Systems

[H]

Just wondering what everyone's thoughts on whether a system has a use by date. What I mean is that I have a system that showed a nice POT over a period of 18 mths but for the last 12 mths has run at loss. Obviously factors such as the speed of the race, different training methods etc that a system cannot generally pick up will impact the figures. But is there a point in time when a system basically becomes redundant??

[umrum]

yep,...........

as soon as you pay for it!

:wink:

cheers
umrum

[H]

The hip pocket is definately burning, and even though it shows all the traits of a bought system it is really a combination of various rules that a few mates and I use. It's close to getting chucked in the bin and starting again. I have only recently started keeping a database so I can't comment on the area that is falling over in. Suppose it is a case of more research.

[Felicity]

I have examined almost every system sold in the past 5 years using a database of 13 years Oz racing.

Just as Ian Barns (RIP) showed year after year in Punter's Choice I have found no system that lasts more than a month or two.

A system is easily generated by going back over the past year or so and finding some "Platinum" characteristic that sits within the 'group' that made up the winners. What the purveyors NEVER do is split the d/b 80/20, run the research on the 80% and then test against the 20% so that the results aren't already contained in the research data.

Roger Biggs calls these short term aberrations "data clumping".

There's no surer way of going broke than following a 'system' over the medium to long term.

[umrum]

gday flick,

is there any particular reason why a system works straight away. I can understand a system having a run of luck and then outs but why would it work early and then drop off.

cheers
umrum

[Felicity]

Umrum, if you get involved in a two-up game in the middle of a run of heads and decide to ride the run ... is this a basis for a 2-up system ?

Have a look at PC and check from yr to yr how many systems disappear without trace. I could quote several but won't because of the sensitive nature of this forums managers.

Roughly 100% of systems depend on place. Now 5th can be anywhere from 0.1 lengths to 30 lengths beaten. Bong .... that's the sound of the penny dropping.

Now I've got to go and get his nibs dinner ready.

[umrum]

doesnt answer my question but thanks anyway.

cheers
umrum

[Bhagwan]

Hi H
I run the TRB GTSi data base & have created a number of profiable systems .

Most successful plans only seem to have 5-12 bets a month betting 7 days a week .
It can be frustrating .
I believe one needs a data base of at least 4 years .
What I do is drop any plan that has a dramatic loss in any 6mnth period.

I run approximatly 100 plans all at once , which is worked out in just 2 seconds after downloading the data for the day.
I have about 400 bets a month , with an average S/R of 26% with a POT between 10-36% each month.

If you feel up to it , I can run your plan through my 4yr data base , with that I can try & improve the results by isolating each rule to find out where it has been falling over.

You can email me on
bhk (at) bigpond (dot)net (dot) au
I have layed this out in this manner so I dont get hit with spam mail, you just rewrite this out in the normal known format.

I have noticed that plans with a low place percentage S/R of less than 44% should be a warning to tread carefully. It means to me the system has got potentialy week legs to pull potential winners.

Another tip is to delete any selections which are 2.9/1 or less, pre-post , the value long term, is not there.

[partypooper]

Bhagwan, not having a go mate, but it seems that you have found the Holy Grail that most of the contributors to thus forum say is impossible! i.e. a winning system (consistantly) not on past results but into the future. Anyway thanks for sharing some of your plans with us.

[Mark]

Sorry guys, can't help myself.
"Blah blah blah, systems don't work", in other words "I can't win therefore no-one can".

:roll:

[Chrome Prince]

Agree Mark.

It seems that most are resigned to the fact that winning using a system is impossible.
It's very possible provided several criteria are met.

a)The rules are logical and apply equally to several systems.
(hypothetical example: eliminate horses drawn worse than barrier 10....if this improves the strike rate and POT in completely seperate systems, then it's got legs, if not then you are probably curve fitting).

b)Your sample size and number of bets are huge.
The greater the sample size and number of bets you have as an example, the more likely the same trends will follow.
I've seen some posts saying things like 134 bets 29 winners - means little or nothing to me.Basing future bets on minimal stats like this is asking for trouble.

Felicity:

Punters Choice do some good work, but they are not the Grail as some seem to think.
Firstly, they monitor only about 10% of the actual services out there. Also they have monitored many systems which still show a profit after 10 years, so I don't know how you arrived at your conclusion.

Your 80/20 rule I use only it's the other way around...I'll run a system for 6 months and then compare it over a different 12 months, then 2 years, then 4 years.

As to never finding a system that wins more than a couple of months - you aren't looking hard enough :grin:


<font size=-1>[ This Message was edited by: Chrome Prince on 2003-12-15 23:48 ]</font>

[crash]

Interesting thread.

