Zwift has shown consistency in policing digital doping and inconsistency.

The immensely popular online training and racing app has realized that as prize money and status increases in its virtual events, the potential for cheating becomes a reality.

In the realm of virtual riding, a rider's power output is used as a baseline to identify possible discrepancies and data violations during digital races Zwift is tough on enforcement. Last year it banned two riders for inaccurate data display.

The latest riders to violate Zwift's strict data protocols are Antonina Reznikov and Selma Trommer. Both were sanctioned through mid-July, and their six-month Zwift competition suspensions went into effect in early January.

Sophisticated software and highly active algorithms allow Zwift to find possible problems in the vast flow of user data. In particular, connectivity and file convergence issues when exporting or uploading race files are some of the most significant problems that Zwift must address.

Reznikov specializes as an Ironman triathlete, not a road specialist, and was suspected of varying the power data used as a reference Zwift's analytics showed that Reznikov's race data was 32% more powerful than her submitted values triggered when she was powerful.

Reznikov initially denied any wrongdoing, but eventually admitted to changing the data.

Selma Trommer is a German road rider whose 9% higher power value submitted to Zwift caught the attention of verification analysts. Trommer solved a problem with uploading files to Garmin Connect using Golden Cheetah.

The use of third-party file conversion software was not considered an adequate excuse, as all of Trommer's output increased at a constant rate. She also admitted that her files had been edited under Zwift's scrutiny.

Reznikov and Trommer were allowed to use Zwift for training and social functions during the ban.