Data Forensics
Your stats can be clean and your data still lying.
£129/mo — dataset certifications quoted per engagement
- Annual term
- £1,290/yrTwelve months of service for the price of ten.
- Billed monthly
- £1,548/yr£129/mo × 12. Monthly billing costs twenty per cent more than the annual term — the uplift is on monthly, it is not a discount on annual.
- The difference
- £258/yrTwo months in twelve. Paying annually saves two months, not twenty per cent — the two are different numbers and only one of them is true.
Dataset provenance, survivorship and point-in-time integrity are dataset-science problems that reach far outside finance. Any field that inherits a dataset inherits its defects. Anticipated academic rate: 50% of the annual list price — £645/yr. The academic rate applies instead of the seat and term bands, never on top of them. Stacked, the three would land below the cost of serving the account, and a rate I cannot honour is worse than one I never offered.
Nothing on this site is on sale. There is no checkout, no cart and no payment link on any page. Prices publish pre-launch so they can be read, compared and checked rather than requested — seat bands, group licences and multi-year terms are set out in full below.
Standing policy — read this beside the figure
- No promise of profit, ever. Hadal makes no performance claims and carries no implied edge. Past measurements describe instrument behaviour — never future returns.
- You own risk management. Hadal cannot control it and does not insure it. Good tools do not fix bad discipline — and this site says so.
- Analytical tools, for discretionary use. Nothing here is investment advice or a recommendation to trade. Every decision, and every outcome, is yours.
The same statement stands in the footer of every page.Terms Privacy
| Specification | Value |
|---|---|
| Catalogue no. | P10 |
| Suite | Backtest Honesty |
| Methodology | How the Backtest Honesty suite measures |
| Availability | Pre-launch — not on sale |
| Anticipated price | £129/mo · £1,290/yr |
| Delivery | Marketplace SKU · hub subscription |
| Published measurements | NOT YET PUBLISHED |
| Provenance artifact | NOT YET PUBLISHED |
Your statistics can be immaculate and your conclusion still false, because every result is bounded by the honesty of the data underneath it. Most backtest post-mortems audit the strategy. Data Forensics audits the substrate: it is the fidelity battery of the Broker Feed Auditor turned away from a live feed and pointed at the dataset you are about to trust.
The distinction matters because dataset defects are quiet. They throw no errors. They produce a smooth equity curve built on prices that were never simultaneously available, and they survive every check a strategy-level review knows how to perform.
What it measures
Data Forensics reports what it finds in a dataset and repairs none of it, so no cleaned file comes back.
- Carry-forward and merge artifacts. Where a stale value has been repeated forward to fill a hole, and where two sources have been stitched into one series at a seam nobody documented. A carry-forward artifact reads as a stretch of perfect calm the market never had — and strategies that trade calm find it unerringly.
- Per-side staleness. The age of each side of the quote inside the dataset, as a distribution rather than an average. Quote staleness that concentrates around events is structure, and a backtest that traded those events traded a price which had already stopped moving.
- Gap structure. What is missing, and where: session boundaries, holidays, outages, and the gaps that appear in one instrument and not its neighbours. Missing data is rarely missing at random, and censoring that correlates with volatility flatters every risk statistic computed on what remains.
- Bar provenance. Whether bars are what they claim to be — how OHLC was constructed, from which ticks, under what timestamp convention, and whether the series is point-in-time or has been quietly revised since. A revised history is a look-ahead channel that leaves no trace anywhere in your code.
Where the dataset cannot support a dimension, the battery publishes the honest null. “Insufficient sample” is a result, and it prints as one.
What it does not do
Data Forensics does not validate your strategy. A dataset that passes every check has established exactly one thing — that the data is what it claims to be — and nothing whatsoever about whether the rule you fitted to it survives contact with the future. That is the Overfit Auditor’s question, which is why the two are bundled. It does not repair data either: it will not reconstruct missing ticks, un-stitch a merged series, or hand back a cleaned file, because a repaired dataset is a modelled dataset and the modelling would then be mine rather than yours. And it does not establish vendor intent — a defect is a defect, not a finding of misconduct.
Who it is for
Anyone about to commit capital on the strength of a backtest run over vendor exports, broker downloads, or a merge of both — and anyone presenting a track record who would rather find the artifact themselves than have an allocator find it for them.
The ship gate
No instrument is sold until it does what this page says it does. Where a page is written in the future tense, that tense is a statement about timing rather than a hedge about capability: the instrument is not finished, so it is not listed as available, not priced as available, and not sold. It waits.
