Systems

How long do you follow a system for before you throw in the towell ??
I've had my first crack at my own system much to pigheaded to follow anyone elses and i'm up a heap in so far 5 weeks...
I'd hate to throw it in if it loses a few and then find out it's made a come back and i haven't been on....
How do you know when you've been just lucky or your onto something..

[kenchar]

I would suggest if it's going that good just keep doing it.
If one day you find you are lining up at the Salvo's then I suggest you change your system. :smile:

[Nag Man]

Maverick I know you cant give away too many secrets if its working that well, but whats the basics is it single bets, staking 2 or 3 in a race, or novelties?

[Chrome Prince]

Maverick,

Work out your strike rate and then the longest losing sequence.

If your system exceeds the expected longest losing sequence given the strike rate, stop betting and have a rethink.

[crash]

Hi Chrome,

I have been trying to work out a method to arrive at an 'expected run of outs' for a [any] system.

Perhaps adding all winner's payouts and dividing by the total number of runners [inc. losers] to come to an average odds for total bets ?

Expected run of outs would require average odds [converted to %] to arrive at an expected run of outs.

Am I way off track here ?

What method would you [or anyone else for that matter] recommend ?

Cheers.


<font size=-1>[ This Message was edited by: crash on 2004-06-18 07:13 ]</font>

Nagman- Its a bit of both..i've detailed it in the racing forum section. Have a look and tell me what you think..

Chrome-prince. I've had my first loss with the system yesterday.
For an example if i lose the next 5 times..
then win on my 6th time..my losing streak would be 6....so are you saying if i lose another 6 in the furure i should stop and rethink the sysytem ?? of course i'd be taking into account the actual stake lost as well..

Thanks for the advice guys ...I'm still refining the process.

[Merriguy]

Mav, Bghawan is great at giving expected run of outs --- so watch for a posting from him :smile:

However your experiment, which I have followed from your thread on the other forum, is complicated in that it mixes straight outs and exotics. I would be doubtful if it would be easy (possible?) to suggest a run of outs in that situation.

[Chrome Prince]

On 2004-06-18 09:09, Maverick wrote:
Chrome-prince. I've had my first loss with the system yesterday.
For an example if i lose the next 5 times..
then win on my 6th time..my losing streak would be 6....so are you saying if i lose another 6 in the furure i should stop and rethink the sysytem ?? of course i'd be taking into account the actual stake lost as well..


Maverick,

This is a little different. It really has nothing to do with your longest loss on a day or your profit/loss at the time.

Unfortunately, we aren't allowed to post links on the forum, but search the internet for a losing sequence mathematical calculator.

Work out your average dividend.
Work out your overall strike rate.

A calculator will tell you what your maximum losing sequence should be.

In my opinion, if the losing sequence goes longer than the average, not the maximum, I'd cut back my bets a little to save the bank.
You may miss a good winner, but it could save you from disaster in the long-run.

[Chrome Prince]

On 2004-06-18 06:53, crash wrote:

Hi Chrome,

I have been trying to work out a method to arrive at an 'expected run of outs' for a [any] system.

Perhaps adding all winner's payouts and dividing by the total number of runners to come to an average odds for total bets ?

[i] Exactly

Expected run of outs would require average odds [converted to %] to arrive at an expected run of outs.

Am I way off track here ?

What method would you [or anyone else for that matter] recommend ?

Cheers.

<font size=-1>[ This Message was edited by: crash on 2004-06-18 07:13 ]</font>


A little off track, only for the longest losing sequence expected, otherwise spot on.
Your calculation will give the expected strike rate, not the longest losing sequence and number of expected losing sequences for each loss.

See my post above to Maverick.

[bren250]

Hello Chrome,
I too am very interested in finding a loosing sequence calculator. Have some systems that all seem to drop off this time of year then pickup in August.
I tried doing a search on the web but found nothing useful.
Any more clues to where I can find someting would be greatly appreciated.

