Principle · Defense and Kaizen

Inversion.

Source: Charlie Munger, vice chairman of Berkshire Hathaway, in repeated public talks (1986-2020), most accessibly collected in Poor Charlie's Almanack (Peter D. Kaufman, ed., 2005). Munger attributes the principle to Carl Gustav Jacob Jacobi, the 19th-century German mathematician, whose phrase was "Man muss immer umkehren" -- "Invert, always invert."

The Principle

To solve a hard problem, do not start by asking how to succeed. Start by asking how to fail, then work backward to avoid every cause. Most thinking moves forward: "How do I get to the outcome I want?" Inversion moves backward: "What would guarantee I do not get to the outcome I want?" Listing the failure modes is often easier than listing the success paths, because failure has fewer degrees of freedom and is more legible.

Munger's argument is that "avoiding stupidity" is a more reliable path to compounding success than "seeking brilliance." Brilliance is rare, hard to predict, and often misattributed in hindsight. Stupidity is common, identifiable in advance, and avoidable. A business that systematically eliminates the most common failure modes will outperform one that relies on occasional brilliant moves, because the failure modes show up reliably and the brilliant moves do not.

Inversion is not pessimism. It is a problem-solving technique. The goal is the same -- a successful outcome. The path is the negative space: identify what would prevent success, then design the avoidance into the system.

Why It Matters Here

Defense and Kaizen exists to find the failure modes the rest of the team is too busy to see. Forward-looking executives are paid to notice the upside opportunities. Defense is paid to notice what would make the company fail and to prevent it before it happens. Inversion is the operating tool for that work. Without it, defense becomes reactive (waiting for breakage). With it, defense becomes anticipatory (asking, in every situation, what would make this fail, and then making sure the answer never appears).

Signals (When to Apply)

A new project, product, or partnership is being scoped, and the planning is dominated by upside scenarios
The team is stuck on "how do we make this work?" and not making progress
A recurring problem keeps reappearing despite attempts to fix it
An audit, retrospective, or pre-mortem is being conducted
A decision has been framed as a yes-or-no, and the cost of "yes-and-it-fails" has not been examined

How to Apply

Reverse the question. Whatever the team is trying to figure out, ask the inverse: "What would guarantee we do not achieve this?" Brainstorm as a list. Treat the list as a checklist of things to actively prevent.
Apply at the start of any project, not just at the end. The pre-mortem is the structured version. Lighter inversion ("what could make this fail?") fits in any planning conversation.
Apply to the operating system itself, not just to projects. "What would make our context cascade fail?" "What would make our decision log become useless?" "What would make our weekly rhythm collapse?" Then design the avoidance.
Look for the "stupidity" patterns first -- the obvious, repeated failure modes other companies in similar situations have hit. Most failure is not novel. Most failure is a known pattern landing on a new company that did not study the pattern.
Distinguish between "could fail" and "would fail." Inversion is most useful when applied to the would-fail items: the failure modes that actually have probability mass, not the theoretical ones.
Pair inversion with action. A list of things that could go wrong is only useful if it produces a list of things to change. End every inversion exercise with at least one concrete fix shipped.
Use it publicly in cross-department conversations. The role-specific cultural contribution Defense makes to the team is the question "what would make this fail?" Asked respectfully and consistently, it sharpens every plan.

Examples

Applied well A team is launching a new product offering. Marketing has the launch plan, Sales has the pipeline forecast, Operations has the delivery process. Defense applies inversion in the planning meeting: "What would guarantee this launch fails?" The list comes back fast: (1) the website breaks under traffic, (2) the first three customers have a bad experience and post about it, (3) the support inbox is unattended for the first 48 hours, (4) the founder is unreachable for the first week. Each item is now a checklist item to actively prevent: load test the site, hand-pick the first three customers and over-deliver, staff the inbox, block the founder's calendar. The launch ships and none of the four failure modes appears. The forward-only planning would have caught maybe two of the four.

Misapplied A team uses inversion to enumerate every conceivable failure mode for every decision, including small reversible ones. The list of risks for each weekly content post grows to 14 items. Risk-mitigation overhead consumes the time that was supposed to go to content production. Output drops by half. The principle, applied to two-way doors at the same intensity as one-way doors, became its own failure mode. Inversion is sized to the asymmetry of the decision, not applied uniformly.

When to Break It

For small, reversible, low-stakes decisions where the cost of inversion analysis exceeds the cost of just trying and learning. Reserve full inversion for the high-stakes situations.
When the team's morale or confidence is fragile and a sustained focus on failure modes would be demoralizing without changing the actions taken. Pair inversion with concrete next steps so it generates progress, not anxiety.
When the question is genuinely about creation rather than risk reduction (designing a brand-new product feature, naming a brand). Inversion is the wrong tool for generative work; it is the right tool for protective work.