Guiding as Leadership Formation

Question. Does guiding develop the guide, or just the operator? Is the leadership pipeline real or decorative?

Analogy. It works like a master taking on an apprentice: keep your own hands on the tools while you teach and naming each move sharpens you, the better get better by guiding; but rise into pure direction, never touch the work yourself, and the feel quietly leaves your hands until the day you reach for the tool and find it slow. The study asks whether guiding develops the guide so the leadership pipeline is real and self-reinforcing rather than decorative, and whether a finding that heavy guiding erodes the guide's own edge forces mandatory rotation between doing and guiding rather than leaving it optional.

What's at stake. The proposal sells one architecture with three co-equal outputs: resilience, innovation, and a leadership pipeline. The leadership claim is that guiding is leadership formation running as a byproduct, because the four-axis reframing the guide does for an operator is the same capacity that changes organisations. The Impact chapter makes this a headline benefit and then concedes the pilot does not measure it. If guiding genuinely develops the guide, the pipeline is real and self-reinforcing (better operators become guides, guiding makes them better operators, who guide better) and it is the retention mechanism that keeps guides in the network. If guiding leaves the guide unchanged or atrophies their edge, the leadership-pipeline output is decorative, the retention model loses a leg, and the proposal should claim two outputs, not three.

The two answers it decides between. Either guiding develops the guide (people who have guided outperform matched non-guides on their own operator work, controlling for total platform time, and the four-axis capacity transfers to the contexts they later operate in), or it does not (no operator-performance difference, or a negative one where sustained guiding erodes the guide's own doing-edge). The controlled comparison, holding platform time constant so the variable is guiding and not tenure, tells them apart, and the sign of the effect decides whether role rotation must be built into the architecture rather than left optional.

What a null result would mean. If guiding does not develop the guide, the leadership-pipeline claim is not supported and the one-architecture-three-outputs case reduces to two outputs: a finding about the proposal's theory, not a platform defect. A negative effect, where guiding erodes operator edge, is more than a null: it would mean the architecture needs mandatory rotation between operating and guiding, an actionable design change, not a dead end.

Why this matters to defence. Leadership-capable personnel who can ask system-changing questions are a named strategic shortage the Impact chapter ties to the Defence Industrial Strategy, Arctic sovereignty, and NORAD modernisation (DRDC Objective 2; the institution and force development). If guide service measurably builds that capacity, the platform is a distributed leadership-formation programme running as a byproduct of its primary purpose, and the guide population is a pipeline DGMPRA can track into the CAF, the public service, and the innovation ecosystem. It changes a concrete decision: whether to value and instrument the guide role as a leadership-development asset, or treat it only as operator support.

How we would run it. Compare the operator-side performance of people who have served as guides against matched operators who have not, controlling for total platform time so the variable is guiding rather than tenure. Measure operator skill on the wargame and the four-axis reframing capacity directly, and test transfer by whether the reframing capacity shows up in the guide's subsequent non-platform contexts, a DGMPRA-scale follow-up across later employers. Include the player-turned-commentator condition, heavy guides who guide far more than they operate, to detect whether sustained guiding erodes the guide's own edge. The naturalistic role-reversal of a real self-paired population, operators who become guides and back, is what a managed cohort cannot reproduce.

Earliest start. Stage 7: the study needs a population of operators who have served as guides and the assessment apparatus to measure their own development.