Semtum

Is your “special” number actually special?

Test your strategy against twenty arbitrary versions of itself. The cheapest high-value check there is, and it works on strategies you bought. Free.

Tool 3 of 5

Every strategy is built around something it claims is special: the first thirty minutes of the day, the 200-day average rather than the 180, the week options expire. This checks whether that thing matters, by testing your rule against arbitrary versions of itself.

Watch: what goes in
What it tells you

Asking "is my strategy profitable?" is nearly worthless. A third of arbitrary date windows are profitable. The useful question is "does it beat arbitrary versions of itself?" If the 200-day average works but 190 and 210 work just as well, you have found a trend effect. If 200 works and nothing near it does, you have found a coincidence that happened to land on a round number.

This is the cheapest high-value test there is, and it is the one test you can run on a strategy you bought from somebody else without knowing how it works inside.

Step 1: what am I supposed to compare? Build the list here

Pick what your strategy's special ingredient is and type the value it uses. This gives you twenty arbitrary alternatives, spread in both directions. Go and run your strategy once against each one, exactly as it is, changing nothing else, then paste the twenty scores into step 2.

Nearly every strategy has exactly one number at its heart. Pick the one yours would break without.

The 200 in “200-day moving average”. Just the number.

Now what? You run your strategy once per value in that list. This is less work than it sounds: it is the same strategy, you are only changing one setting each time.

  • TradingView: open the strategy settings (the gear on the Strategy Tester), change that one input to the next value in the list, read the new Sharpe off the Performance Summary, write it down. Twenty times is about ten minutes.
  • MetaTrader: Strategy Tester → Inputs tab. Set the parameter to Start/Step/Stop so it sweeps the range in one run, then read the results table.
  • Python / Freqtrade: a for loop over the list, collecting the Sharpe of each run. This is the version the course ships.
  • A strategy you bought and cannot see inside: if the seller gives you no way to change that number, you cannot run this test, and that is itself worth knowing. Ask them what happens at 190 and 210. A real edge survives the question.

Score every run exactly the same way. Same data, same date range, same costs, same metric. The only thing that changes between runs is the one number.

Sharpe, from tool 2. Any consistent score works as long as all twenty-one use the same one.

One per line or comma separated. Paste them in any order.

Beaten by
–
Percentile
–
Alternatives' median
–

Where this number comes from

The free first module works through the arithmetic behind this calculator: where the bar comes from, why looking twice raises it, and what to do when your result sits below it. Subscribe for one email every two weeks about what I tested.

Your address is used for this and nothing else. Every email has a one-click unsubscribe.

If a tool here told you something uncomfortable

Why your backtest lied to you

The free first module explains the arithmetic behind these calculators in full: where the bar comes from, why looking twice raises it, and what to do when your result sits below it. No signup, ~14 minutes.

Read it free

The full course is eight modules, nineteen data sources with the trap in each, twelve working download scripts and the searchable ledger of every approach I tested, one of which survived. See what is in it →

These match the Python toolkit. The noise-floor integration is the same calculation as the toolkit’s own noise floor; the real-money bar here uses the exact closed form where the Python simulates it, so it may differ in the second decimal. Everything runs in your browser; nothing you paste is sent anywhere, and the page works with no internet connection.

This is not financial advice and not a recommendation to buy or sell anything. A result that clears every bar on this page is a reason to collect forward data, not a reason to fund it. See Module 8.