[2022] Got the fæstest time possible on Dragon Bash beetle race, check it out!

For the record, when converting from meters to GW2 units ("Tyrian inches" as they are sometimes called) you multiply by 32, not 40. Assuming the measurement in the video is accurate, 2654 meters would turn into 84928 units, which if the max forward speed of the beetle could be maintained would take a minimum of 84928/1800 = 47.2 seconds to complete. It's also possible the tool itself incorrectly assumes a ratio of 40:1 rather than 32:1 when displaying meters, in which case your distance calculation of units will end up correct but you'll be assuming a max speed of the beetle that's lower than it actually is (45 m/s rather than 56.25 m/s). Finally, while the max forward speed of a beetle is listed as 1800 and it won't accelerate past that normally, it can actually go faster to help smooth over some edge cases while drifting etc., and skilled drivers have figured out how to capitalize on that (the real cap is 2500, though I don't know if it's possible to actually hit it).

That being said, this track isn't exactly what you'd call a straight line, so don't let me downplay your time. Certainly it's better than what I could do nowadays, and I used to be the best beetle racer in GW2. Granted that was before it shipped live and I was the only one in the world using it, but that's details.

If this was a few weeks after it shipped it would be one thing, but it's been out almost four years now and I doubt anyone would want to fiddle with it at this point. At the end of the day GW2 is a game, not a simulator, so some dubiousness with the physics is fine and even expected, especially if it makes the common user experience more enjoyable.

For the drift speed thing specifically, my guess there is that it's accidental fallout from the artificial acceleration I gave it during drifting to counteract its natural speed loss from friction and other things I don't remember. I felt that it was very important that I could continuously drift in a circle, because that's cool, and I more or less just adjusted numbers by feel until that worked reliably. I did make sure that you couldn't easily accelerate on flat ground by drifting in a S curve, but I don't recall doing much testing with you also being in the air from jumping. Drifts in air were always deep in turbo jankiness territory but I didn't think at the time that they would come up often. Anyway tl;dr rollerbeetle motion is a pile of partially-fake physics with fixups stacked on top to get the results that I wanted.