- This topic has 132 replies, 20 voices, and was last updated 16 years ago by
.
-
Posts
-
Okay then, looking at the favourites strike rate could be a bad way to look at things. So lets take all of the horses over the last two years, that won last time out and then their next race was 1 class higher or lower next time out.
Class .. S/R .. IV (+/-)
1 .. 19% .. (0.96 / 1.04)
2 .. 17% .. (0.84 / 1.19)
3 .. 21% .. (1.04 / 0.97)
4 .. 23% .. (1.13 / 0.89)
5 .. 19% .. (0.96 / 1.04)
6 .. 19% .. (0.93 / 1.08 )
All classes = 20% (1.0 / 1.0)So, here again .. no advantage concentrating on higher class races. These results are random which means that class has no bearing on the results. This is what I mean by randomness, if I was to continue collecting these results over years, all class groups would average out to a 20% S/R.
I would argue that all you are doing by concentrating on higher class races is narrowing down the number of races you are looking at in an arbitrary manner. That in itself is fine, but it’s not science and it’s not backed up by any statistical evidence.
I am looking for someone to come back at me here and prove that I’m wrong.
Artemis, higher class horses may well be more consistent but this appears to be overcome by the fact that the higher class races are more competitive.
A winner LTO going up or down a class would race against different horses. These horses would be from the same/up/down class – horses going up or down the ladder. Although our winner was the best LTO it will probably not be the best this race. Unless this horse is outstanding or the trainer has carefully entered this race for another win, I think that 20% is about right for a further win.
If there are 5 LTO winners in a race, only one ( or none ) can win twice on the trot.
It this random ?
Artemis, higher class horses may well be more consistent but this appears to be overcome by the fact that the higher class races are more competitive.
Yes, that is what I also thought, because, obviously, higher class races are populated by higher class horses!
Interesting thread, I have just reread most of it and I think what is lacking is a consensus about what is being referred to by the word ‘randomness’. It seems that we have mostly made a jump towards equating randomness with inconsistency, but I’m still not sure that that is correct.
Surely randomness (as inferred in this thread) must be mostly/entirely independent of the characteristics of the participants (ie horses)? Since inconsistency is an accepted (?) characteristic of lower class horses in can, in principle, be allowed for … yes, I know it is nebulous and difficult, but I don’t think it is quite right to call it random. Perhaps what is needed is a better appreciation of variability?
BTW I have looked at the stats for winning 1st favs in UK Flat pattern races (G1,G2,G3 and Listed) over the past 20 years. the numbers are much the same over all four categories, being 33-34%, so again confirming Dave’s earlier findings!
I would say luck is a phenomema about an individual event that occurs within a framework of probabilities that are true over a long series of events. Randomness could be seen how quickly or slowly these individual ‘lucky’ events take to come around to their true probabilities, over time
I find myself quoting myself ..
Dave,
Yes, I was aware that you had said that, it was also part of the reason I previously asked if you would elaborate on :
We also know that horses don’t have a normal dstribution when it comes to races about their relative chances.
So, am I right in thinking that, in the simple case of a series of coin tosses in which ‘heads’ has fallen way behind (so to speak), that you see ‘luck/randomness’ as the process of it eventually catching up? Or have I misunderstood?
There’s an interesting one.
I think that random only looks forward. If this is a fair ( not plotted-up ) coin spin – ‘heads’ should never catch-up.
More or less, yes .. a coin spin might never catch up over say 300,000,000 spins, it might be only 5,000 or so behind. So, although the odds are still true, if your betting at even money, you’ve done your conkers.
Jim, I’m not going to go into [b:8n9onbav]Normal Distribution[/url:8n9onbav][/b:8n9onbav] .. but you can read about it in the link.
Dave,
Ah, I’m with you now (you can probably tell that I haven’t read the book, but it is on order!).
Interesting, I need to think about the consequences, now that I am on the same wavelength.
btw, which ‘normal distribution’ link are you referring too?
Jim
Dave Jay, as a matter of interest, does it change your view on any of what was "discussed" on the infamous VDW thread? Having embroiled myself in that particular waste of time,
Dave Jay,
I’m afraid it was this post by Prufrock that caught my attention in this thread. I mistakenly thought when he talked about randomness it was a quantifiable factor that could be used then assessing a race. Not just an academic theroy that has little or no part in assessing a particular horses chances in any race.
I wish you luck and hope you solve what for me is insolvable. If/when you do, hopefully it will stop the lazy/poor punter from using randomness as an excuse, even more importantly it will stop them using it to make themselves look academically superior to the poor sod who has taken the time analyse a race properly.
Yes, luck, randomness can happen to any horse at any time, but that goes for ALL the runners in any race. Knowing that can’t give anyone an edge, and backing horses hoping it will intervene, is a bigger waste of time than VDW will ever be.