Applying rules that produced winners in the past to horses running in the future is problematic. The horses change as do conditions. Every three or four years we are dealing with all new runners [not exactly but you get my drift] that have contibuted nothing to the results from the past. This fact may help to explain why some systems drop off apart from the usual too many players argument.

I have never used public systems and among the few I do use most still eventualy fall over. Constant re-invention is required. Any system that has worked in the past then failed, still has the potential to work again when future condt. mirror [or nearly] it's past working parameters. Knowing when that might be is the problem. Better to re-invent.
My new systems go straight into play without checking past results using smallish bets which grow if the system performs. If the system is based on sound principals it won't take your leg off. So why not play it and see what happens ?

There is still one system that has stood the test of time that works and that is good old fashioned handicapping. It seems that it is becomming less and less popular due to the application required to turn it into a winning 'system'. It does have the advantage of lasting a lifetime and improving with age as a reward for the personal application required. I can understand why though through time restrictions etc. hadicapping is not everyone's cup of tea.

Cheers.



<font size=-1>[ This Message was edited by: crash on 2003-12-16 07:53 ]</font>

[H]

Totally agree with CP in that it is possible to win using systems. The rules to my system are relatively straight forward and logical, but have not changed since day 1. I think it would be a matter to analyse each rule to dtermine appropriateness to current racing conditions.

[darkydog2002]

FELICITY.Agree wholeheartedly on commercial systems.You,d do better with Rogers golden and platinum pin systems.
Cheers.
Darky.

[partypooper]

PS. I didn't say that "I" don't believe that systems can work, I said that MOST of the contributors to this forum seem to think so. Therefore I'm glad that Bhagwan and others have proven it, as it gives hope to the rest of us!!

PPS. I AM winning, not a lot but I AM wining.

[sportznut]

Well, which is it??? WINNING or WINING??? :grin:

[Mark]

Party

I wasn't having a go at you, look at the times our posts went in, almost the same, and yours wasn't there when I was typing.
I was referring to the newcomer.

[Bhagwan]

I must say that nearly all the commercial systems one has to pay ,from a certain published magazine , have fallen over once put through my data, even when placed over the period that they say it was tested over ,so what does that tell you?

[kenchar]

It tells me they are the low scum I have always said they are.

Preying on people that usually can't afford it, and offering a down payment on a credit card and paid off over some months from the same card.

If I had been in a desperate situation when paying them for all different systems and then losing what I did on those systems I probably would have necked myself.

Actually it would be interesting if that has ever happened.

Cheers

[hermes]

Useful discussion in this thread. My two cents:

I've given up on entirely mechanical systems. I see all "systems" as selection tools and nothing more. I look for systems that throw up a good volume of selections and then I consider the selections in the context of each particular race.

The best "system" is the sound judgement that comes from watching lots and lots and lots of races and learning the races to stay out of and learning what is a good bet and what is not. Nothing can replace that. You must, in my opinion, consider the *context of the race* - purely mechanical systems will fail or draw back to break even.

There are inherent problems in and limitations to predicting the future from the past. The basic problem is that every race - every race! - is a totally unique event. There has never, ever been a race like it before and never will be again. Every Melbourne Cup is a totally unique event, regardless of the family resemblances between all the unique events labelled "Melbourne Cup".

This is why systems should be tested on live races and why, even then, judgement is necessary. It is also why testing a system on 20,000 races may be meaningless - because we are not betting on 20,000 races but on this particular race, and it is a unique event with unique factors.

Often my major system will indicate a good selection but I look at the race and think better of it. I'm getting better at this, I reckon. Same with the commercial system I purchased once. Look at the race. Nice selection, but no.

As for why systems prosper and then collapse as if they were shareware and expired after 28 days, I don't know, but it is an observable thing. I have no idea. Pluto going retrograde in Sagittarius?

But it doesn't happen to all systems. I pulled a little money-spinner from the drawer the other day and dusted it off. Two years ago it was running at a small profit before I got bored. Checking recent weeks, it's still making money. Not much, but consistent. No change in the strike pattern.

What factors make for a consistent system? What factors are eratic long term?

Cheers

Hermes

[Chrome Prince]

On 2003-12-16 22:33, hermes wrote:
There are inherent problems in and limitations to predicting the future from the past. The basic problem is that every race - every race! - is a totally unique event. There has never, ever been a race like it before and never will be again. Every Melbourne Cup is a totally unique event, regardless of the family resemblances between all the unique events labelled "Melbourne Cup".


