“Medium” is another way of saying “Don’t make me choose”

From the Easy in theory, difficult in practice Blog
My musings on project management, project portfolio management and change management. I'm a firm believer that a pragmatic approach to organizational change that addresses process & technology, but primarily, people will maximize chances for success. This blog contains articles which I've previously written and published as well as new content.

About this Blog


Recent Posts

Defeating the multitasking monster

Just because you have information radiators doesn't mean senior stakeholders will review them!

What's the right sprint length for your team?

There can be only one (and it is not YOUR project)!

A camel is a horse designed by a committee

Steve Kay, a Program Manager interviewed in the Closing Credit article of the August 2012 issue of PM Network made an interesting change on his mega-project – he altered the typical 5×5 risk matrix (e.g. very high – very low) to a 4×4 one to remove the “sitting on the fence” option for those participating in the qualitative risk analysis process.

This prompted me to check the risk registers for a few of the historical projects I’d been involved with and I noticed that when there is a three or five rating scale for probability and impact, the medium selection is picked much more often than one of the other choices.  This should not be a surprise since we can draw a reasonable conclusion that probability and impact values follow a normal distribution for projects with an average level of risk.

My concern is that the frequent selection of this rating relates more to indecisiveness or lethargy than to statistics.

The former cause might be tied to the common behavior in some organizations of people being unwilling to take a stand.  With a three point rating scale, providing a low severity for a risk which you had previously identified might incur the wrath of those minimalists who want the focus purely to be on critical threats or opportunities.  Doing the same for someone else’s risk event might put you at loggerheads with them as they might perceive a very different severity for their risk.  On the other end of the scale, assessing a risk event as high might brand you as a “Chicken Little”.

Such behavior might occur if risk analysis participants are lacking knowledge on what the different ratings imply and how to score risks, but more likely it is caused by participants that are uninterested in the process.  I believe that most staff are so focused on their day-to-day operational and project work and the “real” issues that plague them that they wish to minimize their effort spent on risk management which they perceive as being at best, an academic practice.

The recommendation in the PM Network article is a simple way to address the inappropriate use of medium ratings as it forces participants to pick ratings that will be either above or below the point of indecision.  This method could be enhanced by one or more of the following practices:

  • Take the time before qualitative risk analysis begins to clearly explain risk rating scales and illustrate their usage with examples.
  • Make sure that you effectively communicate the linkage between risk scoring and the responses that would be based on the scores as well as the implications of scores on risk reporting.
  • As of a follow up to the analysis session and only if you have sufficient risks to support this, you may wish to construct a graph of the distribution of impacts and probabilities to see if the overall shape of the curve “fits” the team’s perception of the overall project’s risk profile.  If it doesn’t, this could point to inconsistent scoring.

Yoda said “Do or do not, there is no try” – medium might just be the “try” of risk management!

(Note: this high value article was originally published in August 2012 on kbondale.wordpress.com)

Posted on: November 28, 2018 07:56 AM | Permalink

Comments (16)

Please login or join to subscribe to this item
Very interesting, thanks for sharing

Great.. Thank you for sharing

Great points Kiron. You know I never thought about it thsi way but it does make lots of sense.

A 5 point scale is probably the answer. Otherwise we are left with "High" and "Low" which I don't think is much better than omitting "Medium"; it has its own shortcomings.

Interesting. Then how to decipher what is a true medium. Maybe have no 3, if a 5-point. It's either medium-low (2) or medium-high (4).

Thanks Sante & Andrew -

The most honest approach might be to just have two categories - high and everything else! Usually there's insufficient willingness and capacity to respond to the lows and mediums anyway, so this way we will just make things that much quicker to process!


Good points and it really make sense, Kiron and thanks for sharing.

Valid point Kiron.

Very thought provoking. I like "high" and "everything else". Or perhaps "high/extreme" and everything else?

However, I had a look at my risk registers and because mine (spreadsheet based) ask you to estimate likelihood and impact separately and then automatically calculate risk it makes it less prone to people just choosing "medium". Additionally the one I use then applies a mitigation factor based on how good we think the mitigation is "excellent, satisfactory, poor etc" to get the residual risk. This seems to produce a much larger spread.

Thanks Ashleigh - but what happens when folks pick medium for likelihood AND impact :-) ?

Thanks Rajesh!

Good point Kiron!

Good article!!

You bring up a very good point with central tendency. Also, the concept of reviewing definitions with project team members and stakeholders BEFORE the qualitative analysis is important. Perhaps adding in some quantitative analysis as "required" for the medium risks would bring in a mixed method approach to assessing the risk. Our PMIS has the ability to link a risk to a task. We can quantify the risk by affect to schedule or cost --in theory--we haven't gotten there yet :)

Please Login/Register to leave a comment.


"If at first you don't succeed, try, try again. Then quit. There's no use being a damned fool about it."

- W. C. Fields