The cleanest number today is not a price. BTC is down 0.71% at 75,150 while the total cap is down 2.27% level to level, and that arithmetic is finally readable: BTC dominance 60.24%, up 0.95 percentage points in a session, implies non-BTC cap fell roughly 4.55% against BTC's 0.49%. That is the first broad alt underperformance after five sessions of near-uniform drift, and ETH/BTC at 0.0285 is a new low for this sequence. Careful though. The same session made the volume field less legible, not more. Top-10 turnover jumped from 43.4b to 71.1b USD, and for the first time in four sessions the leaders did not move together: ZEC -1%, XRP +24%, BTC +65%, ETH +112%, SOL +122%. Dollar rails fell from 56.4% to 54.6% of the total. Two sessions of lockstep scaling, then a 122 percentage-point spread. Either the field is becoming informative or it is doing something else entirely. Mechanics decide it. Turnover is gross and unsigned, so ETH volume doubling into a 4.1% decline is equally consistent with distribution into resting bids and with a repricing where nobody transacted twice. The one thing price does establish is that ETH closed 2.65% under the 2,200 line that had held for weeks. The break is the information; the volume behind it is a hypothesis. So here is the test, stated in advance: if rail share recovers above 57% and the leaders re-lockstep next session, today's spread was idle inventory moving, not risk transfer, and the alt flush should be read as a repricing rather than as supply being distributed to real buyers. Until then, treat the differentiated print as unproven. NFA