How to Verify a Trading Signal Provider’s Win Rate

Advertised signal win rates typically run 10–20 percentage points above independently tracked performance. Seven checks expose the gap: are losses published individually, is a reward-to-risk ratio quoted alongside the win rate, do timestamps survive a chart check, are stops fixed before publication, how are same-candle ties and unreached targets resolved, is the sample at least 20 decided trades, and is the claim plausible at all — 55–70% is a strong real result, 90%+ is a warning sign.

Almost every signal provider advertises a win rate. Almost none of them publish the data behind it. Where independent trackers have measured providers against their own claims, the advertised figures have run roughly 10 to 20 percentage points ahead of what actually happened — which means the number on the sales page is, for most services, closer to marketing than measurement.

You do not need to take anyone's word for it. Here is the checklist. It works on any provider, including this one.

1. Are the losing trades published?

This is the fastest filter and it eliminates most of the market. A record showing only winners is not a record — it is a highlight reel. Ask to see the losers, at the same prominence, in the same list.

If losses appear only as a summary statistic (“we had a few red trades”) rather than as individual dated entries you can inspect, treat the win rate as unverified.

2. Is a reward-to-risk ratio published alongside the win rate?

A win rate on its own is meaningless, and the people quoting 90% usually know it. You can manufacture almost any win rate by moving the target closer and the stop further away:

The 90% system loses money. The 40% system makes it. If a provider quotes a win rate without the average winner and average loser beside it, you cannot tell which one you are looking at — and that ambiguity is usually the point.

3. Do the timestamps survive a chart check?

Signals posted to Telegram or Discord carry visible timestamps. Take any past signal, note the time, and pull up the price chart for that exact moment.

• Was the entry price actually available when the message was posted?
• Or had price already moved several percent in the signal's favour first?

A “signal” posted after the move has happened is not a signal you could have traded. This is the single most common way a track record gets inflated without anyone writing down a false number.

4. Does every signal define a stop before it publishes?

A signal without a predetermined stop cannot be scored. If the exit is decided afterwards, every trade can be closed at a flattering moment and the resulting “record” means nothing. Entry, target and stop should all be fixed at publication and never moved.

5. How are the ambiguous outcomes resolved?

This is the question that separates a real methodology from a plausible-looking one, and almost nobody asks it. Two cases matter:

• Target and stop hit in the same candle. The candle records a high and a low but not their order, so the outcome is genuinely unknowable from that data. Does the provider score it a win, a loss, or quietly discard it? Only one of those three answers is conservative.
• Neither level reached. Are these dropped from the record? Dropping them silently removes every mediocre outcome and lifts the apparent win rate without a single fabricated trade.

6. Is the sample big enough to mean anything?

A win rate computed on twelve trades tells you almost nothing — one outcome swings it by eight percentage points. Twenty decided trades is a reasonable floor for quoting a figure at all, and a few hundred before you should weight it heavily. Ask how many signals the number is based on, and over what period.

7. Is the claim even plausible?

Sustained win rates in the 55–70% range are a strong result for a real system. Advertised rates of 90% or more almost always mean one of three things: losses are hidden, targets are set so close that nearly everything “wins” while the occasional loss is enormous, or the figure is simply invented.

Treat a very high advertised win rate as a reason for more scrutiny, not less. It is a warning sign, not a selling point — the arithmetic is in why a 40% win rate can beat a 90% one.

Red flags that need no analysis

• Guaranteed or fixed returns. No trading system can promise these. Walk away.
• “Risk-free” anything. The same.
• Screenshots as evidence. Trivially cherry-picked, edited, or posted after the move.
• Artificial urgency. “Only 3 VIP spots left”, countdown timers, escalating DMs from an “admin”. That is a sales technique, and it is uncorrelated with signal quality.

Running the checklist on us

It would be hypocritical to publish this without answering it ourselves, so:

• Losses published? Yes — every resolved signal appears in the public track record and in each asset's own ledger, wins and losses alike.
• Reward-to-risk published? Yes, alongside the win rate, plus the average winner and average loser. Every signal is fixed at 2:1.
• Stops fixed before publication? Yes. Entry, target and stop are written when the signal fires and never moved.
• Same-candle ties? Scored as a loss, always. It is the conservative reading and it can only make our numbers worse.
• Signals that reach neither level? Kept, marked expired, and scored on close-versus-entry, so they still drag on expectancy.
• Sample size? We do not publish a win rate below 20 decided signals; the raw counts are shown instead.
• Machine-readable? The underlying numbers are served as free JSON at /api/signals/stats, so you can check our arithmetic rather than trust it.

We also publish, per asset, how far each trade ran in your favour and against you before it resolved — which exposes how much of a result came from sequencing rather than edge. That is the part most providers leave out, and it is worth asking any provider for.

Educational content, not financial advice. Trading carries risk, including loss of capital.