The weekly spin: Controversy? Welcome to the club, eRacing
The first week of October has already made it a banner month for surreal, post-modern headlines.
“Twitter pulls Nickelback clip tweeted by President Trump over copyright complaint” qualifies. Step back in time a decade and try to explain that one to someone.
“First British eRacing champion banned for using bot to gain access to Tron bike” was another.
Controversy in bike racing? Welcome to the club, eRacing.
When it comes to cycling competition, digital cheating is nothing new. Not long after Strava launched in 2009, tools emerged to “juice” one’s ride files in order to boost their standings on segment leaderboards. Suddenly, data doping was a thing.
On Zwift, the online multiplayer cycling video game and training program, riders are required to enter their weight, so that climbing speeds may be calculated using watts-per-kilogram; the entire Zwift ecosystem is based on the honor system.
And now eRacing has had its first high-profile disqualification for cheating.
Ride it like you stole it. pic.twitter.com/TdslhZJr1v
— Zwift News Network (@ZwiftNews) October 4, 2019
One week after closing the UCI Road World Championships in Yorkshire, British Cycling announced that it had stripped 22-year-old Cameron Jeffers of his eRacing national title, earned on live television at a March 28 at an event held in London. Jeffers was also handed a small fine as well as a six-month suspension from all sanctioned competition, virtual and real.
In all reality, however, this situation is far from black and white, prompting questions about how misconduct is defined, how virtual races are governed, and whether they can ever truly capture the attention of average pro cycling fans.
There is no question, however, that Jeffers cheated Zwift’s upgrade process in order to gain access to the Concept Z1 bike, one of the fastest, most aerodynamic in-game bikes offered, known as the Tron bike due to its futuristic glow. He’s admitted to that.
Prior to the British eRacing championship — and before British Cycling published its eRacing regulations — Jeffers employed a third party to run a simulator on Zwift, faking the performances required to unlock the Concept Z1 bike. The bike is awarded to Zwift users who compete its “Climb Mt. Everest” Challenge, for a total of 50,000 meters (164,042 feet) of virtual elevation gain.
Jeffers, who is active in both Zwift and on YouTube, used that bike in a February 24 qualification race, as well as the British eRacing national championship.
Within days of his national championship victory, British Cycling officials received an anonymous email complaint that Jeffers had broken Zwift’s terms of service in acquiring the Tron bike. And it wasn’t a casual violation; according to DC Rainmaker’s reporting it was a blatant, premeditated workaround that involved simulating superhuman efforts to be credited with elevation gain as quickly as possible, logging in from multiple locations, and then purposefully discarding those rides to avoid detection.
“Jeffers had ridden on multiple occasions at 2000 watts for over 200km with a weight of 45kg, and had also logged on in multiple locations over a short period of time (Cheadle, Plymouth and Denmark,” a Zwift statement read. “Zwift detected a bot in the system to falsify power data and unlock the Tron bike… The historic ride data of the rider in question had not been saved after each activity was completed and therefore did not show on initial checks.“
— British Cycling (@BritishCycling) October 4, 2019
As pointed out in that DC Rainmaker article, using automation or bots in place of human performances is against Zwift’s Terms of Service, so the validity of Jeffers’ Zwift account during the qualifying and national championship races is up for discussion. But his violation had not been identified at that point, and he was free to compete. He actually finished 12th in the qualifying race, but two riders dropped out of the finals, opening up a spot for the national championship.
Jeffers cheated to get the bike, then won the national championship aboard a bike several other competitors were using; presumably they had earned the bike in adherence to Zwift’s upgrade rules. He cheated the system before the event, but won with a legitimate performance on dubiously acquired equipment.
After his victory, ahead of several pros from the Madison-Genesis team, Jeffers said, “I’ve done a lot of course research, a lot of riding and racing on Zwift, and researching the different power-ups. So all of that put together I managed to pull off the win.”
DOES THE PUNISHMENT FIT THE CRIME?
After six months of investigation, and conveniently just after the British world championships had concluded, the sanction came down — disqualification and a six-month suspension, as well as a nominal fine. Jeffers was not penalized by Zwift for violating its rules, but rather by British Cycling for violating discretionary misconduct rules intended to protect the spirit of competition.
On Friday, the same day British Cycling announced his suspension, Jeffers posted a statement, as well as a lengthy YouTube video, acknowledging his wrongdoing while also claiming that British Cycling had published its rules for eRacing after he’d acquired the Concept Z1 bike.
However that’s not quite correct. The rules that British Cycling enforced are from its general rulebook, and not specific to eRacing. Specifically, the federation pointed to Clause 3.2 (h) in its Disciplinary Rules for Cycle Sport, which defines misconduct as “fixing or contriving in any way or otherwise influencing improperly the result, progress or conduct of any event in which the participant is participating in and/or can influence.”
