Simulation

Analysis Paralysis — Clocks, Clouds and the Cost of Waiting for Certainty

Commission more analysis, hold another meeting, run a thousand simulations. Then watch the clock, because not deciding is also a decision.

What this simulation shows

In Monty Python’s philosophers’ football match, Germany field Kant, Hegel and Nietzsche against Greece’s Plato, Aristotle and Socrates. For almost the entire game nobody touches the ball. The players wander the pitch deep in thought until Socrates, with under a minute left, simply runs forward and scores.

The joke works because thinking and doing are not the same activity. Every organisation wants better decisions, every project wants less risk, and every manager wants more information before committing. Those are reasonable instincts. The failure begins when pursuing certainty becomes the objective in itself — when analysis stops supporting action and starts substituting for it.

Karl Popper divided systems into clocks and clouds: orderly and predictable versus turbulent and resistant to precise forecast. It is worth splitting that into three. A simple system has clear cause and effect and stable constraints, so careful modelling genuinely pays off — the system does not move while you study it. A complex system has many interacting parts whose combined behaviour cannot be reduced to any one of them; you cannot model it like a clock, but you can probe it and learn its shape. A chaotic system is so sensitive to small changes that tiny differences cascade into wildly different outcomes.

This simulation lets you manage in all three and see what analysis buys you in each.

How to use it

Start in Clockwork Cup. Spend time on Commission More Analysis and Run Small Experiment before playing. Analysis rewards you here, because the system is stable enough for study to transfer into performance. This is the case where thinking before acting is competence rather than delay.

Now switch to Complex League and repeat the same strategy. Returns flatten much sooner. Small experiments still help — probing works where modelling does not — but exhaustive analysis stops converting into results.

Then try World Cup Cloud, and watch the Analysis Time Curve while you do it. Investment in certainty produces almost nothing, because a single deflection or a mistimed substitution reorganises everything downstream. Meanwhile Advance 5 Minutes keeps moving whether you have decided or not.

Use Run 1,000 Simulations to see the Monte Carlo output, then compare it with what the single match actually does. Over a season, a stronger side creates more chances and that trend is visible. In one match, one moment, it frequently is not.

The Hold Team Meeting and Convene board actions are the honest part of the model: they consume the scarcest resource in the game, which is time.

What to take away

Certainty is only reliably available in simple systems. In complex and chaotic ones, waiting burns the opportunity — markets move, competitors act, and eventually not deciding becomes the decision. This is the argument underneath Lean Startup and most modern product practice: early markets are clouds, not clocks, so you learn by shipping something and measuring what happens.

PHILOSOPHY
MANAGER '94

Greece v Germany. Confucius has the whistle. The philosophers have the ball, theoretically. Manage analysis, confidence, morale and time before the match collapses into metaphysics.

Match Centre

GREECE 0
00:00
0 GERMANY
World Cup Final · Weather: Metaphysical · Pitch: Epistemologically Uncertain · Attendance: 72,431

Competition Mode

Tip: choose decisions as the commentary unfolds. More analysis is not always bad. The mistake is treating every system as if it were the same kind of problem.

Commentary & Reports

Monte Carlo runs 1,000 matches using the current competition mode and your present managerial style. The analysis curve shows where analysis helps, plateaus, or destroys the opportunity.

Analysis Time Curve

Outcome Table

OutcomeCountPercentage

How to use this simulation

Change one input at a time, run the model, and compare the result with the starting state. Then repeat the experiment with a different constraint or strategy so you can see which relationships drive the outcome.

What to look for

Look for trade-offs, thresholds, feedback loops, and points where a locally attractive decision produces a worse system-wide result. The simulation is intended to make the article's idea observable, not to predict a real operation.

Limitations

This is a deliberately simplified model. It omits the data quality, exceptions, human judgement, and operational constraints of a live system, so treat its behaviour as an illustration of a mechanism rather than as a planning recommendation.

Back to Simulations