- This topic has 132 replies, 20 voices, and was last updated 16 years ago by
.
-
Posts
-
Saw a program some years ago about this guy that was using a computer to make money on the roulette wheel. Can’t remember how he was doing it , but I think it had some thing to do with where the guy was spinning the ball from.Nothing is completely random unless you want to get into Quantum Theory and if you understand that you don’t because no one does.If we know evey thing before the race we could no the true odds of each horse. Many say that a horse was unlucky when falling when really this game is all about jumping and a pass check on the horses record will often show it was a dodgy jumper and the horse falling was’nt such a shock.
Prufrock,
The 2:10 at Newbury today was a race that looked as if randomness played a major part. Using this race can you show me how knowing it exists is helpful, and can give one an edge?
Formath on another thread fancied the winner and was happy to put a price against the runners. Looking at the prices he put up I don’t think he had taken randomness into account. However if he wanted to he could have easily added a few points to ALL the runners to cover that eventuality. To cover these unforeseen events the chances of them happening MUST apply to all the runners otherwise it can’t be random.
This race may have shown you randomness is important and does happen, fair enough. Formath had looked for the likely winner, and because of this he could have allowed something in the price for the unexpected. You say this isn’t the way to go, for me it is the only sensible to approach the game.
Be Lucky
‘Unforseen’ occurences may occur in any given race. However, doesn’t the fact that individuals, i.e., trainer, jockey, etc. have a degree of control over the outcome negate the idea that the result of a horse race can be regarded as a random event?
Artemis is surely right here. We know that from time to time there will be incidents that effect the outcome of a race which we cannot specifically predict. The question is, do they happen sufficiently often to negate the forecasts one makes to oneself about the probable outcome when assessing the field.
If one backs on the NH, inevitably a proportion of one’s selections won’t finish, while on the Flat that is very rare. But under either code, the number of horses I back which fail because of specifically unforeseeable events like falls, brought downs, saddles slipping, suffering substantial interference etc is tiny.
I agree with Mtoto that, certainly with better class races, the outcomes are reliably predictable, even if not always predicted. Indeed, if this was not so, betting would be a daft thing to do. There is what to my mind is a pertinent quote in a booklet published in the late 1940s/early 1950s by an accountant who wrote under the pseudonym "Marvex":
"The form book is a record of the past and properly read is a guide to the future – very seldom does the result of a race so confound the form student that he tears his hair; he may be surprised by a result but he has only to get out his form books to find substantiation therein in 99 cases out of a 100."
Thank you, Hensman. Nobody has suggested that the outcome of races is random – just that there are random events which influence results and that we should be philosophical when these unforeseen events occur.
Marvex was right. The vast majority of races can be explained by the form book if you dig deep enough. If that wasn’t so, rational people would not bet or study past form. Occasionally, a result defies form analysis and the explanation for it lies within the race itself. When this happens, we should forget about it and move on.
Scallywag’s point about human influence is valid. Unless we have inside information, we have no way of knowing if(prior to the race) a trainer or jockey wants a horse to run to its best form or whether a horse was given every chance to do its best in the race. When horses run below expectations, we are often very quick to think that the horse wasn’t ‘off’ instead of looking at what happened in the race or if the conditions were not in the horse’s favour. It is another unpredictable thing to ponder.
Artemis
You wrote:
"Unless we have inside information, we have no way of knowing if (prior to the race) a trainer or jockey wants a horse to run to its best form or whether a horse was given every chance to do its best in the race."
I agree with that – we can’t know in the sense of being sure. But we can make reasonable deductions from the evidence, and a point made both by Marvex and VDW which in my experience is spot on is that in many races only a few are there to win. The majority are there as part of their preparation to win in the future. From that perspective race analysis can be seen as a two stage process: first to identify those horses which can reasonably be presumed to be there to win, and second to select from those the one which has the best credentials to do so.
The problem is of course with the class of the race, is that higher class might mean more consistent, but it also means more competitive, these are the favourite statistics for 2007 so far;
Class .. SR .. I/V (+/-)
1 .. 32% .. 0.97/1.04
2 .. 26% .. 0.78/1.28
3 .. 34% .. 1.02/0.98
4 .. 35% .. 1.04/0.96
5 .. 34% .. 1.02/0.98
6 .. 33% .. 0.98/1.02This is looking at all of the favourites, in all of the races and then working out in which class the favourite is most likely to win. The class argument should see the higher class favourites doing better than lower class favourites which is clearly not the case. In fact the opposite is true.
If one backs on the NH, inevitably a proportion of one’s selections won’t finish, while on the Flat that is very rare. But under either code, the number of horses I back which fail because of specifically unforeseeable events like falls, brought downs, saddles slipping, suffering substantial interference etc is tiny.
.. the chance of a horse not completing in a NH race on average is 10%, depending on the type of race and the going.
