CS2 sub-tick explainedWhy the tick rate no longer matters in Counter-Strike 2

Published on August 3, 2026 · last updated on November 4, 2026

What was the tick rate in CS:GO?

The tick rate stated how often per second the server calculated the game state and sent it to the clients. At 64 tick, that was 64 snapshots per second; at 128 tick, twice as many. Every action - a shot, a peek, a throw - was rounded to the next of these fixed points in time. That is why 128-tick servers (FACEIT, ESL) were considered more precise than Valve's 64-tick matchmaking servers.

What does sub-tick do differently in CS2?

Counter-Strike 2 records the exact moment of every input - i.e. the moment between two ticks. When you pull the trigger, the game stores to the millisecond when and where you aimed, regardless of the tick interval. Hit evaluation therefore no longer depends on 64 or 128 updates per second.

The practical consequence: in CS2, there is no sv_tickrate and no startup parameter -tickrate 128 any more. Anyone who still advertises "tick rate 128" for a CS2 server today is describing a technology that no longer exists in the game.

Which server values really count with CS2?

• CPU single-thread performance: CS2 mainly loads one core heavily. A high clock speed keeps the simulation responsive - the most important hardware factor.
• Server FPS: the more constantly the server tick runs, the cleaner the sub-tick evaluation. If the server stutters under load, hit registration suffers.
• Latency (ping): low, even latency is more important than a high tick value. A server location close to your players (e.g. Germany for Central Europe) brings the most here.

That is exactly why we do not state an invented tick rate for our CS2 servers, but rely on CPUs with high single-thread performance and a location in Germany.

What you should still check when buying

If the tick rate is gone as a yardstick, the question remains: how do you recognize a good CS2 server? By three points, all of which can be verified:

• Location. For Central Europe, a server in Germany is the shortest route. Every kilometer costs latency, and latency is the only value that is really noticeable in the game.
• Information about the CPU. A provider who says nothing about processor performance but advertises "128 tick" is selling a number instead of performance.
• Behavior under load. A server that runs cleanly with 10 players can collapse at 20. If you regularly play full rounds, plan the package size accordingly.

What you can safely ignore, on the other hand, are offers promising an "adjustable tick rate" for CS2. The value no longer exists in that form - where it appears, it is a leftover from CS:GO times.

When hits still feel wrong

The most common complaint on your own server is that shots are not registered. In most cases, it is not the server, and the distinction is simple:

• Everyone complains at the same time - then it is the server. Usually load from too many players or a plugin that creates work every round.
• Only individuals complain - then it is their connection. A fluctuating ping affects sub-tick evaluation more than a consistently high one.
• It only occurs in certain situations - for example when many players are in one place. That points to a load-spike effect, which a larger package resolves.

Before you tinker with the configuration, it is worth looking at the plugins: a badly written plugin can cost more server performance than twenty additional players. Via RCON, css_plugins list quickly shows what is actually running - and individual plugins can be unloaded as a test.

Frequently asked questions