In CS2's sub-tick architecture, the stop window for counter-strafing is not about hitting the brakes but about shifting gears. This article breaks down how sub-tick sampling changes the timing of your counter-strafe, why the speed decay curve still dictates a narrow window, and how to train your crosshair reset to align with the new precision.
Introduction
Many players think of counter-strafing as slamming on the brakes. But in CS2's sub-tick architecture, the A and D keys you press aren't sampled at a specific frame; they're timestamped within a tick. The catch: there's an interpolation layer between the moment of recording and the moment the shot is resolved. The stop window is fundamentally a question of shifting gears, not braking distance.
1. What Sub-tick Gives and Takes Away
In the old 128-tick logic, inputs were sampled once per tick, and ballistics were resolved per tick. If you released D right on a tick boundary, the next tick would zero your velocity and your shot would be stable. CS2 decouples input sampling from the tick, so theoretically the millisecond you release is recorded, and the server extrapolates your position based on that timestamp.
Sounds great. But in practice, your spray still drifts, and your first shot still misses "for no reason." The issue isn't sampling precision—it's the velocity decay curve. When you release a direction key, your speed doesn't instantly drop to zero; it decays in the opposite direction of your acceleration. Sub-tick records your release earlier, but it doesn't change the time required for decay.
In other words: the server knows sooner that you want to stop, but your character still slides a bit. That slide is the true width of the stop window.
2. The Stop Window Is Not Zero; It's an Interval
Break counter-strafing into three phases:
- Input reversal: You press the opposite key, and acceleration reverses.
- Velocity crosses zero: Your actual speed goes from positive through 0 to negative. This instant is the "most stable" point.
- Reverse acceleration: If you hold too long, your character starts moving in the opposite direction, and your shot diverges again.
Sub-tick changes the precision of phase 1 being recorded, and phase 2 is simulated more accurately, but phase 3 depends entirely on your fingers. The problem for most players is holding the opposite key too long. They think "holding A is what stops me," but after velocity crosses zero, every extra millisecond adds offset to your next shot.
The real stop window is the narrow gap of a few dozen milliseconds around the zero crossing. Sub-tick calculates this gap more precisely, but it doesn't widen it.
3. Crosshair Reset: The Forgotten "Second Counter-Strafe"
Most people only care whether their character stopped, not whether their crosshair returned. These are two separate things.
When character velocity hits zero, base spread returns to minimum. But the crosshair itself has a visual reset process—your hand is correcting the mouse in the opposite direction, and this correction and character velocity are two independent curves. If the character stops first and the mouse arrives later, your first shot still misses.
Under sub-tick, the server's shot resolution is closer to your "fire" timestamp. This means your mouse correction must complete earlier than your character's deceleration, not later. Previously, tick sampling gave you a little buffer; now that buffer is compressed.
In practice, train counter-strafing in reverse order: first practice crosshair reset, then practice the opposite key. Because the mouse is what you actively control; character deceleration is passive simulation.
4. A Concrete Drill for the Training Map
In a workshop moving-target map, don't rush to kill static targets. Treat the target as a metronome:
- Strafe with D. When the target enters the right side of your crosshair, reset the mouse first, then press A.
- Don't hold A longer than the moment you feel "the character stopped."
- Fire.
Record a demo and slow it down to watch three things: the time velocity crosses zero, the time the mouse stops moving, and the fire timestamp. The closer these three points are, the cleaner your counter-strafe. Sub-tick won't do it for you; it just exposes the error more clearly.
Counter-strafing was never about pinning your character to the ground; it's about aligning velocity, crosshair, and fire in the correct few dozen milliseconds. Sub-tick makes the ruler finer, but you're still the one being measured.