- This topic has 581 replies, 56 voices, and was last updated 15 years, 3 months ago by
.
-
Posts
-
Hensman, someone on another forum summed that up rather succinctly –
[b:28jsqx7t]there’s a formula that goes something like ..
no short-cuts + hard work = not interested
[/b:28jsqx7t]!
I have studied the works of VDW for some years now and concluded that it is somewhat flawed in its method of evaluating races/horses.
First of all, form figures should not be taken by face value alone. You might as well add up the saddle cloth numbers for what useful information that gives out. He also had a suicidal way of conjuring up non-existent form figures, ie, if a horse had only run twice and had form figures of ’32’ then a ‘1’ would be added to make its tally ‘6’. Come on, even someone half as shrewd as VDW (allegedly) would not dream of getting away with such methods of folly in race evaluation!
Prize money: this has some form of feasibility, in particular with 4yo+ horses as there will be enough ‘settled’ information to arrive at a reliable evaluation. I think you would need to tread carefully with younger horses though. Lammtarra is a good example, off the top of my head; prior to it winning the Derby, it had only won one race as a 2yo, worth about 7k.
Where VDW finally becomes transparent, is at the last two rating columns he used. Nearly all his selections were, I think, top-rated in both. However, he would not divulge the method(s) he used to arrive at these ratings – dismissing readers by telling them it was only part of the picture and they could easily substitute it with their own methods. Why, then, say to one reader ‘start talking with your cheque book’ for demanding the ‘secret calculations’ if it was not an important ingredient to the overall method? Imagine myself selling a copy of War and Peace on Ebay with the last chapter missing and telling the disgruntled buyer that it is not important to the book, you can submit you own version of it if you like or I will send you the last chapter in exchange for a fat cheque. Enough said!
Personally, I cannot understand why people are still blowing on this expired ember – the pedestal, in my opinion, belongs to Stewart Simpson, but that’s another story.
Quote from Johngringo – “howling mob”
If by ‘howling mob’ you mean the forum members then yes, I do tend to bow to their wishes.
Damn – I meant to put ‘taxonomical’ on the swear list!
Grand Lodge
Just to correct two factual points (errors of this type tend to get picked up and assumed to be what VDW actually suggested, as we have already seen on the thread):
1) for horses with less than three runs VDW did indeed recommend, as part of his basic method, adding on notional placings, but not as you describe: he recommended adding on the last actual placing, so a horse with two runs, a 3rd and a subsequent 2nd, would, for his consistency aggregate purposes, count as 7 (the actual 3 and 2, and a notional 2). The correct formulation of VDW’s recommendation here in no significant way changes the force of your criticism of it, of course;
2) in the overwhelming majority of his examples VDW did not quote his ratings or indeed any other, though in a few cases he illustrated examples with the Life and Mail ratings, which were probably the two most accessible at the time. (Those wishing to work out how VDW compiled his own ratings should start by comparing them and the Mail ratings for the races in question.)
You are right in respect of the ability rating and younger/less exposed horses. VDW himself drew attention to this issue and offered the sf-based measure of ability as a check, best illustrated in his discussion of the 2000 Guineas won by To-Agori-Mou.
I’m not sure anyone is blowing on this expired ember, as you put it. It is clear from this thread and others that some find at least something of value in VDW’s work and use it in their approaches to race analysis. It is also clear that others are interested in learning a bit about it in order to make up their own minds. Personally, I find the latter encouraging, because it suggests that some, at least, continue to be interested in trying to become competent at analysing races to identify probable winners and have not, like others, given that up as impossible and sought refuge in the intellectually flawed "value betting" approach.
From those who have attempted to unravel the VDW conundrum only a small number ever manage to make it over the first hurdle. Although their interpretation of the method, which surely can only be borne from reading and analysing his letters and selections, is clearly wrong, doggedly claiming that certain elements, and how he suggested they should be used, are irrelevant and of no use.
He gave a numerical picture, which showed, when used correctly, whether a horse was ready to win, or being readied to win. How the trainer then placed the horse would determine whether it was a ‘winner in the race’. There are trainers and there are trainers though.
The majority of VDW’s early selections were being raised in class for the prize. The class form horse seldom fails to take that prize when it is big enough, but it must have been placed to do so. Roushayd filled in the missing blanks regarding how he assessed form, showing us a) how seemingly inconsistent horses were dealt with and b) giving us an insight into important aspects of assessing form, such as probability, in the layout.
Class, in his own words, is kingpin. And again I quote ‘that four letter word ‘form’ is what seems to cause the most confusion’. It would appear, to me at least, that Class is actually the stumbling block judging by recent discussions on here, because without an understanding of class, in the way VDW defined it, the student will be perplexed as far as ‘form’ is concerned. What is form if it is not that one performance is better than another?
An Interesting thread.
It is however quite sad that those who have no interest in the subject continue to post and take the mick out of those who are interested.
Now maybe I am right or maybe I am wrong but I would have assumed that a long standing forum such as this will have grown ups partaking.
I subscribe to Sky but around 55% of the programmes I do not watch as I don’t like them or in the case of the medium shows I do not believe in them, but each to their own. If you don’t like it don’t turn it on, quite simple really.
Grow up gents and smell the coffee and let those who want to debate in proper fashion do so, then let the others out into the play ground and that stands for all forum discussion.
Cheers
Lamby150
Probable winners.
Every runner in a race is, within reason, a probable winner. Assessing an individual horse (or groupof horses) probability of winning against the implied probability suggested by those offering you odds (be it high street bookmaker or an anonymous exchange adversary) and betting when the odds are in your favour is what I take it you mean by ‘value betting’.
Let’s take a logic check.
