What goes wrong when you ignore that the extinction timeline?

The most common reason fails: Treating any resurgence of the old behavior as evidence that extinction has failed completely. Because people expect a linear decline, any uptick — especially a spontaneous recovery episode in week three or four — is interpreted as "back to square one." This triggers abandonment or, worse, a full relapse into the old reinforcement pattern. The timeline model predicts these upticks. They are data points on a non-linear curve, not resets to the origin.

The fix: Select a behavior you are currently trying to extinguish or have recently attempted to extinguish. Draw a simple graph on paper with the x-axis labeled "Days" (mark intervals from 0 to 90) and the y-axis labeled "Urge Intensity" (scale 1-10). First, draw the line you expected — what you assumed the decline would look like when you started. Most people draw something close to a straight diagonal from upper-left to lower-right. Now draw the line that actually happened or is happening — with the burst at the beginning, the uneven decline, any spontaneous recovery episodes, and any plateaus. Compare the two lines. Where do they diverge most? That divergence zone is where you are most likely to misread normal progress as failure. Finally, project the curve forward: based on what you now know about non-linear extinction, when would you realistically expect to reach your asymptote? Write that date down.

The underlying principle is straightforward: Behavioral extinction takes time — weeks or months depending on how established the behavior is.