With the lull in decent handicaps yesterday and today I thought it would be worth posting here on an issue which has been concerning me and others over the last few months, namely the assessment of class.
VDW wrote in one of his early letters that "readers who fully understood my previous letters will know it is the balance between class, form and the other factors which show the good things" (17/01/81, reprinted in "The Golden Years" as item 36), and he emphasised this again in chapter 5 of "Systematic Betting":
"Logically, the relative merit of form must be equated to the quality (class) in which it was achived. This means there has to be at least two elements to jointly equate (sic) when judging the relative merits of one performance against another, FORM and CLASS. Form is what they did and class is the level at which it was achieved. Therefore, when looking at the relative merits of one horse against another, these two elements class and form must be equated along with the other aspects such as distance, going, track etc."
VDW then suggested that a method of assessing the class element which he found "most satisfactory" was to take the win prize money value of races (divided by 100) as a rating, and he illustrated this with reference to Roushayd and the 1988 Old Newton Cup:
the three races Roushayd contested prior to the £17,064 ONC (a class 170 race) had prize money of £5,080, £7,752 and £22,710 respectively, ie were rated 50, 77 and 227, and thus Roushayd was "raised in class in his second and third outings and then dropped fourth time".
However, VDW was writing in inflationary times for race prizes, and generally speaking we are now in deflationary times. Some 2011 races retain their 2010 prize levels, eg the Royal Hunt Cup, but the large majority have lower prizes on offer, sometimes dramatically so:
the Lincoln this year was a class 623 race, compared to 779 in 2010
the Northumberland Plate was a class 925 this year, compared to 1079 in 2010
the Totesport Handicap, run at Windsor on Plate day, was a 187 race this year compared to 280 in 2011.
An important question for the race analyst is what to make of this. Can we, for example, assume that the 2011 renewals of the three races for which I have given figures were all significantly inferior to their 2010 counterparts? Not in my view.
And the matter becomes particularly critical when trying to assess the recent form of horses in current races. To give two examples from winners last Saturday:
in the Totesport Handicap at Windsor, Norville's last three races and the Totesport had the following class ratings: 62, 130, 74 and 187, so it would seem as though Norville was raised in class for his second run, then dropped in class before being raised to what was, on this means of rating, by far the highest class race of the four.
An hour earlier Edinburgh Knight had won a similar 6f handicap at Newcastle. The class ratings for his three previous races and the Newcastle race were 93, 84, 130 and 125. This would suggest he had been dropped in class on the second run, then raised in class and finally dropped slightly in class.
However, if some courses, for some races, are maintaining race prize levels, while in general levels are falling, (a) can we be sure that the ratings can be interpreted as straightforwardly as VDW assumed with his comments about Roushayd's races and (b) how can we compare current race class with what the horses we are assessing have achieved in previous seasons?
Of course VDW was not solely concerned with class in terms of the value of the race, but also with the strength of the opposition, ie the class of the field as well as the class of the race. For some years I have felt that using horses' Official Ratings in handicaps offered a better method of rating the class of a field than VDW's means. There are differences in view about how to use ORs, but my conclusion is that a straightforward average of the runners' actual ORs gives as good a picture as any (though I do leave out from my calculations the ORs of the occasional horse running from more than 7lb out of the handicap).
And for probably three years now I have been using VDW's penalty value rating as the primary measure of class and the average OR as the secondary one, thus covering both race class and field class.
This season, however, I have concluded that it is safer to reverse the priority, ie use the average OR, or measure of field class, as the primary measure and the penalty value rating as the secondary one.
To illustrate, Norville's four races, in average OR terms, were 83.2, 96.9, 89.6 and 94.6, which suggest that, although he was up in class for the Windsor race, it was not the considerable rise in class over the three previous races that the penalty value ratings suggested. Indeed, arguably the second race (21 May, Chester) was the highest class of the four and that compared to that he was actually dropping slightly in class on Saturday.
The figures for Edinburgh Knight were 91.7, 90.1, 95.7 and 88.1, which suggest that far from being dropped in class very slightly on Saturday he was in fact facing by some way the weakest field he had met over the four race series.
I do not suggest abandoning the penalty value rating altogether. I include it in the summary sheets for each race I analyse because I believe that a race with a rating of, say, 280 would be different from one with a rating of 150, even if the fields (and thus the average ORs) were identical. And that is because in handicaps connections are principally running for the money, and a horse may well be targeted to win the class 280, while if the race was only a class 150 it might merely be a prep. race prior to a later, more valuable, prize. But to revert to Edinburgh Knight, I doubt connections saw any real distinction between the £12,952 on offer at York on 13 May and the £12,462 on offer on Saturday. But from the perspective of the horse's chance - assuming it was there to win on both occasions - there can be little doubt that Saturday offered the better prospect, as the competition were, on average, markedly inferior than at York.
In sum, I certainly haven't abandoned VDW's use of the penalty values of races as a measure of class, but this season I have found using the average OR as the primary measure has given me a different perspective and, frankly, better results.