Yes they are independent events as are all races, but look at common denominators.
No sprinters win Melbourne Cups, horses must have had certain number of starts etc etc etc.
Statistics cannot tell you the coreect finishing order for a race, but can tell you wheter your selection is value and a good bet.
Statistics are used in numerous applications in society, they are not meant to predict the exact future as such, but tell you of continuing trends which continue to be profitable year in and year out.

As an example: 9 out of 10 Australians will rely on the Government for retirement income.
This statement comes from statistical research. It does not mean that exactly 9 out of 10 people in your street will be in this position, it simply means that ON AVERAGE this will be the case.

Therefore, lets say I have a system that shows 30% POT with a strike rate of 45%, it doesn't mean I'll win every day or week, but rather come out ahead over a month or year.

As to handicapping, is this not contained within stats anyway?
Or perhaps the two go hand in hand.

Incidentally, there are purely mechanical systems mentioned deep in the archives of this very forum that are still producing profit to this day :wink:

<font size=-1>[ This Message was edited by: Chrome Prince on 2003-12-16 23:58 ]</font>

[crash]

Some good points Chrome. Without stats. Hadicapping would be lost in the woods. Hermes point though about handicapping what a system throws up is a good point too. A system that has a 45% SR might be improved by 5/15% by throwing out selections that for whatever reason handicap as poor value. This information can also be used to fine tune the system rules.
Value is the name of the game and by whatever means is what we should look for. I regularly dump horses that I think will win purely because they represent poor value. In the long haul backing such runners will produce a loss, meaning that an 'any price a winner' mentality is to be strenuosly avoided.

Cheers.

<font size=-1>[ This Message was edited by: crash on 2003-12-17 06:57 ]</font>

[becareful]

Question for the "non-system" folks! One of the main arguments from people who don't "believe" in systems (bah - next you will be telling me you don't believe in Santa!) is that you can't use the past to predict the future or that each race is unique and therefore no system can predict the results. The question is, therefore, why do you think your handicapping would work if a system can't? Surely your handicapping is based on past experience and the results of past races - if these are irrelevant then standard handicapping techniques are just as irrelevant as any system. So either you must concede that past results/experiences/trends ARE important and therefore systems based on these MAY work or we may as well all pack up and go home because this game is impossible!

In my opinion systems DO work, and they can work quite well provided they are based on sensible rules (eg. a system based on backing horses with Taurus in the name is obviously a load of Bull even though it would have worked well last week!).

So why do they stop working? In my opinion there are 2 main reasons:

1. They never actually worked in the first place. Backfitted systems sold by certain publications probably fall into this category! The system itself is not based on valid criteria and the fact it produced profit for a few months is nothing other than mere statistical co-incidence (or simply downright lying in the promotional material). Systems that rely on loss-chasing staking also fall into this category - they only show a profit because they haven't hit a losing streak of a certain length yet that will break the bank.

2. Overuse. Any system or handicapping relies on finding a big enough edge to overcome the bookie/TAB take and then make a profit. As we all know (I hope) when too much money is placed on a particular runner the price falls so the edge you have is reduced. If too many people are backing the same horse then the edge will eventually disappear. Of course they don't all have to be using the same system but if various systems are targetting similar characteristics then they will come up with similar selections (eg. lots of systems target last start winners so there is often no value left in these selections).

Combine these two reasons together and you will find very few published/sold systems work for any length of time. If they were valid in the first place and show a good profit then word-of-mouth will probably result in them becomming popular and their profitability will reduce.

Bottom line is if you want a really profitable system you will probably have to develop it yourself and then keep your mouth shut about how you are doing it!!!

[crash]

Becareful, I agree with what you are saying except for one point. Your belief that punters can be devided into Hadicappers and system users. It is a grey area surely. Most hadicappers also use systems and vis-a-vis. I have my system account and my handicapping account [system account this year is killing my hadicapping account] and often due to time restrictions, my system account is the only one used.
I maintain hadicapping [a system in itself but not in the sense that it is purely mechanical like most 'systems'] due to the intellectual challenge I need to maintain interest. Profit is the the no. 1 goal but not the whole aim as otherwise racing is reduced to nothing more than counting money [accounting]. The sport of Kings ? Hardly.

[stebbo]

Hello All,

Becareful's point about handicappers using the past is a very valid point. Even those people who "do the form" are using past events to try to predict the future. Even in a first start 2yo, "form analysts" use past events to predict the outcome... Which horse has the better breeding, which has the better trainer, which looks the best, etc, etc.

