> For the complete documentation index, see [llms.txt](https://tonify-1.gitbook.io/tonify-docs/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://tonify-1.gitbook.io/tonify-docs/start-here.md).

# Start here

## How to read this

The study has three layers, and you can stop at any of them.

**Plain language.** [Results in plain language](https://tonify-1.gitbook.io/tonify-docs/theory/plain.md) states what each of the four lemmas means for an artist, with no formulas. That page is generated from the italicised route inside the paper itself, so it cannot drift from the mathematics it summarises.

**The paper.** [Full text](https://tonify-1.gitbook.io/tonify-docs/theory/full.md) carries the statements, proofs, assumptions, and the correction ledger. The canonical version is English; the Russian variant of this site carries the longer source text.

**The code.** Every claim has an executable check. [Simulations](https://tonify-1.gitbook.io/tonify-docs/simulations/sim1.md) documents what each simulation asks, what gates it must pass, and how to run it.

**What to do.** [The policy reading](https://tonify-1.gitbook.io/tonify-docs/policy/what-next.md) is the decision that sits on top of the four lemmas: an acceptable formula reform, and why the cash register should still move off the play onto the fan. It is not a theorem.

## Proved, measured, open

**Proved** — statements about any fixed play matrix, independent of how the data arose: the ratio identity and its four corollaries (L1), pass-through invariance with its exact criterion (L2), tail preservation and the atom at zero (L3), the full stochastic-dominance blockade (L4). Each was attacked twice by an adversarial red team that did find errors; the retractions are on [this page](https://tonify-1.gitbook.io/tonify-docs/reference/retracted.md).

**Measured** — one quantity, on public Last.fm logs: the intensity-size coupling, +0.101 overall, +0.232 at the head, +0.023 among beyond-mainstream listeners. Details and caveats: [Measurement](https://tonify-1.gitbook.io/tonify-docs/measurement/gamma.md). Whether anyone measured it before us: [Prior art](https://tonify-1.gitbook.io/tonify-docs/measurement/prior-art.md) — the answer is yes, repeatedly, since 1969.

**Open** — five questions we did not close: equilibrium versions of L1/L2 (the matrix itself moves when the rule changes), the sign of the Gini difference in general, the exact condition under which user-centric thins the tail, dominance for recurrent payments with churn, and an axiomatic characterisation of the direct mechanism.

## Four levels of claim

The document mixes four kinds of statement, and they carry very different weight. Nothing below the first line is a forecast.

| Level          | What is claimed                             | What it depends on                                          |
| -------------- | ------------------------------------------- | ----------------------------------------------------------- |
| Theorem        | An exact property of the payout operator    | A fixed play matrix $$p$$ and the listed assumptions        |
| Simulation     | The size of the effect in a synthetic world | The generator, the calibrations, $$\gamma$$, the tail model |
| Measurement    | The behaviour of Last.fm logs               | Sample, window, scrobbling, historical period               |
| Policy reading | What might happen at a streaming service    | Listener and label behaviour, recommendation, royalties     |

This matters most for the crossover table on the front page. The *sign* and the *direction* — user-centric takes from the head harder than the independence assumption predicts — sit at the theorem and measurement levels. The *magnitude* travels the whole bridge b → $$\gamma$$ → simulation outcome, and every joint of that bridge is a modelling choice.

## What this study is not

The lemmas and simulations are accounting statics: the play matrix is held fixed and the operators are compared on it. They are not a product pitch. A separate [policy reading](https://tonify-1.gitbook.io/tonify-docs/policy/what-next.md) states the decision that follows — labeled as policy, not as proof. Every equilibrium effect — behaviour changing because the rule changed — is outside the frame and listed as open. Numbers from the simulated world are calibration, not forecast; the one empirical number is dated (Last.fm, 2005–2014) and its sign is domain-bounded, as [Prior art](https://tonify-1.gitbook.io/tonify-docs/measurement/prior-art.md) shows.

## Conventions

Gates print **before** conclusions, and a failed gate exits non-zero — a run that fails cannot report results. Seeds are fixed (42). Every status is one of three: proved, refuted by counterexample, or conjecture. When something we published turned out wrong, it is retracted in the open, on [its own page](https://tonify-1.gitbook.io/tonify-docs/reference/retracted.md), not in a footnote.


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://tonify-1.gitbook.io/tonify-docs/start-here.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.
