Select to expand quote
Macroscien said..
Things become even quite complex when you try to fairy decide order on winning list.
Handing rewards and prizes at Burrum sometimes you need to say
Here is 75% probable winner in 500 meters competition.
But knowing exactly limitation of our gps devices and other elements effecting recording we coukd theoreticly design system to be 100% fair, not just statistically probable.
It mean that grey area or errors qualify competitors to be even , no winner or looser if results gets too close.
Such a system is already designed and used
WGPSSRC records are 'Claimed' speed. That is: The average of two side by side GPS's, and then subtract the reported error for that run. You then have a figure which you can claim with 99%+ probability that you did AT LEAST the claimed speed.
A few issues for GPSTC:
1. Not everyone wants to always run 2 identical GPS's side by side.
2. People like to have a higher number, even if there is a 50% probability that the number you did is a bit lower. (Remember, that there is the same 50% probability that you went slightly faster as well - within the reported error range.)
For the GPSTC competition we think that the accuracy we have now is good enough, and it is certainly a
lot better now than when we started many years ago.
Team rankings on a day are not very often within the error margins of the devices used, and if they are we can have a great time sledging each other about who was really faster

or agree to call it a draw.

Also, if we adopted the 'claimed speed" approach, there would be race among the competitive sailors to get the best, most accurate GPS device, with the lowest error figures that they could find... Hmmm...