Nothing described in this catalogue is a placeholder that will quietly disappear. An instrument that turns out to be wrong gets a kill-ledger entry, not a deletion — which is the only version of that promise anyone can check.
Commercial terms
Published in full, pre-launch, so they can be read and checked rather than requested. Every figure is an anticipated indication I have set and not yet ratified, and every derived figure is the arithmetic of the one above it — shown, not asserted. Nothing here is purchasable: there is no checkout on this site.
Why Data Forensics is priced the way it is
- What the figure buys
- The recurring licence covers the fidelity battery pointed at datasets you supply: carry-forward and merge artifacts, per-side quote staleness reported as a distribution rather than an average, gap structure and where the gaps cluster, and bar provenance — how OHLC was constructed, from which ticks, under which timestamp convention, and whether the series is point-in-time or has been quietly revised since. Where the dataset cannot support a dimension, the battery prints the honest null; insufficient sample is a result and renders as one. The boundaries are firm and stated on the page: it does not repair data, so nothing is reconstructed, un-stitched or returned as a cleaned file; it does not validate a strategy, which is the Overfit Auditor's question; and it does not establish vendor intent, because a defect is a defect rather than a finding of misconduct. Dataset certification sits outside the subscription and is quoted per engagement.
- Why it is priced this way
- There are two units because two different things are being sold. Continuous checking is subscription-shaped: datasets are extended, re-merged and revised, so a clean result decays and the battery has to run again over whatever arrived since. Certification is engagement-shaped because that work is finite and its scope is not knowable in advance — one named dataset at one version, whose size, format and number of stitched sources decide the effort, which is why it is quoted after looking rather than listed before. Neither unit counts seats: the subject is the dataset, and pricing by how many people read the findings would measure the audience instead of the substrate.
- What the alternative costs
- Some of this you can do yourself, and you should know that before you buy. Repeated values held flat across a hole, gaps that land on session boundaries, and a timestamp column that does not behave are all findable in a spreadsheet or a short script, and that first pass is worth making. What the paid battery adds is the tedious remainder: staleness resolved per side of the quote and reported as a distribution, provenance and point-in-time checks that require the series to be compared against a record of what it claimed to be earlier, and a published null wherever the sample cannot carry a claim — the step that has to be built in deliberately, because a script that finds nothing prints nothing. A vendor's own quality documentation describes the pipeline as designed, which is different evidence from the file you actually received; and doing nothing defers the discovery rather than removing it, since an artifact that is present is found later either by an allocator reading your track record or by the live positions themselves, and which of those finds it first is the whole difference.
Every figure on this page is an anticipated indication awaiting ratification, and nothing here is purchasable. The reasoning above is published for the same reason the arithmetic below is: a price you can interrogate is worth more than a price you have to accept.
The two-SKU split
| Route | What it is | Anticipated |
|---|---|---|
| Marketplace SKU | Where this instrument ships through a platform marketplace, that SKU is the fully-functional standalone tier — the complete battery, running natively on your platform, no external account required. Never paid-but-crippled. | NOT YET LISTED |
| Hub subscription (this site) | The deepening: hosted runs, the published methodology behind them, and the content-hashed artifacts that let a stranger re-derive the result. This is the tier the figure on this page prices. | £129/mo · £1,290/yr |
The hub tier is priced at or below its marketplace equivalent. That is a standing rule the pricing table is rendered against, not an offer: a marketplace platform takes a percentage of every sale it processes, and that saving goes to the hub subscriber rather than to Hadal, so a platform fee can never invert your margin. No marketplace SKU has been listed yet, so the cell above reads NOT YET LISTED and the rule stands as a commitment rather than a comparison you can run today.
The commitment ladder
| Term | Months paid per year | What it means | Anticipated |
|---|---|---|---|
| Monthly billing | 12 | £129/mo × 12. Twenty per cent more than the annual term — that is the uplift for paying monthly, not a discount for paying annually. | £1,548/yr |
| 1-year term | 10 | Twelve months of service for the price of ten. This is the annual rate every band below is taken off. | £1,290/yr |
| 2-year term | 9 | Ten per cent off the annual rate, held at that figure for the whole term. | £1,161/yr |
| 3-year term | 8 | Twenty per cent off the annual rate, held at that figure for the whole term. | £1,032/yr |
Eight months paid per year, across three years, is 24 months paid for 36 months of service — one year in three carries no charge. On this instrument the three-year term totals £3,096. That is 24 × £129 = £3,096, and two years billed monthly is £1,548 × 2 = £3,096. The same money. It buys three years instead of two, and the arithmetic is on the page so you can check it rather than take it.
