Looking for advice on TCP vs UDP in Esports Infrastructure and Hybrid Transport – any experienced users? OFFICIAL BOARDS

Hey man,

I dealt with this recently. The core issue usually comes down to:

Early esports titles (e.g., Quake, Counter‑Strike 1.6) relied on raw UDP for player input and state updates to minimize latency, accepting occasional packet loss as tolerable. As competitive scenes grew, the need for reliable lobby chat, matchmaking, and stat tracking prompted the addition of TCP‑based services alongside UDP game traffic, creating a dual‑stack hybrid model. The late 2000s saw the emergence of engine‑level reliable UDP frameworks

Best way forward is:
Scenario 1 – Large‑Scale FPS Tournament (LAN) The event deployed a hybrid stack: UDP‑based game state (player position, aim, hit‑registration) using a custom reliable UDP layer with selective repeat ARQ and FEC (10 % overhead). All non‑latency‑critical data (lobby chat, matchmaking, stat logging) ran over TCP with TLS 1.3. Measurements showed median one‑way game latency of 0.7 ms (99th‑percentile 1.2 ms) versus 2.3 ms TCP‑only baseline, with < 0.

Hope this helps!