I’m calling the William Hill events API from a Python script that relies on the requests library. Even with a fixed-interval delay between calls, the endpoint still answers with HTTP 429 “Too Many Requests.” My goal is to eliminate this error so the script can run reliably again. Current setup • Language: Python 3.x • Library: requests (no async) • Throttling already in place: simple sleep interval between calls What I need from you • Diagnose why the existing delay fails to satisfy the API’s real limits. • Design and implement an effective solution—whether that means smarter back-off logic, header manipulation, session reuse, token bucket logic, or another proven approach. • Update my script and document any new dependencies or environment variables so I can reproduce the fix locally. • Provide a short note explaining the API’s actual rate-limit behaviour so I understand what changed. I’ll share the current code and sample credentials as soon as we start.