If I make a profit from betting on horses winning races then it logically follows that I must be betting horses at odds/probabilities which are greater than their actual probability was of winning. Everyone must agree with that, right?
Therefore successful (and by that I mean profitable) followers of a methodological approach as has been described on these many pages are doing no more than identifying and bettting on horses whose probable chance of winning is greater than implied by the odds on offer.
In fact, using the definition of ‘value betting’ outlined in the first paragraph, they are ‘value betting’, even though they may not care to term it as such
“If I make a profit from betting on horses winning races then it logically follows that I must be betting horses at odds/probabilities which are greater than their actual probability was of winning. Everyone must agree with that, right?”
Cormack,
Not necessarily.
If any form of retrieval is employed the profit is derived from the staking strategy rather than the selection process.
Johngringo are you saying that a staking plan can turn a loss into a profit because I don’t think that is’nt possible at least not in the long run.Hensman you surprise me by saying you think that value betting is flawed. Why do you think that?I think VDW teaches value betting and indeed with out it its impossible to win in the long run.
Maggsy,
Very, very, simply put, yes a staking plan can put a level stakes losing method into profit.
Imagine that you only bet evens to 2/1 shots but double your last losing stake, as long as your bank can follow the progression of losses you will always come out ahead.
By using partial retrieval most ‘professionals’ look to keep ahead over a longer period of results by using a gentler progression of stakes.
Any reasonable selection method can be made profitable in this manner.
A few years ago I took all non-hcp horses that were rated 10+ pts clear by the RPR and backed them for a place, retrieving after a loser. I once went 29 consecutive winning bets but never went worse than two consecutive losers.
It might seem clever to some, but having been there I can assure you that it is only stupid gambling, and boring as hell to undertake once the mechanics have been tamed.
Cormack
You wrote: "Every runner in a race is, within reason, a probable winner."
Not on the normal sense of the word – more likely than not.
Maggsy
The principal reason the "value betting" approach as advocated by some is intellectually flawed (though not necessarily in practice unprofitable) is because, unlike the Stats.101 textbook standbys of tossed coins, spun roulette wheels, rolled dice, etc., no objective probabilities can be derived for the individual runners in horse races against which the actually available odds can be compared.
We all need to feel that the price of a horse we want to back is fair (or, better still, generous) but that is a quite different matter.
For all his qualities, VDW was no statistician – as demonstrated by his notion that if there were three horses in a race each having won its last three races, it was 99% probable that one would win. That "logic" tells us that if there are four (not unknown in Cheltenham novice hurdles) then it is 132% probable that one will win!
Maggsy,
Any selection method that is profitable in the very long run can be enhanced by using a staking plan, provided the plan is not too ambitious, is related to the average strike rate and never puts the entire bank at risk.
This can be demonstrated using mathematical models that simulate thousands of trials(millions, if necessary).In the normal run of events, the expectations from level stakes and various staking plans are about the same. The increased prospects of profits from increasing stakes is almost equally counterbalanced by the increased chances of losing the entire bank or a significant portion of it.
The punter’s dilemma is that when you start out on a method, you never know how it will fare in the very long run, so you can never be sure that your staking method will increase or decrease your profits compared to level stakes.
I would say that for all practical purposes, progressive staking methods are no better than level stakes methods in terms of profitabilty, but they are more interesting to apply and do stimulate discussion.
Johndringo that system would probably work most of the time probably for months or even years, but sooner or later you would hit a losing run and than if you were doubling up you would probably end up losing all that you had won very quickly.Abit like winning the lottery in reverse. the chances of picking the 6 right numbers are 14 million to one and people play the lottery for years with out winning.
Hensman I think VDW was talking about averages. The average horse with form figures 111 will win about 33% of the time . If 3 horses in a race have these figures then there would be 99% chance the winner will come from these 3 all things being equal. Although of course they never are equal because you would have to look at the class of the last 3 races and many other factors that would affect the probabilities of the horses winning. I think what he was trying to put across was how stats can be used to narrow the vision to where there are alot of winners.
Hensman I think VDW was talking about averages. The average horse with form figures 111 will win about 33% of the time . If 3 horses in a race have these figures then there would be 99% chance the winner will come from these 3 all things being equal. Although of course they never are equal because you would have to look at the class of the last 3 races and many other factors that would affect the probabilities of the horses winning. I think what he was trying to put across was how stats can be used to narrow the vision to where there are alot of winners.
I agree with the last sentence.
If you had a race with 3 (111, 33%) horses and 6 (122, 25% say) horses then on "average" you have a total percentage of
3×33 +6×25 = 249%.
Putting that back to 100% means that the "average" (111) horse has a 33/2.49 = 13.3% chance and the "average" (122) horse a 25/2.49 chance = 10%.
So the combined chance of a winner coming from one of the (111) horses is 3×13.3% = 39.9% (40%), which is a long, long way from the 99% statement. The chance of a winner coming from one of the 122 horses is 60%.
A serious error in VDW’s thinking but not something that can’t be forgiven in comparison to the other real insights given. Of course, the detractors jump on these errors to rubbish the whole method (or series of methods) that they do not have a prayer of ever understanding.
Robert99 that would’nt be an average race that would be a competitive race. If you look at the races at the Cheltenham Festival probably the most competitive meeting of the year you get alot of horses with 111 form figures, but these are not average races.At the Festival a horse with form figures 111 wins only 15.85%( according to RSB) of races because the races a far above average and very competive.In an average race 111 wins 33% of races on the Jumps according to RSB. If you look at other races you will see less competitive races where 111 might have a 50% chance of winning because all the other horses have 000 or 478 figures. All this has to be related to class and many other factors though the figures can never be taken at face value as VDW made clear although many still seem not to understand this.
- You must be logged in to reply to this topic.