How long does it take to judge whether a system is working?

[FrogThimble]

Day 8:

Today's racing isn't over yet but I'm not going to try to play the last few races tonight because I've got other stuff I need to do. I played 11 races today bringing my total trades up to 111.

Today I was taught a brutal lesson by my own stupidity. My strike rate today was 10/11 yet I only basically broke even on the day. This was because I didn't have the confidence to let my Servants get on with it... so I intervened manually and blew 9% of my bank (28% of my profits) in one stupid trade... Lesson learned!!! After that I sat on my hands and let the Servants do their thing for the remaining 7 races that I played. So I ended the day on just 0.6% profit when I would have had 6.3% profit if I'd behaved myself properly today...

Page 8.jpg

Day 8.jpg

Today's strike rate: 91% (10/11)
Overall strike rate: 74% (82/111)
Today's profit: 0.6% growth of my bank (£14.08 to £14.17)
Overall profit: 41.7% growth of my bank over 8 trading days (£10.00 to £14.17)

[FrogThimble]

Day 9: 13 trades today taking the total up to 124 trades. The 1p win at the end of the day was an error - that was me exiting pre-off a race I didn't really want to trade on. I've included it in the stats though as it was still a trade.

page 9.jpg

Day 9.jpg

Today's strike rate: 77% (10/13)
Overall strike rate: 74% (92/124)
Today's bank growth: 5.5% (£14.17 to £14.95)
Overall bank growth: 49.5% over 9 trading days (£10.00 to £14.95)

[FrogThimble]

Day 10:

Almost no data to add today. I was out all day involved in the local elections campaign putting leaflets through doors... in the rain... Great fun.

So I only played the races from 6pm onwards and only 2 of the evening races met my criteria. So that nudges up my total trades up very slightly to 126.

List 10.jpg

Day 10.jpg

Today's strike rate: 100% (2/2)
Overall strike rate: 75% (94/126)
Today's bank growth: 1.3% (£14.95 to £15.14)
Overall bank growth: 51.4% over 10 trading days (£10.00 to £15.14)

[FrogThimble]

Day 11:

15 trades taking the total up to 141. Still early days but I am now past the equivalent point at which my previous system collapsed anyway.

List 11.jpg

Day 11.jpg

Today's strike rate: 80% (12/15)
Overall strike rate: 75% (106/141)
Today's bank growth: 5.1% (£15.14 to £15.91)
Overall bank growth: 59.1% over 11 trading days (£10.00 to £15.91)

[FrogThimble]

Day 12:

15 trades today bringing the total up to 156 trades.

list 12.jpg

Day 12.jpg

Today's strike rate: 87% (13/15)
Overall strike rate: 76% (119/156)
Today's bank growth: 7.4% (£15.91 to £17.08)
Overall bank growth: 70.8% over 12 trading days (£10.00 to £17.08)

[FrogThimble]

Final update:

Day 13... 19 trades today and I totally messed up.

Day 13.jpg

The big drop was due to another stupid intervention on my part rather than the system... but the system can't possibly work if I can't make myself play by my own rules. So I'll go and think about my self-discipline (or lack thereof)...

Anyway, that's the public part of my experiment over with.

[xitian]

Can't you automate it somehow? A system isn't really a system if there's discretion involved. At the same time if discretion is not require for it to work then in theory you should be able to convert it to a set of rules which you can try to automate?

Even if it needs a bit of configuration at the start of the day or whatever, if at least the execution can be automated then that would be worth it.

[FrogThimble]

Can't you automate it somehow? A system isn't really a system if there's discretion involved. At the same time if discretion is not require for it to work then in theory you should be able to convert it to a set of rules which you can try to automate?

Even if it needs a bit of configuration at the start of the day or whatever, if at least the execution can be automated then that would be worth it.

It was automated... but... that didn't stop me interferring. Even though the system was (mostly) producing a slight profit I couldn't stop myself from sometimes making manual interventions if I felt something was going against me. All of the biggest dips and biggest leaps in my graph were due to interventions. All of the small upwards and downwards movements were due to the automation doing what it should have done.

The problem was with me rather than with the system. It's no big deal with a £10 bank but image how I'd have got on with a bigger bank that I actually cared about? More interventions and more significant losses, I think.

Anyway, over the last few days I've mostly been studying the movements of the markets and I think I've noticed another huge pitfall to my approach... the biggest flaw possible (one so big that I'm amazed I didn't spot it before) really so I'm rather glad I didn't persue this any further. I'm looking at the strengths and weaknesses of a totally different in-play idea now but just in the early stages of testing it.

