# Risk impact can be modified?

#### qualprod

Trusted
Hi guys

I have evaluated risks in a simple manner in 9001 2015.
and had the idea that when risk is evaluated, PxI,
the probability was the only that can be minimized.

However lastly have read somewhere that also impact can be modified.

Risk evaluation steps is as follows:

I´ll try to imagine a case, please tell me if is right or wrong.

Case:

Key person near to retirement
Risk value = probability x impact

range 1-5
probability is very high, (in six months is leaving the company)

Impact= well, here it could be measured in several viewpoints
normally, image, insatisfaction to clients, financial loss, because this person knows all the company, clients, business operation,etc.

So RV = 5 x 5 = 25, very high value

Actions for mitigation:

To hire a person with similar competency and be trained properly.

Ok, now new person is in place and is performing very well.

Then we need to evaluate residual risk.

RV = PxI
Probability is very low = 1, and the Impact value?

In my viewpoint, is the same because if this person (Position) is absent
the impact value remains the same, or can be changed?

A vague idea to reduce the impact could be, for example in the same case
(financial loss) that I may get some funds to reduce the impact, or the poor image to client, I could provide additional services or products to improve
satisfaction.
this way impact value, could be changed and lowered.

Rv = 1 x 2= 2

very low risk value.

Could you give some comments ?

Thanks

#### Bev D

##### Heretical Statistician
Staff member
Super Moderator
Well "I reject the premise of your question"

you have mitigated the effect of the event. good. have you put in systemic processes to ensure that the event won't happen again? if so you really haven't changed anything - just kicked the can down the road. after all the replacement you hire has a 100% probability of getting old and retiring or suffering an early death.

the bottom line is that the tired old RPN or Probability * Severity math is not relevant or useful. it is simply a ritual.

The intent of risk based thinking is to identify the risks and mitigate them as appropriate.

The calculation of mathematical formulas is no substitute for thinking...
.

#### Attachments

• 873.5 KB Views: 103

#### qualprod

Trusted
Thanks Bev D

well, I ´ve been very short time in this fields (quality, QMS)
So your response is for me somewhat not clearly understood.

According to what article you shared, it seems maths in these cases, maybe wouldnt apply properly to risks.

So, in simple words, what approach would you suggest to address
risks according to Iso 9001?

Could you share some simple ideas?

Thanks

#### dsanabria

##### Quite Involved in Discussions
Hi guys

I have evaluated risks in a simple manner in 9001 2015.
and had the idea that when risk is evaluated, PxI,
the probability was the only that can be minimized.

However lastly have read somewhere that also impact can be modified.

Risk evaluation steps is as follows:

I´ll try to imagine a case, please tell me if is right or wrong.

Case:

Key person near to retirement
Risk value = probability x impact

range 1-5
probability is very high, (in six months is leaving the company)

Impact= well, here it could be measured in several viewpoints
normally, image, insatisfaction to clients, financial loss, because this person knows all the company, clients, business operation,etc.

So RV = 5 x 5 = 25, very high value

Actions for mitigation:

To hire a person with similar competency and be trained properly.

Ok, now new person is in place and is performing very well.

Then we need to evaluate residual risk.

RV = PxI
Probability is very low = 1, and the Impact value?

In my viewpoint, is the same because if this person (Position) is absent
the impact value remains the same, or can be changed?

A vague idea to reduce the impact could be, for example in the same case
(financial loss) that I may get some funds to reduce the impact, or the poor image to client, I could provide additional services or products to improve
satisfaction.
this way impact value, could be changed and lowered.

Rv = 1 x 2= 2

very low risk value.

Could you give some comments ?

Thanks
How about mitigating the process by cross training or hiring another individual to be trained.

Note: your impact is way off - you are making assumptions please don't share this formula - there are too many variables to come to a conclusion.

Risk in this case is dynamic and unpredictable. Maybe the position is not needed and it is replaced with a robot.

Maybe the customers are not aware and really don't care as long as they get their products.

Your formula is not workable due to many variable - please do not overthink it and take people throught this path - keep it simple - KISS formula.

#### qualprod

Trusted
How about mitigating the process by cross training or hiring another individual to be trained.
I did it in this way to mitigate the risk.

Note: your impact is way off - you are making assumptions please don't share this formula - there are too many variables to come to a conclusion.
I have seen this method used everywhere, why not to use it?
It so common.

