EntropySink
Technical & Scientific => Programming => Topic started by: Mike on November 18, 2013, 01:20:37 PM
-
This is more of a stats question that programming but...
I have events that produce either A or B. Each day all the events are reset and the user can redo them. Each event has a different chance of producing A (otherwise it produces B). Trying to figure out the chances of A being produced once for a given day. I know how to calculate the chance that at least 1 A is produced but not sure how to check for a particular amount.
Right now my chances are 10%, 10%, 20%, 30%, 40%, 50%, 60%, 70% which almost guarantees at least 1 A but it is producing too many As.
Anyone know the equations?
-
The chance of one and only one A? It's the prob event i produces A times the prob that all others produce B, summed over all i.
-
Something like this?
http://www.mathsisfun.com/data/probability-events-independent.html
or maybe better
https://www.khanacademy.org/math/probability/independent-dependent-probability/independent_events/v/compound-probability-of-independent-events
-
The chance of one and only one A?
Correct
It's the prob event i produces A times the prob that all others produce B, summed over all i.
Ah, so no nice easy spreadsheet equation? Gotta do it bit by bit?
Something like this?
http://www.mathsisfun.com/data/probability-events-independent.html
or maybe better
https://www.khanacademy.org/math/probability/independent-dependent-probability/independent_events/v/compound-probability-of-independent-events
I think it's ~98.367% likely of getting A to happen.
Yeah, already got that part. But that is the probability of A happening not the probability of A happening only once.
What I'm seeing from actual results is many people are getting A three times in a day which makes sense as the last three events have good odds of producing A. I want to try and manipulate the numbers so that the chance of A happening is good while limiting the number of times A happens.
Note: Forgot one event with 80% probability which takes the probability of A happening at least once to over 99%
-
ok, so looks like there is a 3.39% chance of A only happening once. Hmmm need to relook at these numbers.
edit: Spreadsheet error gave wrong value
-
I get 10.45%.
Oh, wait.. you added one more with 80% chance. Yeah, 3.397%.