The thousands of bets refers to one criteria, to give better odds that a mechanical system will not fall over.

This has to be part of the structure not all.

That's not to say a few hundred bets won't work, but how many systems based on a few hundred bets will be around in 5 years time.

There are no mechanical systems, that I know of, that have failed which satisfied all the criteria.

There are hundreds which failed to satisfy just one part of the criteria, and wound up as toilet paper.

__________________
RaceCensus - powerful system testing software.
Now with over 412,000 Metropolitan, Provincial and Country races!
http://www.propun.com.au/horse_raci...ng_systems.html
*RaceCensus now updated to 31/12/2024
Video overview of RaceCensus here:
http://www.youtube.com/watch?v=W821YP_b0Pg

To suggest that a large statistical sample indicating a trend is no evidence that the trend will continue is absurd. Show me a roulette wheel which in 10,000 spins turns up 9500 blacks and I'll take even money black for the next 10,000.

Why?

Because if you're not "retro-fitting" (term used earlier, never heard that one before but I like it) then if something happens over a sample that large it happens for a reason. In the case of the roulette wheel the reason will be because it is biased. In the case of form, it'll be because those form factors work, whatever they might be.

If anyone actually does believe that the 10,000 results is meaningless, let me rig up a roulette wheel here at home, come over and give me even money black. Oh yeah, bring a lot of cash.

Give me a method proven over thousands of bets any day over one that has been tested on 200 or less....


To me it's bit like the "Global warming" debate...we measure temps for less than 200 years and decide we have a problem..its complete b.s..200 years, in the context of the time this planet has been alive, is less than the blink of an eyelid.

Aside from the thought police comment which could be considered derogatory to your good selves this question from stebbo is a legitimate one.
The forum members who contribute freely what must be considered the drawcard to this site have an interest in threads they are generating. That's what makes them worth reading. I know I've come back after 24 hours eager to see how something has developed and been quite annoyed to find all trace of it gone.
I have no argument with your right to remove posts or threads to protect your legal position or the integrity of the site but for the sake of us poor saps who have an interest in what's going on I would ask you to do it with some sensitivity. I've known a few interesting threads removed in their entirity because someone got out of hand towards the end.
Personally I would find an a post from the moderator with a few explanatory words along the lines of - several posts removed because of xxxx - much more satisfying. Or maybe a post saying - xxxxx given a yellow card for insulting language - would be entertaining as well as informative.
Probably people who breeze in would find it entertaining too if you show a little creativity instead of merely brandishing the rule book.

Now standing in the hall with an excersise book down the back of my pants waiting for my turn in the naughty room.

KV

I guess of all of us you would be the most qualified to make a set of rules like this. I wonder if you could quantify them at all. I realise the number of bets must depend on the size of the divvy etc but I would like to see some figures. How many bets for what average price using how many rules would you say would guarantee a system is going to keep winning.
Aside from the obvious benefits of knowing your answer to this question I've always seen backfitting as a fun project to try to undertake and a few figures would give a target to aim at. Like is it possible to devise a system that would win over a thousand bets just using the letters of the horses names as a silly example.
Come to think of it a good example of retrofitting like that would serve as a good lesson to would be system creators to show how easy it is to fool yourself into thinking you have a winning system.

KV

Hi KV!
I have checked it out, and it works! Never followed it myself but if you are interested I can look up the results and post it.
I actually went through the alphabet and checked the name beginning with the first letter ,second letter etc.

Sounds interesting Marcus if it's easy to find.
I did one myself once on all the races in the last 10 years at some obscure WA track (it's on the forum somewhere). First letter t or something had a great record but only about 30 or so races. Funny thing is it fell over at the next meeting. :-)

KV

Hi KV,

I am not sure about qualfied, but thanks for the compliment.