Risk in this case is dynamic and unpredictable. Maybe the position is not needed and it is replaced with a robot.
No comment.

Risk in this case is dynamic and unpredictable. Maybe the position is not needed and it is replaced with a robot.
no comment

Maybe the customers are not aware and really don't care as long as they get their products.
who faces The risk, is aware of it.

Your formula is not workable due to many variable - please do not overthink it and take people throught this path - keep it simple - KISS formula.

Thanks

#### Marc

##### Retired Old Goat
Staff member
How about mitigating the process by cross training or hiring another individual to be trained.
I did it in this way to mitigate the risk.
Your assumption that in "key" personnel changes "risk" can be reduced is questionable. Please list positions which are "key" personnel (in your world).

Note: your impact is way off - you are making assumptions please don't share this formula - there are too many variables to come to a conclusion.
I have seen this method used everywhere, why not to use it?
It so common.
Yes it is. The link pretty Bev gave much says (correctly) "If you don't understand statistics, talk to a statistics expert." Too many people give "false" information, typically because they do not actually understand, for example, statistics (but who think they do).

Risk in this case is dynamic and unpredictable. Maybe the position is not needed and it is replaced with a robot.
No comment.
Case:

Key person near to retirement
or not?

Risk in this case is dynamic and unpredictable. Maybe the position is not needed and it is replaced with a robot.
no comment
Case:

Key person near to retirement
or not?

Maybe the customers are not aware and really don't care as long as they get their products.
who faces The risk, is aware of it.
This of course, depends upon the role and function(s) of the "key person". As noted herein, "key personnel" often have assets which one can not quantify, such as in sales and design. Reputation and "friends" (connections) is only one of the assets you can not quantify.

Your formula is not workable due to many variable - please do not overthink it and take people throught this path - keep it simple - KISS formula.

Thanks
Here is my take...

First of all, no - There is no simple "KISS" approach which would apply. As Bev says - There is no replacement for "thinking".

Bev gives her "down vote" for FMEA's, and I agree with her for the most part. I go back many years and in critical missions and products we didn't do a typical FMEA. We brought teams together, multifunctional teams, to discuss potential failure modes. We had to think together in Teams which always included a reliable, knowledgable statistician.

I was involved with a DoD program back in the Iraq War 1 of 1991. While I can not relate details, I can tell you that the project was critical on a new "product" design. We had a weekly team meeting. We discussed many aspects. We "thought". Every week's output was the next weeks input. No one ever even said "FMEA". We did have a team statistician whose input was invaluable.

I also did a lot of work in automotive "airbag restraint" systems, not to mention explosives and automotive breaking systems. Statisticians were always key project personnel.

This discussion started out with
...Case:

Key person near to retirement...
You can not apply a mathematical / statistical formula to evaluate risk because, as has been said - There are too many variables.

For example, I know people who are well known in their industries. There is no training or other way to assess what the loss of their personal reputation and "connections" comes into play on many levels. Comments such as
...Risk in this case is dynamic and unpredictable. Maybe the position is not needed and it is replaced with a robot.
no comment...
make no sense other than at the level of line personnel. Sales cannot be replaced by robots. R&D personnel can not be totally replaced by a robot. Many positions can not be replaced by a robot.

That said - Neither AI (artificial intelligence) nor robots can replace many (most?) personnel at this point in time/history. While both are being highly hyped these days, what do you see in reality? A McDonalds somewhere has paid for equipment which will cook burgers (equipment cost ~US\$60,000). But - It still requires a human to apply certain toppings. Order clerks being replaced with kiosks.

I believe in the long term. I have lived a lot of years and seen changes I could never have imagined. We can only guess as to what the future will bring. As I age, I remember my father telling me, when I was very young, about the first time he saw an airplane fly over him. From self driving cars (a dream of mine considering the miles I have driven) to AI. I am old enough to realize that there are many things I would not have considered possible, to actually become possible. "Feature phones" today have more computing power than many computers sold 10 years ago.

However, the "future" is not now. One can dream of tomorrow, but one has to live in "today".
***********************
Having said all of this - I'm not a professional statistician, nor am I a GD&T expert. But throughout my career of about 35 years, I have always called in an expert when necessary.

And - There are a number of experts here whose participation, helping people in the discussion forums, I seriously appreciate their input in discussions.

Thank you all.