I passed the results from 1000 races to stats expert to see if the strategy I was using had a margin that was statistically significant. His response was quite interesting hence the post...
The margin in this case was 1% from 1000 races.
He reckoned you could be only 60% sure it was a genuine margin. To get up to over 95% certainly he recommended a sample size of 100,000.
I suspect this is because the margin was so small (harder to detect)... Anyhow thought it was interesting so decided to share.
How do you know if you have a margin?
It depends a lot on the median pricesionascaig wrote: ↑Thu Apr 08, 2021 8:59 amI passed the results from 1000 races to stats expert to see if the strategy I was using had a margin that was statistically significant. His response was quite interesting hence the post...
The margin in this case was 1% from 1000 races.
He reckoned you could be only 60% sure it was a genuine margin. To get up to over 95% certainly he recommended a sample size of 100,000.
I suspect this is because the margin was so small (harder to detect)... Anyhow thought it was interesting so decided to share.