Peeker’s Advantage Physics: Round-Trip Latency and Angle Geometry: Fundamental Mechanics and Deep Mathematical Modeling

[Zoe Rivera]

I’ve been trying to nail down exactly where the peeker’s advantage actually originates, and I keep hitting a wall reconciling the network model with what we observe in-game. When a peeker crosses the angle, the defender’s client is rendering a delayed snapshot based on the interpolation buffer—usually 100ms of buffered history. But here’s what confuses me: how do we properly partition the round-trip latency between the two parties? If the peeker’s position updates propagate with their own upload latency, and the defender reconstructs via interpolation plus extrapolation, isn’t the effective advantage roughly half the RTT plus the interpolation delay minus whatever extrapolation buys the defender? I’m looking for anyone who’s actually modeled the exact split rather than the hand-wavy “50% of ping” rule of thumb everyone repeats.

Second, how are people handling the extrapolation artifacts when the buffer underflows mid-spray? When the server sends no new packets, the client extrapolates velocity vectors, but any change in the peeker’s movement direction creates a rubber-band snap that can make them visible for a frame before the real packet arrives. Does anyone have data on whether that artifact actually favors the peeker or defender depending on angle geometry? And practically

"Dota 2 Immortal Bracket | Position 1 Carry • 7.5k MMR"

[Author]

Posts