A matching hash confirms file integrity, not app safety.
Actual emulator captures. Synthetic data.
New in 0.8: a fragility grid on your CSV. Test EMA periods around your setting and fees raised by 10 and 25 bps per side. Nine sequential, cancellable calculations; no automatic winner selection.
Keep a reference A and compare it with the current experiment B: net return, post-cost drawdown and elapsed time in cash. Different datasets or engine versions are marked as non-equivalent.
Export the report, source CSV, reference and matching fragility grid in JSON. Everything stays on your device until you export it; no automatic upload.
60 to 5,000 bars, up to 1 MiB. Comma or semicolon, explicit timestamps, optional volume. Invalid rows are rejected with a reason; the previous valid report is kept.
Persistent Workshop, Result and Charts navigation keeps the main views within reach, including on small screens.
Standalone offline Rust CPU engine. The bundled synthetic cases are compared on PC and Android; personal imports are retrospective calculations, not independently validated strategies.
EMA 6, 12 or 24 and three baseline fee settings. The engine reseeds its EMA on an N-bar window: this is not TradingView ta.ema parity. Cash when the entry signal ends.
Buy & Hold comparison with the same starting window and fees; measured fee impact and trade replay.
Interactive 3 × 3 sensitivity matrix: nine sequential native calculations, compared by net return, drawdown and difference from Buy & Hold. Explore cash and positions bar by bar.
Three comparison bars show the strategy, Buy & Hold and interest-free cash on one scale. Inspect the decline from each capital peak in the replay.
The last two experiments stay on your device. Compare changed settings, capital and drawdown; changing both settings is not a one-variable experiment. A bounded workshop, not a general Pine importer.
A matching hash confirms file integrity, not app safety.
Actual emulator captures. Synthetic data.
Compact episode selection and a direct return to the current session.
3 synthetic missions, each with 240 bars.
Select 2 of 3 rule-based profiles plus cash; the initial allocation stays fixed.
4 reveal chapters with interactive replay: inspect capital, cash and exposure at each revealed bar.
Jump to recorded entries and exits. The debrief separates absolute return, performance against the reference, drawdown budget and time spent in cash.
Positive and negative contribution bars show what each profile added or lost. Activity details distinguish closed trades, time in position and idle capital.
Compare another cash allocation without changing the original: capital, drawdown and their tradeoffs are measured separately. Compatible saved results survive interface-only app updates.
Tested scope, clear limits
Android 8.0+ (API 26). Both APKs are debug-signed prototypes. Tested on an x86_64 Android emulator, API 35. ARM64 builds exist, but have not been verified on a physical phone. Compatibility with your device is not guaranteed.
Both apps work without an account, Internet permission or billing. Fixed rules and synthetic examples, plus personal CSV import in Pocket Lab. No live market feed or AI calls, broker connection or real orders. Imported data has no verified market identity; calculations assume close execution, fees per side and no slippage. No holdout or future-return claim.
Before installing an APK
Updating an existing installation? Open the new APK and choose Update. Do not uninstall the old app or clear its data: experiments are stored locally. Both apps can be installed side by side.
Install only if you trust Strategy Arena and the source of this file. Android may warn that the app comes from outside the Play Store and request installation permission for the browser or file manager opening it. This is not a Play Store release or a safety certification.
Keep Play Protect enabled. If Android blocks the app or flags it as harmful, stop and report it; do not bypass the protection. You can revoke the browser or file manager's installation permission afterwards.
