Every year the announcement of the International Classifications and the Longines World's Best Racehorse Rankings and those of the best races in the world are greeted with close to universal approval and their findings are widely accepted in a non-critical manner.

Their acceptance is, above all, proof of the lasting appeal of a good heuristic or a rule of thumb coated in scientific language.

The classification and the rankings are drawn up by experts from 17 different countries, expressed in numbers and have been gradually developed over nearly 50 years, and so they must be sensible and positive.

And yet the rankings of the best races are devised with an illogical system of averages which is both nonsensical and contrary to the experiences of all of those trying to win these races, and the whole system of Group races upon which both the rating and the rankings are based has both undermined and obscured the appeal and attraction of racing for years.

It makes no sense at all to judge the quality of a Group or Grade 1 race by taking the average rating of the first four home, and a system devised on this basis is sure to favour races for older horses and in particular sprints in racing programmes where there are many older horses and geldings competing.

Under this system it is inevitable that the best races restricted to three-year-olds will be undervalued, and the best races in regional and national pools with a relatively small horse population unfairly penalised.

A logical system would be one in which races are graded by adding together 50 per cent of the winner's rating, 25 per cent of the second's, 15 per cent of the third's and ten per cent of the fourth's.

This is of course how the money is distributed and then Group 1 races for three-year-olds are designed to find out who is the best, not to produce close finishes between evenly matched horses.

The Classic races are the races used for selection for breeding because every year only a handful of horses have the ability and maturity to compete successfully in these races.

Group 1s for older horses, particularly sprints and mile races, are more evenly matched as many more horses are ready to compete at the highest level as older horses or geldings.

And then the connections of any high-class three-year-old are trying to find the best opportunities to win money and prestige while its ability and comparative advantage lasts. Nobody is trying to finish fourth in a Group 1, indeed to do so is a failure as there are always alternatives and it makes more sense to look for an easier Group 1 or drop back and try to win a Group 2.

Classic races are at their best as a spectacle and a test when a Frankel is able to win by a wide margin. The average rating of the first four home in Frankel's 2,000 Guineas was 115, and if their ratings are adjusted as above the overall rating is 123. Which seems the fairest judgement of the race to you?

This leads to the central problem with the current system of Pattern races. The endless multiplication of alternatives and the ranking of races by average ratings has meant that it is now achieving the reverse of what was intended. A system that was designed to promote competition and to make it clear to the wider public which were the best races now has precisely the opposite effect.

Every sport needs its grand slams, its majors, its playoffs or championships in which even the most casual of observers can recognise competition between the very best. In racing. this hardly ever happens as there are always alternatives and it is logical for the owners and trainers of every good horse to try to maximise its earnings and prestige by avoiding competition wherever possible.

The ranking of races by average ratings has also had an unforeseen effect of weakening regional or local events where the excellence of the best is undermined by the lack of competition for the places.

It is surely time to reconsider the whole system. What are, or should be, the aims of the Pattern race system, and is the current Pattern the best way of achieving these?

Aside from changing the way in which the best races are ranked there are numerous other possibilities including, most obviously, drastically cutting back the number of Group and stakes races and giving both black type and more money to the first six home in the ones retained.

If the aim is to reward the best five to ten per cent of the horse population with a marked status in pedigrees and sales catalogues doesn't it make sense for them to earn this by competing with each other rather than by travelling around looking for the easiest opportunity?

Is it time to reconsider the status of handicaps? For older horses doesn't it make more sense to encourage all but the very best to run against each other in valuable handicaps rather than in uncompetitive Group races with small fields?

Above all, isn't it time the outside view is taken into account? Surely the aim should not be to fill sales catalogues with horses in black type but to put on a coherent spectacle which is of interest for the betting public and a wider audience?