The Ripple Room

Name the loop, then mind the delay

Every structure under Meridia is a feedback loop of one of two kinds. A reinforcing loop (R) amplifies itself — more leads to more, a vicious or virtuous cycle. A balancing loop (B) pushes back toward a goal — it resists change and seeks equilibrium. Click a dynamic on the left, then the label-plus-consequence on the right that fits it. Link all four, then check.

Mind the delay

The council finally approves 2,000 new homes — enough to genuinely ease the shortage. Here is the question that breaks most decision-makers: from the day approval is signed, how many months until renters actually feel rents stabilise? Drag to your best guess, then reveal the true curve.

Run the system: the overshoot trap

You are now Meridia’s transport chief. The commute is 45 minutes; the target is 30. Each round you set one control: how much road capacity to add (or remove). Here is the cruel part — capacity you add this round only takes effect next round (construction delay), and every time roads feel faster, induced demand pulls more cars on until they fill up again. Slam the lever and you will watch the meter overshoot past the target and rebound the wrong way. Drive for five rounds.

Trace the ripple

1. Spot the hidden loop

Here is the causal-loop diagram the council used to justify widening the Harbor Road. It treats widening as a simple open chain that ends in “shorter commute.” That is exactly why the road re-congested in nine months. Click the two links that are missing or wrong — the ones that, once added, turn this open chain into the induced-demand balancing loop that erased the gain. Then check.

2. Pick an intervention — watch the ripple

The mayor wants one intervention on Meridia, funded now. Pick one and watch its first-, second-, and third-order ripples unfold — then try the others. The meter shows where the commute lands 18 months later. It can move the wrong way.

Climb the leverage ladder

1. Which lever is stronger?

Two proposals land on the mayor’s desk, both aimed at the commute. One tweaks a number inside the existing system. One changes what the system is for. Pick the higher-leverage intervention — then check your reasoning.

2. Order the leverage ladder

Donella Meadows ranked the places to intervene in a system from weakest to strongest. The surprise: the levers that are easiest to reach and argue about — fees, buffers, numbers — are the weakest. Order these six Meridia interventions from weakest leverage (top) to strongest (bottom), then check.

3. Find the wrong-direction push

Meadows’ deepest warning: people often find the high-leverage point — and then push it the wrong way, “with a great deal of enthusiasm.” Here are six items on Meridia’s budget. Select every one that aims at a real leverage point but shoves it in the wrong direction (making the system worse). The skill is exhaustiveness.

Design the intervention, then carry it out

1. Build a counter-loop

Now you design the intervention package. Toggle structural elements on and off and watch the system-health meter respond. Structural moves — adding a balancing loop, shortening a delay, changing the goal — lift it. Parameter tweaks barely move it, and one seductive option lowers it: feeding the reinforcing sprawl loop. The meter can fall.

2. Your diagnosis — a fresh system

New case, same moves. Read it, then write three things: the loop driving it, the archetype it matches, and the highest-leverage move — not the obvious one. Then self-assess against the checklist and compare with a model answer.

3. Carry it out

The whole point of the Ripple Room is what happens after you leave it. Here is your protocol on a card — then name one system in your own work where an AI or a KPI is busily optimising a symptom, and state the higher-leverage point it’s missing.