PDA

View Full Version : Why do the majority of betting systems fail?


Marcus
6th November 2001, 07:22 PM
1986 I think it was I really got sucked in. Bought this load of rubbish after seeing an ad and being sent blurb about life on a tropical island.
They sent me a place betting system with a load of ridiculous, stupid rules. It basically was back a horse for a place if it starts in the booies' ring at 7/4 or shorter. Go all up a few times to improve your odds. Your results will be better if you do your own form to decide which of the horses at 7/4 or shorter you eliminate.
I was a sucker and paid several hundred dollars for that.

Lescoach
7th November 2001, 03:19 AM
Hi Marcus,
The load of rubbish i purchased was $6,400
of place betting crap. Certainly gives the buisness a bad name. I was surprised not to see it on the scam list.
For information the product is PROTEC 2000
distributed by Harper Menzies & Associates
from Queensland DON'T BOTHER

darkydog2002
10th October 2002, 11:17 AM
they fail because they are not based on class/weight ratings.

becareful
10th October 2002, 11:38 AM
Darkydog - A system doesn't have to be based on class/weight ratings to be profitable!

As to the question of why purchased systems are not profitable there are a few different reasons (in my opinion):

1. A large number (but by no means all) are simply scams where the people have developed something with the sole intention of flogging it off to the suckers to get their money.

2. Some systems are developed/sold by people with good intentions who honestly believe they have found a working system but either they don't understand statistics well enough to realise that it will fail in the end or it is developed using a flawed methodology (most commonly designing the system rules to fit past data)

3. A small number of systems do genuinely work but whether they continue to do so will often depend on how many people are using the system. All systems will have a maximum level of betting they can sustain - if this is exceeded then they will start to lose money.

_________________
"Computers can do that????" - Homer Simpson

<font size=-1>[ This Message was edited by: quapi on 2003-01-07 20:59 ]</font>

Rain Lover
11th October 2002, 09:44 PM
There's another possibility that you guys haven't really addressed. The fact that any "system", which is historically successful, can turn out to be an absolute dud in the future, and usually does.
This is simply because past results can't guarantee any future results, unless there is an underlying connection between the "system" and the horses winning. Invariably there is not, as most of these "systems" are invented by sifting through masses of old race results, trying to find a common thread among the winners.
I can understand why people attempt to sell this crap, but I can't understand why people fall for it.

Equine Investor
11th October 2002, 10:56 PM
There is no problem creating a system from past results PROVIDING the parameters and filters you set are based on fundamental principles and research.

For example: 95% of horses cannot carry 69kg and win over 3200m. That is not likely to change in the future unless horses are allowed steriods or the weight scale changes dramatically.

Betting on TAB numbers etc are not fundamental principles, but doing research on the way horses perform in general and the impact of extra distance and weight etc are.

Not all systems are scams. Not all people who provide a service to punters are scam artists either.
The problem is sorting out the honest and legitimate ones from the thousands out there.

darkydog2002
14th October 2002, 01:05 PM
becareful there has never been any legend of the turf to my knowledge who has attained that status on the back of a racing system.on the contrary from don scott to warren block and others they have consistently warned against them.if you want to win over a long period of time then you must use class/weight ratings.all iam trying to do is put people on the right track.

The Phoenix
14th October 2002, 01:36 PM
I suggest that a class/weight ratings approach is simply another (of many) systems. The goal is to be able to select horses on which one can obtain a higher price than their real chances of winning the race, ie make a profit. Whether the system is based on weight/class, particular form lines, particular breeding, a preference for one track rather than another is not the issue, only whether over a long period of time they do or do not make a profit.

On the other side of the ledger, the bookies want to do the opposite, offer horses at prices which are under their true chance of winning and therefore ensure that the punter makes a loss and they make a profit. Prior to the running of a race, the difference is simply opinion and the bookie is no more sure that their opinion is right than the punter is.

