This is an earlier experiment. Read the amendments below before treating its findings as current.
The question
In the 70° wind the craft without priority ended trips far left of its route. Does the offset carry over, and does wind pick the side?
What was tested
A switch that re-centres the craft at each trip restart; force records in both winds; fresh layouts at 3×.
What was found then
The offset carries over: put back each trip, its typical end-of-trip distance fell from 633 to 132 m in one setting, and similarly in the other. That did not clearly help safety (others' losses rose without a link and fell with one), and trips were 1.5–1.7 times slower. 'The first full trip always goes left' failed on fresh layouts (3 of 8 went more than 100 m left). Early leftward drift appeared in both winds, cause unknown. At a five-minute snapshot all 16 runs sat on the wind's side, an association only.
What changed
A switch that re-centres the craft on its route at each trip restart was added, off; stage 4 closed without adopting anything.
Numbers and source references
- 8
- fresh runs (two seeds × two link settings × two winds)research/design/prio_first_lap.md
- 633 → 132, 730 → 90 m
- median lap-end distance from route, before → after re-centringresearch/design/prio_plain_drift_close.md
- 24 → 36, 96 → 70
- ordinary lost, no link / 1 s link, with re-centringresearch/design/prio_plain_drift70.md
- 1.5–1.7×
- slower laps with re-centringresearch/design/prio_plain_drift70.md
- 3 of 8
- fresh first full laps more than 100 m leftresearch/design/prio_first_lap.md
- 16 of 16
- runs on the wind's side at the 300 s snapshotresearch/design/prio_plain_drift_close.md