[Chrome Prince]

Guys,

Perhaps to keep safe with management, post your strike rate and average dividend (shouldn't give too much away), and I'll post the expected stats here.

[crash]

Thanks Chrome,

90 bets [count as 1 unit each]
27 wins
130.6 units return

Longest run of outs has been 9, but that is really meaningless.

Cheers.

<font size=-1>[ This Message was edited by: crash on 2004-06-18 17:45 ]</font>

[Chrome Prince]

On 2004-06-18 17:42, crash wrote:

90 bets [count as 1 unit each]
27 wins
130.6 units return

Longest run of outs has been 9, but that is really meaningless.



<pre>For a Win % of 30.0% over 1,000 trials, the expected LLS are ...

Exp # of LS of length 3 is 103 Prob of one or more LS is 100.00%
Exp # of LS of length 4 is 72 Prob of one or more LS is 100.00%
Exp # of LS of length 5 is 50 Prob of one or more LS is 100.00%
Exp # of LS of length 6 is 35 Prob of one or more LS is 100.00%
Exp # of LS of length 7 is 25 Prob of one or more LS is 100.00%
Exp # of LS of length 8 is 17 Prob of one or more LS is 100.00%
Exp # of LS of length 9 is 12 Prob of one or more LS is 100.00%
Exp # of LS of length 10 is 8 Prob of one or more LS is 99.98%
Exp # of LS of length 11 is 6 Prob of one or more LS is 99.76%
Exp # of LS of length 12 is 4 Prob of one or more LS is 98.47%
Exp # of LS of length 13 is 3 Prob of one or more LS is 94.54%
Exp # of LS of length 14 is 2 Prob of one or more LS is 86.83%
Exp # of LS of length 15 is 1 Prob of one or more LS is 75.70%
Exp # of LS of length 16 is 1 Prob of one or more LS is 62.76%
Exp # of LS of length 17 is 1 Prob of one or more LS is 49.84%</pre>

[bren250]

Hello Chrome,

Some of my systems. Data more than 1 year. Been following for 6 months.

Name Str/rate Av Win Div Av div
1 35 $3.31 1.16
2 43 3.59 1.55
3 46 4.04 1.85
4 31 4.41 1.37
5 40 4.83 1.97
6 35 3.88 1.33
7 24 4.90 1.15
71 33 4.19 1.39
72 30 4.37 1.30


Some don't have many selections each year but some have quite a few.

Would appreciate it if you could tell me what search words to use to find a calculator on the web.

[Researcher]

Losing Sequence Formula

Number of bets=1/(1-strike rate)n
n= number of losing sequences.
Strike rate is expressed as decimal.
/ means ¡§divide by¡¨
How many bets am I likely to have before I experience a losing sequence of 5. Strike rate is 25% (ie 0.25)
Bets=1/(1-0.25)5
=1/(0.75)5
= 1/(0.75x0.75x0.75x0.75x0.75)
=1/0.237
=4.2
ie 4 bets

There is also a formula for the longest losing sequence, but can¡¦t find it at the moment.

[crash]

Thanks Chrome,

Looks like a run of 9 outs per 100 bets is roughly typical around my strike rate.

I can live with that and might start upping the anti on bet size a little, though not nearly enough bets yet for the strike rate to be taken too seriously.

cheers.

<font size=-1>[ This Message was edited by: crash on 2004-06-18 20:26 ]</font>

[Dan]

The formula I use is as follows,It's easy to use once you've done it a couple of times.

Predicting the number of losing runs of a given length, against a given strike rate:

X = N x S^2 x (1 - S)^L,

Where:-

X = Numner of losing runs of length L
N = Number of results.
S = Strikerate (as a decimal fraction).
^ = To the power of.
L = Length of losing run you wish to investigate.

