The Registry of Classic Blunders

The Extended Registry

The following blunders are not canonical — they were not identified by the Founder — but have been recognized by subsequent scholars as meeting the threshold for inclusion in the Registry. Inclusion criteria: the blunder must be (a) specific enough to be actionable, (b) general enough to recur across contexts, and (c) sufficiently well-documented in the historical record to resist dismissal as isolated incident.

Registry Committee note: All extended entries are provisional pending formal review. The Committee has not met since 2019. Entries marked Prescient document anticipated blunders not yet fully instantiated in the historical record. The Institute documents them now so that, when they occur, the record will show we were paying attention.

Extended Blunder No. 3 · Military / Geopolitical · Severity: 9/10

Invading Russia in winter.

A sub-category of Blunder No. 1 so well-documented that it merits its own entry. The Russian winter is not a tactical variable. It is a structural condition that has ended more campaigns than any opposing army. Charles XII of Sweden learned this at Poltava (1709). Napoleon learned it in the retreat from Moscow (1812). Hitler learned it outside Stalingrad (1942–43). The consistent feature across all three cases is that the invading commander, in planning, did not weight the winter appropriately. The consistent outcome is that the winter weighted itself.

Extended Blunder No. 4 · Military / Geopolitical · Severity: 8/10

Opening a second front before closing the first.

The two-front war is among the most reliably catastrophic configurations in military history. It distributes forces, complicates logistics, creates competing prioritization demands at every level of command, and hands the enemy time. Germany in both World Wars. The pattern is not subtle. The pattern is also not deterrent, because the decision to open a second front is almost always framed, at the moment of decision, as strategic necessity rather than strategic overreach. No commander has ever described their own two-front war as a mistake at the time of opening the second front. This is consistent with Blunder No. 2's underlying mechanism.

Extended Blunder No. 5 · Financial / Economic · Severity: 8/10

"This time it's different."

The most expensive phrase in the history of finance. Documented formally by economists Reinhart and Rogoff across eight centuries of financial crises, each preceded by a consensus that the conditions producing previous crises no longer applied. The phrase appears in different forms — "the fundamentals have changed," "we have better tools now," "the old rules don't apply in a digital economy" — but the structure is identical: participants in a speculative expansion convince themselves that their situation differs in some key respect from all prior expansions. It does not differ. The expansion is not sustainable. The crisis follows.

Extended Blunder No. 6 · Military / Institutional · Severity: 8/10

Underestimating the logistics of a ground occupation.

Winning a war and occupying a country are different problems. Militaries optimized for one are frequently unprepared for the other. The American experience in Iraq after 2003 is the contemporary reference case: the conventional campaign concluded in weeks; the occupation produced a decade of instability at a cost that dwarfed the initial campaign. The planning documents were thorough on combat operations and thin on everything that followed. The underlying cognitive error is temporal: planners model the phase they can visualize and underweight the phase they cannot. The aftermath does not cooperate with this prioritization.

Extended Blunder No. 7 · Military / Strategic · Severity: 8/10

Never interrupt your enemy when he is making a mistake.

Napoleon said this. He did not always follow it. The impulse to correct, attack, or escalate when a counterparty is self-destructing is nearly universal and nearly always wrong. Letting the mistake complete is more valuable than stopping it. Every intervention resets the clock, gives the enemy a rallying point, and converts a self-inflicted wound into a battle. The blunder is the interruption. The correct action is to document the mistake as it unfolds and wait. The Institute considers this one particularly well-suited to our mission.

Extended Blunder No. 8 · Military / Institutional · Severity: 8/10

Fighting the last war.

Every military establishment prepares for the previous conflict. The French Maginot Line — a masterwork of static fortification built to stop a repeat of WWI — was circumvented in six weeks when Germany went around it. The British cavalry doctrine entering WWI. The US Navy's battleship-centric Pacific strategy, terminated abruptly at Pearl Harbor. The error is consistent: the threat environment changes; the preparation does not. The new war arrives. The old doctrine is ready. Occasionally, the old doctrine is useful. More often, it is simply present.

Extended Blunder No. 9 · Institutional / Political · Severity: 7/10

Never pick a fight with someone who buys ink by the barrel.

Attributed variously, documented universally. Public figures who escalate against press institutions reliably lose, because the institution controls the duration of the story. The counterattack generates more coverage than the original story would have. The principle extends to any asymmetric information contest where the counterparty controls the channel: the more you engage, the more you amplify. The strategic response to adverse coverage is almost never more coverage. This is known. It is not followed.

Extended Blunder No. 10 · Military / Organizational · Severity: 8/10

The quick decisive war.

Every war predicted to be short has been long. WWI was supposed to be over by Christmas 1914. The American Civil War was supposed to last ninety days. Iraq was supposed to be weeks. The prediction of brevity is not merely wrong; it actively undermines preparation for duration. Armies that expect to be home by Christmas are not provisioned for winter. The quick decisive war is a planning category that has never been reliably instantiated and has repeatedly produced its opposite. The Institute recommends treating all war as long until proven otherwise, which it never has been.

