwift, there are a number of key differences between the two TTTs, but also a handful of comparable similarities. First, there is the timing and intensity of each pull and how it relates to the average speed of the group. In both Zwift and in real-life, the top teams seem to use ~30-45 second pulls for their strongest TTT riders, and perhaps a little bit less for their weaker riders. The intensity of each pull is also similar, with 6.5 – 7w/kg being the target range for most top-tier efforts.

Küng (Team BMC) – full 2018 Tirreno Adriatico TTT:
Time: 22:19
Average Power: 448w (5.5w/kg)
Normalized Power: 475w (5.8w/kg)

The average speeds between the ZRL TTT and Tirreno-Adriatico TTT vary greatly because of the drafting, wind, and rolling resistance algorithms of Zwift versus real-life factors in Italy. During the ZRL TTT, teams were only allowed to use road bikes, not TTT bikes; while in real-life, the wind-tunnel-tested and scientifically engineered skinsuits, aero helmets, TT bikes, and disc wheels were all on display in the Tirreno-Adriatico TTT, helping explain the massive difference in average speed (50kph vs. 57.8kph) between the two TTTs.

Overall, Zwift and WTRL have done an incredible job forming a realistic TTT platform in virtual cycling. Mastering the nuances of the game, drafting, and algorithms can be overwhelming at times, but with all things considered, Zwift’s TTT mode closely mimics the physical efforts of real-life TTTs, with all the suffering that you would expect and more.


Zwift Power ZRL Season 2 Round 2 Official Results

Power Analysis data courtesy of Strava and Zwift Power

The team at ZADA (Zwift Accuracy and Data Analysis) helps ensure fair competition for all, and verification for all riders participating in the Premier Division, and a select number of riders and teams throughout the Community Divisions. 

***

Men’s riders on Strava: 

Stefan Küng

Damiano Caruso 

Zach Nehr

Ashton Lambie