Some of the greatest winners at the track did not follow class/weight ratings at all. Their methods could better be desribed as a set of rules or axioms which they had determined, probably by trial and error and making losses, that gave them enough winning bets to make a profit. They are or were for their time perfectly valid systems albiet probably peculiar ones so anyone who claims that class/weight ratings are right and everything else is wrong in my opinion miss the main point which is MAKE A PROFIT.

regards

The Phoenix

becareful
14th October 2002, 03:07 PM
The Pheonix,

That was my point exactly - Class/Weight ratings is simply one approach but people who believe that it is the ONLY approach are closing their eyes to other possibilities. Someone else mentioned elsewhere on this forum that you need to find your own approach to make a profit in punting - if you just do what everyone else does then you will lose money (like everyone else does!). If everyone is doing class/weight ratings then maybe looking elsewhere is not such a bad idea.

Personally I am currently using two different systems in my betting (and working on or testing several others) - one of them is based on class/weight and one is not.

Dr Pangloss
14th October 2002, 04:32 PM
An excellent thread and don't disagree with anything anyone has said thus far. However, it is interesting to note the 'How to Win' debate always seems to gravitate toward the selection process.

A long term professional punter on another forum ranked the importance of the selection process third behind STAKING and PSYCHOLOGY.

Yet another wisened old pro signs off his posts with the reminder,

"An average handicapper with superior betting skills will outperform a superior handicapper with average betting skills."

Nevertheless I'm not expecting the topic of most posts to stray too far from selection methodology anytime soon.

becareful
14th October 2002, 06:38 PM
Dr Pangloss,

Personally I think the selection process is the most important. I would rank Psychology 2nd and Staking 3rd. My reasoning is that no amount of Psychology or Staking can turn a losing system into a winning one (although either one can turn a winning system into a losing one) so the selection process has to be right to start with. If you do have a winning system then Psychology does become quite important - I know I have missed out on quite a few good wins in the past due to being too cautious.

Maybe you should start another thread with your thoughts on the matter - I will be happy to contribute to it (and I am sure others will too). I just don't want to divert this thread from the main subject matter.

Dr Pangloss
14th October 2002, 09:17 PM
TOPIC: Why do the majority of betting systems fail?

Response: Because of an unfailing and persistent disregard for -

(a) staking methodology
(b) Psychology

It might be the case that the majority of betting system users fail at (a) or (b), or both.

I respect the view expressed in your post be careful and can not say you are wrong. As for being off topic - I think not.

Equine Investor
14th October 2002, 10:12 PM
1. Psychology.

This has to be ranked number one in my book.
If you don't possess the self discipline, patience and faith in your own system, you'll never win. Providing your system is proven and sound, you will win - but not if the right mentality isn't there.

Staking is ranked last.
Your system has to win at level stakes betting. Correct staking will only maximise profits, but if your system doesn't show a level stakes profit, it will maximise losses.

darkydog2002
15th October 2002, 01:44 PM
becareful .what would you say about the power of ten then,.if you want to make money at the races year in and year out you cannot achieve this with flat stake betting.the margins on win betting are too tight.bettors must look for more innovative ways of betting.in todays enviroment the only way to make consistent long term profits is through effective staking strategies.winning at racing is acombination of picking enough winning selections and an appropiate means of betting the selections.

becareful
15th October 2002, 02:45 PM
Ok, heres my more detailed thoughts/ramblings on the subject of Selections v Staking v Psychology. Obviously doing any one of them badly can ruin you so they are all important, however if I had to rank them I would do so as follows:

1. Selection System/Methodology
I rank this first as it is the foundation of becoming a successful punter. As has been said numerous times before no amount of creative staking can save a system that does not show a profit on the basic selections. Similarly it doesn't matter how good your psychology is you still will not win if you can come up with selections that show a profit. Also I believe it is the hardest one of the three to develop so it probably deserves to be the area you should concentrate on initially (no point wasting time working on the other two if you can't crack this one - find another hobby/job instead).

2. Psychology
Once you have a selection system/method that works then the next most important thing is getting the psychology right (I am still working on this). To a certain extent the psychology required is similar to that needed for share trading. I think you need to be able to ignore the actual money involved and just focus on your plan/system - it shouldn't matter whether you are $1000 up or $1000 down for the day you need to treat each bet on its merits. As Equine Investor said you need patience, self discipline and faith in the system otherwise you can cost yourself a lot of money through either over aggressiveness (not enough patience/self discipline to wait for the right bet) or being too cautious (not having enough faith to place a bet your system has indicated is a valid bet).

