A Synopsis of Course Layout and the effect on results
Course layout can have a dramatic effect on the relative fairness of our handicap system. Fairness in this instance means the correlation of handicap to the final corrected time.
Ideally there should be no correlation of corrected finish time to the handicap number ("fast" boats with low handicaps should be no more likely to win than "slow" boats with high handicap numbers if equally sailed and equipped). The purpose of the handicap process is to attempt to make that happen.
In the real world many things can prevent this from happening. One fact is that large boats generally tend to sail proportionally faster upwind than smaller boats. Downwind however the opposite is usually the case and smaller boats tend to sail proportionally faster. Thus, while the large boats tend to "make time" to windward in the handicap system they tend to lose it back to the smaller boats on the downwind legs. Reaches tend to be neutral. When boats race on a closed course (start and finish in the same place) they tend to get a relatively balanced compromise (approximately equal amounts of downwind and upwind sailing) whether it be with triangular or windward-leeward courses.
Handicap fairness problems will arise when race organizers set courses that contain higher percentages of either upwind or downwind legs. This will change the relative fairness between the faster and slower boats. In my opinion, windward-leeward courses (equal) provide the best course type from a handicap standpoint. First the upwind and downwind legs are equal tending to cancel any advantage attributable to boat size and more importantly it allows a smaller (slower) boat, who is at a disadvantage in the windward leg, to select an optimum downwind strategy and course (triangular races tend to be a parade on the off-wind legs with little opportunity for course strategy since boats tend to sail the rhomb line).
Given enough time, our system of adjusting handicaps from the race performance, should tend to compensate for the course as well as boat differences but practically speaking that might never happen. One of the reasons I asked for up to date race results this year is that I'm interested in the relative fairness of the system as a whole now that we have had quite a few years to adjust out bias. Years ago I ran some studies that checked the relative handicap fairness within fleets and found that there was indeed a bias towards larger boats. In my experience that still does seem to be the case at Genesee Yacht Club and I'm curious if that is still the case on a lake wide basis.
The system PHRF-LO uses to convert time-on-distance numbers to time-on-time multipliers contains a scaling number (known as the "Q" factor) that allows adjustment of the relative handicap difference between small and large boats. We have used a Q factor of 0.045 since the inception of PHRF-LO to generate the time-on-time multiplier. My previous analysis showed that 0.22 was the optimum value (on average) and for a time we used that value to calculate the time on time multiplier for our races at Genesee Yacht Club. For small handicap differences within a fleet (6-15 sec/mile) the difference is really not significant. However in smaller clubs where boats in a division might have handicap differences of 50 sec/mile or more adjustment of the Q factor can provide increased fairness (we eventually reverted back to the PHRF-LO multiplier however to avoid confusion in interclub races).
In conclusion: I support (actually encourage) the use of a windward-leeward course that provides equal amounts of windward and leeward sailing. Triangular courses are also acceptable from a handicap perspective provided that they provide an approximately equal amount of upwind and off-wind sailing. Port to port (long distance) races should never be used for handicapping purposes because they contain an unknown and likely uneven mix of upwind and downwind sailing.