General Betting
Welcome to Live View – Take the tour to learn more
Start Tour
There is currently 1 person viewing this thread.
13 Jan 12
19:19
Joined:
Date Joined:
20 Jan 04
I have been looking at the 'will both teams score?' market just recently and have come up with a prediction method that after 100 games has a strike rate of 62% and a net profit of about 14 points.
I haven't been putting money on, just collecting the stats and getting the odds available on here.
My question is, how likely is it that the next 100 games will produce a strike rate of above 54% ? (which is the percentage to break even at the sort of odds I've been getting.)
I know it's not a large sample, but I want to try and gauge when (or even IF ) to start putting money on this.
Thanks in advance to anyone willing to help or offer advice.
I haven't been putting money on, just collecting the stats and getting the odds available on here.
My question is, how likely is it that the next 100 games will produce a strike rate of above 54% ? (which is the percentage to break even at the sort of odds I've been getting.)
I know it's not a large sample, but I want to try and gauge when (or even IF ) to start putting money on this.
Thanks in advance to anyone willing to help or offer advice.
Show More
.
http://www.surveysystem.com/sscalc.htm
I don't have time to do much explaining right now, have a look at that site though and see if you can wangle some numbers out of it.
http://www.surveysystem.com/sscalc.htm
I don't have time to do much explaining right now, have a look at that site though and see if you can wangle some numbers out of it.
Thanks Lori, I have tried that , and if what I have done is correct, I can be 95% sure that the next 100 predictions will have a strike rate of between 55.26% and 68.74%.
I could also be be 99% sure the figure expected would be between 53.12% and 70.88%.
If I have got this right....it seems I have reason to be optimistic!
Again, thanks for your help...and if you get the time could you please confirm that my conclusions are correct please. If it turns out I have used it right....what a great calculator that is!
I could also be be 99% sure the figure expected would be between 53.12% and 70.88%.
If I have got this right....it seems I have reason to be optimistic!
Again, thanks for your help...and if you get the time could you please confirm that my conclusions are correct please. If it turns out I have used it right....what a great calculator that is!
I've just had a thought. It's not the strike rate for the next 100 games it is predicting is it? It is the TOTAL games ie. 200.
Therefore if the SR was at the lower end of the expected range (say 54%) for the whole 200 then that must mean that the second 100 was considerably lower than this as the first 100 was 62%. By my calculations (in my head) this would be 46%.
My head now hurts. Can someone please tell me if I am on the right track please?
Therefore if the SR was at the lower end of the expected range (say 54%) for the whole 200 then that must mean that the second 100 was considerably lower than this as the first 100 was 62%. By my calculations (in my head) this would be 46%.
My head now hurts. Can someone please tell me if I am on the right track please?
Rather than look at the strike rate I would recommend putting the 100 different profit/loss outcomes in Excel, calculating the standard deviation and average. P&L is more interesting than strike rate because it reflects the kind of odds you're betting at.
You could use your stats to estimate the likelihood of making 14 points profit by chance. If the probability of making more than 14 points profit by chance is less than 1% then you can be confident that your method has not just been lucky.
Given the sum of independent events tends to a normal distribution a standard z-test would be appropriate and straightforward. Have a look at http://en.wikipedia.org/wiki/Z-test
You could use your stats to estimate the likelihood of making 14 points profit by chance. If the probability of making more than 14 points profit by chance is less than 1% then you can be confident that your method has not just been lucky.
Given the sum of independent events tends to a normal distribution a standard z-test would be appropriate and straightforward. Have a look at http://en.wikipedia.org/wiki/Z-test
Post Your Reply
Please login to post a reply.
Wonder
Instance ID: 13539