3. Staking
In my book staking comes a long last to the other two. As I have mentioned many times before I don't believe any staking plan can improve your POT in the long run so the main purpose here should be to insure your bets are sized appropriately to maximise the return. This means making sure your bets are not too big otherwise you will destroy your bank if you hit a run of outs. If the bet size is too small you will miss out on potential profits from your winners. I personally use a 1% of bank rule (so if my bank is $10000 then the maximum bet is $100 - if my bank falls to $9000 then bet is $90, if it increases to $11000 then bet is $110). The 1% comes from the fact I am generally betting on 10-1 chances to win - obviously you need to work out a % appropriate for your type of betting. Also if you are betting with TAB with larger bets you also need to consider how much you can bet into the pool without affecting the odds too much.

In my opinion the reason most betting systems fail is due to factor 1 - the wrong selections. A system itself cannot fail due to Psychology (although an individual punter may lose money with a winning system due to bad psychology). Many systems try to use elaborate (or simple) loss chasing staking plans to make a profit and inevitably fail in the long run, however the real cause of the failure is not the staking system but rather the fact the selections do not make a profit at level stakes (the staking plan just hides the problem until the inevitable crash).

becareful
15th October 2002, 02:55 PM
DarkyDog,

I have never seen the details of "The Power of 10" but I totally disagree that you cannot make a profit at level stakes in the long run. It may take a lot of work to keep your system/method fine tuned but I believe it is certainly possible to average 10% POT or higher year after year. Mathematically speaking staking plans will give you the same POT in the long run as level stakes betting (the only exception is if you increase bet size on bets that have a higher expected return - ie. the larger the "over" the larger the bet) so I just don't see the point.

The Phoenix
15th October 2002, 07:47 PM
In principle, I tend to side with Becareful.

The most important element is the Selection System/Methodology. If the Methodology does not make a level stakes profit, no matter what one does with a staking plan there is still the potential - ie risk - of being wiped out due to a losing seqence of bets. If the Methodology makes a level stakes profit, provided care is taken in terms of level of bets you WILL make a profit. In addition, as PPM have highlighted on a number of occassions, staking plans do not improve the POT.

If you can not make a level stakes profit over a long sequence of bets, it is my opinion that you will never make profit using a staking system, all you will do is delay the inevitable realisation that the method is wrong.

In order to improve my own methods, I review them without fail every week, ie they self learn. As a result, some of my methods have returns in excess of 44 % POT over tens of thousands of dollars of wagers.

The second element and just as important at improving the profitability is the Psychology. The key to this is still related to methodologies and in saying this I mean the thought process/psychology of chosing whether or not to lay the particular bet. An example that comes to mind was Avon Jewel in Brisbane two or three weeks ago. My methods picked the horse as a bet however given that the horse had run at the track 5 times for no wins and no places, I left the bet alone. The horse won but over the long term those final choices whether to bet or not have saved me far more money that they have cost me. In fact, over the past 9 months, only two of the possible bets left alone have won for a saving far exceeding the dividends.

The Psychology issues seem to often be approached in terms of "will to bet" whereas I approach them more from the point of "reason why my selection should be bet". In saying that, for me a good bet is a method/system selection that seems to have a good chance of winning either because it has run a good race in a fast time already at the distance (my preferred final review) or for some other potential positive reason. A bad bet is one that afterwards you can look back as say "I was not confident that it had a good chance because xxxxxxx" and in Psychological terms, we should always learn from those decisions.

The staking for me is a non issue as I always total the days results in terms of profit or loss however treat each method as a totally seperate entity on level stakes.

regards

The Phoenix

Bhagwan
16th October 2002, 03:49 AM
If you are a real systems buff there is a service call Winform GTS by Marcel Plant & accociates.

