── ── Decision-making · Trace backward from failure
Inversion
Inversion flips a problem: instead of asking 'how do I succeed?', you ask 'how could this fail catastrophically?'—then build the plan around systematically eliminating the failure paths that matter most. Avoiding the ways to lose is often more tractable, and more decisive, than chasing the ways to win.
How it works
Start from the outcome you most want to avoid and reason backward to the conditions that would produce it. Enumerate the specific failure modes, rank them by how likely and how fatal each is, and direct your effort at neutralizing the few that could actually kill the plan.
Inversion works because catastrophic outcomes usually have a small number of identifiable causes, while success has many diffuse ones. Removing the known killers is concrete and verifiable, whereas 'do everything right' offers no clear target to act on.
When to use it
- Pre-mortem on a launch, fundraise, or major hire before committing
- Designing a system where a single failure is unrecoverable (security, payments, data loss)
- A plan everyone is excited about and no one has stress-tested for downside
- Allocating scarce attention—deciding which risks are worth engineering against now
When not to use it
In open-ended creative or exploratory work where the goal is to expand possibilities—fixating on failure modes too early can kill ideas before they have room to develop.
Worked example
Apollo 1 FMEA, 1967
After the Apollo 1 cabin fire killed three astronauts, NASA leaned hard on failure-mode-and-effects analysis: systematically asking, for every component, how it could fail and what that failure would cause. Rather than only asking how to reach the Moon, engineers catalogued the ways the mission could kill its crew and redesigned around eliminating those paths—the discipline that made later flights survivable.
Why it matters for founders
Most founder energy goes into visions of how things go right, while the company actually dies from a short list of avoidable failures—running out of cash, a co-founder split, a single dependency with no fallback. Inversion forces you to name those killers early and engineer them out while it's still cheap. deciqAI's agents run the failure analysis before acting, not as a post-mortem after the damage is done.
Install this skill (free, MIT)
npx skills add deciqAI/knowledge-skillsFAQ
How is inversion different from a risk assessment?
A risk assessment often catalogs risks as a checklist; inversion makes avoiding catastrophic failure the organizing goal of the plan itself, working backward from the worst outcome to the design decisions that prevent it.
When does inversion beat thinking about how to succeed?
When failure is concentrated in a few identifiable causes and is costly or irreversible. Eliminating known killers is more actionable than the diffuse, many-paths question of how to win.