
Post by
BooBill | 2020-08-17 | 00:14:19
Didn't want to leave it where it may appear that I was suggesting there might be a bug in the optimizer. After much analysis, manual calculation and running an alternate router, my conclusion is that the optimizer was running correctly.
Using the same GFS data, that was the right route (slightly). Still leaves the possibility that the data was bad. Fastseas router, which is much more simplistic with 1 hour time steps, and also using GFS came up with something much more normal looking and close to what the current leaders were sailing.
There was significant differences in ECMWF, GFS and ICON forecasts, so sailing negative VMG based on GFS alone, not going to happen on my boat.
We'll see at the finish line. Some big names took a 40 mile hit.