General Betting
Welcome to Live View – Take the tour to learn more
Start Tour
There is currently 1 person viewing this thread.
10 Jan 10
11:17
Joined:
Date Joined:
14 Jul 09
which are the most relevant years?
last 10, 5 or 3 years maybe
for instance backing handicap chasers, you have a profit at a course over 5 years, but a slight loss in last 3. which group of years would you take?
or at another course the last 5 years have shown a slight loss, but a nice profit in last 10.
Thanks.
last 10, 5 or 3 years maybe
for instance backing handicap chasers, you have a profit at a course over 5 years, but a slight loss in last 3. which group of years would you take?
or at another course the last 5 years have shown a slight loss, but a nice profit in last 10.
Thanks.
Show More
obviously I would re-assess after each year. thinking of going with last 3 years in the portfolio.
That was just an example kenilworth, just wanted an opinion on optimum years to use, given above stats, cheers.
I'd have thought that you should be asking yourself why particular years have shown a profit and other years a loss.
Personally I wouldn't waste my time with these type of stats as there are a thousand and one things that can skew the figures.
As an example, years ago, 18th February 1993 to be precise, there was a jumps meeting a Leicester. Looking at the trainer's table that day I noticed that David Nicholson, although one of the top trainers in terms of the number of winners at the course over the previous five years, was actually showing a deficit based on a £1 level stake bet on all of his runners.
In a five runner novice hurdle race on the card he ran a horse called Stylus ridden by a 7lb claimer. It was the rank outsider and it won at 150/1.
This single big priced winner transformed Nicholson's level stakes loss at the course to a thumping profit and anyone looking at the trainer's table at the next Leicester meeting may well have been misled by the bald figures.
I would suggest that your time would be better spent investigating how certain trainers go about getting their unexposed horses handicapped to win at a low level in terms of class thereby opening up the opportunity to stay ahead of the handicapper for a few races.
Personally I wouldn't waste my time with these type of stats as there are a thousand and one things that can skew the figures.
As an example, years ago, 18th February 1993 to be precise, there was a jumps meeting a Leicester. Looking at the trainer's table that day I noticed that David Nicholson, although one of the top trainers in terms of the number of winners at the course over the previous five years, was actually showing a deficit based on a £1 level stake bet on all of his runners.
In a five runner novice hurdle race on the card he ran a horse called Stylus ridden by a 7lb claimer. It was the rank outsider and it won at 150/1.
This single big priced winner transformed Nicholson's level stakes loss at the course to a thumping profit and anyone looking at the trainer's table at the next Leicester meeting may well have been misled by the bald figures.
I would suggest that your time would be better spent investigating how certain trainers go about getting their unexposed horses handicapped to win at a low level in terms of class thereby opening up the opportunity to stay ahead of the handicapper for a few races.
although if you've a 1500 race sample with a ROI of 115% I think you have something to work on.
Anything of any merit is factored into the price of the runner. I suspect in the long run your better off selecting horses from stables that you trust are competent and dont show a level stakes profit.
The trouble with this sort of analysis (and it may be back-filling) is that you can never be sure if you have discovered any real insight into the course, trainers, jocks, horses, etc - or just found out that some local betting markets are smarter than others - or that there are statistical blips that should be of no concern. And while it's easy to tweak the past to make a profit (esp at SP), it's far more difficult to truely understand the future and beat the market on here.
Good luck anyhow.
Good luck anyhow.
Many thanks for that interesting comment REM.
What's the biggest sample of results required, to define a system as not backfitted. 500, 1000?
There must be a loose figure that when arrived at, can overcome the backfitting issue.
What's the biggest sample of results required, to define a system as not backfitted. 500, 1000?
There must be a loose figure that when arrived at, can overcome the backfitting issue.
mitchell d
Your original post mentioned handicap chasers.
By my reckoning, since 2005 there have been around 4200 such races with a total of around 40,000 runners.
As an extreme example of a 'system', using just three filters, backing Venetia Williams trained runners at Aintree in races of between 16-40 runners would have yielded a profit at SP of over £100. That's the good news.
The bad news is that the profit was achieved from just 2 winners and only 20 qualifiers! The 2 winners equate to just 0.05% of all winners of handicap chases, and the 20 qualifiers to around 0.05% of all runners in handicap chases.
It's impossible to say whether or not this 'system' will be profitable in future years but I would guess that one would need to work on a much higher percentage figure in terms of number of winners/qualifiers as an indicator of the utility of any filters one may use.
What these figures might be, has to be a matter of personal choice.
Your original post mentioned handicap chasers.
By my reckoning, since 2005 there have been around 4200 such races with a total of around 40,000 runners.
As an extreme example of a 'system', using just three filters, backing Venetia Williams trained runners at Aintree in races of between 16-40 runners would have yielded a profit at SP of over £100. That's the good news.
The bad news is that the profit was achieved from just 2 winners and only 20 qualifiers! The 2 winners equate to just 0.05% of all winners of handicap chases, and the 20 qualifiers to around 0.05% of all runners in handicap chases.
It's impossible to say whether or not this 'system' will be profitable in future years but I would guess that one would need to work on a much higher percentage figure in terms of number of winners/qualifiers as an indicator of the utility of any filters one may use.
What these figures might be, has to be a matter of personal choice.
Yes, all good answers to lead you in the right direction.
Don't be thinking of 10 years, you may be using a criteria that has shown a loss for 5 consecutive seasons or more
I would add that the Sporting Life annual guide used to run a portfolio which pretty much made a profit every year. I think it was as a straightforward as identifying trainers at courses in certain races, criteria being decent strike-rate, decent profit, and to have done so in each of the last 3 years. Cant go too far wrong on that.
But that was in days when all such information was just not available on a daily basis.
And be warned, I once assessed some sire stats for 20 sires based on previously profitability. No kidding the next year it started with a straight run of 57 losses.
If you've got the stats it would be wise to identify your logic based criteria then run it for results if you'd started operating it in 2008 2007 2006 etc, then not to be tempted to amend it until you show a profit
Don't be thinking of 10 years, you may be using a criteria that has shown a loss for 5 consecutive seasons or more
I would add that the Sporting Life annual guide used to run a portfolio which pretty much made a profit every year. I think it was as a straightforward as identifying trainers at courses in certain races, criteria being decent strike-rate, decent profit, and to have done so in each of the last 3 years. Cant go too far wrong on that.
But that was in days when all such information was just not available on a daily basis.
And be warned, I once assessed some sire stats for 20 sires based on previously profitability. No kidding the next year it started with a straight run of 57 losses.
If you've got the stats it would be wise to identify your logic based criteria then run it for results if you'd started operating it in 2008 2007 2006 etc, then not to be tempted to amend it until you show a profit
Post Your Reply
Please login to post a reply.
Wonder
Instance ID: 13539