JH
Jonathan Haber

Understand why systems oscillate — and stop overcorrecting

Delayed feedback loops and overreaction to perceived gaps cause the boom-bust cycles in your own system.

Key takeaways

  • What it is: Delayed feedback loops and overreaction to perceived gaps cause the boom-bust cycles in your own system.
  • Why it works: Oscillation in a system arises when a balancing loop has a significant delay between action and feedback, combined with a decision-maker who overreacts to the perceived gap. The classic example is a shower where the temperature takes time to respond: people who turn the knob too far (overcorrect) end up oscillating between hot and cold. Meadows identifies this as a fundamental dynamic in any system with delayed feedback, and the fix is the same: smaller corrections, longer patience, and explicit recognition of the delay.
  • Evidence: Backed by observational / correlational evidence.
  • Avoid: Diagnosing oscillation as an external problem — "things keep going up and down" — rather than as a result of your own correction behavior interacting with the delay.

Why it works

Oscillation in a system arises when a balancing loop has a significant delay between action and feedback, combined with a decision-maker who overreacts to the perceived gap. The classic example is a shower where the temperature takes time to respond: people who turn the knob too far (overcorrect) end up oscillating between hot and cold. Meadows identifies this as a fundamental dynamic in any system with delayed feedback, and the fix is the same: smaller corrections, longer patience, and explicit recognition of the delay.

How to do it

  1. 1When you notice a boom-bust pattern in your own behavior or results, ask: "Where is the delay between my action and the feedback?"
  2. 2Estimate the delay time.
  3. 3Make smaller adjustments and wait for the full delay to elapse before making the next one.
  4. 4Resist the urge to overcorrect — the oscillation is driven by correction magnitude, not by the underlying trend.

What the evidence says

Observational

Oscillation from delayed balancing loops is mathematically derived in control theory and system dynamics, and is demonstrated in simulations (beer game, inventory management). Human overcorrection in delayed-feedback settings is well-documented in judgment research.

Honest caveat: Damping oscillation through smaller corrections requires accepting slower convergence; in genuine emergencies, faster response may be necessary despite the oscillation risk.

References
  • — Sterman (1989), "Misperceptions of feedback in dynamic decision making," Organizational Behavior and Human Decision Processes — overcorrection and oscillation in delayed feedback systems

Common mistake

Diagnosing oscillation as an external problem — "things keep going up and down" — rather than as a result of your own correction behavior interacting with the delay.

IX Coach identifies oscillating patterns in your tracked behaviors and helps you redesign your correction strategy — smaller adjustments, longer evaluation windows — to damp the cycle rather than amplify it.

Practice this with IX Coach →

Related practices