Drift Tengoku — October 2013

For Round 4, Kobashi planned to set the angle quickly in the final corner: build angle sharply on the Sector 1 straight before entering Sector 2. For the fourth corner (between Sectors 4 and 5), he thought it would score better to snap the car around before the corner and take a straighter line through it, rather than gradually increasing angle and wrapping around. But initiating a sharp angle on the straight could make the car slide toward the front rather than toward the judges, so he wanted to tuck it in slightly…

…small mistakes can easily cost about 0.3 points. So they changed the setup to increase cornering speed and raise average speed in Sectors 3 and 4. In the first run, Kobashi was a little cautious on speed and his score was mediocre. But on his second run he increased exit speed from the final corner and scored highly in Sector 1. Sector 5 was also decent. ‘The way I set the angle on the straight and drove Sector 5 worked well. But I was a little unsettled, so in Sectors 3 and 4…’

…he recalled that he hadn't scored well there. Suenaga, meanwhile, was fast and smooth from his first run, and earned judges' bonus points for a high score of 8.27. Looking at his sector scores, there was no single standout sector, but neither were there any large deductions—the result matched his plan to score consistently throughout. He seemed pleased: ‘I got the run I was aiming for, and I think I adjusted it to suit DOSS.’ Kobashi finished tenth in the solo event and Suenaga sixth. Concerns about DOSS. Returning to Saturday's qualifying, Kumakubo, who was eliminated, felt a gap between his own impression and DOSS's score. He asked the D1 office to publish, if possible, the percentage weight given to angle and speed in each sector. Many competitors probably felt the same. Even before introducing DOSS, the D1 office had said it would not publish the algorithm. The D1 Corporation hoped to profit by selling the system overseas, and the algorithm is its most important part, so they did not want it copied easily. But I have serious concerns: if the algorithm is secret, outsiders cannot detect cheating. Trust may be enough in Japan, but what about countries where cheating is common? Will people trust the system, or could it become a source of abuse? DOSS is designed to approximate the old human judges' 100-point scoring method. Is that really the right approach? Wouldn't it be better to use a simpler scoring method and publish all the details? So far, DOSS judging seems to be working well in Japan's D1 series. But if it is meant to become a global standard, I have doubts that the current system will be accepted worldwide…

…the office has stated that it will not disclose the system's formula. D1 Corporation hopes to earn money by selling the system overseas, but the algorithm is DOSS's most important component and they do not want it copied easily. Still, I have major concerns. If the algorithm is not public, a third party cannot tell whether the system has been manipulated. In Japan, trust may be enough, but what about countries where cheating is widespread? Will people trust this system, or could it become a breeding ground for abuse? DOSS is meant to approximate the traditional human judges' 100-point scoring. Is that really desirable? Perhaps scoring should use a simpler formula and disclose all of its components. At present, DOSS scoring seems to be going well in Japan's D1 series. But if it is intended to become a world standard, I have a feeling the current approach may not be acceptable internationally…