Team of Teams

McChrystal's task force was better trained, better equipped and better resourced than its opponent, and was losing. His diagnosis is that the opponent was faster, more adaptable and organised as a network rather than a hierarchy.

His conclusion is that the structure built for efficiency in a predictable environment was the wrong structure for an unpredictable one — and that the resource advantage could not compensate.

Better trained, better equipped, better resourced, and losing. The structure was the variable.

Devmark’s reading — not a quotation from the book

The book's most useful conceptual distinction. A complicated system has many parts and is predictable — an engine, a factory. A complex system has interdependent parts whose interactions produce emergent, unpredictable behaviour.

Their claim is that management practice is built for complicated systems, and that applying it to complex ones produces confident plans that are obsolete before execution.

Their attack on the Taylorist inheritance: that optimising each component for efficiency produces a system that is fast at what it was designed for and incapable of anything else.

Their alternative is deliberately trading some efficiency for adaptability — accepting duplication, slack and redundancy because they are what permit response to the unforeseen.

The first half of their solution: systematic, radical transparency of information across the whole organisation, so that anyone can understand the overall situation.

Their mechanism was a daily operations and intelligence briefing involving thousands of participants — expensive, apparently wasteful, and justified on the grounds that the cost of everyone understanding was lower than the cost of nobody being able to act without permission.

The second half, and the one that requires the first. Decision authority is pushed down to people who now have the context to use it well.

Their sequencing is the argument: empowerment without shared consciousness produces incoherent local decisions, and shared consciousness without empowerment produces a well-informed organisation that still cannot act. Both, together, are the mechanism.

Empowerment without shared context produces chaos. Shared context without empowerment produces informed paralysis.

Devmark’s reading — not a quotation from the book

McChrystal's reframing of his own role: not directing moves but creating the conditions in which good decisions can be made by others.

His account of deliberately withdrawing from decisions he was capable of making well, because making them was preventing the organisation from learning to make them, is the most personally candid material in the book.

The book is about a military counter-terrorism campaign, and it discusses organisational effectiveness in a context with serious moral and political dimensions that it largely brackets.

A reader can take the organisational argument while noticing that the book does not engage with the wider questions about the campaign, and that this omission is a choice.

It is a single organisation's transformation, narrated by its commander, with outcomes attributed to the restructuring.

The transparency prescription is also genuinely costly, and the book underweights the cases where information sharing has real risks — security, privacy, competitive exposure — that a military task force could largely disregard internally.

It alternates between the operational narrative and broader material on systems theory, organisational history and complexity science, some of which is thin and some of which is padding drawn from secondary sources.

The historical digressions — Taylorism, the Prussian general staff, the Apollo programme — are the weakest part and the least necessary. A reader losing patience can follow the task-force narrative alone and skip them entirely without losing any part of the argument, which is a fair indication of how much work they were doing.

The complicated/complex distinction, applied as a test: does your planning process assume the environment will hold still long enough for the plan to be executed?

And the sequencing insight — that you cannot delegate authority to people who do not have the context, and that building the context is the expensive part everyone tries to skip.