Forums

General Betting

Welcome to Live View – Take the tour to learn more
Start Tour
There is currently 1 person viewing this thread.
Chosen®
07 Feb 10 04:53
Joined:
Date Joined: 03 May 09
| Topic/replies: 603 | Blogger: Chosen®'s blog
If a set of criteria returns 8% winners, can anyone tell me what the expected highest run of consecutive losers would be in 4000 trials.

To make it a bit more specific, what would be the highest number of consecutive losers in 4000 trials with at least a 50% chance of occurring?

If anyone can tell me how they made the calculation, that would be great too. I've thought about it but found that my probability statistics isn't what it used to be.

Thanks in advance.
Pause Switch to Standard View Statistics question
Show More
Loading...
Report farraflash February 7, 2010 9:14 AM GMT
The expected is about 76.

The 50% mark is about 73, you have a 50% chance it will be 73 or less and 50% chance that it will be 73 or more.
Report top2rated February 7, 2010 10:03 AM GMT
Via a spreadsheet......

Label cell A1 Trials

Label cell B1 StrikeRate

Label cell C1 LLR

Enter 4000 in cell A2

Enter 8% in cell B2

Enter formula =LOG(A2)/(LOG(1-B2)*-1) in cell C2
Report Huggy February 7, 2010 10:09 AM GMT
Statistics mean nothing

arecent survey found out 42% of people dont believe in them
Report Chosen® February 7, 2010 1:29 PM GMT
Very useful, thanks farra/top2
Report Chosen® February 8, 2010 4:47 AM GMT
Tpo2rated,

I plugged in the labels, formula, and parameters but came up with 99.471 rather than farra's answer of 76, which is intuitively (and hopefully) closer to the mark.

Sorry to bug you with this, but possible typo in the formula?
Report farraflash February 8, 2010 10:19 AM GMT
That formula estimates an outer limit to the maximum LS which is different to the mathematical expected value (76) of the distribution of all possible maximum LS's.

I make a maximum LS of 99 a 0.6% chance of occurring and 99 or less 93%.
Report Chosen® February 8, 2010 2:27 PM GMT
Thanks farra.
Report Hayden February 8, 2010 10:32 PM GMT
Statistics don't give answers they merely raise more questions.
Post Your Reply
<CTRL+Enter> to submit
Please login to post a reply.

Wonder

Instance ID: 13539
www.betfair.com