- This topic has 581 replies, 56 voices, and was last updated 15 years, 4 months ago by
.
-
Posts
-
I believe Bentner and his syndicate made no profits until they incorporated unexpected variations in the market odds into their model.
This is what I referred to as the possible ‘missing link’ in my recent post.
Pru,
I read your post that starts… I am a form and time man (I left out the sectionals part as I’m still trying to get to grips with it) with great interest. The interest is based on if I had been asked to discribe my outlook on racing I could have written it. I hasten to say perhaps not as eloquently, but certainly with the same content. All of these things are part of the VDW thinking and anyone who has studied it would agree VDW covered them in some detail.
I too watch as much racing as I can, and form my own opinions, but I do like those opinions to be backed up with hard figures. I too accept every horse has a theoretical chance of winning a race, but here the difference is the theory part doesn’t happen often enough to make it worthwhile.
I have also read about Bill Banter and the bit that stuck in my mind is the fact his method only works because he is only dealing with a small pool of horses on a few race tracks. As much as I admire anyone that make money from backing horses, I feel his method has little to do with form reading, thus just making it number crunching.
Be Lucky
Thanks for the link. This looks similar to the method I have been developing over the last 5 years but only got 28 variables in current model. A lot more have been tested and dropped. Combing over 30 years of traditional form study and a first class SQL developer is proving useful.
As for VDW, this is a good starting point for beginners but is not a method for serious punters looking to earn a living from the game.
Ah, the missing link. A thread that could go on for years.
It seems to suggest that there is some magic formula waiting to be discovered. But, alas, this isn’t so.
I think Bentner knew that no matter how much data was put into his analysis, there would always be certain information that was known only to those dealing with a horse on a daily basis. Most astute punters know this instinctively. They look for signs from the market that horses are doing their best and are expected to run well and also signs that suggest the opposite.
But just how important is this information? How can you quantify it and allow for it in your model? That’s the problem.
As for VDW, this is a good starting point for beginners but is not a method for serious punters looking to earn a living from the game.
.. my point exactly Wallace, everyone has to start somewhere.
Possibly worth starting a new thread on this one, Artemis
.. if your a value punter Pru, as you seem to intimate, then how are you avoiding being shark food and not backing loads of losers?
I think vdw type analysis ultimately ends up with a Benter type model. It does need to be done properly though as you say Pru.
Every factor has a value, it depends on the level of accuracy that you want to operate under and what you are happy with that makes the difference. Saying a horse won last time out doesn’t really tell you much, except it won but it still has a value .. speed/weight handicapping will give you a more accurate measure of the win, but everyone knows this, so how much more value does it really have??
Artemis, I did some research recently where I only looked at horses that were half of their last starting price 5 minutes before the off, which was quite interesting .. !
Hi all,
I have read through some of the comments here with interest.
Re ‘the missing link’
I think the first reference to it was a letter from G Hall who claimed to see the ‘Method’s key’ from the first seven/eight selections given by VDW
Later it (or something else?) was referred to as the missing link. The whole problem is that those who have claimed to have found it, will not give it away. (The Key/missing link)
People have rubbished VDW on the basis that you are searching for a Holy Grail that does not exist. Others still search.
Perhaps it will go on for ever!
.. if your a value punter Pru, as you seem to intimate, then how are you avoiding being shark food and not backing loads of losers?
I don’t see why it should be a surprise that a value punter can make it pay.
For what it’s worth, I started off with a fairly deterministic view of things, thinking that all I needed to do was get better and better at predicting winners – a sort of VDW outlook, possibly – did OK but no better and ended up handicapping for a living.
Being a professional handicapper is a bit like having a bet in every race (without it costing you anything other than your job at the end of it if you are no good). After 100,000 or so repetitions (a bit like the aforementioned McSim) it does begin to sink in that pretty much anything can, and does, happen if you wait long enough for it. Black Swans do exist.
Oh, and it helps to be slightly sceptical about your own prowess and to look gift horses in the mouth occasionally.
"Artemis":vj29qqnj wrote: .
I think Bentner knew that no matter how much data was put into his analysis, there would always be certain information that was known only to those dealing with a horse on a daily basis. Most astute punters know this instinctively. They look for signs from the market that horses are doing their best and are expected to run well and also signs that suggest the opposite.
But just how important is this information? How can you quantify it and allow for it in your model? That’s the problem.
The Benter method does not give two hoots about "certain information that was known only to those dealing with a horse on a daily basis". It looks only at past performance to estimate a price and compares that with the PMU board. It has to avoid putting too much on so that the syndicates end up winning their own money back, less the takeout.
You can qualify that type of information by seeing the difference between taking account of it and ignoring it completely.
Not my understanding of how the syndicate worked, robert.
I got my information from the appendix in Nick Mordin’s book, Winning Without Thinking and Mordin’s own comments about the method in that book. I don’t have the book anymore, so I’m relying on memory.
I’m fairly sure that the syndicate did factor in the opinion of the crowd(the market), because their initial analysis was not profitable enough to make their work worthwhile.
This had a certain appeal to me because I also feel that it is not possible to make money from horse racing by relying purely on the published form book and extracting data from it.
I exclude private handicapping from the above generalisation. I know that you and quite a few other forumites use your own judgement, watch most races and as a result, you are not relying on the published form that most punters use.
The collective opinions of the private handicappers and shrewd judges, together with other inside information, must find its way into the market in any country, and it is this money that causes ‘unexplained’ movements in prices.
I don’t think you should confuse the Benter model with the syndicate operation Artemis. The Benter model can be used for multivariate analysis, but the syndicate operation worked in pool betting, probably based on the herfindahl index, exploiting the fav/longshot bias imo.
Black Swan’s Pru .. tell me about it ..
It is a surprise that you can make it pay value punting, to me anyway. It seems that the horses that appear to be the worst value (drifters) would be the ones that are most likley to lose.Winning On Betfair For Dummies, by Jack Houghton:
A Betfair study looked at a sample of 65,000 horses competing in 2005.
19% of the horses that drifted significantly in the betting went on to win. The result is only slightly less than the 22% of horses whose odds had shortened dramatically in the betting.
Most importantly, if you had placed £10 on all the horses, you’d have lost £3,000 on the horses that had shortened in the betting and yet won over £2,000 on the drifters.
Unless we know specifically how a "significant" or "dramatic" drifter or shortener is definied and at what price the £10 is placed, those statistics mean very little.
If the tenner is placed at the absolute top offer on a drifter and at the bottom offer on a shortener it is hardly surprising that the top offers show a profit and vice versa.
It was for "Dummies" (not for you, obviously), which is why the methodology was not explained.
I believe that the determining price was the Betfair Weighted Average Price in the final minute before the off and that "significant" was clearly defined as a relative % before the study was undertaken.
I can ask Jack for more details and then post it on here if you would like.
- You must be logged in to reply to this topic.