A reliable, privacy-focused tennis scoring app for Apple Watch — with iPhone companion for live viewing, announcements, and match history
DeuceMate is a lightweight, native Apple Watch app for scoring tennis matches in real time, bundled with an iPhone companion app for live score viewing, umpire-style announcements, and a complete match archive with detailed statistics. Built from the ground up with SwiftUI, it delivers a fast, reliable scoring experience without the bloat, crashes, or privacy concerns of existing alternatives.
 |
 |
|---|---|
| Gesture-first scoring, right on your wrist. | Watch it live on your phone — same score, real time. |
 |
 |
| See your momentum — overlaid with heart rate and movement. | Pro-level analytics — 20+ stats, side by side. |
 |
 |
| Filter to your winning points — see exactly where you took control. | Filter to lost points — pinpoint where the match slipped away. |
 |
 |
| Discover the heart-rate zone you win in. | Five court-inspired themes. |
 |
 |
| One tap to send your match data to an AI coach. | A structured prompt — with full context — ready for any AI. |
 |
 |
| Complete stats analysis included automatically. | Every point, in order — so the AI spots patterns you missed. |
 |
 |
| Share a match as one self-contained web page — no app required. | The full stats breakdown, filterable by set, right in the browser. |
After trying multiple tennis scoring apps that crashed mid-match, collected unnecessary data, or required subscriptions with intrusive ads, I built DeuceMate to be:
- Reliable - Core scoring works offline with optimized battery usage
- Private - No developer data collection, analytics, or tracking
- Simple - Clean interface focused on the essentials
- Free - No subscriptions, ads, or in-app purchases
This project also showcases my ability to ship production-quality native iOS/watchOS apps from concept to App Store — built end-to-end with an AI-agent development workflow (see Development).
Gesture-first scoring on Apple Watch
| Gesture | Action |
|---|---|
| Swipe up | Award point to you |
| Swipe down | Award point to opponent |
| Double tap | Toggle second-serve context |
| Swipe left | Undo last point (full game-state rollback) |
| Swipe right | Open live stats view |
- Complete tennis scoring — games, sets, tiebreaks, deuce/advantage, automatic server rotation, break-point detection, and side-change prompts with a compass court-end badge. Six match formats; singles and doubles with full service-order management.
- Bulletproof state — full undo stack across game and set boundaries; in-progress matches survive backgrounding, notifications, and overnight pauses.
- iPhone companion — stadium-style live scoreboard, umpire-style spoken announcements, optional swipe-to-score iPhone Input, manual match entry, and an unlimited on-device archive.
- Point outcome tracking — tag every point with a cause and ending shot to unlock serve, return, break-point, pressure, rally-depth, and score-state analytics, filterable per set or whole match.
- Coaching insights — RecCoach (eight data-driven observations per match) and PulseCoach (heart-rate-correlated insights), plus an interactive points-momentum graph with HR and steps overlays.
- AI coaching exports — one tap builds a structured coaching prompt (including the raw point-by-point table) for ChatGPT, Claude, Gemini, and others, with a dual perspective mode so your opponent gets coached too.
- Shareable interactive web report — export any match as a single self-contained HTML page (momentum chart with toggleable outcome/shot scatter, the full Me-vs-opponent stats with Stats/Points tabs and per-set filters, and a built-in AI-coach launcher) that opens in any browser, fully offline — no app, account, or server.
- Health & fitness — optional HealthKit workout with live heart rate, calories, steps, distance, and per-set timers.
- Five court-inspired themes, momentum strip, and point-confirm highlight — all synced between watch and iPhone.
📖 User Guide — the full reference: every gesture, format, statistic, coaching rule, and export mode, plus battery numbers, padel compatibility, and a refresher on tricky tennis rules.
The developer does not collect or receive data from DeuceMate and operates no backend, analytics, advertising, or tracking. Match data is stored on your Apple devices and, when available, in an automatic archive in your personal iCloud Drive that excludes HealthKit measurements. Apart from that archive, data leaves DeuceMate only when you explicitly export or share it, including with an AI service you select. See PRIVACY_POLICY.md (also published at emrahman.github.io/DeuceMate/privacy.html).
- Apple Watch Series 4 or later (including SE and Ultra), watchOS 9.0+
- iPhone running iOS 17.0+ (the watch app is bundled with the iOS app)
- App Store: submission in progress — coming soon
- 🌐 Support site: emrahman.github.io/DeuceMate/support.html
- 📧 Email: mail@ehsanrahman.com
- 🐛 Report issues: GitHub Issues
- 📖 Help & FAQs: SUPPORT_PAGE.md
Native SwiftUI on both platforms with zero third-party dependencies — Apple frameworks only (WatchConnectivity, HealthKit, AVFoundation, CoreMotion/CoreLocation, Swift Charts). All portable logic — models, the pure scoring engine, stats derivation, the sync wire format, and merge policy — lives in the shared DeuceMateCore Swift package with comprehensive unit tests. The watch is the source of truth for live matches; the iPhone is the durable read-only archive, synced over WatchConnectivity with queue-and-deliver semantics.
AI-agent workflow: the repo is deliberately structured for agentic development — CLAUDE.md is the operating guide coding agents follow (architecture map, multi-site change recipes, known traps), docs/architecture/ holds the component, sync, and match-lifecycle diagrams agents keep current, and docs/features/TECHNICAL_DEBT.md is the prioritised backlog they work from. Changes land as agent-authored PRs held to the rules in CONTRIBUTING.md.
Architecture documentation: docs/architecture/ — component topology, sync and data-flow, match lifecycle, and a complete file inventory.
Contributions are welcome! Please feel free to:
- Open issues for bugs or feature requests
- Submit pull requests with improvements
- Share feedback from using the app
See CONTRIBUTING.md for conventions.
This project is licensed under the MIT License - see the LICENSE file for details.
Ehsan Rahman 🌐 ehsanrahman.com 💼 GitHub
Built with ❤️ for tennis players worldwide 🎾
This project showcases production-quality native iOS/watchOS development. Available on the App Store soon.