Use the 10-year horizon as a regret test

Ask: in 10 years, will I regret not doing this more than doing it?

Why it works

Jeff Bezos has described a similar regret minimization framework: project yourself to age 80 and ask which choice you would regret more. The mechanism is identical — vividly imagining a future self reduces temporal discounting and shifts weight from immediate discomfort to long-run values. The 10-year framing is shorter and more operationally useful than age-80 projection for most decisions.

How to do it

1. At the 10-year step, ask specifically: "Will I regret not doing this?" and "Will I regret doing this?"
2. Articulate which regret feels larger or more likely.
3. Distinguish between regret about outcomes ("it didn’t work") and regret about the choice given what you knew ("I shouldn’t have tried").
4. Let the regret test break ties when the other horizons point in opposite directions.

Evidence

Regret research (Gilovich, Medvec) suggests people more often regret inactions than actions over the long run, while the reverse holds in the short run — the time horizon shifts the dominant regret type. This is consistent with the 10-10-10 framework. Construal-level theory offers a mechanism for the shift: the distant 10-year frame recruits abstract, values-level thinking, where forgone opportunities loom larger than concrete missteps. (observational)

Regret research focuses on retrospective report; prospective regret minimization has plausible validity but fewer direct studies.

Sources

• Gilovich & Medvec (1995), "The Experience of Regret: What, When, and Why", Psychological Review
• Gilovich, T., & Medvec, V. H. (1995). The Experience of Regret: What, When, and Why. Psychological Review, 102(2), 379–395.
• Trope, Y., & Liberman, N. (2010). Construal-Level Theory of Psychological Distance. Psychological Review, 117(2), 440–463.

Common mistake

Imagining the 10-year outcome and then discounting it emotionally because it is far away — which is precisely what the exercise is designed to counter.