cryptochartics
Home Blog The audit
Press kit · Updated May 30, 2026

Press resources

Cryptochartics is the indie crypto app that measured its own engine's accuracy (35.5%) and published the data, then pivoted to honest divergence detection between crowd sentiment and technical indicators. This page has everything press needs: founder bio, app screenshots, brand assets, quote-ready stats, and direct contact.

The pitch (in one sentence)

Cryptochartics is the only crypto app that publishes its own engine's measured accuracy openly — 35.5% across 3,767 trades — and uses that finding as the case for pivoting to honest divergence detection between crowd sentiment and technical indicators, rather than selling unprofitable signals. — Suggested lede for product profile

The 30-second backstory

Founder Murat (UAE-based, solo) spent two years building Cryptochartics — a crypto signals app running 12 technical indicators across major coins and timeframes. After fixing a long-standing R:R bug in May 2026 and running a complete audit on 3,767 evaluated trades, he found:

  • The engine had a 35.5% directional win rate — well below the 50% needed for 1:1 R:R to be profitable after Binance's 0.20% round-trip fee.
  • The "confidence percentage" displayed to users was anti-correlated with outcomes (75-80% confidence won only 27% of the time).
  • One indicator (Williams %R) was strongly reverse-predictive — when it disagreed with consensus, the trade won 60.5% of the time.
  • Per $100 deployed on every signal, users would have lost approximately $1,470 cumulatively.

Rather than hide the data or keep selling, Murat published the audit publicly, dropped the "confidence percentage" from the UI, renamed the "Returns" screen to "Engine Accuracy", and repositioned the product around divergence detection — surfacing the moments when crowd sentiment and technical indicators pull in opposite directions. The pivot ships in v1.1.0.

Quote-ready stats

Use these directly in articles. All numbers are reproducible from the audit data. See the methodology section in the audit if you need the raw computation.

35.5%
Engine win rate
Across 3,767 evaluated trades since the 2026-05-18 R:R fix. Need >50% to be profitable at 1:1 R:R after Binance fees.
−$1,470
Per $100 per signal
Cumulative loss if you'd traded every signal at $100 since the bugfix. Real financial cost of the engine's underperformance.
60.5%
Williams %R inverted
When Williams %R disagreed with the engine's consensus direction (n=152), the trade won 60.5% of the time. Reverse-predictive at scale.

Founder quotes (use freely)

"I spent two years building a crypto signals app. Then I measured every prediction it ever made. The engine had a 35.5% win rate — academically that's exactly what you'd expect from a 12-indicator ensemble per the Wei (2024) paper on 7,846 crypto TA rules. The choice was: hide the data and keep selling, or publish the data and rebuild around something honest. I picked the honest one." — Murat, founder of Cryptochartics
"The most damaging finding wasn't the win rate. It was that our 'confidence percentage' was anti-correlated with outcomes. Users were reading 75% confidence as 'this is a strong signal' when the data said the opposite. That's not a math bug — that's a trust bug. We removed it from the UI in v1.1.0." — Murat, founder of Cryptochartics
"Crypto signal services have a survivorship bias problem: the apps that publish their accuracy data tend to be the ones whose data is good, and the apps whose data is bad never publish. We're flipping that — publishing our bad data publicly, because the alternative is letting users keep believing TA is more predictive than it actually is." — Murat, founder of Cryptochartics

Founder

M

Murat Chary

Founder + sole developer · Cryptochartics

UAE-based independent developer. Built Cryptochartics solo over two years using React Native (Expo) for the mobile app, Supabase for backend + auth, and a Python TA engine running on DigitalOcean. Lives in the UAE; the company was originally registered under Arabcell FZCO and is in the process of transferring to a personal Apple Developer account. Open about everything: the app code is publicly viewable, the engine accuracy is published openly, the pivot rationale is in writing.

github.com/muratchary @cryptochartics hello@cryptochartics.com

Brand assets

Right-click any preview to "Save image as…" or copy the inline SVG source from the page. All assets are CC0 — use freely. Higher-resolution PNG renders available on request via the contact form.

Logo mark (color)

The 12-bar bell pattern. Each bar = one indicator. The two off-color bars tell the brand story — mixed and bearish readings exist.

SVG · Color

Logo mark (mono)

For dark-on-light or single-color contexts. Pure white version included separately for light backgrounds.

SVG · Mono
50% breakeven 5m 15m 1h 4h 1d 35.5% overall WR 3,767 trades · all timeframes net-negative cryptochartics.com/blog/we-measured-ta-failure

The 35.5% chart

Win rate by timeframe. The headline visual for any article about the audit. Inline SVG — view-source to copy.

SVG · Chart
41.4% 40.2% 27.3% 55-60 65-70 75-80 Confidence paradox higher confidence ≠ higher accuracy

The confidence paradox

Win rate by confidence band. Higher confidence trended worse — the most damaging finding in the audit.

SVG · Chart
App screenshots

App screenshots

iOS + Android. Pulse tab (divergence hero), Signal cards with crowd chip, Engine Accuracy screen. Email for full set.

Request via email

App icon

1024×1024 PNG. The brand mark in the iOS/Android app icon container.

Request via email

The full audit · 8-page PDF

Same content as the live blog post, formatted as a stand-alone deliverable for journalists, editors, and academic reviewers. CC0 — quote freely.

cryptochartics-audit-2026-05.pdf
8 pages · 80 KB · reproducible methodology included
Download ↓

One-page fact sheet

The quick-reference card. Print this if you're writing a sidebar.

Product

  • Name · Cryptochartics
  • Tagline · See when crypto crowd sentiment and technical indicators disagree
  • Category · Crypto intelligence / fintech / iOS + Android mobile
  • Version · 1.1.0 (released May 2026)
  • Pricing · Free tier + Pro tier (crypto payment, TRC20 / Polygon / BSC)
  • Platforms · iOS, Android (universal)
  • URL · cryptochartics.com

Founder + Company

  • Founder · Murat Chary (UAE)
  • Role · Sole developer + operator
  • Founded · 2024
  • Headquarters · United Arab Emirates
  • Funding · Bootstrapped, no outside investment
  • Stack · React Native (Expo), Supabase, Python TA engine on DigitalOcean
  • Open source · github.com/muratchary/cryptochartics

The audit (May 2026)

  • Trades analyzed · 3,767 evaluated since the 2026-05-18 R:R fix
  • Win rate · 35.5% (wins / wins+losses)
  • Confidence-paradox finding · 75-80% band won 27.3%; 55-60% band won 41.4%. Anti-correlated.
  • Williams %R finding · When it disagreed with consensus (n=152), trades won 60.5%. Reverse-predictive.
  • Academic context · Sullivan/Timmermann/White (1999), Bajgrowicz/Scaillet (2012), Wei (2024 IJFE)
  • Full audit · cryptochartics.com/blog/we-measured-ta-failure

Contact

For interviews, additional data, or to verify any of the audit findings.

Murat Chary · Founder

Email (preferred)
hello@cryptochartics.com
Twitter / X
@cryptochartics
GitHub
muratchary/cryptochartics
Response time
Same business day (GST / UTC+4)

Murat is happy to do interviews, walk through the methodology, or share additional cuts of the audit data on request. He's also willing to be on-camera (YouTube, podcasts) — including for skeptical/adversarial framings. The data is the data.