Handicap - A Final Solution?

Wrote this in the other thread oops!

Physical testing of dinghies "in the field" would be impossible to do in a scientific and meaningful manner whereby you could quantify every aspect of the aero and hydro characteristics accurately enough.
Big boat IMS (and I believe subsequently IRC) handicapping sought to handicap the boats based on their VPPs - i.e velocity prediction programmes - maths models of the aerodynamic and hydrodynamic components of the yacht which can resolve the equilibriums for any speed / heading / sail set to effectively produce polar charts from which theoretical times around theoretical (perfect triangle or ww/lw or point to point) courses. Most polars that you get on leadmines to steer and trim to are derived from the VPPs that have been created rather than physical testing (far, far more expensive given that you'd have to do extensive wind tunnel and towtank / CFD analyses to populate all the physical models).

There is no reason why you couldn't run every dinghy class through a vpp, apart from effort and expense. Trouble is then, what windspeed would you decide to take the relative handicaps at? Thats a can of worms to start with. Also, the VPP number will have to be given for whatever course config (i.e. ww/lw; specific point to point etc) that has been fed into the computer. I don't need to overstate the disparity in performance between most classes depending on time spent on which leg of the course or wind strength.

Then the aspect that could never be simulated in a computational environment, except by means of a massive subjective fudge, is that an empirical scheme like the PY is the only way to capture the differences between how a boat should perform on paper compared to how it performs in practice - how easy it is for the performance to be realised by the crew etc.
An RS 100 may be a rocket at VMG running in moderate breezes, easily captured in a maths model for steady state, but how do you quanitify and account for the ballache of raising and hoisting the kite singlehanded on a 70 yard run on a pond?

Believe me, as one who has thought about this more than most, PY especially of late with all the effort going into it, is by far the least imperfect way of divining some comparisons between disparate craft in disparate circumstances competing in the sport with the largest set of variables.

You can put any technology you like into generating the numbers, but the PY system fundamentally is based on the idea that the time taken by boat A should be linearly proportional to the time take by boat B.
That used to work reasonable well when comparing say Solos with Enterprises, or Graduates and GP14's.
It works OK for RS400's vs Laser 4000 up to F4.

It does not work very well when you have disparate boats like Wayfarers and RS400's. A well sailed Wayf will nearly stick with us upwind and down in very light winds, but in a breeze we completely blow them away. The PY tries to average that.
Each boat has a different speed vs windspeed curve.


If people stuck to sensible PY groupings, like asy, una rig, and similar size two man boats, there would be a lot less need for adjustments. Thank god the cats have mostly gone off and done their own thing.
When boats are within 10% or so of each other's PY, they see roughly the same tide. As you increase the spread, it can quickly get ridiculous. I did a race the other year, Where I did 3 laps before better sailors sailing slower boats did 1 lap. I was going 10% faster than the tide, they were sailing on the spot quite a bit. Luckily we were out there for fun.


Of course it is great fun to mix in Mirrors, fast boats and keelboats a few times a year, but you have to be relaxed about the results.

Technology is no substitute for common sense and having a clear idea about what you are trying to calculate.

Originally posted by Daniel Holman Wrote this in the other thread oops! Physical testing of dinghies "in the field" would be impossible to do in a scientific and meaningful manner whereby you could quantify every aspect of the aero and hydro characteristics accurately enough. Big boat IMS (and I believe subsequently IRC) handicapping sought to handicap the boats based on their VPPs - i.e velocity prediction programmes - maths models of the aerodynamic and hydrodynamic components of the yacht which can resolve the equilibriums for any speed / heading / sail set to effectively produce polar charts from which theoretical times around theoretical (perfect triangle or ww/lw or point to point) courses. Most polars that you get on leadmines to steer and trim to are derived from the VPPs that have been created rather than physical testing (far, far more expensive given that you'd have to do extensive wind tunnel and towtank / CFD analyses to populate all the physical models). There is no reason why you couldn't run every dinghy class through a vpp, apart from effort and expense. Trouble is then, what windspeed would you decide to take the relative handicaps at? Thats a can of worms to start with. Also, the VPP number will have to be given for whatever course config (i.e. ww/lw; specific point to point etc) that has been fed into the computer. I don't need to overstate the disparity in performance between most classes depending on time spent on which leg of the course or wind strength. Then the aspect that could never be simulated in a computational environment, except by means of a massive subjective fudge, is that an empirical scheme like the PY is the only way to capture the differences between how a boat should perform on paper compared to how it performs in practice - how easy it is for the performance to be realised by the crew etc. An RS 100 may be a rocket at VMG running in moderate breezes, easily captured in a maths model for steady state, but how do you quanitify and account for the ballache of raising and hoisting the kite singlehanded on a 70 yard run on a pond? Believe me, as one who has thought about this more than most, PY especially of late with all the effort going into it, is by far the least imperfect way of divining some comparisons between disparate craft in disparate circumstances competing in the sport with the largest set of variables.

I wouldnt be so quick to suggest that class racing will never come back, just think of all the teenagers currently sailing Oppies, Toppers, Fevas, Teras, etc. They are all enjoying class racing and as they get older I expect they will gravitate towards class sailing. Some will be lost from the sport but if they come back in later years I expect they will remember it was the class racing that they enjoyed and will pick a local class, that is certainly what I have observed locally.

Originally posted by Daniel Holman Trouble is then, what windspeed would you decide to take the relative handicaps at?

Force 2-3 about where it is at the moment.

Originally posted by Daniel Holman Also, the VPP number will have to be given for whatever course config (i.e. ww/lw; specific point to point etc) that has been fed into the computer. I don't need to overstate the disparity in performance between most classes depending on time spent on which leg of the course or wind strength.

3 legs should do it. 1000secs upwind, 1000secs dead down and 1000 at whatever the polar suggests is optimum for the given craft.

Originally posted by Daniel Holman Then the aspect that could never be simulated in a computational environment, except by means of a massive subjective fudge, is that an empirical scheme like the PY is the only way to capture the differences between how a boat should perform on paper compared to how it performs in practice - how easy it is for the performance to be realised by the crew etc. An RS 100 may be a rocket at VMG running in moderate breezes, easily captured in a maths model for steady state, but how do you quanitify and account for the ballache of raising and hoisting the kite singlehanded on a 70 yard run on a pond?Believe me, as one who has thought about this more than most, PY especially of late with all the effort going into it, is by far the least imperfect way of divining some comparisons between disparate craft in disparate circumstances competing in the sport with the largest set of variables.

Whatever.. there should have been a base line set before this latest attempt to 'put things right' I'm not saying give up the returns system, I'm saying that factual base lines need to be created about which alterations by classes that do alter can orbit.

Certain known speed and performance types need setting in stone, then the returns system could function a lot better than it does at present.