Home › Open data

Open data: the full signal ledger

Updated 2026-08-06 · 330 signals tracked

• 57.9%Win rate
• 1.36Avg R
• 11TP2 hits
• 330Signals

A win rate you cannot recount is a claim, not a result. This page exists so that anyone — a sceptical reader, a researcher, a competitor — can take the underlying records and check every number published on this site.

Download

• drishti-signal-history.csv — every resolved signal, oldest first, 319 rows. This is the file every statistic on this site is computed from.
• signals.json — the live feed, including signals that are still open. Updated roughly every 15 minutes.
• api/stats.json — headline statistics only, if you want the aggregates without the rows.
• /live — the same current data as a plain HTML page, no JavaScript required.

What is in the file

One row per resolved signal. A signal is resolved when it reached take-profit 1, hit its stop-loss, or ran out its 24-hour window. Signals still open are not in the CSV; they are in the JSON feed.

How the outcomes are decided

Every signal is scored against Delta Exchange India mark prices on one-minute bars. If price touches take-profit 1 before the stop-loss, the signal is a win and the stop moves up to TP1, after which the position rides toward TP2. If the stop is touched first, it is a loss. If 24 hours pass with neither touched, it expires.

Expired signals appear in the file with status expired and are excluded from the win rate. 79 of 319 records fall into that category. Including them in either the numerator or the denominator would change the headline number, which is exactly why the rule is stated rather than assumed.

Recomputing the win rate

Filter to rows where status is won or lost, then divide the count of won by the total. That is 139 ÷ 240 = 57.9%. Everything else on this site is a slice of the same file: by asset, by direction, or by time window.

Reproducing every figure on this site

Each headline number is a filter over the same file. In order:

1. Win rate. Keep rows where status is won or lost. Divide the won count by the total. Expect 57.9%.
2. Outcome breakdown. Group by close_reason. tp1_floor and tp2_hit are wins, sl_hit is a loss, expired_no_hit is excluded.
3. Per-asset rates. Group the same filter by asset. Sample sizes fall to the low tens, so the per-asset intervals are wide.
4. Long versus short. Group by direction. Expect 132 long and 108 short resolved signals.
5. Rolling windows. Filter closed_at to the trailing period. This is what the daily, weekly and 30-day report pages do.

If a count of yours disagrees with a number on this site, the file is the authority. Tell us and the page gets corrected.

How the file is produced

Signals are written to a public feed the moment the model generates them. A separate tracker reads one-minute bars and marks each signal won, lost or expired against its published levels. The CSV is regenerated from that resolved history on every deploy, so it always reflects the current state of the ledger rather than a periodic export.

Nothing is edited after the fact. A signal's entry, stop and targets are fixed when it is published, and only its outcome fields are filled in later.

Licence and attribution

The ledger is published under a CC BY 4.0 licence. You may republish, analyse, chart or redistribute it, including commercially, provided you credit the source and link back. Suggested attribution:

Drishti Pro signal ledger, https://www.drishtisignals.in/data/, CC BY 4.0.

If you publish an analysis of this data we would rather you did it independently than cleared it with us. The point of releasing the file is that our own summary of it carries no special authority.

Known limits

• The record begins on 22 July 2026. It is a real ledger but a short one, and a short window cannot separate a working system from a favourable market.
• Outcomes are scored on mark prices, not on fills. They do not include slippage, fees or funding, so they describe whether the market reached a level, not what a particular trader would have netted.
• Sample sizes differ by asset. Per-asset rates are published with their counts attached so a thin slice is visible rather than hidden.
• The file is regenerated on every deploy, so a download taken today and one taken next week will differ by the signals resolved in between.

For what the numbers mean, see the verification page and the win-rate benchmark.

See Drishti Pro's live signals Get the app

Educational & illustrative only — not investment advice. Drishti Pro publishes AI-generated trade ideas and their public track record for information. Crypto is volatile and you can lose money. Nothing here is a recommendation to buy or sell any asset. Do your own research.