As a physicist, science writer, and collaborator of Stephen Hawking, Leonard Mlodinow shows in his book “The Drunkard’s Walk” (2009), through examples from statistics and everyday observations, how strongly coincidence and human perception distort our thinking.

These cognitive patterns can also be observed in projects.

Inspired by this perspective, the following article uses Mlodinow’s ideas as a foundation to reconsider uncertainty and decision-making in a project context, and to translate them into a mindset for modern project leadership.

———————————————————————————————————————————————————————————-

From Plans to Probabilities

Project management has historically been shaped by the belief that good planning leads to good results.
Clear goals, clean schedules, clearly defined responsibilities.

In practice, however, we observe something different:
Despite good planning, projects evolve very differently.

Some fail without an obvious mistake. Others succeed despite clear risks.

The reason is often not a lack of competence, but the fact that projects do not behave deterministically.

———————————————————————————————————————————————————————————--

Coincidence Is Not a Disturbance, but Part of the System

In “The Drunkard’s Walk,” Leonard Mlodinow describes how strongly coincidence shapes our everyday live and how much we tend to ignore it in hindsight.
We prefer to explain outcomes through clear causes rather than probabilities.

The same happens in project work:

• Success is attributed to skill

• Failure is personalized

• Coincidence is ignored

  
  

At the same time, projects are constantly influenced by factors that cannot be fully controlled or planned:
market dynamics, dependencies, people, technology, timing.

———————————————————————————————————————————————————————————-———————

We See Patterns Even Where None Exist

Even in pure randomness, the human brain is a pattern-seeking machine.
If a project performs poorly for several weeks, we quickly say:
“This project is cursed.”

Yet fluctuations are normal. Not every sequence is a trend, and not every low point is a structural problem.
Probabilistic thinking helps distinguish random variation from real signals.

———————————————————————————————————————————————————————————-———————————————

Small Samples Are Deeply Misleading

From very few observations, we often draw surprisingly strong conclusions.

Two successful sprints are often enough to hear statements like:
“Our new approach is working.”
Statistically, this is barely reliable.

Early successes or failures are usually based on too few observations and data to make robust statements about stability. Ignoring this often leads to overreaction.

———————————————————————————————————————————————————————————-———————————————Regression to the Mean Is Often Misinterpreted

An escalated, highly problematic sprint or project phase is often followed by a noticeably calmer one.

This does not necessarily happen because agreed measures start to take effect,
but because extreme deviations rarely persist over time from a statistical perspective.

Nevertheless, we tend to attribute the improvement to our intervention.
This creates misleading success stories and leads to measures that may be unnecessary or even counterproductive next time.

———————————————————————————————————————————————————————————————

The Real Issue: Our Mental Models

Many project approaches act as if uncertainty could simply be “organized away.”
But projects are not machines they are complex systems.

What is missing is a more conscious way of thinking that:

• makes uncertainty explicit

• allows for multiple scenarios

• evaluates decisions in terms of probabilities

  
  

Not: What will happen?
But: What could happen—and how likely is it?

———————————————————————————————————————————————————————————————————————————

In Hindsight, Everything Appears Logical

After a failure, we often hear:
“We should have seen this coming.”

Beforehand, the situation was usually unknown.
Only in hindsight we construct a seemingly compelling story.

This hindsight bias distorts learning and decision evaluation.

—————————————————————————————————————————————————————————————

Good Decisions ≠ Good Outcomes

A central idea of the book is that outcomes and decisions must not be conflated.
Someone can make the right decision and still have bad luck.
Likewise, a questionable decision can turn out well by chance.

Project leadership should therefore focus less on outcome evaluation alone

and more on how decisions are made under uncertainty.

A decision is of high quality if:

• the information available at the time was considered

• uncertainty was explicitly acknowledged

• alternatives were weighed

• risks were consciously accepted or reduced

• assumptions were transparent

even if the outcome later turns out poorly.

The key question is not only:
Did it work?

But:
Was the decision reasonable given the information available at the time?

—————————————————————————————————————————————————————————————-

My Understanding for Modern Project Leadership

I do not see project leadership as a promise of certainty,
but as the creation of clarity.

Clarity about:

• which scenarios are realistic

• how likely they are

• what can be influenced

• and where chance remains part of the system

  
  

Not every deviation is a failure.
Some events are simply part of an uncertain system.

Maybe God rolls the dice.
But we can learn to lead more consciously with probabilities.