CS2 grenade lines are validated by first-contact frame geometry—not HUD crosshair screenshots. After small map or engine tweaks, teams should re-verify release ticks, first-touch world coords, and second-bounce endpoints before trusting old defaults.
Lead
Many teams treat default utility lines like permanent IP. In CS2, what actually settles a grenade arc is the surface normal and material tag on the first contact frame, not a frozen crosshair screenshot on your HUD. When the engine or clip brushes move even slightly, the same jump-throw can remain mechanically legal while the landing shifts half a tile—onto a smoke lip, window sill, or crate corner. In info fights, that half tile often decides who gets the first legal peek.
First contact: trajectories are collision events, not drawn lines
During flight you train muscle memory; at first contact you train whether map geometry still matches your practice server. Smokes, molotovs, and HE share one collision response stack: angle of incidence, face normals, and sometimes seams between clip brushes. A minor patch that thickens an Ancient ramp, trims a Mirage window bevel, or rounds an Inferno channel mouth can turn an old “kiss the edge and bounce in” line into a ground roll for fire, or an HE that third-bounces into dead space.
Practical QA: Split every default line into three gates—release frame (are you still locked to the same tick band for hold duration?), first touch (mark world coordinates with demo or training commands; do not trust the screen cross alone), terminal state (does the smoke still sit on the info boundary you want?). After any build, run a first-touch drift table before you tweak aim or stance—not before the whole roster memorizes old POV clips.
Bounce chains: the second landing is the tactical finish
Player notes say “aim this corner,” but the round is often decided by velocity after the second collision. Edge kisses, box-top pops, and ramp extension bounces are brutally sensitive to normal error: one or two degrees can move a second landing from outside site into inner plant, or from a safe pocket into a common pre-aim. If a default relies on double bounce into site, re-test the second contact frame after a patch—not just whether the first hit still looks right in replay.
Maintain lightweight version tags: bind each line to “verified on build xxxxx + map revision.” When Major map pool overlap shifts, pull mandatory-map bounce lines to the front of scrim blocks. Steam Workshop drills can rebuild timing; they cannot replace official collision acceptance. Skins from C5GAME or the Steam Market do not change physics—but do not swap in third-party “one-click spots” for your own recorded first-touch coords; those tools rarely track clip changes across minor builds.
Why “re-record per version” beats micro-adjusting crosshair
Nudging crosshair pixels bets on drift being one-dimensional. Patches are usually multi-variable: throw speed tiers, subtick release edges, prop hull tweaks stacked together. Moving the reticle two pixels might shift both release frame and first-contact angle; post-mortems become unparseable. Safer workflow: keep the old line as a control group, then on the new build record three parallel lines from zero (quick throw, full pin, jump-throw) under one acceptance rubric—smoke rear still blocks A main, fire still seals jump spot, HE still clears the angle you care about—then commit to the playbook.
In late rounds, utility is time currency. Re-validating defaults each version buys fewer “still thought it was legal” wasted throws—one fake smoke saved is one extra frame of duel sovereignty. Pay down trajectory debt in patch week before you invent new sets; the scoreboard cares about that order.