I also agree with Crash that the line between systems and handicapping is very thin. I use Price Predictor. A good proportion of my systems are based on either one of it's ratings systems. Surely this is handicapping? I'll agree that it's purely mechanical handicapping, because I mostly back the selections blindly, but it's still handicapping.

As for why most systems fail, I wonder whether MOST systems do in fact fail? I suspect that MOST systems will hit a terrible run of outs, and that if the betting bank isn't structured correctly they will break the bank, but does the system actually fail in the long term?

Earlier this year I came across a free system, which I particularly liked. I started betting it, and after 56 bets it had a horrible strike rate, and I had lost half my bank. As of last weekend, it's strike rate is 35%, flat stakes POT is 26%, and with judicial use of a staking plan, the bank is 2.5 times what I started with. (actual POT 35%)

So, after 3 months and 56 bets, I would have considered this "system" a failure. Today it's easily my best performing system - the one that I put my largest bets on each Saturday.

I have done a lot of analysis on the "cash flow" of my systems - graphing the bank has been extremely useful. I've come to the realisation that most systems will be "down" for long periods of time. Even my best system has already had two 3 month stretches where it has been below it's previous bank high. Not a long way below, but below all the same.

I can be a very stubborn fellow at times.. when I start a system I generally bet it to the last dollar.... This has saved 3 of my systems so far, and these three continue to perform well at present.

Cheers,
Chris.

[partypooper]

Re: mechanics. Approx. 30% of favs win. 60% placed, this figure doesn't seem to alter much (so they tell me) whether Sat only or any day. Also the same figures seems to hold up with actual favs. or pre-post favs. Now presumably the odds are arrived by a mechanical handicapping technique involving all known parameters. The resulting loss seems to be equally mechanical i.e. approx 9% on t/o (depending who you talk to)improved slightly with using top fluc.

So Crash it seems you are right, we're sunk without that "personal" judgement, especially regarding "VALUE"

[becareful]

Crash,

I certainly never meant to imply that punters are either handicappers or systems users. If most people looked at the way I pick my bets they would classify me as a system user BUT the systems have been developed using (in my mind at least) sound handicapping techniques. The system is just an automated way of perfoming that handicapping approach.

There are, however, some people who seem to believe that a fully automated selection system can never produce long-term profits because the system is based on previous events. My question was therefore how do they perform their handicapping without any reference to the past????

Several poeple have also mentioned using a system to produce selections but then using a manual process to decide if they should bet or not. My main betting now is all from systems I have developed and if the system says a selection is a bet I put the bet on without question. In the past trying to examine each selection manually to improve on the system has cost me far more than it has saved when horses I assessed as poor chances romped home at double-figure odds!

[hermes]

To clarify my contribution, stats on 20,000 races might show that, for example, last start winners from barriers 1-3 are good things. And of course there is a certain predictability to this that you can rely on. But every race is a unique event all the same, and while such runners might be good things statistically they might not be in any given race. That's my point.

The question to ask about any stat is: in what context? It is the context that is unique. You can narrow it down but the best instrument for accounting for the full context is human judgement, in the end. I reckon.

My main point is that any mechanical system can be improved by the application of wise judgement, especially judging which races to shun.

At least I see it that way because most mechanical systems I devise either fall over or break even eventually and the best way to push them into the black is not to throw more stats and filters at it but to apply some judgement. Departing from your "system" you might decide, as the race shapes up, scratchings are made and real money is laid down and the context changes, to increase a bet. Or pass the race. These deviations from the rules are where I make the money that pushes the mechanical system into profit.

I regard mechanical systems as guides. Most of them break even at best, long term. For me. There comes a point in the development of a system where throwing more statistical power at it won't achieve much more than drastically reducing the volume of bets. That's the point at which I think its best to drop the science and apply some well-honed human judgement which is best able to comprehend and synthesize the unique context in which the mechanically chosen selection is situated.

Hermes

[crash]

In two words Hermes, spot on !

The missing ingedient of all mechanical systems is 'reasoning'. Reminds me of when automatic cameras came on the market [yes I was there way back then]. Basicaly the rhetoric was 'it does the thinking for you'. No more exposure problems. However when confronted with a white cat in a black coal mine type situation, what does the automatic brain do, expose for the cat or the coal mine ? Reasoning is why Pro. photographers still use manual cameras and why even though I have a great automatic camera with all the bells and whistles, my old Nikon F1 manual camera and my hand held light meter are the first things I reach for to take a serious photo.

'Systems are great value but their final 'filter' with this punter anyway, will always be reasoning.

Cheers.

[becareful]

So Crash where was your "reasoning" yesterday at Sandown :lol: :lol: :lol:

Sorry - couldn't resist.

