Many Simple Models over One Complicated Model

I see it again and again.

When they have invested time and energy in a model (tool, framework, method), people have a tendency to make their models more and more complicated. “Let’s add another dimension.” “Let’s deepen the domains.” “Let’s add some columns or swim lanes.” “Let’s draw an extra diagram.”

But complexity thinkers know better. They understand you need different approaches for different contexts. It is better to apply different models to different problems.

Each systems approach is useful for certain purposes and in particular types of problem situation. A diversity of approaches, therefore, heralds not a crisis but increased competence in a variety of problem contexts.

– Michael C. Jackson, Systems Thinking

This means it makes more sense to use multiple simple models instead of one complicated model. Having a toolkit of methods and frameworks, which each fail in their own way, is a smarter approach than relying on one method or framework to deal with all situations.

Complexity itself is anti-methodology. It is against "one size fits all."

– Tom Petzinger, Interaction of Complexity and Management

Of course, it’s very human to hold on to the model you had already invested in. In behavioral economics it’s called the endowment effect. We value more what we already have, because we own it. It’s irrational, but natural. And especially those who have created their own method or framework, will usually cling on to it like Bashar Hafez al-Assad to his presidential chair.

The more work you put into something, the more ownership you begin to feel for it.

– Daniel Ariely, Predictably Irrational

It makes sense to realize that holding on to your favorite method or framework is predictably irrational. The sensible thing to do is to invest in multiple models, and multiple approaches.

Don’t extend your diagram to accommodate an extra dimension. Just erase it, and start from scratch.

(image: House Democrats' Health Plan)


  • Join My Next Book: Management Workout
  • I Don't Care About "Agile"
Related Posts
free book
“How to Change the World”
  • Tonino

    We may want to complicate a model because we think that the additional “complication” is needed to make it more consistent, forgetting that it may be true under specific assumptions that may not apply. Worst is wen we don’t have any check phase to see if the new complication is helping or not, or refusing to drop it because “otherwise we had worked for nothing” (a case of sunk cost fallacy). A provoking example is in this graph, modelling relations and agents in the context of the Afghan war receiving the comment “when we understand that slide, we’ll have won the war”

  • Kelby Zorgdrager

    You make a great point. Developers spend so much time creating their framework that it’s almost painful to admit that maybe, just maybe, there’s a better way.

  • Tonino Lucca

    p.s. sorry I haven’t posted my complete name: Tonino (or Antonio) Lucca (@tonyxzt).

  • campaign tracking tool

    Simpler models is their relative transparency when it comes to their ultimate use.

  • Roger_review

    Many simple models vs one complicated model. I sidestep this question by ‘playing the joker’ …
    The Joker is the wild card that you can play at any time.
    The Joker does not take any model too seriously.
    The Joker gives the system a human face.
    The Joker keeps you alert to contradictions.
    The Joker challenges procedures.
    The Joker is sharp, quick and perceptive.
    The Joker brings fresh perspectives.
    The Joker is alive and dynamic.
    Above all, the joker lets you trust your judgement and play your own game.
    Every game needs a joker.
    Every model has exceptions.

How to Change the World - free Workout - free