Consistency

[wesmip1]

I just thought I would post an example on how important consistency is in system development. Here is an example of a bad system :

Month 1 : 177 Bets with a Profit 20%
Month 2 : 133 Bets with a Profit 25%
Month 3 : 125 Bets with a Profit 22%

From here a lot of people would start betting the system

Month 4 : 115 bets Loss of 5%
Month 5 : 126 bets Loss of 4%
Month 6 : 110 bets Profit of 15%
Month 7 : 150 bets Loss 50%
Month 8 : 117 bets Loss 30%
Month 9 : 105 bets loss 40%

Alternatively I also have a system that does the following :

Month 1 : 5 bets with a Profit 400%
Month 2 : 7 bets with a Profit 300%
Month 3 : 9 bets with a Profit 122%
Month 4 : 3 bets with a Profit 105%
Month 5 : 5 bets Loss of 4%
Month 6 : 5 bets with a Profit 240%
Month 7 : 6 bets with a Profit 350%
Month 8 : 4 bets with a Profit 100%
Month 9 : 5 bets with a Profit 520%

I prefer to use this system altought it doesn't have many bets becasue it is more consistent on the results. ie 8 months in profit out of 9 whereas the other is only profitable 4 out of 9 months.

I would say you need a minimum of 2 years testing before you can declare a system consisitent.

Good Luck.

[crash]

Your demonstrating [correctly] that the more bets there are the higher the loses ....very 'consistently' !

The confusion punters make is they see the 'odds' they take for a bet as true odds and then jump to seeing them as the same as true chances. Of course they are not the same and the more bets they have the worse the posibbility of profit. Why? Because odds ain't odds.

At best, if there are 10 starters in a race true odds are really 10/1 before they jump. True chances are actually evens [a bet wins or it loses]. I'll take a system with few selections over one with many every time and take my 'chances'.

[Mr. Logic]

As someone who has an understanding of statistics, a betting system purely based on automatic selection rules that has shown a profit after 435 bets has a better chance of succeeding long term, that is over thousands of bets, than a betting system that has shown a profit after just a paltry 51 bets.

Does it matter that the 435 bets are spread over three months or two years? I think not. Is the second method statistically better because the 51 bets are spread over nine months? I think not.

In addition, in the second and preferred example given below, I would know that it is a statistical fluke, as making hundreds of per cent profit in a negative expectations game in markets of 120%, which betting on horses is, is IMPOSSIBLE over the long term.

Mr. Logic.

I just thought I would post an example on how important consistency is in system development. Here is an example of a bad system :

Month 1 : 177 Bets with a Profit 20%
Month 2 : 133 Bets with a Profit 25%
Month 3 : 125 Bets with a Profit 22%

From here a lot of people would start betting the system

Month 4 : 115 bets Loss of 5%
Month 5 : 126 bets Loss of 4%
Month 6 : 110 bets Profit of 15%
Month 7 : 150 bets Loss 50%
Month 8 : 117 bets Loss 30%
Month 9 : 105 bets loss 40%

Alternatively I also have a system that does the following :

Month 1 : 5 bets with a Profit 400%
Month 2 : 7 bets with a Profit 300%
Month 3 : 9 bets with a Profit 122%
Month 4 : 3 bets with a Profit 105%
Month 5 : 5 bets Loss of 4%
Month 6 : 5 bets with a Profit 240%
Month 7 : 6 bets with a Profit 350%
Month 8 : 4 bets with a Profit 100%
Month 9 : 5 bets with a Profit 520%

I prefer to use this system altought it doesn't have many bets becasue it is more consistent on the results. ie 8 months in profit out of 9 whereas the other is only profitable 4 out of 9 months.

I would say you need a minimum of 2 years testing before you can declare a system consisitent.

Good Luck.

[partypooper]

Whilst I have come close to a "run in" a couple of times with Mr. logic, in this this case he absolutely "SPOT ON" say no more we'll just wait ey?

The other question of splitting the sytem results down to "Months" is an interesting one, i.e. an argument could be made that certain months do not favour certain methods of selection etc etc.

But overall is the way to go I reckon, warts an' all if there's still a profit,.... ok

[crash]

Mr Logic is basically saying [?]: if you toss a penny often enough you'll end up with about an even score for both heads and tails because the odds are evens. It's only with relatively few throws where the scores can be wildly different as in the absurd profits shown above for the 'few bets system'.

What many throws [bets] demonstrate in a negative expectation game such as horse racing [because the odds overall are against the punter], only one result can possibly occur ...LOSS!

The more bets we have the more probability kicks in and favours the odds [against].

Naturally, in a negative expectation game the less bets we have the better. Why? Because although there is also a chance of showing a loss, a chance of showing a good profit exits too, even if the odds overall are against us.

The challenge for the long term System, is to produce circumstances where overall the odds are genuinely in our favour [turning horse racing into a positive expectation game]. Not and easy task at all !

[wesmip1]

I only posted the second system for the first 9 months of the first year I tested.

I have tested it over 30 months on paper and 5 months so far live.
Out of the 35 months I have had only 4 losing months, most months have been over 300% profit and most have had between 5-8 bets a month.

The point was you can have a system that has 200 bets a month and it can be profitable over 3 months and people think 500 bets is enough of a test base. But you also have to take into consideration the time over which it was tested as the testing period may have had an anomoly (especially in only 3 months).

Overall results are below :

272 bets
WIN $911.70
PLACE $308.50

Betting $100 a unit has returned $91,170 in the past 3 years ... not a bad effort if you ask me.

Good Luck

[wesmip1]

partypooper,


But overall is the way to go I reckon, warts an' all if there's still a profit,.... ok

This can be a problem if you want to live off the cash. You need it to be consistent as well. Also us investors hate to lose so we normally throw away a system when it has a bad run...

Good Luck.

[crash]

Overall results are below :

272 bets
WIN $911.70
PLACE $308.50

Betting $100 a unit has returned $91,170 in the past 3 years ... not a bad effort if you ask me.

Good Luck

272 bets at $100 a unit is $27,200 outlay and for a return $91,170 is well over 300% POT [or over 100% POT consistently per year for 3yrs. running].

'Not a bad effort' [?] wesmip1, you are being modest. Your the greatest punter the world has ever seen!

What system are you about to try and sell us?

:-) :-) :-) :-) :-)

PS. You should join the 'sudden death' comp. Winning will be a walk in the park.

[wesmip1]

crash,

Regardless if you believe it or not this is a system i follow and I am making good money on it.

I didn't start this thread to argue my results, I started it to show people that a system needs to be tested over a longer time frame than 500 bets as it all depends on how many bets you have a day. You need to test it over a long time period as well.

Once again the non-believers who can't do it argue the irrelevant points.

Good luck.

[crash]

Ok, I believe you.

A system you have devised that produces over 100% profit per year [I'd increase my stakes per bet If I was you as your barely living at $20k odd profit a year]. How much for the system ?

I want in !

[wesmip1]

Crash,

Thanks for your input but obviously I am not sharing (nor was that the point of the thread)....

Good Luck.

[moeee]

Betting $100 a unit has returned $91,170 in the past 3 years ... not a bad effort if you ask me.

Good Luck

Have you still got the money Wesmip or did you buy yourself something nice?

[wesmip1]

moeee,

I have only been doing the indicated system live for the last 5 months but it is up a lot since then and the money is in the account.

Hoping to get to 5 digit profit this year and on target so far. I follow numerous systems I have developed myself.

Good Luck.