About
Evidence, systems, and application.
An independent human-performance knowledge platform that turns primary research into transparent, practical decisions — without pretending the evidence is cleaner than it really is.
Mission
To make the state of the evidence on human performance legible, and then usable. Most of what circulates about training, recovery and nutrition is a conclusion with the reasoning stripped out — which makes it impossible to tell a strong claim from a confident one. Ground State publishes the reasoning and the strength alongside the conclusion. The evidence shows up in the verdicts, the limitations and the sources.
What the name means
In physics, a ground state is the lowest stable energy configuration of a system — the reference against which every other state is described. Nothing about a system means anything until you know its baseline.
The same is true of a person. Sleep, load, stress, history and habit form a baseline, and an intervention only makes sense relative to it. Understand the system first. Then improve the outcome.
Principles
Strength before conclusion
How good the evidence is comes before what to do about it. Always in that order.
Uncertainty is content
“We don’t know yet” is published as readily as a conclusion. It is the more useful answer more often than people expect.
Source-traced
A claim is traced back to the evidence it actually rests on, not to whoever repeated it most recently.
One variable at a time
Systems are changed deliberately, so that when something works you know what it was.
Long horizons
Decisions are judged over years. Most things that work quickly stop working quickly.
Practical when justified
Every synthesis states what the evidence changes in practice — including when the justified action is to change nothing.
Source-led, not personality-led
Ground State Research is intentionally source-led rather than personality-led. A claim is evaluated through its evidence and its provenance — what the sources are, what they measured, and how far they reach. Who is making the claim is not one of those inputs.
That is a statement about method, not about other people. Plenty of good work is published under a name. The point is narrower: an author’s identity should never substitute for checking the source, and a reader should not have to trust anyone to verify a record here.
Nor is it anonymity as a shield. Human review remains part of governance — a person reviews what publishes and is answerable for it. Facelessness removes the shortcut, not the accountability.