This is looking at all of the favourites, in all of the races and then working out in which class the favourite is most likely to win. The class argument should see the higher class favourites doing better than lower class favourites which is clearly not the case. In fact the opposite is true.
Dave Jay,
Why do you think just because a horse is the favourite it stands a better/higher chance of winning? In most/many cases in these better class races hype is the only reason some of these horses are made favourite. Even more so in many of the class 2 hcps. The record of the favourites may prove something to you, but for me it is just another stat that proves little/nothing..
I agree it maybe harder to isolate the "true" class horse in higher grades of racing and easier in the lower classes. It is the fact that finding the class horse that highlights the likely winner, not being the favourite.
Be Lucky
Mtoto .. how else would you measure a series of races for randomness?
There have been loads, nay millions of posts, all over the place, stating categorically that higher class races are more consistent (and less random) than lower class ones, without one shred of data to back it up. Its either a fact and can be supported by some sort of evidence or it’s a myth and it can’t.
dave jay
The incidence of non completing horses in better class NH handicaps since March 2005 (the earliest records I keep) is 22.4% from a pool of nearly 6,100. And as you suggest the rates for subsets are very differential. For example, of the 2,400 hurdlers in the 6,100 just over 88% completed, while only 70% of the 3,700 chasers who set out finished.
Mtoto .. how else would you measure a series of races for randomness?
Dave Jay,
As it is something I have never considered important I haven’t given the matter much thought. I just can’t see how the fate of a certain set of horses be it first, second, third favourites or rank outsiders can be used to show it. Some of the winners from these categories must also have benefited from the results of randomness as well as the losers. I don’t really see how randomness can be measured, what is it? Is it the same for everyone, I look at a race/horse and see a possible weakness, you can’t see it. When/if it happens something that is random to you, isn’t for me.
The nearest I come to looking at the truly unexpected result is when my rating fail to include the winner of a race. I only work in the highest class races so when I say good horses conform to their profiles I can only talk from my own experience. It would be fair to say it is a rarity when a race is won by a horse that didn’t figure in the ratings. As I have no intention of explaining how I formulate the ratings any stats I give would be meaningless as they can’t be checked. However even with these unexpected results very few are down to randomness.
Be Lucky
I think it is well accepted that horses of moderate ability tend to be less reliable and more inconsistent in racecourse performance. So, to assume that better class races are more consistent, meaning that the participants are more consistent, seems to me to be a reasonable assumption.
But certain aspects of randomness, such as being brought down in a race, are mostly not class sensitive, I think. That said, I guess you could argue that fallers are more likely in lower class races (anyone got that stat?) and that you first need a faller in order to be brought down.
The favourite is the market interpretation of the most likely winner of a race, but even if it is true that lower class races have more randomness I am not surprised that the favourite win % is roughly the same for all classes.
The evidence that the horse with the best perceived chance in the race(ie the favourite) wins the highest proportion of races is overwhelming.
It is also the convential wisdom(although not proved) that the better class races produce the more reliable form lines.
So, it is quite logical that you would expect favourites in better class races to perform better than they do in lower grades.
Dave’s figures for 2007(quite a large sample) contradict the above. I am inclined to conclude that the conventional wisdom concerning better class races is a myth, as Dave suggests.
If an analysis over a longer period yielded the same results, it would be proof positive to me that this convential wisdom (which I have held myself) is plainly wrong. Nice work, Dave.
You have to remember that there are many false favorites every day so studying favorites isn’t going to prove anything. Many factors go in to what makes a horse favorite including people sticking pins in the paper,women liking the name of the horse and horses being hyped in newspapers etc .Just because a horse is favorite doesn’t mean its the horse most likely to win. You have to find out if its a legitimate favorite or a false favorite.As for the better class horses being more reliable all I can say is my experience and betting records show this to be true.
maggsy,
The statistics(facts) tell a different story. Horses win races(on average) in proportion to their odds; so, 4/6 chances will win about 58% of all races.
Granted these are average figures from a very large sample of races and certainly contain ‘false’ favourites, I would say that these favourites are typically horses with the best form credentials in the race.I would have expected that 4/6 chances in Group 1 races would be marginally more profitable than 4/6 chances in Group 6 races.
My assumptions are:
1. That horses in the higher grade are more consistent and reliable.
2. The betting margins in the higher grade are more favourable to the backer.If the stats show there to be no difference between profitability between the grades, then at least one of my assumptions is wrong. I’m fairly certain that assumption 2 is correct, so assumption 1 must be false.
As I stated in my last post, if Dave’s figures are correct, the better class horses are, in fact, no more reliable than those in lower grades.
This has been suggested to me on several occasions in the past by various forum members, but I never really believed that to be true. Until now.
- You must be logged in to reply to this topic.