I have studied system viability over many years, and the rules are something that emerged after studying WHY systems fall over.
In fact there are many systems which can be turned into winning ones, with a little tweaking.
Firstly you have have to ditch the rules which have no logical basis other than snaring longshots.
Then you have to apply rules which have a logical basis and don't just throw one or two longshots.

Admittedly, with any system, the profits come from a small number of winners.
If the small number of winners are very long prices, then you are relying on horses with very little winning chance and are bound to not repeat, as the same horses are not running over the next set of bets.

To quantify the rules, if you want a system with a very good chance of not falling over, deduct at least 1% of the winners which are the longest price, very few systems can withstand this deduction and inevitably fail anyway.

If a system can withstand this sort of deduction, chances are it will hold up in the future, but of course there is no guarantee. It is about increasing your chances of maintaining success.

I cannot really supply exact ratios and figures, because every system throws different patterns and strike rates and average dividends.

But I can give you an example:

One of my (many) systems

Selections: 612
Winners: 299
Strike Rate: 48.86%
Profit: $56.60
POT: 9.25%
Average Dividend: $2.24

Now let's dissect this...

1. There are not thousands of bets to qualify this system, which means I cannot be confident of a future profit, BUT these are the reasons it works....

1. All the horses are well supported in the market.
2. The profit does not come from a handful of longshots
3. I started following this system around three years ago, and consistently tested in real time with real bets, and it has held up well with only very small bad runs.
4. It always bounces back after a loss (to date).
5. Before testing in real time, I ran a system test over 12 months, then a seperate test over the previous 12 months, then I contacted someone with a very large database going back to the 1980's to test it on those races independently, he also confirmed consistent profit.
6. The MAXIMUM win dividend is $2.90, divide that by the profit and you have 19.5 longshots making up the profit.

This summarises how this system is viable in short.

__________________
RaceCensus - powerful system testing software.
Now with over 412,000 Metropolitan, Provincial and Country races!
http://www.propun.com.au/horse_raci...ng_systems.html
*RaceCensus now updated to 31/12/2024
Video overview of RaceCensus here:
http://www.youtube.com/watch?v=W821YP_b0Pg

Hi KV,
Being a magpie I knew I had it tucked away someplace, well here it is.
It should be self explanatory, the name of the speadsheet tells it all.
FirstLetter.xls is where the horse's name begins with the corresponding letter, SecLetter.xls where the second letter in the name corresponds with the letter in the line etc.
I can't remember why I done it (somebody must have asked for it) but it proves one point, if any, mechanical systems can work if you stick to them long enough, but first you must test them (retrofit??)
There is about four years of data involved, so it should be OK.
I personally could not be bothered with it because my rating works well enough, if not for the jocks and trainers I would be a wealthy man.
Cheers

Attached Files

	Fifthletter.xls (17.5 KB, 647 views)
	FirstLetter.xls (17.5 KB, 669 views)
	FourthLetter.xls (17.5 KB, 603 views)
	Secletter.xls (17.5 KB, 642 views)
	ThirdLetter.xls (17.5 KB, 624 views)

Hi Marcus,
Just an observation, I notice the letter "R' has been omitted.

I beleive that one has to have a min of 4 years of stats to see if it has any long term legs to it. Even then there is no guarantee. But it is a stronger guarantee than something based on just one year .

I get confused where certain people use the word retrofitting as a derogatory term , the opposite to this is creating systems based on the future with results unknown as yet.
Past history is an excellent starting point when creating systems .

I run a number of plans all at once , the SR is there but the divs fluctuate markedly, thus , the POT changes month to month.
E.g. 21-25% SR and a POT from 1%-40% in any given month.

The suggestion that a large data base is no more affective than a small sample , is a mathematical abserdaty.
All data companies rely on large data samples to help obtain a truer picture of probability towards the future. e.g. engineering , space research , medical research , science and the running of many simulations , not just one or two.

The first question of any statistical analysis that is always asked is , what size data sample was used?
The larger the data sample , the more credability it is given, rightly or wrongly.

Cheers.

__________________
Cheers.