Below is a very slightly edited version of a post I made on another forum in response to one from Mtoto which may be of some help on the related issues of consistency and probables.
"With regard to the Prominent King letter, I think it is the *s there that have created much of the difficulty we've all faced and probably many still do.
The letter invites us to put together two apparently simple facts: position in betting forecast and last three race placing totals. VDW then, through the words "respective horses" and the fact that he only names the first five in the forecast, shows us that he means us to focus on the last three placings of those in the first five of the forecast, not the field as a whole. So by the time one reaches the end of the second paragraph of the letter, one is looking for him to highlight (and he seems to choose *s to do so) the qualifiers from that simple procedure, and we expect the numbers to be the placing totals.
But of course neither of these expectations is fulfulled. Instead of highlighting Decent Fellow, Beacon Light and Mr Kildare, he highlights Beacon Light, Prominent King and Mr Kildare, and several of the numbers are not what we should expect, most notably Beacon Light 3 instead of 4 and Prominent King 5 instead of 8. But at this stage the word "probables" doesn't worry us - after all, that would be an entirely reasonable word to use if VDW was indeed wanting to refer the the group of horses we were expected him to highlight - the three within the five which actually do have the three lowest placing totals.
Then come the years of difficulty, and we see and indeed contribute to discussion of several possible explanations. Decent Fellow, although being shown as 7, was really treated as 14 because in his last race he came last of three runners and in a later article VDW said that for some horses who finished last we should count their placing as 10. If so, that would make Prominent King rather than Decent Fellow one of the three from the five with the lowest placing totals, but how logical would assigning Decent Fellow a 10 be, and if that is what he did why didn't VDW say so, as he did explain the simple counting on procedure used for Mr Kildare? The numbers for Beacon Light and Prominent King are put down as mistakes. The number for Monksfield, also a mistake, is viewed as understandable because we assume (rightly in my view) that VDW used the Mail for the forecast and they gave Monksfield's last three placings as 3/3/0 when in reality they were 3/3/6 with the 6 not being last. Etc etc.
If one thinks about it - which eventually, largely due to the insistence of Pro, I did - that is an awful lot of errors from someone we take very seriously and are inclined to see as a genius among race analysts. Because if the explanation for Decent Fellow is right, and that for Monksfield is accepted, the numbers ought to have been 14, 4, 16, 8, 3, ie three out of five given wrongly! How likely is that?
That of course led me to think that VDW had adopted a sleight of hand here and had introduced an additional idea, over and above the two that seem to have been presented in the second paragraph, that of probables being potentially different from the horses found by the two ideas clearly presented in the second paragraph. So the task then is to find a way of rating the horses which does indeed generate the numbers as printed - and, logically in my view, a way that is consistent with all VDW's other examples. (He does, after all, write in the March 1981 article that each of the elements of his method "has proved highly successful and consistent for a considerable number of years" - my emboldening.)
So there one has it. Had the only error in the numbers been Monksfield, I'd have had no problem and at a stretch could even have accepted the quite widely held view about Decent Fellow. But the Beacon Light and Prominent King numbers wrong, when we are dealing with (a) a horse he specifically said had had a hard race last time against Sea Pigeon and therefore surely can't be unaware that he came second, and (b) his selection for the race and whose form he specifically said he had checked? No way. In the end I came to agree with Pro that this was pushing things too far, and reached the view that VDW was, albeit very slyly, introducing us to a third idea here.
From my position, therefore, the *s in the Prominent King are not, unlike those in the "Most consistent" columns of the tables in the March 1981 article, there to show the three most consistent horses from the first 5/6 in the forecast, but to indicate the probables in a different sense, and as you know I think they indicate the horses left after VDW applied another rating method to the totality of the consistent horses (what I refer to as "automatics" and "discretionary") if there are more than three.
As regards identifying the consistent horses, I think that for the "automatic" ones that is done simply on numbers and without regard to anything else, except in two very specific circumstances - where one of the placings reflects an unfinished race (as per the Uther Pendragon letter where one needs to assess whether a count back is appropriate - virtually always with falls and unseateds but pulled ups are more difficult), or re a horse which has one placing in the three that is out of line (as with the Gaye Chance example, when class needs to be considered to see if for that run the horse was out of its depth). With the potential "discretionary" consistent horses, I think VDW took a little more time and did run a quick eye over the class and form of the last three runs before making his judgements (but if I am right about the criteria he used, as per this thread, this doesn't take more than seconds a horse and is in no way comparable to the "study [of] the form of all concerned" referred to under the Little Owl table of the March 1981 article).
Since getting to the position I now occupy re consistency and probables, the more I realise why "VDW" remains a mystery to most and no one has marketed what they claim to be the full VDW method. Because if I am right, before one understands how VDW got to his probables, one must first have as full as possible understanding of how he got to the totality of consistent horses, and outside this forum I don't think anyone has got very far with that. To do requires study of all the examples and a way of identifying the "difficult" ones (Love from Verona and Son of Love, of course, but also others less widely recognised as difficult, including Prominent King, Righthand Man and Roushayd, none of whom was an "automatic" consistent horse). Of course, quite often just taking the three lowest placing totals from the first 5/6 of the forecast will find all the consistent horses, and if then there are only three horses, all the probables. But quite often falls some way short of always."