Within British Cycling’s rules on eRacing, published on March 8, there is no specific mention of in-game equipment. Perhaps that’s because it’s essentially impossible to verify who was riding when certain equipment was unlocked — and why the federation instead cited its pre-existing, all-encompassing rulebook.
Jeffers also claimed that he wanted to access the Tron bike because it “looked cool.” That’s not quite right, either. Previous comments he made online, and his prominence within the Zwift community, make it clear he was aware it was the fastest bike Zwift offers. So even in his acknowledgement and apology, Jeffers hasn’t been forthright.
That said, whether or not the punishment fits the crime is open to interpretation. Some have argued that the penalty is too harsh; Jeffers was the best rider on the day, and any federation rule he might have broken is undefined, and discretionary. Others have argued the penalty was not harsh enough; clearly he went to great lengths to cheat Zwift. What other lengths might he go to in order to get an advantage?
“I accept this practice was unethical and unsporting and I have fully cooperated with [British Cycling] on their investigation,” Jeffers wrote in his statement. “I fully believe in eSports and its part in cycling’s future.”
WHOSE RACE IS IT, ANYWAY?
And what this means for eSports, and its part in cycling’s future, is perhaps the bigger story.
On September 29, during the Road World Championships in Yorkshire, the UCI announced that, in partnership with Zwift, it will host its first ever eSports World Championships in 2020. That means they’ve got less than a year to make sure issues like this don’t come up.
Ultimately, it seems the sanctioning body — British Cycling, or the UCI — must be responsible for policing the equipment the riders use in virtual competition. As the British Cycling example demonstrates, Zwift was unable to flag a competitor who so blatantly cheated their system. If they aren’t able to tightly monitor their own database to identify fraudulent performances meant to unlock equipment, that equipment should not be used at high-stakes events. More on that below.
In the scenario of a national or world championship, it seems Zwift should play the tole of a race organizer; they provide the infrastructure, the venue and the route. As it is, Zwift’s role is nebulous, providing the platform as well as the ability to access advantageous equipment, with its own rules surrounding how that equipment can be accessed.
I reached out to both Zwift and the UCI for comment on whether access to equipment should be placed behind lock codes for a national or world championship. Is it fair that some competitors should be on faster bikes than others, or should they all compete on the same standardized bikes to level the playing field?
Virtual racing is not real-world racing, I realize, and equipment choices in the pro peloton are dictated by team sponsorships, but it is interesting that UCI rules prohibit the use of equipment that is not available to all competitors. Rule 1.3.006 states, “Equipment shall be of a type that is sold for use by anyone practicing cycling as a sport.”
Of course availability is subjective, and in the case of the Tron bike, it was available to anyone who chose to log 50,000 meters of elevation in Zwift.
The UCI replied with a prepared comment, sent via email. It doesn’t address the issue of standardized bikes, likely because they don’t yet have an answer.
“We can only deplore this cheating attempt, but we congratulate British Cycling on the handling of the case, both in the detection of the fraudulent material and the disciplinary procedure. Ensuring the integrity of our sport is a fundamental part of our strategy, as demonstrated by our investment in anti-doping and the fight against technological fraud.
“On the governance and development of cycling eSports as a new cycling discipline and the introduction of the first UCI Cycling eSports World Championships in 2020, the UCI and Zwift will collaborate to ensure the sporting credibility of the cycling eSports competitions. This will include measures to ensure that manipulation of both hardware and software by competitors is counteracted, as well ensuring that rider profiles are verified ahead of competition.”
No eSports regulations have been published by the UCI. There’s also been no definitive decision made as to whether or not participants in eSports will be put into the WADA drug-testing pool.
I reached out to Zwift eSports CEO Craig Edmondson to ask about issues such as standardized equipment, identifying fraudulent performances meant to complete challenges, and Jeffers’ violation of Zwift’s terms of service. He sent several detailed replies, via email.
On the issue of standardized bikes, Edmondson wrote that equipment choices are part of the game, and that all bikes on Zwift are “available to all riders who are willing to put in the time and effort to earn these bikes through the accumulation of more [experience points].”
“The performance differences between the equipment in game are quite subtle, just like the differences in gear in real life,” Edmondson continued. “Like real-world athletes in traditional sport fairly deploying equipment that gives them an advantage, utilizing a different bike in Zwift is similarly part of the rider strategy.”
Asked if equipment advantages have a place in a national or world championship, Edmondson wrote that Zwift is still working out these details with the UCI.
“We haven’t yet announced the race formats, rules or regulations for the UCI Cycling Esports World Championships or the qualifying events,” Edmondson wrote. “We are working with the UCI to develop these. Fair competition and good governance are integral to the establishment of this new cycling discipline.”
