Note: This was previously posted on the wrong board...
I've run 3 monte carlo simulations using Crystal Ball for a product in development. Each of the 3 simulations represents a different patient group that could potentially use this product.
Typically, I would run the results of these 3 simulations (mutually exclusive of each other) through yet another simulation (kind of like a simulation of the entire portfolio for this product).
Here is the atypical situation. Instead of running this last simulation of the 3 together, I'm going to add together the results for the best case revenues for the 3 simulations, the 3 base cases, and the 3 worst case revenues. What this 'roll-up' essentially does is make the best case extremely high and the worst case extremely low. It's really NOT a realistic scenario; it's what I call the Perfect Storm scenario. If you've seen the movie or read the book, you'll know what I mean.
My question is: is there a technical term for what I've done with the 3 scenarios? I'm sure there is a statistician out there who can tell me what to call my 'perfect storm' scenario. I need to put in on a slide for presentation purposes. Thanks!
Update: Since my original post, I've gotten some more insight. In my original 3 scenarios, what I called the 'best case' above was really the 90%-tile case, and the 'worst case' was the 10%-tile case; 'base' was 50%. When I added the 10%-tile stream of revenues for all 3 cases to create this perfect storm scenario, someone pointed out that this is NO LONGER a 10% case. Since each patient group is mutually independent of each other, what I've really created is a 10% * 10% * 10% = 0.1% case!!! In other words, the chance of the best case (or worst case) actually happening or exceeding is NOT 10%, it's really 0.1%.
I've run 3 monte carlo simulations using Crystal Ball for a product in development. Each of the 3 simulations represents a different patient group that could potentially use this product.
Typically, I would run the results of these 3 simulations (mutually exclusive of each other) through yet another simulation (kind of like a simulation of the entire portfolio for this product).
Here is the atypical situation. Instead of running this last simulation of the 3 together, I'm going to add together the results for the best case revenues for the 3 simulations, the 3 base cases, and the 3 worst case revenues. What this 'roll-up' essentially does is make the best case extremely high and the worst case extremely low. It's really NOT a realistic scenario; it's what I call the Perfect Storm scenario. If you've seen the movie or read the book, you'll know what I mean.
My question is: is there a technical term for what I've done with the 3 scenarios? I'm sure there is a statistician out there who can tell me what to call my 'perfect storm' scenario. I need to put in on a slide for presentation purposes. Thanks!
Update: Since my original post, I've gotten some more insight. In my original 3 scenarios, what I called the 'best case' above was really the 90%-tile case, and the 'worst case' was the 10%-tile case; 'base' was 50%. When I added the 10%-tile stream of revenues for all 3 cases to create this perfect storm scenario, someone pointed out that this is NO LONGER a 10% case. Since each patient group is mutually independent of each other, what I've really created is a 10% * 10% * 10% = 0.1% case!!! In other words, the chance of the best case (or worst case) actually happening or exceeding is NOT 10%, it's really 0.1%.
