This yacht rating system uses a set of mathematical models that draw on a range of well-established analytical and empirical techniques to estimate the resistance of a hull, and the power available from the sails, based upon a set of hull and sail measurements. These are fed into a velocity prediction program (VPP) which balances all the forces and calculates the speed potential of the yacht - on various points of sail and when pointing hard upwind, for a variety of weather conditions.

The VPP generates a set of polar performance data (boat speed vs wind speed / angle) from which a weighted average - simulating the conditions encountered in a race, or series - is used to calculate a rating. The rating is given in the familiar form of a Time Correction Coefficient (TCC).

Using a VPP to calculate performance-based ratings

The rating calculator is partitioned into the following main components:

	aerodynamic and hydrodynamic models a stability model which estimates the stiffness of the yacht the measurement data set - and a number of routines which pre-process the data independent variables describing the yacht's state (speed, heel angle, leeway, sail trim, etc) the solution subroutine - the heart of the VPP - which is driven by ... a wind speed / angle 'task matrix' defining the polar plot (which includes best VMG) a seed generator which guides the starting point of the solution subroutine

Every boat on the system has an independent data file - from which the measurements are read when a rating is required. Estimating and interpolating routines then compensate for the restricted number of measurements and any missing data - and generate a validated and coherent data set.

The validated data is translated into a set of conventional parameters used in naval architecture and these are used to configure the behaviour of the aerodynamic, hydrodynamic and stability models. In turn, the stability model allows the heel angle to be calculated, which is one of the parameters fed back into the aerodynamic and hydrodynamic models.

The VPP subroutine runs once for each of the entries in the task matrix and works by iteratively adjusting the independent variables - until all of the governing equations show the aerodynamic and hydrodynamic forces acting on the yacht to be in balance. This corresponds to the point at which the yacht settles for the prevailing conditions - and the speed of the yacht is recorded in the task matrix to create a polar plot.

The seed generator uses simple performance measures to estimate the independent variables for each entry in the task matrix; these estimates give the solution subroutine a good place to start each speed-polar calculation. This has no bearing on the final solution, but considerably shortens the calculation time by reducing the number of iterations needed for each run of the VPP.

Some fundamental differences ...

VPRS is distinct from most other rating systems, because it uses a true VPP to calculate the ratings without needing a hull lines diagram, or a large number of measurements. To achieve this, commonly available yacht measurement data is translated into parameters recognised by the VPP. As a result, VPRS calculates ratings from measurements similar to those needed by other rating systems.

There are no direct allowances or adjustments to any calculated rating; these are not needed as the reconfigurable models, from which the yacht performance is determined, were designed to use the entire data set. In addition, formulae characteristic of traditional rating rules are also not needed, but they are used by the 'seed generator' to reduce the calculation time (see above).

That VPRS is a performance-based (rather than a concession driven) system is readily apparent when comparing 'no-spinnaker' ratings. When the aerodynamic models are reconfigured for sailing without a spinnaker, the ratings usually fall by 7% - 10% and represent the true loss of averaged performance expected on the water. For boats competing in strict 'no-spinnaker' classes, these are the fairest ratings to use. When racing against fully canvassed boats however, only a small proportion of the reduction is appropriate, and this is the only such 'class management' concession made in the system.

Yachts are not explicitly assigned a 'boat type' in order to classify the hull form - as occasionally seen. Whilst this appears to be an intuitive means of differentiation, it turns out to be a rather deceiving over-simplification which is very hard to apply consistently. Instead, the set of measurements is sufficient to allow the 'boat type' characteristics to be deduced - and hence accounted for in a more rigorous way. In addition, the complete independence of each set of measurements allows significant changes made during a class production run to be taken into account - and new designs and custom builds can also be rated without delay.

Traditional rating systems often place a lot of emphasis on measurement specification - and then use a simple rule to calculate the rating. Unfortunately, this seems to encourage attempts to defeat the measurement rules which become ever more complex. VPRS takes the alternative approach - a complex VPP is used to calculate the ratings, whilst the measurements are more simply defined. This makes the measurements much easier to understand - and the system is more accessible. The measurement guide will of course be revised when necessary, but the emphasis on keeping it simple will remain.

Optimising for VPRS

Optimising for a rating system is rather contrary to the spirit of fair racing - as it conveys advantage by exploiting system weaknesses. This is one of the reasons that VPRS isn't a simple rule-based calculator. It also accounts for the absence of boundary cases and complex conditional clauses in the measurement definitions; clarity and brevity promote wide accessibility and remove opportunities for exploitation.

The aim in racing should always be to work the boat harder than the rating system!

Feedback - and reviewing system performance

Whilst the model has in part been designed to ensure that the ratings are in line with observed performance, no subjective adjustments are made to any calculated ratings; these depend on measurement alone. The model is however continuously under review and will be updated between seasons in the light of racing experience. The resulting changes will be applied to all yachts - and new ratings generated in time for the next season.

The prototype VPRS software, successfully trialed by PYRA in 2010, was producing ratings that were on average within 1% of those issued by IRC. The 2011 system software was significantly more sophisticated, and as a result of recent feedback, there are a number of minor improvements included in the 2012 software.

When evaluating a rating system remember that close racing results are the goal of handicap systems (these include rating systems which apply corrections—overtly or otherwise). This is not the case for pure rating systems, like VPRS, which 'remove the boat from the race' in order to yield results reflecting crew performance and tactical decisions. The distinction is clearly illustrated by one-design class events—where the racing is frequently seen to be anything but close.

You are welcome to give feedback on any aspect of the system - our contact details are here.

The aim: Fair racing

VPRS aims to meet the aspirations of club racers. It was developed in close co-operation with the racing community in Poole Harbour, with the model considerably refined in response to feedback from the 2010 trial. Whilst it uses a VPP for performance-based ratings, it has also been designed to work with a manageable number of measurements.

In the interests of fair competition the certificates for all rated yachts, which have the measurement data used by the system, are freely published. However, the design of the computer models and estimating algorithms will not be made public, in order to discourage owners or designers from attempting to gain unfair advantage.

To accommodate visiting yachts, a number of estimating routines allow a rating to be issued when some of the measurements are unavailable - although the estimating bias here aims to ensure that no advantage is conveyed.

If you are interested ...

... then have a look at the measurements page - which provides all the material needed to complete an application. As well as the written guide, there are pdf diagrams to help with the measurements and there are pdf and Excel application forms.