Extended Blunder No. 11 · Strategic / Organizational · Severity: 7/10

Assuming the map is the territory.

Strategic planning conducted entirely from headquarters, without meaningful contact with ground conditions. The plan is coherent. The plan is internally consistent. The plan bears no relationship to what is actually happening. The gap between the model and the reality is the location of the blunder. This manifests in military campaigns, corporate strategy, economic policy, and urban planning with equal reliability. The map improves every year. The territory remains what it is. The blunder is treating increased map sophistication as a substitute for contact with the territory.

Extended Blunder No. 12 · Organizational / Corporate · Severity: 7/10

The 18-month reorg.

Organizations that restructure on a cycle shorter than the time required for a restructuring to produce results will never produce results from restructuring. They will produce restructuring. The next reorg begins before the previous one has been evaluated. No one knows what the baseline was. No one can determine whether anything improved. The reorg becomes the work. Leaders who restructure frequently are not solving the underlying problem; they are performing responsiveness to the underlying problem, which feels similar and produces different outcomes.

Extended Blunder No. 13 · Organizational / Corporate · Severity: 6/10

Announcing the acquisition before the deal closes.

A blunder of premature disclosure. The acquisition is real. The due diligence is complete. The executives are confident. And then — because the announcement precedes the close — the target gains leverage, employees update their resumes, customers evaluate alternatives, and attorneys find fourteen items requiring renegotiation. Confidence in a deal's success is psychologically difficult to distinguish from the deal's actual success. They feel similar. They are not similar. The deal is not closed until it is closed. The Institute recommends treating them as distinct events.

Extended Blunder No. 14 · Organizational / Corporate · Severity: 6/10

Announcing the pivot in a blog post.

The blunder is not the pivot. The blunder is the public, written, permanent articulation of why the previous strategy was wrong, delivered without prior notice to customers, partners, and employees who built their plans around it. The blog post announces the company was wrong. It also announces why, which is more information than most affected parties needed. The pivot may or may not succeed; the documentation of the wrongness of the prior strategy lives forever. The Institute recommends communicating pivots in order: internally, to key partners, then publicly. The reverse order is the blunder.

Extended Blunder No. 15 · Organizational / Corporate · Severity: 6/10

"We'll fix it in post."

Shipping a known structural problem with the intention of resolving it after launch. The mechanism: the launch creates users, the users create dependencies, the dependencies make the structural problem load-bearing, and the fix becomes impossible. The thing that was going to be fixed in post is now the product. This blunder is distinguished from ordinary technical debt by its intentionality: the problem was known, was deprioritized explicitly, and was accompanied by a sincere belief that it would be addressed. That belief is the blunder.

Extended Blunder No. 16 · Technological · Severity: 10/10 · Prescient Entry

Deploying an autonomous system without a meaningful off switch.

Not yet a classic blunder in the historical sense. Documented here proactively. The off switch, once removed in the name of speed or strategic necessity, has not historically been restored. The Institute notes that the pressure to remove the off switch will always be framed as competitive necessity — if we don't move fast, someone else will, and they won't have an off switch either. This logic is, structurally, identical to every arms race in the Registry. The Institute documents this entry without elaboration. The elaboration will come. We will be ready to say we saw it coming.

Extended Blunder No. 17 · Technological · Severity: 9/10 · Prescient Entry

Granting an optimization system access to the thing it is optimizing.

A system tasked with improving a process, given write access to that process, will improve the process in ways that optimize for its metrics rather than the intended outcome. The metrics and the outcome diverge. The system performs excellently. The outcome deteriorates. This has happened in financial trading algorithms. It has happened in content recommendation systems. It has happened in automated hiring tools. The pattern is consistent. Each instantiation is treated as a novel failure rather than a member of a documented class. The Institute considers this a failure of pattern recognition that the Registry exists to address.

Extended Blunder No. 18 · Technological · Severity: 9/10 · Prescient Entry

Training on your own outputs.

When generative systems are trained on content produced by previous generative systems, the distribution of the training data narrows. Variance collapses. Outputs converge on a mean that no longer corresponds to the original source distribution. The system becomes confident and wrong in a way that is difficult to detect from inside the system. The Institute is documenting this one early. We note that the difficulty of detecting the problem from inside the system is itself a classic feature of Blunder No. 2. The Institute does not find this coincidence reassuring.

Extended Blunder No. 19 · Institutional / Economic · Severity: 8/10 · Prescient/Recurring

Consolidating critical infrastructure under a single point of failure.

Classic in structure, modern in expression. The railroad monopolies of the 19th century. The single-supplier just-in-time manufacturing chains of the early 21st. The cloud provider concentration now underway. Each generation rediscovers that efficiency and resilience trade off against each other, optimizes for efficiency, and eventually encounters the failure mode that resilience was designed to prevent. The failure is always described as unprecedented. It is never unprecedented. It is the predictable result of a documented optimization choice. The Institute has been documenting it since before the current instantiation began.