So for example, if you have 100 results and a strikerate of 30% and you want to know how many losing runs of 4 you should have, the calculation becomes:-

X = 100 x 0.3^2 x (1-0.3)^4
= 2.16

and for losing runs of 10,

X = 100 x 0.3^2 x (1-0.3)^10
= 0.25

This last calculation means that you have a 25% chance of 1 losing run of 10 in your 100 results, so if you continue with this system you would expect to have 1 losing run of 10 at 400 results.

It's interesting to note that it doesn't matter how high your strikerate is (except for 100%), if you work to enough decimal places you will always get a positive value, thus indicating that however small the probability, there is no length of losing run your results cannot theoretically contain.

[Mark]

"The academics have taken over the assylum", expected run of outs????, fantasy stuff. I'd be more interested in expected run of ins. Useless information. What do you do when you come to the limit of your so called expected run of outs?, plonk what's left of your bank & confidence on the next selection? (does it know about your expected run of outs?), and if it gets beaten, what then ???

[davez]

On 2004-06-19 13:08, Mark wrote:
"The academics have taken over the assylum", expected run of outs????, fantasy stuff. I'd be more interested in expected run of ins. Useless information. What do you do when you come to the limit of your so called expected run of outs?, plonk what's left of your bank & confidence on the next selection? (does it know about your expected run of outs?), and if it gets beaten, what then ???



I have to agree with mark here, what actual value do these calculations have?

Any realistic punter knows he's going to hit a bad streak now & again, some worse than others, & no amount of planning or calculations can avoid this.

I would suggest instead of focusing on how best to protect your bank when the inevitable happens, with for example a 'stop loss' of some sort, that is suspend betting once a certain amount of losers in a row has been struck & start again once a winner strikes & the horror is over.

ps - which was advice I did read on this forum some time ago.


<font size=-1>[ This Message was edited by: davez on 2004-06-19 14:00 ]</font>

[Chrome Prince]

On 2004-06-19 13:08, Mark wrote:
"The academics have taken over the assylum", expected run of outs????, fantasy stuff. I'd be more interested in expected run of ins. Useless information. What do you do when you come to the limit of your so called expected run of outs?, plonk what's left of your bank & confidence on the next selection? (does it know about your expected run of outs?), and if it gets beaten, what then ???


Hi Mark,

Never been called an academic or a lunatic before but I'll take both on board :lol:

I don't think it's useless information, but a GUIDE - that's all. I'd be dumping my system long before I hit the longest run of outs, not plonking on the next selection.

We all know that horseracing is not like other games of chance, there are variables to the chances, so the LRO is a guide only to what the mathematical expectation should be. Therefore if my system hit anywhere near the average losing sequence - I'd dump it. But that's just my take on it.

[Dan]

The reason for me is it gives me an idea of what percentage of my bank I can safely put on each selection. I'll take the maximum expected run of outs and then at least double it and divide it into my bank to give me an idea of percentage to risk. There's alot of tipping services that talk about risking 5-10% of your bank on each selection, which to the beginner sounds ok but to us, we know is just a disaster waiting to happen.

[Mark]

CP

I wish lunatic was the worst I'd ever been called :lol:.

Dan

Yes you are right, use it as a guide then double it, as it's still only a guide. Thanks for the input.

[crash]

misprint

<font size=-1>[ This Message was edited by: crash on 2004-06-21 05:48 ]</font>

[crash]

Mark,

I think Chrome was pointing out the usefulness of mathematical 'probability' as opposed to 'possibility' [what you where giving examples of] as a punting tool.

Expected runs of outs for a particular SR is probability at work which can then be used as a punting guide far more profitably than relying on possibility [the 100/1 shot].

The same 6 winning Tatslotto numbers might possibly come up twice [or more] in a row, but it is highly improbable.

Cheers.

<font size=-1>[ This Message was edited by: crash on 2004-06-21 06:01 ]</font>

[Mark]

Maybe that's why my current "run of ins" is 53 days in a row. I think differently to others, expecting to win & not thinking about what ifs of losing runs.