Algorithm for Club

[DarinK]

To start, I don't think this algorithm would be good for a general purpose recalculation of handicap.  I'll post below what would be more interesting in my opinion.

As for club, we use the algorithm created by Kevin Box that basically tries to move everyone closer to a Median finish time.  Penalize the leaders and gifts to the back of the pack.  For example, here is the current results for this week:
Tuning Factor   
28
Course Length   8.6
Median Time Behind   11.5 min
Median Elapsed   2.01 hr
Boat   Elapsed   Time Behind   Rating   New Rating   Delta
Boat 1   1.92 hr   0.0 min   191   187   -4
Boat 2   1.88 hr   1.6 min   165   161   -4
Boat 3   1.87 hr   3.0 min   150   147   -3
Boat 4   -1.10 hr   3.8 min   49   46   -3
Boat 5   1.88 hr   3.9 min   147   144   -3
Boat 6   1.82 hr   4.5 min   118   115   -3
Boat 7   1.98 hr   6.5 min   173   171   -2
Boat 8   1.86 hr   6.5 min   121   119   -2
Boat 9   2.06 hr   7.8 min   198   197   -1
Boat 10   1.93 hr   8.9 min   133   132   -1
Boat 11   2.30 hr   10.2 min   280   280   0
Boat 12   2.09 hr   10.4 min   191   191   0
Boat 13   2.24 hr   11.2 min   248   248   0
Boat 14   2.13 hr   11.2 min   204   204   0
Boat 15   2.11 hr   12.2 min   187   187   0
Boat 16   2.25 hr   12.8 min   239   239   0
Boat 17   2.27 hr   13.0 min   248   249   1
Boat 18   2.18 hr   14.4 min   201   202   1
Boat 19   2.12 hr   14.6 min   174   175   1
Boat 20   2.25 hr   15.8 min   221   223   2
Boat 21   2.24 hr   16.0 min   216   218   2
Boat 22   2.11 hr   16.3 min   160   162   2
Boat 23   2.25 hr   16.8 min   214   216   2
Boat 24   2.40 hr   16.9 min   274   276   2
Boat 25   2.52 hr   29.2 min   240   247   7
Boat 26   2.68 hr   32.5 min   282   290   8

Generally, this doesn't make sense for people who are racing near their handicap.  They will be punished for being consistent, and gifted for being slow.  New folks will move the median back or forth, and should someone not race during a big movement, they will be inaccurately gifted/penalized the next race.

It is fun though.  You get more variety of winners.  It is just unfortunate that the consistent sailors will always be mid pack.

[STuma]

Great job. I think if the method is described where people can refer to it, they will be accepting. I had asked the question about the method and was never given an answer. It doesn't matter to me, but when I'm trying to describe the method and sell the idea to a new sailor, they seem to feel it is a magical unknown factor. The idea is to help group everyone to a more competitive level. Unfortunately, it does not do much for raising people up to a higher level, it brings the mean down (this is where I'm mixed). The thing I wasn't able to do, was recalculate these results with the given factor - to see where the boats end up with the new rating and same elapsed time. There is no correct answer, it's just how much do we want to pull the leaders back and bring the bottom up?

[BrianL]

The spread on this last adjustment is only about a minute and a half. The more people that race more often, improves the results. Also as your experience improves your rating should drop, bringing the crowd closer together. Having a fixed course that reverses, evens out the differences in the legs.  If you actually peak out on boat handling then you should be in the mix and it becomes more tactical skill to win. In the last race, I had new crew and messed around with the spinnaker which let some folks catch up. Then on the last leg I did not sail the boat to its potential upwind, which costs us the most time.