Hello to you all

I would like, through this post, to share my problematizations concerning the analysis of outcomes after a quantitative risk analysis using Crystal Ball PROJECT MACRO. (oracles’ add-on to microsoft project for risk analysis)

Above you can see the outcome of the simulation.

What I find hard to understand is why I get a different outcome when I set “certainty” at 67% (typical for 1 std. dev.) and the value depicted in the automatically generated equivalent to 1 std. dev. (text in blue context on the right of the blue vertical line)

Your help would be appreciated

Than you in advance

Ilias

### Like this:

Like Loading...

*Related*

Responding to myself after pondering….

Is it different because the fitted distribution is not a normal one but a lognormal one (small sample data according to CLT)?

LikeLike

I can’t quite see what you think this is telling you but just a shot in the dark seeing the percentages associated with the various markers on your graph – are you and Crystal Ball getting confused between one sided and two sided expressions of variation?

With a Normal distribution, roughly 15% of outcomes fall above a value one standard deviation above the mean and 15% below a value one standard deviation below the mean, so roughly 30% of outcomes fall more than one standard deviation away from the mean. Not knowing how Crystal Ball operates, perhaps it’s using ‘certainty’ in a two-sided way and you are using it in a one-sided way or vice versa.

LikeLike

Hi Mr Gray!

Thank you so much for the response! I guess, that is why they say “think out of the box”.

All i had in my mind was the bell curve, with the typical 1, 2 & 3σ at each side, and i was failing to interpret the real results.

You are correct, the application depicts 2 sided expressions of variation.

Thank you for pointing that out!

Regards

Ilias K.

LikeLike

It’s easy to get locked in. Happens to all of us.

LikeLiked by 1 person