What Tick Rate Actually Costs, Measured Across Forty-One Games

Every argument about netcode that happens in public is an argument about a number nobody in the conversation has measured. This log is our attempt to put the number on the table so that the argument can move on.

The arithmetic

A server running at a rate of N updates per second resolves the world once every 1000/N milliseconds. Between resolutions it knows nothing new. Add the round-trip time of the connection and you have the window during which two players can hold contradictory beliefs about the same event.

Peek advantage — the practical name for that window in a game with sight lines — is approximately the tick interval plus the round-trip time. It is not exactly that, because interpolation and buffering vary, but across forty-one games our measured figures sit within twelve per cent of the simple sum.

Where the floor is

The useful finding is not the arithmetic, which anyone can do. It is where the perceptual threshold sits, and for that we needed a large sample of testers describing games in their own words without being told the tick rate.

Across nineteen thousand four hundred matches, the word players reach for changes at roughly eighty milliseconds of peek advantage. Below it, losing feels like being outplayed. Above it, losing feels like being cheated. The transition is not sharp but it is remarkably consistent, and it is independent of genre.

Eighty milliseconds is the difference between a hard game and a dishonest one, and it is entirely within the studio's control.

Read the table against that threshold and the design implication is blunt. At twenty ticks you are at the floor on a perfect connection and over it on anything else. At thirty you have room for a moderate connection. At sixty you are safe for most of the world. This is why Verge Squad was accused of cheating for eight months and then was not.

The eleven predictions

Since establishing the threshold we have measured the tick rate of eleven games before playing them properly and predicted whether players would describe them as unfair. Ten predictions held. The failure was a game at twenty-four ticks that nobody complained about, and the reason turned out to be that its combat has no instant hitscan at all — every weapon has travel time longer than the peek window.

That exception is instructive rather than damaging. The threshold applies to games where an event is resolved instantly. Give a weapon a flight time and you have bought yourself the tick interval back, which is precisely why siege engines in Banner Fields are disputed at nine per cent while its melee is disputed at forty-four.

What this does not say

It does not say that a high tick rate makes a game good. Bazaar Run runs twenty ticks and scores 7.9 because nothing in it is resolved instantly. Spore Depths runs thirty and scores 8.2 for the same reason.

It says that if your design resolves an event faster than your server can arbitrate it, no amount of art, writing or balance will stop players from concluding that the game is lying to them. The required rate is a function of the shortest decision in the game, and it should be calculated during design rather than discovered in the forums.

Findings

• Peek advantage is approximately tick interval plus round-trip time, within twelve per cent across forty-one games.
• Testers switch from unfair to difficult at roughly eighty milliseconds of peek advantage.
• The threshold is consistent across genre and independent of how good the game otherwise is.
• Weapons with travel time longer than the peek window are exempt.
• Eleven pre-play predictions, ten correct, one instructive failure.
• Twenty-seven per cent of the archive runs below twenty ticks.

Status

Published. The finding survives the removal of any single game from the sample, including Verge Squad, which supplied the cleanest evidence. We will revise the eighty-millisecond figure if the tester sample crosses thirty thousand matches and the threshold moves by more than ten milliseconds.