Formula E Zone can reveal that fanboost, one of Formula E’s prime interaction tools, has been manipulated by fake email addresses and automated “bots”, negating the fans actual choices for the London ePrix.
It has been speculated throughout season two that the fanboost system was not giving an accurate representation of who the fans were actually voting for, with some drivers frequently venting their frustration on social media.
Their dissatisfaction can now be justified, as evidence will show some drivers have received thousands of votes within a short time frame, taking them from lower down the table and comfortably into the top three.
Approaching the fanboost deadline for both Saturday and Sunday’s London ePrix’s, some drivers were gaining between 5-10,000 votes in the final few minutes before the power boost was awarded. Therefore, questions have to be asked if any fanboost results have been chosen legitimately by the fans.
It is important to note that the evidence shown does not suggest that the drivers, teams or Formula E have done any wrongdoing to modify the outcome of the fanboost vote, but instead indicates cheating from somewhere is involved in manipulating the fanboost standings.
Formula E Zone was able to find the information that showed the number of votes that each driver had received by looking through the fanboost site’s source code. Once the data was found, the votes were monitored close to every hour of the day excluding between midnight and 6 am.
Two Ways of Cheating the System
A source Formula E Zone was working with during our investigation, had expressed fears that the ‘one vote a day’ system did not exist on Formula E’s website or app. It was believed that you could write anything in the email section as long as it including the ‘@’ symbol and ended in any country’s domain.
We conducted a short experiment by creating the most obscure email we could think of and tried voting for a few drivers that had a low amount of votes.
The votes we casted were indeed counted, and there seemed to be no limit to how many fake emails you could have used. Instead, it was merely our level of patience doing it.
However, putting fake emails in and voting manually is very much the long way round when it comes to manipulating the fanboost results.
Therefore, we investigated the use of online websites which enable anyone to purchase a set amount votes such as 1,000, 5,000, or even 10,000 at any one time to be placed on a specific individual.
We contacted a number of those websites, but the majority did not provide us with the information as it was a breach of their customer privacy rules.
Two of the websites we asked did confirm that they had customers who have purchased votes for the fanboost site. Due to their privacy agreements, they were not able to say much about the purchases, but they could reveal how many times they had worked on the fanboost site.
“We have worked on this platform many times before. We have had eight customers purchase votes for that site,” said the first bot website we contacted.
“Due to our privacy agreement with our customers we can’t supply all the information you have asked for, but we can confirm we have worked on that website before,” said the other website that replied to us.
Over the 13 days of fanboost voting ahead of Saturday’s London race and then the extra day for Sunday’s ePrix, we have seen the standings for fanboost altered significantly by bots almost each and every day. We won’t go into detail about every bot situation, but instead, we will tell you about the most significant data, with the all the data available to see for yourself by clicking the download button at the end of this article.
It did not take too long for us to notice the first signs of bot activity. On the second day of fanboost voting, the graph and data shows Simona De Silvestro gain 5,608 votes between 10-11pm. The Swiss driver only had 129 votes at 9 pm, but ended the day with 5,738.
The data also showed the rise of Ma Qing Hua and Daniel Abt. Ma, who was in single figures for the majority of the day, gained 2,000 votes in 45 minutes and continued to gain a further 1,000 votes by 6 am the following morning.
Abt gained 3,474 votes within 40 minutes from a starting position of 84 votes. The data for the German driver goes over two days as he ended the second day with 1066 votes before a 30-minute climb until 12:30 am on day three saw him with 3558 votes.
Day five saw Antonio Felix Da Costa jump to the top of the fanboost standings as he received 8,000 votes between 8 am to 10 am after gaining 1,000 votes between 6 am and 10:45 am two days prior. Sebastien Buemi and Stephane Sarrazin also benefitted from 3,000 votes each.
Jumping to the final day before the first London ePrix, the data shows a wide range of movement between drivers at the top. Nick Heidfeld gain over 32,000 votes to take him to the top of the leaderboard, contrast that to the 408 votes the German driver had at the end of day nine.
In the closing minutes of fanboost for the first race, Heidfeld gained 10,000 votes, which took his total over the final four days to over 60,000 votes to be the ‘fans’ number one choice.
The same can be seen for Lucas Di Grassi, Buemi and Sarrazin on the final day. Di Grassi gained 40,000 on the final day, with Buemi and Sarrazin gaining a further 25,000 votes.
Each driver’s final vote counts were heavily manipulated by bots in the final four days before the first fanboost deadline which you can see by downloading the data.
The votes for the final race of the season were also heavily manipulated by bots. De Silvestro again received the biggest push early on with 1,000 votes in first few hours of the fanboost reopening. Most of the other drivers were still in early teens or still in single figures at that time.
However, it was not long until the drivers affected by bots for Saturday’s event were affected again. Buemi, who had gained 3,000 votes, overnight gained a further 5,000 votes between 6 am and 8 am with Sarrazin gaining 3,000 votes overnight as well.
Buemi and Sarrazin were on 8,087 and 3,357 by 8:10 am, but by 9:42 am Buemi had 14,019 and Sarrazin moved up to 10,159.
Heidfeld gained 2,500 votes between 10:35 am and 11:47 am from being on 161 at 10:35 am. At 1:19 pm Buemi was moved up to 14,356, Sarrazin had 13,667; Heidfeld had 6,983 votes. Di Grassi had 4,615 votes, but his votes had been consistent throughout, so it was deemed unlikely he had been affected by bot votes.
In the final 20 minutes until the fanboost deadline for Sunday’s race activity picked up once more. Buemi, who had gained 6,000 votes between 1:19 pm and 3:52 pm went on to gain a further 16,000 votes, with Sarrazin also gaining 16,000 votes in the final 20 minutes.
Heidfeld, who had gained a further 10,000 votes between 1:19 pm and 3:41 pm, went on to gain a further 5,000 votes in the final 20 minutes before the deadline.
Di Grassi gained 25,000 votes in the final 20 minutes of fanboost to move him ahead of Heidfeld in the standings. The Brazilian had also gained 5,000 votes between 1:19 pm and 3:41 pm.
In the closing stages, the drivers that were not affected by bots only gained a small number of votes with some gaining over ten votes in the final 20 minutes
Who should have won Fanboost?
By disqualifying those who received artificial votes, the drivers that should have received fanboost for Saturday’s race are: Sam Bird, who had 6,229 votes, Jean-Eric Vergne, who had 1,298 votes and Bruno Senna, who had 671 votes.
While Sunday’s race should have been Bird with 2,474 votes Oliver Turvey with 202 votes and Jerome D’Ambrosio with 131 votes.
An interesting side note, regardless of bot votes, Di Grassi would have been in the top three on both days taking away any artificial interference, meaning he would have got fanboost fair and square.
Fanboost is a great way for fans to interact with the sport and for the sport to interact with the fans, but it is clear that the voting system needs to be addressed going into next season.
Di Grassi said during the London ePrix that fans should only be able to vote through for social media for fanboost, which would take away the uncertainty of whether some drivers are being affected by bot votes.
In addition, voting via social media is currently the only reliable method where cheating cannot take place. We tried to vote twice on the same day on Twitter and Instagram, and the votes were NOT counted. On the other hand, the website and app allow anyone the chance to vote multiple times as long as the email address or computers IPs are not being monitored.
Fanboost is important to the identity of Formula E, but drastic changes in the rules of how fans vote need to be addressed as it is not clear if any of the fanboost results correlate to who the fans are actually voting for.