Drawdown routes, it doesn't score — the Gate/DD authority correction.

What drawdown is

A capital-location metric, not a score.

The correction starts with a definitional point. EV is a score — it grades performance quality and belongs in a scoring layer. Drawdown is not: it tells the system where the account sits relative to its most recent equity peak and how much capital compression is currently in force. It's a location, measured from peak down through the six gate bands to the −40% boundary the system will tolerate. Location metrics route; score metrics grade. Treating drawdown as a scoring modifier — Growth adds tier points, Buffer subtracts them — misclassifies its nature and produces a specific, damaging arithmetic error.

The three jobs

Gate/DD identifies the state, selects the row, and enforces the ceiling.

Properly bounded, gate authority does exactly three things. First, it identifies the active capital state from the drawdown bands — Growth, Recovery, Buffer, Floor, Deep-Floor, Ground-Floor, or System Lock past −40%. Second, it selects the pool-budget row that all tier lookups will use, which is why T4 Mid resolves to 29% pool at 7.25% per trade in Growth, 22% at 5.50% in Buffer, and 18% at 4.50% in Floor — same tier, different envelope. Third, it enforces the gate cap: in Buffer, the system may not exceed the Buffer ceiling. What it never does is say 'you must use T4.' The tactical tier still comes from daily EV, the weekly structural read, Category 6, and diagnostic overlays operating inside the arena the gate defined.

The double-count

Why a tier penalty on top of a gate row is the same punishment twice.

Here's the mechanical harm the loose wording invited. The gate row has already compressed the pool values — Buffer's T4 authorizes materially less capital than Growth's T4, by construction. If the engine additionally applied a tier penalty for being in Buffer, the account would be demoted to a lower tier and that lower tier would then be priced in the already-compressed Buffer row. The same drawdown fact would have reduced deployment twice through two different mechanisms. The result isn't conservative, it's incoherent: deployment in defensive gates collapses faster than the ladder's designed shape intends, and the system's actual risk gradient stops matching the gradient documented in the pool panel.

• Gate row compression is the drawdown response. A second tier penalty re-applies a response already delivered.
• The corrected engine role: capital-state router and hard ceiling — does not promote, does not select the tactical tier.
• Gate/DD retains highest authority precisely because it defines the risk universe, not because it picks inside it.

The surviving authority

Highest authority, narrowly exercised — and still absolute.

Nothing in the correction weakens gate supremacy. The gate remains the layer no diagnostic strength can overrule: a brilliant structural read in Floor does not unlock Growth deployment, exceptional expectancy does not raise a ceiling, and a favorable benchmark comparison grants no permission at all. The System Lock boundary past −40% remains terminal. What changed is only the mechanism of that authority — it governs by defining the arena and capping its height, not by voting on the tier. Narrow authority, absolutely enforced, is stronger than broad authority loosely specified, because it's the version that can be implemented without contradicting itself.

The key idea

Precision about authority is what keeps hierarchy from becoming hand-waving.

'Outranks everything' feels like strong governance and is actually a specification gap — it doesn't say through what mechanism, which is exactly where implementation errors hide. Naming drawdown a capital-state router closes the gap: one entry point, three jobs, no scoring role, no double-count, and a hard ceiling nothing else may lift. This is the level of precision the whole authority hierarchy is written at, and the correction is a good illustration of how it gets there — by an operator noticing that a phrase everyone agreed with didn't actually describe what should happen.