Blog
Notes from Foundation Nova Diem
Writing on the evidence behind Lux, the standard we hold our own claims to, and how our cohort-evidence programme grows over time. Not news for its own sake, just an honest record of the work.
Why we wait for n≥500 before we say a cohort is calibrated
The evidence gate that decides when a community's results are ready to anchor Lux's next layer of analysis, and why we won't move it for anyone.
Confident, not overstated: how we decide what we're allowed to say
Why Foundation Nova Diem leads with what Lux does well, and still refuses to round a claim up past the evidence that backs it.
The evidence behind Lux, in plain terms
What a 19-layer scoring engine actually buys you, and why we treat reliability figures as a selling point rather than a disclaimer.