🎾 Tennis Grand Slam Tracker

Live: danpune.github.io/tennis-slams-tracker

All four tennis majors in one place, season after season — every draw live, official highlights, match times in your timezone, market win-odds, champions history and world rankings. One self-contained page. No ads, no cookies, no tracking, no build step, no dependencies, no API keys.

Sibling project of worldcup2026 — same playbook (static page + scheduled fetch committing JSON snapshots), independent codebase.

flowchart LR
    subgraph sources ["Data sources (free, no keys)"]
        ESPN["ESPN public tennis feed<br/>scores · draws · rankings"]
        PM["Polymarket gamma API<br/>match-winner prices"]
        YT["Official YouTube channels<br/>@Wimbledon · @usopen · …"]
        WM["Wikidata + Wikimedia Commons<br/>freely licensed player photos"]
    end

    subgraph actions ["GitHub Actions — every 30 min"]
        F["fetch_data.py"]
        H["build_highlights.py<br/>oEmbed-verified official clips"]
        P["build_photos.py<br/>matched by ESPN id or unique name, credited"]
    end

    subgraph repo ["Committed JSON snapshots"]
        D["data.json<br/>(current Slam, live)"]
        C["champions.json<br/>(evergreen roll of honour)"]
        E["editions.json<br/>(completed Slams, QF→F)"]
        HL["highlights.json<br/>(verified clip + match context)"]
        PH["photos.json<br/>(player id → licensed photo)"]
    end

    Pages["GitHub Pages<br/>index.html — one file,<br/>inline CSS + vanilla JS"]
    B["Your browser<br/>⭐ follows & timezone in localStorage —<br/>nothing ever leaves it"]

    ESPN --> F
    PM --> F
    YT --> H
    F --> D & C & E
    WM --> P
    H --> HL
    P --> PH
    D & C & E & HL & PH --> Pages --> B

The page is static; all data work happens in the scheduled fetch. The live feed only carries the current tournament, so anything worth keeping is written to append-only files the moment it happens — champions to champions.json, each finished Slam's business end to editions.json. Those files are the site's permanent memory.

• Fail-safe everywhere. Every writer is merge-only and never replaces good data with a failed or empty fetch. If a source breaks, the site degrades gracefully — it never breaks.
• Never fabricate sports data. Champions, scores, dates and rankings are fetched and verified or not shown at all. (There's no free head-to-head source, so there's no head-to-head feature — rather nothing than something invented.)
• Trust, but verify the source. Highlight links are checked against YouTube's oEmbed author_url so only the tournament's official channel can ever appear — author_name is spoofable; the URL isn't.
• Privacy as a feature. No cookies, analytics, accounts or tracking. The browser's only external requests are player photos (ESPN, Wikimedia Commons), highlight thumbnails (YouTube, embeds load on tap via the cookie-less domain), ESPN's public scoreboard (only during a Slam, when the saved snapshot is behind) and one anonymous visit-counter ping.
• Boring on purpose. One HTML file, system fonts, no framework, no build. The whole site can be read in one sitting and hosted anywhere.

python3 fetch_data.py            # refresh data.json (+ champions/editions when finals land)
python3 build_highlights.py      # fill verified official highlight links
python3 -m http.server 4600      # open http://localhost:4600

Backfill a Slam that finished before the site existed (ESPN serves historical snapshots):

python3 build_editions.py 20260201 20260607   # final-weekend dates

Results, draws and rankings via ESPN's public tennis feed (unofficial); player photos from ESPN or, where ESPN has none, Wikimedia Commons under each photo's free licence (credited in the photo view); schedules per the official tournament sites; win probabilities are Polymarket market prices — informational only, not betting advice. Not affiliated with the ATP, WTA, ITF, any tournament, ESPN, YouTube or Polymarket; all trademarks belong to their owners. Nothing is hosted or re-uploaded here — highlights link or embed from the tournaments' own channels.

• US Open 2027 dates when the USTA publishes them (the card reads "DATES TBA" until then), then 2028

Code is MIT-licensed. Match data comes from the public sources credited above and stays subject to their terms.