Good Luck
Harsh.
I personally cannot quantify randomness because it is, erm, random and I’m not smart enough to land it on the head of a pin.
I know that it exists, in horseracing as in other fields, and my punting improved enough to make a living from it when I accepted this fact and stopped trying to be quite so deterministic.
I bet on thousands of races in recent years and saw randomness at work all the time. How can you possibly predict or explain the processes that go into horses leaving the stalls and taking up positions – possibly on slower or faster going, chasing a slower or faster pace, or finding themselves blocked in or getting a clear run – in a Flat race?
I still like to think that I "take the time to analyse a race properly", just like the "poor sod", who in this instance I invite to sod off.
Be random.
I’m only halfway through the book, so won’t add much.
However, I don’t think the key is in attempting to deal with ‘randomness’ within each race. I think it more likely that it is in one’s staking strategy/fund management.
From what I’ve read so far, the main impression is that the financial traders the book is focusing on had poor or non existent risk management strategies. If there’s a gambling analogy to be drawn it would probably be to those who do their gonads on betfair on 1:01 shots, or dare I say it, to the currently highflying but risk-insane Harry Findlay.
So, am I right in thinking that, in the simple case of a series of coin tosses in which ‘heads’ has fallen way behind (so to speak), that you see ‘luck/randomness’ as the process of it eventually catching up? Or have I misunderstood?
JimF,
The randomness is that the heads and tails have not fallen in equal numbers. The heads can be behind, the precise number of falls behind can vary randomly, and exactly the same can apply for the tails. You can predict the "probabilities" of heads being behind and behind by a certain number for a 50% true chance each and every spin, but you cannot predict the exact future state.
The thread to me is about being "fooled" by randomness.
You cannot be certain that a historic strike rate or profit persists into the future even if you then continue to use the exactly the same method that gave those past statistics, all else remaining equal.However, if you have a historic 50% strike rate it is how closely that would follow the possible future pattern of heads and tails in a coin toss that determines if your past success is still performing as a true 50% chance – the coin toss.
In the coin toss example, all the possible events are exactly known but it still behaves randomly. In horse racing, we don’t have sufficient knowledge to even predict all the possible events and behaviours – so we might add a bit onto the fair price (ignoring these possibilities) for luck. This might be called a double randomness – so there is an even higher chance of being fooled.
It would have a quite different pattern of possibilities from tossing a 6 on a dice (probability 1/6). So if you were just told the toss results blindly you could tell from just the results, if a dice (3 points each 6) or a coin (1 point each head) had produced the results, or even if they had been mixed.
Exactly tooting .. finally someone who has grasped the thrust of the debate, fooled by the 100% book. Its easy to think you have an edge because of randomness, over many bets.
Harry Findlay is probably right, the unexpected is not so unexpected at all.
Prufrock the way I was thinking of trying to calculate luck was by using an adaptation of the formula is Steven Skiena’s book ‘Calculated Bets’. I don’t know if people are familiar with this methodology?
- Total Posts 17716
But DJ class, used properly, does give an edge over randomness, despite what your earlier table suggests.
Using the 1st 3 in the market, all races past 7 years, shows the following:Dropped in class
class 1
class 2 26.3%
class 3 25.0%
class 4 25.6%
class 5 24.8%
class 6 22.4%Raised in class
class 1 18.8%
class 2 17.9%
class 3 24.2%
class 4 22.1%
class 5 21.6%There are many other ways too, to reduce ‘chance’ in one’s betting, but it is fortunately still an art, rather than a science, to combine them effectively on a race by race basis, and will always confound those that believe horse racing to be unpredictable.
There are many other ways too, to reduce ‘chance’ in one’s betting, but it is fortunately still an art, rather than a science, to combine them effectively on a race by race basis, and will always confound those that believe horse racing to be unpredictable.
Agreed RH
Isn’t the unpredicatable element in racing – that part unable to be forecast by historical modelling – precisely why it is a weakly-efficient betting market able to be exploited by the ‘artist’ rather than the ‘scientist’
Some would say the advent of betting exchanges has moved the market from weak-form to semi-strong
See:
Click on the words in my post.
LOL Many thanks, but I don’t think that Wikipedia can tell me anything I don’t already know about the ‘Normal Distribution’. It was your words "… when it comes to races about their relative chances" that had me confused.
Robert99,
Thank you very much for taking the time to provide me with such a lucid explanation, I am now fully on board and Amazon have just emailed me to say that both books are on their way to me!
My first reaction is on the same lines as ‘tooting’ has just said, namely, that "one’s staking strategy/fund management" is likely to be important here. Personally, I subscribe to your "we might add a bit onto the fair price", which I currently apply heuristically!
- You must be logged in to reply to this topic.