I ask this question because ratings are the first step in evaluating a race and they have a great affect on pricing. The most influential are Timeform’s and the Racing Post’s figures but there are plenty of others.
The one thing that we are never given is how these ratings have performed over a long period. Have they improved , i.e. do they point to a higher % of winners than they used to, or a lower %? Are they more or less profitable to follow?
The only freely available site that I know of where you can get an answer is the excellent Adrian Massey site at http://www.adrianmassey.com/index.php.
I looked at his figures for his top-rated horses since 1991. OK, they are his ratings using his methodology but a rough check shows that there is a strong correlation with what you would expect from other ratings such as RPR’s, so I will use his results as a proxy for them.
Since 1991 there has been a consistent decline in the performance of the top-rated horse both in terms of win% and in terms of NP%.
In 1991 to 93 the win% were 28,29, 29 and the net profit% were +9% in each year.
By 1999-2001, the win% had dropped to 26,26,25 but the net profits % held at +9%.
For the last 3 full years 2007-9 the win % dropped further to 24,22,24 and the NP% also declined to losses of –10%,-13% and –8%. So far, 2010 is showing a win% of 25 and a net loss of –8%.
Of course, AM’s figures may not be representative. He may have changed his approach, I don’t know. But, if we take them at face value, and for the present that’s what I am inclined to do, and draw general conclusions for ratings as a group, then the findings are very interesting. The obvious conclusion is that ratings are indeed losing their importance.
Why is another question.
Closer inspection shows that it is in handicaps where the decline is greatest, at least in profitability falling (in Flat h/c’s) from +15,+15,+19 in 1991-3 to –10%, -17 ,-5%.in 2007-9. However, the win % remained more or less constant at 20%. Over jumps, the results for H/C’s (hurdles & fences) showed that the results showed significantly worse declines in both win% and NP%.
I don’t believe that average field sizes have changed so one possible explanation is that the official handicapper has got much better at his job and handicaps are more of a lottery than they were 20 years ago. Given that average over-rounds per runner have been forced down over the period (thanks to BF) one might have thought that the NP% might not have been as badly affected. Seemingly not.
Strike rate should always be looked at in the context of what would be achieved by random. That is, in this respect, the field sizes involved.
I suspect the drop off in figures referred to above has more than a little to do with improved standards in official handicapping. Back in 1990 it was a piece of p1ss to do a better job than the Jockey Club handicappers (jumps especially).
Ironically, this could be taken as proof that "traditional" ratings - at least in terms of what is now known as the BHA - have increased importance.
As posted over on ante-post betting:Strike rate should always be looked at in the context of what would be achieved by random. That is, in this respect, the field sizes involved.I suspect the drop off in figures referred to above has more than a litt