For $1000 you get back data of 20,000 races .
With this you can start testing any criteria & most systems criteria that one may want to check , to see if it`s got legs or not .
It has a provision to greate & save your own system criteria ,if you subscibe to their daily form data service , you can have your system scan the entire days various meetings to find your systems selections within seconds.
It will show ,profit & loss , strike rate , etc.within seconds, it can do more in one hour than someone could do in a month.

For another $3000 a year they supply a daily form service plus update past results to add to you data .
If one makes money out of any system idea ,then I guess its not expensive , no matter what system you come across they all have winning $ loosing months.
You have to be mentally prepaired for that .

Dr Pangloss
16th October 2002, 09:08 AM
Very interesting (and different) contributions all round. It seems that:

SELECTION methodology will determine expectancy (ave win div./strike rate etc)

STAKING (position sizing) will determine net profit

PSYCHOLOGY will determine the degree of integrity with which SELECTION and STAKING are applied.

Bhagwan

The TRB product is the most impressive system testing tool I have had the good fortune to trial. However, excellence has its price, and the software itself was somewhat difficult to manage for a computer dunce such as myself.

In searching for alternatives I came across http://www.ird.com.au a little while back and was particulary impressed. Wondered if you or anyone else had any experience with this product.

Thanking you in anticipation.

<font size=-1>[ This Message was edited by: Dr Pangloss on 2002-10-16 10:11 ]</font>

Tony
21st October 2002, 09:10 PM
Dr Pangloss, I am presently using the two day trial of the CIRA system tester you mentioned above. Perplexing thing is that I can't replicate results found with the same rules, or so close as to be not funny, ex the Price Predictor program.
Other comments are that it is a bit slow and laborious only giving the option of scanning 3 month blocks at a time and the lack of a spread sheet facility (unless I'm missing something) for later, more in-depth analysis is a minus. Interesting though, thanks for the link.

Dr Pangloss
21st October 2002, 09:33 PM
Hello Tony

Attempted to contact Mike Mayo last week without success. Thought the project may have expired so glad to hear the two day trial is still available.

There is no doubt CIRA is inferior to TRB (GTSi) but the cost comparison compensates. The integrity of the back test results compared to PPPro is a worry. Don't know what to think. (Data sample difference?)

What version is CIRA up to now? Does it have the TAB update screen working as anticipated February last?

TIA

Tony
21st October 2002, 09:43 PM
Using version 1.03. January 2002 is as recent as data gets and no, no sign of TAB update screen.

Dr Pangloss
22nd October 2002, 08:04 AM
Sounds like it may have drawn the last breath.

How does the parameter/filters settings that are available with CIRA stack up against PPPro? When I had a look at PPPro it didn't appear nearly as comprehensive as either CIRA or TRB software.

Tony
22nd October 2002, 09:52 AM
PPPro has nearly all the things I would be interested in. There are no end of extra parameters covered in this CIRA program, and winline too from your experience, but to my quick perusal I think most are so pedantic and specialised as to be maybe not too long term relevant. ie Easy to make a system that has historically worked with all this stuff, but inversely difficult to make it happen in real time.
The problem with PPPro, apart from my inability to generate profit/time graphs from excel, is that they have only made available a little over 12 months historical SP price data which is not enough to base any thing on.

darkydog2002
12th December 2002, 11:34 AM
the killer of all is plain and simple greed.

Equine Investor
12th December 2002, 09:17 PM
On 2002-12-12 12:34, darkydog2002 wrote:
the killer of all is plain and simple greed.


darkydog2002, couldn't agree more!

7th January 2003, 06:36 PM
Reply to becareful

re: 2. Some systems are developed/sold by people with good intentions who honestly believe they have found a working system but either they don't understand statistics well enough to realise that it will fail in the end...

I agree with the statement that many systems are developed by picking the best criteria from a few years of stats and then are passed off as the best system ever, but please don't imply that that this sort of system would obviously have to be a losing system.
Almost every system has to rely on past stats, specially those that are based on form, because after all, form is based on past results and stats.

<font size=-1>[ This Message was edited by: quapi on 2003-01-07 21:01 ]</font>

woof43
7th January 2003, 07:46 PM
Anyone that has developed a system on past data, would split the data in half then create the proposed system then run that system on the other half of the original untouched data and compare results....