Gaussian Handicaps

as it isn’t possible to know exactly what yardstick should be applied in a race (both because they vary year on year and no race is exactly average), then the right thing to do is to set PY ranges for a class e.g. Laser is somewhere between 1085 and 1105. Assume a Gaussian distribution and then calculate the scores in a probabilistic way. Ie if boat A beats boat B by 3 seconds it currently scored as 1 and 2, but in reality we may only be 60% confident boat A actually won (as its PY May be too generous), so it may be fairer to score as 1.4 and 1.6.



Sorry, not explained very well because I’m on a phone, but hopefully you get the gist.

The theory is sound. I'm of the opinion that at Utopiaville SC many more results would be recorded as ties. If boats are 1 second apart on corrected time then that's actually a greater resolution that the average club timing is capable of measuring. How ever the complications of several boats together render it impractical: if you say that corrected times 1087 and 1088 secs should be a tie, what do you do if 4 boats correct out at 1087, 1088, 1089 and 1090? One feature of any good scoring system is that you should be able to explain it to *all* the competitors.

The other factor, of course, is, when you calculate a set of real series results based on your concept, does it actually make any difference? I once diligently worked out a scoring system that was supposed to combine the best features of high point and low point scoring, wrote a spreadsheet to calculate results using it, and even started drafting a little article about what a wonderful system it was. Then I plumbed a couple of sets of old club results into it, and was mortified to find that not only did it make no difference beyond swapping a few places around mid fleet, but also I wouldn't be able to put my hand on my heart and say I was confident the different results were definitely better...

Edited by JimC - 09 Dec 18 at 9:21pm

Sometimes, I have to blow the hooter when the boat passes through the transit and note the time in my head, then write the time I have remembered on the sheet, hopefully next to the correct boat, sometimes several boats can finish almost together, that gets interesting, sometimes to further complicate matters I could have a new member with me, showing them the ropes.

Edited by 423zero - 09 Dec 18 at 10:03pm

perhaps race timers should only count to to the nearest ten seconds or perhaps not even show seconds? 

How about a points system, a bit like in the decathalon/heptathlon?