When is it time to scale up?

[zippus]

Morning all.

Just looking for some advice please...

I've been working on a value betting strategy that is showing early signs of promise. I've been testing across all UK and IRE racing markets over the past few days with £2 stakes and have seen a decent ROI of +16%. This is based on a small sample of 211 races.

My question to the forum is around when to know if it's time to scale up. I'm very aware that these results could simply be randomness working in my favour and was thinking of maybe scaling up to £5 per race should I have similar results after 1,000 races.

Any thoughts as to how many races are likely to be needed to iron out randomness?

[Trader724]

A few thousand data points should give you confidence that what you see is EV.

[zippus]

A few thousand data points should give you confidence that what you see is EV.

Many thanks!

[Anbell]

Morning all.

Just looking for some advice please...

I've been working on a value betting strategy that is showing early signs of promise. I've been testing across all UK and IRE racing markets over the past few days with £2 stakes and have seen a decent ROI of +16%. This is based on a small sample of 211 races.

My question to the forum is around when to know if it's time to scale up. I'm very aware that these results could simply be randomness working in my favour and was thinking of maybe scaling up to £5 per race should I have similar results after 1,000 races.

Any thoughts as to how many races are likely to be needed to iron out randomness?

It depends very much on the % chance of the outcome. If you are backing 20.0 chances then you a much, much larger data set than if you are backing 2.0 chances.

[Anbell]

Morning all.

Just looking for some advice please...

I've been working on a value betting strategy that is showing early signs of promise. I've been testing across all UK and IRE racing markets over the past few days with £2 stakes and have seen a decent ROI of +16%. This is based on a small sample of 211 races.

My question to the forum is around when to know if it's time to scale up. I'm very aware that these results could simply be randomness working in my favour and was thinking of maybe scaling up to £5 per race should I have similar results after 1,000 races.

Any thoughts as to how many races are likely to be needed to iron out randomness?

It depends very much on the % chance of the outcome. If you are backing 20.0 chances then you a much, much larger data set than if you are backing 2.0 chances.

Also, scale slowly! If you have an actual edge, you don't need to be in a rush. There's a golden rule that you'll experience negvar the instant you scale up.

[zippus]

Morning all.

Just looking for some advice please...

I've been working on a value betting strategy that is showing early signs of promise. I've been testing across all UK and IRE racing markets over the past few days with £2 stakes and have seen a decent ROI of +16%. This is based on a small sample of 211 races.

My question to the forum is around when to know if it's time to scale up. I'm very aware that these results could simply be randomness working in my favour and was thinking of maybe scaling up to £5 per race should I have similar results after 1,000 races.

Any thoughts as to how many races are likely to be needed to iron out randomness?

It depends very much on the % chance of the outcome. If you are backing 20.0 chances then you a much, much larger data set than if you are backing 2.0 chances.

Thanks Anbell. That makes total sense. The odds are typically at the lower end.

[ODPaul82]

Football-data has a easy to use p-calc which I put my results into to ensure that anything that I'm getting profit from is giving random variance
https://www.football-data.co.uk/blog/P- ... lator.xlsx

Basically the closer you can get the p-value to 0, the more likely it is that it's not just down to chance you're being profitable

[Crazyskier]

It depends very much on the % chance of the outcome. If you are backing 20.0 chances then you a much, much larger data set than if you are backing 2.0 chances.

THIS

[gazuty]

Any thoughts as to how many races are likely to be needed to iron out randomness?

Good question. Part of the answer can be using https://stattrek.com/online-calculator/binomial.aspx

Discussed an example of how to use the calculator here - viewtopic.php?f=5&t=10865&start=90

Scaling up is a topic here - viewtopic.php?f=17&t=23699

[Kai]

Discussed an example of how to use the calculator here - viewtopic.php?f=5&t=10865&start=90

Btw you can directly link to a specific post by clicking on the header of that post, example : viewtopic.php?p=91628#p91628

[zippus]

Football-data has a easy to use p-calc which I put my results into to ensure that anything that I'm getting profit from is giving random variance
https://www.football-data.co.uk/blog/P- ... lator.xlsx

Basically the closer you can get the p-value to 0, the more likely it is that it's not just down to chance you're being profitable

Thanks for the spreadsheet. Looks very interesting. Much appreciated!

[zippus]

Any thoughts as to how many races are likely to be needed to iron out randomness?

Good question. Part of the answer can be using https://stattrek.com/online-calculator/binomial.aspx

Discussed an example of how to use the calculator here - viewtopic.php?f=5&t=10865&start=90

Scaling up is a topic here - viewtopic.php?f=17&t=23699

Cheers for the information gazuty. This is very helpful.

[gazuty]

Discussed an example of how to use the calculator here - viewtopic.php?f=5&t=10865&start=90

Btw you can directly link to a specific post by clicking on the header of that post, example : viewtopic.php?p=91628#p91628

[jamesedwards]

When scaling remember that you risk value declining as you increase your stakes. I made the mistake of assuming small increases in stakes would have no difference on return when it was turning my automations into negative every time.

[Crazyskier]

When scaling remember that you risk value declining as you increase your stakes. I made the mistake of assuming small increases in stakes would have no difference on return when it was turning my automations into negative every time.

Same here with both dogs and horses.

What works wonders sub £2, is often sufficient to 'spook' the market when raised above £2, and therefore the value or edge is eroded. Greens become reds.

Thus I am resigned to never getting rich, only having nice treats once in a while.

CS