Prediction or Predictability

[David Kerridge <dfk@rsc.co.uk>]

Some time back (it is in the Digest no 69) Rajeev Tatkar made
a very interesting and important point:

> I would like to make a distinction between prediction and
> predictability. ..................
>.........
>WED's philosophy on the other hand was more about predictability.
>................
>The route was through making the system predictable. If the system
>is very predictable the need to predict the unpredictable greatly diminishes.

To complete the picture we should note that WED always avoided
concentrating on the obvious problem. In fact he concentrated on
seeing the *whole* problem. This is very rare.

In an uncertain world there are many links in the chain that leads
to achieving the aim of the system:

1  Make the system more predictable

2  Observe current state - get accurate information

3  Interpret using theory

4  Predict the results of alternative actions - also using theory

5  Choose best option

6  Carry out action

It is very easy to spend all our time improving the action stage,
(as with some TQM programs?) because it is the most obvious:
but the strength of the chain is the strength of the weakest link.
*None* of these steps is trivial.

But then WED introduced something corresponding to the
zero link:

0   Clarify the aim

And we should add a seventh link

7  Check outcomes and modify theory if necessary.

Improvement at the zero stage is the most effective of all.
But we must be careful not to fall into the error of putting
all our endeavours into *any* one stage. The obvious is
often wrong - "Best efforts".

It is useless to have an aim if we have no means of achieving
it. And it is harmful to make the system predictable if that
conflicts with the aim. We make a system totally predictable
if we close it down.

More information may lead to more tampering, or information
overload. Or at the other end of the chain, if we have the wrong
aim, it is better for actions to be ineffective. This may be why
some businesses survive.

Improvement in any *one* of the links, except clarification
of the aim, can do harm if pursued in isolation, especially
by short term methods. The Deming Philosophy shows how
to make long term systemic improvement without doing harm.

This is why the Deming approach appears strange, indirect,
and counter-intuitive - and why it is so effective.

David Kerridge
Director of Research, British Deming Association
dfk@rsc.co.uk
===================================================================