Cumulative Probability

I write for me…mostly. It would be disingenuous of me to say that I do not care if anyone else enjoys this. I also struggle with the “is this really worth writing about????” issue. Translation: Is this something others would care to read. And then I had an espresso-bean fueled epiphany; I will write what I would want to read about. There is a subtle difference between “writing for me” and this ideal. We’ll see how this goes…I’m also incredibly lazy so frequency is also an issue.

So here we go…a blog dedicated to beer and sports is writing about cumulative probability charts. Namely S-curves.


If at this point you are lost, stick around and allow me to bore you further. If however, these S-curves are a real part of your life, please suffer with me.

Many organizations use these curves to illustrate the “cumulative probability” of a decision or course of action. These curves are often restated as P10-P50-P90* figures.

What can you glean from this curve? Many managers will look at this and rightly assume that the result of the action will fall between the P10 and P90 points on the curve. However, when you decide to soldier forth with the decision you are not agreeing to accept the P10-P90 range as your probable outcomes. You are buying the entire S-curve. That means that 20% of the time you are going to end up outside of that range.

People lock in on that P10-P90 (or god forbid they just lock in on the P50) and are just dumbfounded when the outcome is outside of this. I watched a guy with a PhD lose his mind when, after around 200 deployments of a thing (gotta be vague) he got something in the P3 area. After grilling me for around 8 minutes straight, I simply asked him “What part of a curve do you not understand?” As the discussion progressed I realized that even though he understood the core concepts he had been so socialized through conversations about the P10-P90 that he believed those to be the only potential outcomes.

If you see a cumulative probability chart like an S-curve, you are buying the entire curve.

Garbage In….But wait! There’s more! This entire S-curve is based on inputs or sources outside of yourself (or they better be). Everything driving that curve is based on the assumptions and factors collected and calculated in your model. So if your subject matter experts who tell you that they predict an outcome could be anywhere from 100 barrels to 1000 barrels, anything outside of that range will fall completely off of the curve. Point being that you need to be extremely careful in (1) who you choose to get inputs from and (2) how you frame the discussion around what ranges you need. …Garbage out.

*Brief explanation of what a P90 is: There is a 90% chance that the outcome will be at or below the P90 point. Contrariwise, the P10 is the point on the curve where there is a 10% chance that the outcome will be at or below this point.

Published by


Born and raised here in Houston. I enjoy all things Houston. If they keep a score I probably watch it.

Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google photo

You are commenting using your Google account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s