Cheltenham Festival
There is currently 1 person viewing this thread.
this is a mathematical certainty what I post below. It's fundamental statistics.
----------------------------------------
So last 18 years of RSA winners and beaten horses in the top 5 in the betting who hadn't run in that year.
Bostons Angel - Time for Rupert 7/4f (6l 5th when horse was sick)
Weapons Amnesty - None
Cooldine - What a Friend 13/2, Gone To Lunch 7/1
Albertas Run - None
Denman - None
Star de Mohaison - Darkness 11/2
Trabolgan - Comply or Die 3/1 (came 2nd was off track since bad fall in nov 04)
Rule Supreme - None
One Knight - None
Hussard Collonges - None
Brief Gale - Lady Cricket 5/1
Nahthen Lad - None
Hanakham - None
Florida Pearl - None
Looks Like Trouble - None
Lord Noelie - None
Monsieur Le Cure - None
I think I see a pattern regarding where this is going.....
To update the plain English "If you backed every horse in the top 5 in the betting for the past 18 years at the SP that hadn't run in the calendar year what is the % chance you'd not have had a winner of the RSA?"
Convert odds to % chances of winning, take away from 1 to give chance of not winning, then multiply them all together to get your % chance.
(1-0.133)* (1-0.125) * (1-0.15) * (1-0.25) * (1-0.167) = 40% Chance, Based on SP's meaning it's probably more likely you would not have backed the winner than you would have backed the winner.
Note: TFR was excluded as the horse was sick, which is a contaminating factor. We don't know what he would run like if he wasn't sick and the sickness was likely a very big contributing factor to his performance.
I'm assuming if I continue back into the past there will be very few if any horses in the top 5 in the betting who hadn't run that year. It's hard to get data back prior to 1993.
----------------------------------------
So last 18 years of RSA winners and beaten horses in the top 5 in the betting who hadn't run in that year.
Bostons Angel - Time for Rupert 7/4f (6l 5th when horse was sick)
Weapons Amnesty - None
Cooldine - What a Friend 13/2, Gone To Lunch 7/1
Albertas Run - None
Denman - None
Star de Mohaison - Darkness 11/2
Trabolgan - Comply or Die 3/1 (came 2nd was off track since bad fall in nov 04)
Rule Supreme - None
One Knight - None
Hussard Collonges - None
Brief Gale - Lady Cricket 5/1
Nahthen Lad - None
Hanakham - None
Florida Pearl - None
Looks Like Trouble - None
Lord Noelie - None
Monsieur Le Cure - None
I think I see a pattern regarding where this is going.....
To update the plain English "If you backed every horse in the top 5 in the betting for the past 18 years at the SP that hadn't run in the calendar year what is the % chance you'd not have had a winner of the RSA?"
Convert odds to % chances of winning, take away from 1 to give chance of not winning, then multiply them all together to get your % chance.
(1-0.133)* (1-0.125) * (1-0.15) * (1-0.25) * (1-0.167) = 40% Chance, Based on SP's meaning it's probably more likely you would not have backed the winner than you would have backed the winner.
Note: TFR was excluded as the horse was sick, which is a contaminating factor. We don't know what he would run like if he wasn't sick and the sickness was likely a very big contributing factor to his performance.
I'm assuming if I continue back into the past there will be very few if any horses in the top 5 in the betting who hadn't run that year. It's hard to get data back prior to 1993.
sort by:
Show
per page
Replies: 12
|
By:
Say bookies run a 120% book scale the 40% up by multiplying by 1.2 and you get 48% so basically it's 50:50 if you would have backed the winner.
|
|
By:
yup so ignore the stat
|
|
By:
Said in one of the many threads that I really don't like simplistic statistics applied to just a single race. There are logic based exceptions, like the stat for experience being essential in the CH - would pass your test above comfortably with loads of fancied horses failing in the last 2 decades.
I much prefer to use wider based stats, cut off point of 84 days absence works very well in the main festival non handicaps (4 champ races+ryanair/cathcart, novice hurdles, novice chases(not 4miler)). In the last 8 years just 2 winners at 17/2 and evens failed the stat and there were 33 losers at 10/1 and under and a further 22 at 20/1 and under. |
|
By:
You can use your cut offs and guides, whatever suites you.
But the simple point I'm proving here is. If someone says have horses who have not run in that calendar year fared worst than they were expected to in the RSA. The answer is no. I proved the exact same thing last year with regards Montjeus and the festival. What I didn't differentiate in my analysis is that I would place a big difference in horses who were off the track injured and horses who were off the track by choice. For example First Lieutenant is off track by choice of Mouse, Comply or Die was off track because of a very bad fall. What I'm getting at here is that I will exclude the calendar year stat from my RSA analysis as it not backed up by the evidence. So grand Crus, First L and Join Together are not excluded from betting decisions as that would be erronious. |
|
By:
I used the same logic to break my rule and back Somersby in the Arkle 2 years ago, won't make the mistake again. The 84 days is based on being just long enough to accommodate rest pattern horses (who typically need 42 days+ between races) and is actually a cut off point I have used for many many years. Only rules out JT and not the other 2 anyway.
Do you have to be so snotty even when I'm agreeing you CV? Just trying to be helpful to the people who's stat you are debunking, no more no less. |
|
By:
I wasn't being snotty
 sorry if it came off that way in my post. |
|
By:
Fair enough.
|
|
By:
With regards to the above, I believe the stat is last 48 winners have ran in same calendar year so is there a further 30 years of analysis to be undertaken?
Of the 6 in the top 5 of the betting, how many were placed? None won, regardless of falling lto, sick, whatever. 0%. I am also not sure you need to be in the top 5 of the betting to win an RSA? I think a few of the winners mentioned above were outsiders? Stats are not the be all and end all but would I wish to back a horse in an RSA who hasn't ran since January 1st? No. |
|
By:
ballydoyle...that just about sums it up
 |
|
By:
People can make their own betting decisions and people did last year with regards Montjeu stat.
I'm just showing there is no statistical evidence to back up the claim about it being a disadvantage to not have run in the calendar year. It's simply normal randomness. There is no statistical link, it's actually indisputable as it's mathematics and no opinion is involved. So people would be stupid to discount Grand Crus, First Lieutenant and Join Together based on the stat. As for outside the top 5 in the betting being excluded, it was a time factor, I had to research each horse individually. But the odd 16/1 or 25/1 shot would not effect the result significantly, a 16/1 should would only adjust the outcome % by 1.5%. Meaning it's negligable. Also the stat does go back for 48 years but it's clear that given only 1 horse - Lady Cricket - out of the first 10 years examined (1994-2003) fit the criteria, it would be logical to conclude especially given how much racing has changed, that there would be possibly 4/5 horses that fit the bill, given horses raced far more in the past. what is evident is how few horses run in the RSA when they have not run in the calendar year. The frequency with which it is happening now reflects changes in training methods. It's possible these days to get a horse in excellent condition without the need of a prep run, a stark conrtast to previous years. Like many of the stats that have been around for awhile and have been broken in recent years, it is logical to conclude this one will be debunked soon enough. |
|
By:
CV ...have you backed JT, FL or GC by any chance?
|
|
By:
cant see what difference it makes if a horse last run on say 31st of december would it have less chance than if it ran on 1st of january ?
|