Erin Question

[JennyK]

Regarding VDW's examples, T Tufnel, it is a matter of working at them with provisional hypotheses in mind and, as one does, sharpening up those hypotheses as one understands more.

From the last month or so's work, I can now see I was premature in thinking that the highest ability-rated "form" horse was the class/form horse, ie from VDW's point of view the most likely winner. It often is, because often it will be a consistent "form" horse, but what if we have in the same race a "form" horse with a higher ability rating than a consistent "form" horse?

It was clear from his 1988 Mackeson evaluation that VDW occasionally ruled out a class/form horse because of the "other factors" but what did he mean by "other factors" and how did he consider them in practice? We know from the Mackeson one that they included distance and going, and from other examples course type. Weight? Lack of key horse support for class droppers? On the first he said a little but on the second as far as I can see nothing directly.

The above questions can only be answered by understanding VDW's examples.

The good news is that Lee was undoubtedly correct when he wrote that the more one studies the examples, the more one sees. Roushayd is relatively easy, being the highest ability-rated horse in the race. Pipsted less so, as only fifth in the ranking. As you discover more about how VDW assessed "form" status it will I think be clear to you that Pipsted was the only "form" horse in the top five, (and a consistent "form" horse), significantly ahead of the other "form" horses AND with very much better data (laying it out as per the Roushayd tables) than the others. Because of the presence of higher ability-rated horses in the field, which might return to form, the terms in which VDW compared Little Owl ("racing certainty") and Sunset Cristo ("almost a certainty") would have been in my view a better way to compare Roushayd and Pipsted than the one he chose.

Re taking a break, that sounds like a good idea. I have found it difficult analysing current races while hopefully making significant research progress with VDW's method, not just because doing both one's time is split but also because one has the dilemma of whether to modify how one analyses a current race in the light of what one thinks one is learning from the research. As a consequence, I've analysed fewer current races over the last few weeks than I otherwise would have and won't be focusing solely on current races for several more weeks, when hopefully I'll have more solid answers to the questions I referred to above. I am happy to cut back on current race analysis because with solidly based answers to those questions, I think the <20%, >80% issue will be resolved. And I think it was because he had made greater progress than Guest on that which explains the much higher strike rate Lee achieved with pre-race posted selections on the Gummy forum.

 

[gerry]

This is just a thought from what you just said JennyK about ability horse and consistent horse and form horse i thought about what you wrote and this came to my mind now it might be nonsense but i will type it anyway as could help thoughts. this is to extreme to show point.
Say you have two horses in a race one horse A has won six times and the other one B won once a small race.
But both horses have run 50 times and horse A never placed in a other race horse B was in a place 40 times in the races. So we have a clear ability horse and a clear consistency horse what is the class/form horse for today and what horse wins.
First thing is those two things wont matter if thats the only facts we have so horse A would in theory most likely win as other one B had chances but does not win.
The only way we could get the class and form horse is by what races they competed in horse A won 4 8k races and 2 10k races and all its unplaces since or in between those where 10k races and higher horse B won a 7k race and was placed a lot in races between 4k and 7k.
Now we know something solid clearly the class form horse is A and B will be placed again his usual consistent form but class should tell because this is a 7k race a big drop for A for first time and a level B has won but now consistently placed in a lot lesser races.
So the biggest question was is A going to return to winning form or is B going to improve to get 2nd win the answer is clear here A will beat B
Thats how i found pierte contra other week if thats right spelling it was clear class horse and the race it was in IE: clear winner in a race.
Now were looking at ability horse consistent horse class horse and the well being lately horse would help now we have found our horse is it the right distance GOING track for a bet.

 

[JennyK]

gerry

What distinguishes VDW's approach is not that he used the terms ability (and class), form and consistency, but that he did so in tightly defined ways and for the latter two, idiosyncratic ways A simple example; many would I think take the view that a horse who won lto was in form. He might be raised in class next time out so that he had no realistic chance, but that was a matter of ability (class), not form. VDW viewed form differently and his means of rating form status could mean a lto winner was not a "form" horse as he saw it. (That wasn't the case with the one race I looked at today - 4.10 Ascot - but I have had reason to look back to the Portland, 12/09/26, and one of the more fancied horses, Rosario, had won last time out a fortnight earlier but in my view was not one of the VDW "form" horses.)

In your example, it is quite clear that horse A had the higher ability rating, which is a plus.

What we don't know from the data you've provided is whether either (or indeed both) horses were "form" horses in VDW's sense, and if they were, whether they were just "form" horses or consistent "form" horses. It is only when one has a list of runners in ability order, and their "form" status and whether they were also consistent horses in VDW's sense that other factors come into play.

The evidence of his examples suggests that VDW was not concerned whether a horse was proven on the going or over the distance of a current race, but was if the evidence showed one or more was clearly wrong for the horse, ie one could say with confidence that a horse does not act on, for example, heavy going. Arguably the most successful VDWer, based on pre-race posted selections, initially thought that conditions didn't matter - a trainer who had a horse ready to win would only run him if he or she were confident that the conditions the horse faced were all right. He later modified that view, partly perhaps as a result of more detailed study of the VDW examples or maybe because he realised that VDW's view that sometimes horses were not running to win, but to prepare them for a future win. (I rather hope Tiriac yesterday is such a horse.)

One could fairly say that VDW's approach was broadly on the lines by which you found the winner you mentioned, but in some respects taken further.

 

[Tufnel]

Regarding VDW's examples, T Tufnel, it is a matter of working at them with provisional hypotheses in mind and, as one does, sharpening up those hypotheses as one understands more.

