CS2 sub-tick explained: what changed for peeker's advantage and what did not

What exactly did CS2's sub-tick change, and was it a peeker's advantage fixer as sold? The short answer is no. CS2's new "sub-tick" system enabled capturing the exact moment of movement, shooting or grenade throws, reducing player's input latencies. However, sub-tick does not affect lag, packet loss, or major sources of peeker's advantage in CS2.

Let's break it down.

The Mechanic Valve Says has Changed

Valve's CS2 sub-tick architecturally allows the server to record player inputs—as data received by the server—down to the millisecond. These player inputs are then utilized for player movements, firing, throw and angle attributes.

Prior to CS2, input timing was recorded at every discrete tick. In Source-based games, a server sends updates, like player positions, to all clients, in discrete blocks/intervals called "ticks".

Now, the sub-tick system records the exact moment of player movement, shot, or grenade throw. That means, rather than recording the closest tick, the server can record player inputs during the intervals between ticks too. It precisely captures the moment milliseconds between ticks, irrespective of tick rate.

What tickrate used to do

Source-based games, like CS:GO, feature tick rates. Every game sends out server updates over discrete time blocks, called ticks. For every 15 milliseconds, the server "ticks", sending out a server update to all clients. Thus, slower servers naturally send out player position updates less often, affecting game accuracy and precision.

64 ticking servers send server updates every 15 milliseconds. 128 tick servers send server updates every 7.81 millisecond, almost twice as often. Therefore, with 128 ticks, servers could provide a more seamless and responsive gameplay. Conversely, performance too depends on hardware amongst server and clients. Twice as many ticks mean twice as much work for the system to calculate at every tick, not just for server.

CS:GO primarily uses 64 ticks by default, but hosting servers can theoretically set any tick rate within a supported range. Thus, higher tickrates shorten the interval between ticks. Now, CS2 elegantly solves this not by diminishing error between ticks at higher tick rates, but removing tick errors by capturing precise input timing in-between ticks.

Why the old 64 vs 128 tick debate shifts

Using sub-ticks is different from just increasing the tick rate. Intervals between slices of game time are still retained, what's key is it allows inputs to be recorded at any temporal point, which means higher precision. Now, CS2 knows the exact temporal variables of a player's action, thus removing variability around their inputs. A peek is easier, faster movement is faster, and grenade throws are also optimized well. CS2 reduces latency by recording precise input timing as compared to capturing positions before a tick, in-between ticks, or after a tick.

Therefore, neither gaming clients or servers suffer under the heavier tickrates or server load. Response to player input is faster, so tickrates have become less relevant, particularly around movement, shooting and throwing.

However, this does not obviate tick rates. Lower tick servers still send updates less frequently, reducing updates for server player positions, leaving gaps that netcode then interpolates or extrapolates using a client side solution. This is a vector, amongst several, for overcoming peakers advantage in online systems.

Sub-ticks are also not the same as lowering latency or removing interoperability. There's more we need to know.

Conclusion: What This Does Not Prove About Peeker's Advantage

What Peeker's Advantage actually refers to can vary per game and definition. Generally, it exists when players are partially exposed during a peek, so the defender that faces them, has a disadvantage. Peeker's Advantage is therefore a combination: The attacker is responding faster (essentially moving & aiming). The defender gets hit first (when an attacker starts responding to them). The defender exposure is later (since they haven't started responding yet).

Now, sub-ticks certainly accelerate the defender's response, making the attacker start responding faster. That does lessen initially peeker's advantage against them. Conversely, sub-tick does not remove the defender's later exposure advantage. Nor has Valve any published claim saying sub-ticks reduce peeker's advantage. Faster defender response after a peek is not the same as removing defenders' later exposure and hit disadvantage. Sub-ticks alleviate latency, not peeker's advantage.

Therefore, while CS2 optimally captures input timing, it does not out-of-the-box resolve every source of peeker's advantage. Tick rates still matter, as client side and server side delays still exist, particularly for server player positions as evidenced by snapshots on entities, interpolation and extrapolation. Furthermore, axes are not symmetrical: Defender is still exposed and gun and hit-registrations lag, whilst attackers move first. Sub-tick changes timing capture, not network delay.