What you are seeing
The Lorenz system produces qualitatively different behaviours depending on parameter values. The scenarios on the left are pre-set configurations demonstrating five distinct regimes — from stable convergence to full chaos. Load them in sequence, then use the sliders to explore the boundaries between regimes.
The key observation: the transition between regimes is not gradual. The system looks stable — until it doesn't. The threshold is not visible in any single parameter reading.
The parameters
| a | Rate of internal response — how quickly the system reacts to imbalances between its variables |
| b | Intensity of external pressure — the critical parameter. Contains the bifurcation point |
| c | Damping — the system's capacity to dissipate energy and resist departure from equilibrium. Must reach ~8–10 to overcome chaos at b=28 |
Parameters are intentionally abstract. The structure of the behaviour is the message — not the specific metaphor applied to it.
The Lorenz System
In 1963, meteorologist Edward Lorenz was studying atmospheric convection — the rolling circulation of fluid heated from below. He derived the simplest possible equations that could describe this phenomenon:
dy/dt = x · (b − z) − y
dz/dt = x · y − c · z
Three variables. Three equations. Entirely deterministic. And yet, for certain parameter values, the system produces behaviour that is sensitive to initial conditions, non-periodic, and unpredictable over time. This was one of the foundational discoveries of chaos theory.
The critical bifurcation occurs at approximately b ≈ 24.74 (with a = 10, c = 8/3). Below this value, all trajectories eventually settle. Above it, chaos is the only attractor.
Five Behavioural Regimes
1. Stable equilibrium (b = 8). All trajectories converge to one of two fixed points. The system has a global point attractor. This is the assumption embedded in most linear planning frameworks: systems return to equilibrium after perturbation.
2. Spiral convergence (b = 16). The fixed points are losing stability, but the system still converges — slowly, through a widening spiral that eventually settles. The system is not at rest, but it is bounded. A useful analogy for organisations under moderate pressure: oscillating, but not yet escaping.
3. Edge of chaos (b = 24). Near the bifurcation threshold. Orbits are wide and irregular. The system has not yet crossed into full chaos, but the attractor is becoming strange. Small perturbations have disproportionate consequences.
4. Full bifurcation (b = 28). The classic Lorenz attractor. Trajectories jump unpredictably between two lobes, never repeating. Long-term prediction is structurally impossible. This is the butterfly effect.
5. High damping (b = 28, c = 10). The same external pressure as full chaos — but with high damping. The system converges. This is the key contrast: with c = 2.67, b = 28 produces chaos. With c = 10, the same pressure produces convergence. Resilience is not just a delay — at sufficient levels, it changes the regime entirely. But the threshold in c is high. Moderate resilience (c = 5) is not enough. This is the uncomfortable message: half-measures in resilience-building may delay collapse without preventing it.
Bifurcations and Tipping Points
A bifurcation is a qualitative change in the structure of a system's behaviour as a parameter crosses a critical value. The change is not gradual — it is discontinuous. Below the threshold, the system behaves one way. Above it, it behaves in a fundamentally different way.
This is structurally different from linear degradation. In a linear system, doubling the pressure doubles the response. Near a bifurcation, a small additional increment of pressure shifts the system into an entirely different regime. The response is disproportionate to the cause.
The bifurcation threshold is not visible in any single variable. The system looks stable until it isn't. This is not a failure of measurement sophistication. It is a structural property of non-linear systems.
Sensitivity to Initial Conditions
In the chaotic regime, two trajectories that start arbitrarily close together diverge exponentially. After a short time, they are completely uncorrelated. This is the butterfly effect — not a metaphor, but a mathematical property.
The practical implication is a hard limit on prediction: beyond a certain horizon, forecast is meaningless not because we lack information, but because the structure of the system makes it structurally impossible. Deterministic does not mean predictable. The equations fully determine what will happen — but sensitivity to initial conditions means that any measurement error compounds exponentially.
Applications and Honest Limits
The Lorenz equations were derived for atmospheric convection. They have been applied with rigour to lasers, chemical reactions, electrical circuits, and population dynamics. Applications to social and organisational systems are more contested — suggestive but not empirically robust.
This simulator is not a model of your organisation. It is a model of a class of behaviours — tipping points, bifurcations, disproportionate responses, sensitivity to initial conditions — that are well-documented in complex systems and plausible in organisational contexts. The equations are the vehicle. The behaviour they produce is the argument. We use this model to make a structural claim visible, not to make predictions about specific systems.
An Interpretive Example
The parameters a, b, c are intentionally abstract. But it helps to anchor them in a concrete system. One plausible reading is a governance system under pressure — where the three state variables are:
y — social pressure (demands, expectations, mobilisation)
z — capacity to respond (resources, coherence, mandate)
Read through this lens, the equations say something recognisable:
dx/dt = a(y − x) — Legitimacy adjusts toward the level of social pressure. When pressure exceeds current legitimacy, the institution is forced to adapt or lose ground. The parameter a controls how quickly this happens — a slow bureaucracy or a responsive one.
dy/dt = x(b − z) − y — Social pressure grows when legitimacy is high but response capacity is low: people believe change is possible but it isn't happening. It dissipates naturally over time (−y). The parameter b is structural pressure from outside — crisis, economic shock, climate disruption. It is the parameter the institution cannot control.
dz/dt = xy − cz — Response capacity builds when there is both legitimacy and pressure simultaneously — when the institution has mandate and reason to act. It erodes naturally (−cz). The parameter c is institutional resilience: memory, culture, the capacity to absorb tension without fragmenting.
Under this reading, the five scenarios describe recognisable governance situations. A stable democracy with moderate external pressure (scenario 1). A system absorbing rising demands but still converging (scenario 2). A system near its institutional limits (scenario 3). Chronic instability — oscillating between crisis and partial recovery, never settling (scenario 4). And a system with the same structural pressure as scenario 4, but deep enough institutional resilience to hold (scenario 5).
This is one possible interpretation of the abstract variables — not a validated model of governance dynamics. The equations were not derived from governance theory. We are borrowing the structure of the behaviour, not claiming that governance systems follow Lorenz dynamics in a strict sense. The value is in the questions it prompts: where is your system in this phase space? What is your b? How high does your c need to be?
Further Reading
Lorenz, E.N. (1963). Deterministic nonperiodic flow. Journal of Atmospheric Sciences, 20(2), 130–141. — The original paper.
Strogatz, S. (2015). Nonlinear Dynamics and Chaos. Westview Press. — The standard graduate text. Chapters 9–10 cover the Lorenz system and the full bifurcation sequence.
Gleick, J. (1987). Chaos: Making a New Science. Penguin. — The accessible account of the discovery of chaos theory.
Meadows, D. (2008). Thinking in Systems. Chelsea Green. — Best practitioner introduction to feedback, tipping points, and non-linear behaviour in social and ecological systems.
Bateson, G. (1972). Steps to an Ecology of Mind. University of Chicago Press. — The philosophical foundation for applying systems thinking to human behaviour and organisations.