Group licences, across entities
Licensing here is per account or per desk, not per seat, so a seat band does not apply and none is offered — applying one would be a category error dressed as a discount. The scaling axis here is entities: where the same instrument is run by more than one legal entity, desk, fund or network inside a group, the licence is negotiated as a single group licence rather than replicated entity by entity. Multi-network clients are exactly the case this exists for, and the commitment ladder above applies to a group licence on the same terms it applies to a single one.
Commercial routes
Pricing scales on entities and on term — one negotiated group licence across desks, funds and legal entities, never a seat band.
Licensing — seat bands, group licences and multi-year terms in full
Support
- TierPriority at this instrument’s base contract — ticket, prioritised, first response targeted at one business day. Support tier follows the annual contract value, not the price of a single unit — more seats, a suite licence or a group agreement raise the contract value and can raise the tier with it.
Support — the tiers, the ticket-only channel model, and what a target does and does not promise
Trial mechanics
The trial runs on the hub tier — thirty days, a full natural proof cycle, disclosed in plain words before you start it and cancellable in one step; none of it is live yet. Trial mechanics in full, by delivery class.
Read before you commit
The documentation is published ahead of the product on purpose — intended behaviour is only a commitment if it exists first. Start with installation and first run, then the limits: the conditions under which this instrument refuses to produce a number are the part worth reading before you pay. The full centre is at /docs/, and the support model states what a ticket does and does not cover.
Questions and answers
Answered from what this instrument publishes about itself. Nothing below is attributed to a customer, because there are none yet.
If the dataset passes every check, is my backtest sound?
No. A clean dataset establishes one thing — that the data is what it claims to be — and nothing about whether the rule you fitted to it survives contact with the future. That is the Overfit Auditor’s question, which is why the two are bundled.
Will it repair the data it finds defects in?
No. It will not reconstruct missing ticks, un-stitch a merged series, or hand back a cleaned file. A repaired dataset is a modelled dataset, and the modelling would then be mine rather than yours.
Is a dataset defect a finding against the vendor?
No. A defect is a defect, not a finding of misconduct. The instrument does not establish vendor intent.
Can I buy this instrument today?
No. Nothing on this site is on sale — there is no checkout, no card capture, and no product account to create. Every figure on this page is an anticipated indication I have set so it can be read and compared, not an offer, and final pricing awaits my ratification. The launch list is the only thing you can join today.
Change log
Data Forensics has not shipped, so there is nothing to record. When it does, every version lands here — dated, append-only, written by a person, and including the changes that removed a capability rather than added one.
Where this sits
Data Forensics is one of the instruments in the Backtest Honesty suite. How that suite measures — the per-instrument battery, and the receipts each measurement will carry — is set out in the Backtest Honesty methodology, part of the site-wide measurement methodology.
Also in the Backtest Honesty suite
Data Forensics shares the Backtest Honesty suite with two other instruments.
- P2Overfit AuditorMathematical bounds on how hard you tortured the data.
- P18Reproducible Verdict KernelSame inputs, same verdict, byte for byte — or it is not a verdict.
Research behind this instrument
Data Forensics draws on eleven research notes on this site.
- Does my regime filter use future data?Asked as: "does my hmm regime filter use future data in a backtest"
- Do brokers hunt your stop losses?Asked as: "do brokers hunt stop losses by widening the spread"
- Do round numbers matter in forex?Asked as: "do round numbers matter in forex trading"
- Does a wick rejection mean anything?Asked as: "does a wick rejection actually mean anything"
- Is my volatility regime just telling me the time?Asked as: "is my volatility indicator just measuring time of day"
- The level that remembered nothingAsked as: "do price levels remember what happened there before"
- Does my backtest have survivorship bias?Asked as: "does my backtest have survivorship bias"
- Why do correlated pairs decouple on low timeframes?Asked as: "why do correlated pairs decouple on low timeframes"
- Why do two data feeds give different backtests?Asked as: "why does my backtest give different results with different data feeds"
- Why do backtests fail in live trading?Asked as: "why do backtests fail in live trading"
- Why does my backtest use data that did not exist?Asked as: "why does my backtest use data that did not exist yet"