[FrogThimble]

Can't you automate it somehow? A system isn't really a system if there's discretion involved. At the same time if discretion is not require for it to work then in theory you should be able to convert it to a set of rules which you can try to automate?

Even if it needs a bit of configuration at the start of the day or whatever, if at least the execution can be automated then that would be worth it.

It was automated... but... that didn't stop me interferring. Even though the system was (mostly) producing a slight profit I couldn't stop myself from sometimes making manual interventions if I felt something was going against me. All of the biggest dips and biggest leaps in my graph were due to interventions. All of the small upwards and downwards movements were due to the automation doing what it should have done.

The problem was with me rather than with the system (up to a point... as the system itself has a major flaw that I spotted today). It's no big deal with a £10 bank but image how I'd have got on with a bigger bank that I actually cared about? More interventions and more significant losses, I think.

Anyway, over the last few days I've mostly been studying the movements of the markets and I think I've noticed another huge pitfall to my approach... (alluded to in my previous paragraph) the biggest flaw possible (one so big that I'm amazed I didn't spot it before) really so I'm rather glad I didn't persue this any further. I'm looking at the strengths and weaknesses of a totally different in-play idea now but just in the early stages of testing it.

[FrogThimble]

Apologies for duplicate posts above due to poor editing of my reply to xitian.

[dragontrades]

I have been testing a straight betting strategy in practice mode for a while now so a few days ago I started running it for real on small stakes.
It had a bad day today which has got me worried.
I know you need to run these things for a long, long time to get a good big enough sample but when is the right time to stop and adjust some parameters or kill of the strategy?
And how large does the sample size need to be before making a judgement?
The strike rate is around 1in10 and I am staking by book % so how many races would be enough to draw some conclusions? Is there a formula for figuring the sample size out?
Hopefully someone can point me in the right direction

[Derek27]

...so a few days ago I started running it for real on small stakes.
It had a bad day today which has got me worried.

Personally, if I was moving a strategy from practice mode to live, I would start my P/L records afresh, as it's very much a different ball game, and it goes without saying that a few days isn't enough (especially if it's a straight betting strategy), and risking small stakes is nothing to get worried about.

In the absence of any obvious reason why practice mode would be an inappropriate testing platform (for example, large stakes) I think any strategy that succeeds in practice mode deserves a go in real mode, to small stakes.

[mcgoo]

If it helps I have run automation and seen returns of 10+% for 95 bets(punting system) over 2 weeks before seeing it all return to nothing in 4 days A trading system I have going now is doing 7% ROI over 32 trades (29% on exposure though) but I wouldn't be confident(after years of me trying this stuff) until I saw those returns over a much larger sample set(perhaps 1000 trades) but that's me -Not got a great maths brain

[marksmeets302]

The strike rate is around 1in10 and I am staking by book % so how many races would be enough to draw some conclusions? Is there a formula for figuring the sample size out?
Hopefully someone can point me in the right direction

The required sample size depends on how sure you want to be of an edge. Start recording the returns, and every time calculate the mean and standard deviation. Excel will do this for you. You will see that with more samples the standard deviation will tend to become lower: most returns will resemble the mean and you will have relatively fewer outliers. If you are aiming to be 95% sure of an edge, that translates to 2 standard deviations. If you deduct 2 standard deviations from the mean and this is a positive number then you can be 95% sure that you have an edge (a positive expectancy). This is under the assumption your returns are normally distributed, which they usually are. Unless you consider really long time periods where the market starts to change. If you only want to be 68% sure (that's one standard deviation) you deduct only 1 standard deviation from the mean. Of course for this you need much less samples. 3 standard deviations is like 99.7% and you will need a lot more samples for this.
Note that you can never be 100% sure.

[Euler]

If it helps I have run automation and seen returns of 10+% for 95 bets(punting system) over 2 weeks before seeing it all return to nothing in 4 days A trading system I have going now is doing 7% ROI over 32 trades (29% on exposure though) but I wouldn't be confident(after years of me trying this stuff) until I saw those returns over a much larger sample set(perhaps 1000 trades) but that's me -Not got a great maths brain

I'd add that you should cut the data in a number of different ways. I've found niches in very tiny cross sections of the market that if I lumped it all together would have just averaged out and lost due to the commission.