2022 Round 3 - Lancaster. Sunday 22nd May

Kim wrote: And I've provisionally updated the points tables using 50% weighting of the two rounds.  Unless it should be something else...?

I'm told by Heather, who is the closest we have to an organiser for Lancaster, that 50/50 is correct.

Many thanks to Kim and Natalya (Timelord in training?) for what look like some pretty pristine results, especially for picking through the digital debris of the crash in race 3.

The crash wasn't really a problem - it was clear enough from the missed reads.  (I think the system was working so well that *all* the missed reads were actual stoppages.)

The real issue was picking up riders on the far side of the track.  Filtering by minimum lap time on its own wasn't sufficient, as there was enough variation in riders' speeds that at a couple of points either read was plausible, and if it picked the wrong one CrossMgr's filtering would then follow the far-side read for subsequent laps.

Natalya suspected this was happening at the time, so the first thing we did was inspect the video for all of Slash and Lee's supposed lap times (which is quick and easy in CrossMgrVideo, though we did have to go frame-by-frame at one point to spot the Beano neatly hiding behind my tailbox/helmet) and delete a couple of recorded times where Slash was on the wrong side of the track.  At which point the algorithm did the Right Thing and all the subsequent lap times were back in phase.

With correct times for the leaders, we could tweak the minimum lap time to the second and everything else fell into place.  Lovely.