He also denied rumors swirling online that some of the pro riders using the Tron bike at the British championship had reportedly been “gifted” the bike by Zwift. “No, this is not the case,” he wrote. “Zwift has never ‘given’ the Tron bike to any rider, professional or otherwise, without completing the Everest Challenge.”
Edmondson acknowledged that Jeffers was able to exploit a loophole within Zwift’s ecosystem, and said that systems have been put into place to prevent something similar from happening in the future.
“Since Jeffers’ bot-led rides were never saved, and were out of competition, his behavior wasn’t detected by the typical means used by Zwift to flag suspicious performances,” he explained. “To identify the irregular behavior, we (Zwift) had to look into server history to uncover the hidden ride logs. After this was discovered, the same checks were run against all other finalists, and will be a part of the verification process moving forward. It’s important to note that the charges brought against Jeffers relate to British Cycling’s code of conduct. Jeffers has not been found guilty of cheating during a race on Zwift.
“Zwift is now implementing code in the game to detect physically-impossible performances produced by bots and other simulators. That code deletes any [experience points] earned for those specific activities and those specific activities alone meaning it is no longer possible to complete challenges or unlock equipment in this way. Moving forward, Zwift will be investing more heavily in machine learning to help combat cheating — a method that is commonly used, and being developed by, other competitive esports platforms.”
I also inquired if Jeffers would be allowed to continue to compete in Zwift races moving forward, having admitted to violating its terms of service? As it stands, he cheated Zwift, but has been punished by British Cycling, rather than Zwift. And he plans on continuing with unsanctioned virtual racing, telling Zwift Insider, “Of course I’ll be taking part in races with my WahooEsportTeam mates. I love Zwift racing, and I wouldn’t want something like this stop me from racing.”
Edmondson said that, at this point, Jeffers remains free to compete within Zwift.
“There is no Zwift-specific punishment in place for Cameron Jeffers for violating the terms of service,” Edmondson wrote. “Zwift eRacing rules state that punishment for violation of National Federation rules or standards is at the discretion of the Federation. Punishment for manipulation of data in a Zwift race would result in a one year ban from Zwift Racing.”
Whether the UCI, or national federations, should be involved with virtual racing at all is another question altogether. I’m a Zwift user and advocate — over the weekend I was telling my brother-in-law, a recreational cyclist, about what a game-changer it is — and I’m also a sometimes-bike racer. But I have no real interest in watching virtual racing as a non-participant. And I know I’m not alone.
After skimming British Cycling’s statement on Friday, I errantly posted on Twitter about this, assuming the “unsporting conduct related to manipulation of pre-race data gain an unfair advantage” referring to falsifying rider weight. I quickly corrected my mistake with a subsequent tweet, but left the original up because of the quality and diversity of the comments.
A lively debate about the merits and future of eRacing emerged between a few commenters, which I present here, edited for clarity.
Commenter 1: “Must be my age but… who cares? Has nothing to do with cycling as I regard it — surfaces, wind, rain, sun, scenery, motion, landscape, tan lines, etc. Is there skill and athletic ability? Yes. Do I care who wins? No. Would an eRacing winner’s jersey be a cherished object? No.”
Commenter 2: “I don’t disagree with you, but if Zwift gets people on bikes, keeps people on bikes, and broadens cycling in general then eRacing is a good thing. Some people, because of work, life, kids, or location, don’t have access to the outdoor crits and races. This notion of panning eRacing is petty.”
Me: “I’m not panning it. I just don’t care about it.”
Commenter 3: “I’m going to respectively disagree [with Commenter 1]. I think it is a generational thing. I think I’m on the cusp of that generation change, too. But look at their numbers, they are athletes, they are riding their hearts out, and for them, they are cherished wins and jerseys.”
Commenter 1: “I don’t deny the numbers, but it’s pretty much going to the gym and a lot of interactive hyper-marketed graphics.”
Commenter 3: “I agree many of the riding skills from the road and track are absent. I don’t believe it is meant to replace any form of cycling. I just think it deserves a bit more time and less negativity.”
That’s a snapshot of where eRacing is at the moment. Some cycling fans criticize it, some support it, some simply don’t care about it — and some participants are already gaming the system to get an advantage.
There is real money to be made in eSports, in multiplayer games like Dota 2 and Fortnite, though I’m not sure that qualifies to virtual bike racing; Jeffers and Rosamund Bradbury each received $500 in prize money for winning the British championship.
I know that Zwift received $120 million in funding late last year, with much of that earmarked towards eSports. Great Britain, Australia, and now the UCI have partnered with Zwift to offer official national championships. Personally, I have a hard time imagining it growing as a spectator sport, but I’ve been known to be wrong before.
Then again controversy is interwoven into the well-worn tapestry of professional cycling. Maybe this is all just part a rite of passage for eRacing. Next up, the biological passport?
Let the virtual games begin.