From the last month or so's work, I can now see I was premature in thinking that the highest ability-rated "form" horse was the class/form horse, ie from VDW's point of view the most likely winner. It often is, because often it will be a consistent "form" horse, but what if we have in the same race a "form" horse with a higher ability rating than a consistent "form" horse?

It was clear from his 1988 Mackeson evaluation that VDW occasionally ruled out a class/form horse because of the "other factors" but what did he mean by "other factors" and how did he consider them in practice? We know from the Mackeson one that they included distance and going, and from other examples course type. Weight? Lack of key horse support for class droppers? On the first he said a little but on the second as far as I can see nothing directly.

The above questions can only be answered by understanding VDW's examples.

The good news is that Lee was undoubtedly correct when he wrote that the more one studies the examples, the more one sees. Roushayd is relatively easy, being the highest ability-rated horse in the race. Pipsted less so, as only fifth in the ranking. As you discover more about how VDW assessed "form" status it will I think be clear to you that Pipsted was the only "form" horse in the top five, (and a consistent "form" horse), significantly ahead of the other "form" horses AND with very much better data (laying it out as per the Roushayd tables) than the others. Because of the presence of higher ability-rated horses in the field, which might return to form, the terms in which VDW compared Little Owl ("racing certainty") and Sunset Cristo ("almost a certainty") would have been in my view a better way to compare Roushayd and Pipsted than the one he chose.

Re taking a break, that sounds like a good idea. I have found it difficult analysing current races while hopefully making significant research progress with VDW's method, not just because doing both one's time is split but also because one has the dilemma of whether to modify how one analyses a current race in the light of what one thinks one is learning from the research. As a consequence, I've analysed fewer current races over the last few weeks than I otherwise would have and won't be focusing solely on current races for several more weeks, when hopefully I'll have more solid answers to the questions I referred to above. I am happy to cut back on current race analysis because with solidly based answers to those questions, I think the <20%, >80% issue will be resolved. And I think it was because he had made greater progress than Guest on that which explains the much higher strike rate Lee achieved with pre-race posted selections on the Gummy forum.

I think the only ways I can view Gorgeous Strike as a non form horse for Haydock is the last 3 runs showing he had to be dropped in class to collect and that the horse he beat a neck had been dropped in class last time and failed to win again having been 2nd in a slightly higher race before.

VDW used the phrase "dropped to collect last time" in several places and of course he often struck that horse off his list or out of his shortlist for the race.

Pipsted had been disappointing in the 2 runs before Ascot. Both on firm ground and once at the shorter 1m 2f. But clearly VDW saw the Ascot win as an obvious improvement including recording his best speed figure on soft going. This was enough to absolve Pipsted of those 2 previous runs. Which is another reason why I've said before that I saw the form status process as a 3 stage flag raising issue. When a run shows as form the flag is raised and if it gets lowered in the next run it has to be an improved run for the following run to raise the flag again.

The only issue with that though is in relation to consistent form. If every run has to be a form run, how can the 2nd and 3rd last races that Pipsted gave be form runs in their own right?

My guess would be it's one of those occasions where you have to consider a 4th last run. Starting at that point makes sense and I can see the flag is never lowered.

One of VDW's letters not long after the Spells It All Out letter is the one suggesting a crosscheck of 3 highest ARs from the 5 most consistent runners in the field. He made a point of saying "..given other factors anyone not selecting Saher should start asking questions of themselves".

I've largely thought the weight differences between that horse and those who beat him were a significant factor, given the horse had no winning class rating in his last 3 runs. Plus he'd already been dropped in class from his 3rd last run. But if you go back further to his 2 wins before that, the flag didn't drop afterwards.

Like yourself JennyK I find trying to research and uncover new things in the examples does not sit hand in hand with trying to sort out a race due off tomorrow or even later on the day itself. And part of that is trying to put together a fresh puzzle without really having all the facts on why an older one was solved.

And I allowed myself to get side tracked this week on ratings issues again. But I'm fairly sure I'm close to cracking the Handicap Hurdle puzzle and ratings apart from class don't seem to play a part in that one. So going down a blind alley seems a bit daft now.

 

[JennyK]

"Pipsted had been disappointing in the 2 runs before Ascot."

Look again, from the basic VDW tenet "what is form if it is not that one performance is better than another?", T Tufnel.

In what sense could Pipsted's performance on 25/05/87 not be seen as a better one than his performance on 06/05/87?

In what sense could Pipsted's performance on 18/06/87 not be seen as a better one than his performance on 25/05/87?

Try not to add your own value judgements or look at the situation as you might if you are analysing a race tomorrow. If your aim is to work out VDW's approach, suspend all that. Work on the assumption that every one of the mainstream VDW selections (about 80) was the class/form horse in its race, work through them with, in respect of form status, the VDW tenet in mind, and look for reasons why, not reasons why not.

Some, like Pipsted, are straightforward, others much less so and one or two extremely difficult. If/when you can't see how a given selection could be the class/form horse, just leave it while you work through others. Then, hopefully, as you learnt more, the difficult ones will become possible.

There are distractions in VDW's body of work, in the sense of other avenues he explored not related to the way he analysed the mainstream eighty or so. The handicap hurdles method is one, the 1995/6 is another. Both of those are probably worth cracking, not that I have, but the mainstream ones are where the main VDW approach was used and from which one can hope to find out how he worked.

ps

There are no blind alleys - just ways of finding out how VDW did not work. I am confident that every serious VDWer has at times thought they'd made a breakthrough only to find the following day that it stumbles on another example. But that is the only way to complete the journey. The hope is that when one has taken things as far as one can, one will be able to emulate VDW's claimed and Lee's proven strike rates. But the first goal is coherence - does one have hypotheses which explain ALL the mainstream examples?