Hermes - I guess your human judgement is better than mine! As I said in the past when I have tried applying my judgement to my selections I end up missing good winners.

I am interested in your point that you use judgement to decide which races to ignore - surely in the majority of cases you could incorporate rules into your system to cull most of these for you anyway? eg. exclude races with fewer than X runners, exclude maidens, jumps, etc, exclude the race if more than X horses have a rating higer than Y, etc. All I am saying is that a lot of what people call judgement can be incorporated into a system quite easily. Of course there are the "gut feeling" type decisions that are harder - but unless you have been punting for lots of years (and I haven't) then gut feelings are often mistaken.

[crash]

Well Becareful I have been laid bare, I really can't find a 'reasoned' answer to your post now can I ?

Cheers.

[BettyBoop]

Oh BC you've done it again :roll:

[gumboot]

Having read this forum many times silently over a few months I finally open my big mouth.

The way to win on the horses is to have an edge over the mug punter..that edge is a knowledge of how races are conducted and how horses are assessed ...The Bookies know what they are doing when they set their prices..they have information that allows them to assess a horses chances of winning and set their prices accordingly.
That information is not top secret undercover stuff...most of it is available to us as well..if we knew what to take notice of!.
Forget horse names and numbers or which way the wind is blowing ..that's for the mug punters who kindly lose $ so we can win $.

Statistics can be twisted to suit any purpose so the only value in statistics is to use them to find the important factors.

The guys who come up with the data for the form guides and the ratings know what they are doing as well and I'm not talking about the tipsters in the form guides..I'm talking about the info that may seem dull and not important.

Why reinvent the wheel ?..the hard work has been done and the results of that work presented to us in the form of a form guide and a rating.

The guy who sets the weight for the horses does so according to a well tested formula and the guy who comes up with the ratings does the same...but they are human and any computer programs they use were designed by humans;so there is a human bias there that needs to be counterbalanced with facts.
Those facts are simple....how has the horse performed over its past few races and has it won in todays conditions or at that distance or at that track..use those simple facts to check on the rating and you are halfway to having a system that works well enough to pull a profit.
But wait ..there's more!...What types of races are good to bet on ?...maidens..not likely...filly and mares...I have enough trouble with human female erraticness without betting on a female only horse race.
Short races don't give the good horses enough time to get good position and long races give plently of time for upsets to occur..
Races with only a few horses don't pay places and races with lots of horses lower your chance of picking the winner.

Main venues attract better horses than provincial venues do and the better horses are more reliable in their performance.

As to why systems often fail after a certain length of time...well nothing stays the same ( thank god ),but goes in cycles..those systems that were thrown out the window because they failed for a few months may well work again right now.

The biggest problem with any system that intially works is the user...we find it so hard to be disciplined enough to stick to the system and betting strategy and we go and bet on unsuitable races and over-ride the system with our emotive hunches.

Who here is not guilty of that ?!... I am.

Any system that relies on past statistical factors such as barrier number,number of days since the last win or spell,last finishing position being 7th etc etc will show over a long period of time a cyclic pattern..just like tattslotto does.
Stick to the relevent facts and expert assesments of the horses and the results will be more consistant...as long as you are as well !.
Cheers to all.

[Merriguy]

For mine there is a lot of wisdom and practical nous in what you have posted, 'Boot. Thanks.

Does anyone know if the bookies all receive some sort of "centralised" info.? Whenever I go to the races they all seem to be relying on some such rather than individually setting a book. Is IAS on Saturdays the only place that we common folk can access that or similar info?

[crash]

Merriguy,

Bookies more or less use the tote prices from a little TV screen to set their book nowadays. Less than tote and the smaller the meeting the worse their prices. Bookies are quickly becoming irrelevant as there is no longer the punter numbers for them to set a competitive book. Pity.

cheers.

<font size=-1>[ This Message was edited by: crash on 2003-12-30 04:52 ]</font>

[Merriguy]

Thanks for the reply, Crash.

So what you are saying is that they more or less follow the 'common' wisdom of the weight of investment on the particular race?

While there are variations across the three TABs, and these obviously become of more interest as the race start time approaches, is there an agreed common starting point as it were? It would be great as a punter to have that indication (I know IAS gives that info, or similar, on Sats fairly early around noon); but is there a similar starting point for other days/races? Anyone?

[crash]

Merriguy,

No there is no agreed starting point, but the bookies watch each other. Their prices go up once they see the tote trend.

Most of my on course punting is at country tracks now, and I can only say forget the bookies in the country. I see prices that are half the tote price regularly, especially cup days when there are plenty of inexperienced punters around.

Cheers.