Is it Time for Time-on-Time?
For at least the past 100 years, most sailboat races in the US have been scored using the so-called time-on-distance method. This scoring system, at least in the form we are now familiar with, can be traced to Nathaniel Herreshoff. When he was a student at the Massachusetts Institute of Technology in 1868, he developed a time-on-distance allowance table for use by members of the Boston Yacht Club.
Today the most popular Time-on-Distance allowance system is PHRF. It works quite well for all boats over a narrow band of wind strength, around 10kn, and for narrow bands of rating over a wider range of wind strengths for boats of like design, but like all single number systems it does not work well with a wide range of boats and a wide range of wind speeds.
In light winds all boats travel slower and the race takes longer. The separation in finishing times becomes greater.
As an example, let's assume that we have three boats rated 100, 150 and 200. In 10kn, their average speed around the course would be 5.5, 5.1 and 4.76 kn. With a course of 10 miles the low handicap boat takes 6545 seconds, the mid handicapper takes 7058 seconds and the high handicapper takes 7563 seconds. After allowances the corrected times are 6545, 6558, and 6563 an extremely close finish.
Now lets assume that they race the same course but that the wind speed has dropped to 6kn. The average speeds drop to 5, 4.65 and 4.32 respectively. The elapsed times increase to 7200, 7742 and 8333 respectively. The corrected times becomes 7200, 7245 and 7333. The race is no longer close; the slower boat does not receive enough handicap allowance. The result in such races, as we have all seen, is that fast boats win slow races.
Similarly, when the wind pipes up to say 15kn the same boats complete the course faster and all the finish times become compressed. The slower boat now receive too much allowance with the result that we have all seen, slow boats win fast races.
In order to try to avoid these problems a different form of correction based on the duration of the race has been developed. Under this system, the allowance is based on the time it takes for the races to be sailed instead of the distance. In practical terms, this means that the slower the race is, the larger the corrections will be, no matter what the course length. The opposite will be true for fast races. It also means that the exact placement of the marks and calculation of the length of the course no longer matters, it doesn't figure into the calculation of the results.
Time-on-Time scoring is nothing new. It has been used in the United Kingdom and Nova Scotia for years and is rapidly spreading in the US. I have been using for about the last 15 years.
Time-on-Time handicapping uses a time correction factor, which is a function of the PHRF handicap. The elapsed time is simply multiplied by the correction factor to give the corrected time. That's all there is to it.
The factor most generally being used in the US is
Time correction factor = 650/550+PHRF
Now if we go back and look at our three sample boats and races the factors would be:-
In the first race the results would have been:
PHRF      Elapsed       Corrected
100         6545          6545
150         7058          6553.8
200         7563          6554.6
Still a close race and the same finish order.
In the second race:
PHRF       Elapsed       Corrected
100         7200          7200
150         7765          7210
200         8333          7222
Now we have a race that stays close even in the lower wind. All boats feel that they have a fair chance at winning.
If we think of the perfect handicap system it would have all boats, equally well sailed, correcting to identical times. We can therefore consider the spread of corrected times to be a function of the efficiency of the handicapping system as well of how well the different boats are sailed.
I have gone back over several year's results for a couple of clubs and recalculated the results using Time-on-Time and then calculated the standard deviation of the corrected times under this system and compared it to the standard deviation under time on distance scoring. In almost every case the variation in the corrected times was reduced.
In all but one race, the standard deviation of the results was reduced using TOT. Initially I didn't know what was different about that race but I subsequently found out that there was something unusual about it. The race was started in very little wind and was initially very slow indeed. Boats were virtually standing still. Then a strong sea breeze kicked in and the course was sailed quickly in the new wind.
In no race was the winner changed although there would have been changes in the finish order of three of the races. Clearly, the results became closer showing that the handicapping was less of a factor in the results.
As a result of my analysis, I have become a strong proponent of using time on time handicapping.