Bottom line: Valve's 18 March 2026 reload update cut the AWP to 15 shots per round, and the game's most prominent AWPer asked for the gun to be deleted. Four months of professional match data say the nerf did not land: AWP kill share is statistically unchanged in every slice we can cut, and the elite AWPers now put a larger share of their kills through the weapon than before the patch. The change that actually rewrote pro CS2 is the one nobody argued about. The M4A1-S fell from 82% of M4-platform kills to 56% in a single quarter, and 24 of 68 teams swapped their primary rifle. Every figure is computed from the EsportsOdds CS2 data API as of 20 July 2026.
The patch, and the player who hated it
The reload update did two things at once. Reloading now discards every bullet left in the magazine, and reserve ammunition became a hard per-round budget: most rifles kept three spare magazines, the M4A4 got four, and the AWP was cut to two, a total of 15 shots per round. The M4A1-S, with its 20-round magazine, ended up with 80 total bullets against the M4A4's 150. Valve's stated intent was that "the decision to reload should have higher stakes".
The discourse went straight to the sniper rifle. m0NESY's reaction, as reported at the time: "Another AWP nerf. Just remove it from CS2 already." donk reportedly welcomed the change. Three months later m0NESY won the IEM Cologne Major with the AWP in his hands, and donk's own weapon record shows what the patch really did, which was nothing to do with the AWP at all.
One chart carries the whole story. Both lines are drawn on the same fixed 0-100% scale, so neither can be exaggerated by a zoomed axis:
The flat grey line is the weapon everyone argued about. The collapsing lime one is the weapon nobody defended.
The AWP's kill share didn't move
Throughout this piece, "before" means 1 October 2025 to 17 March 2026 and "after" means 6 April to 30 June 2026. The fortnight in between is excluded: the patch shipped mid-March but the first big LAN played on it was PGL Bucharest, starting 4 April, so those weeks mix both versions of the game. July 2026 is excluded because our depth ingest for the current month is still filling in and would skew the sample toward top events.
Within those windows, the AWP's share of all recorded kills:
There is no slice, by tier or by venue, in which AWP kill output fell. The absolute version of the metric agrees: at S+A-tier events the AWP produced 71.1 kills per 100 rounds before the patch and 73.0 after, against a stable ~664 all-weapon kills per 100 rounds. A share metric alone could mislead here, because kill share is zero-sum and the M4 upheaval moves every other weapon's denominator; the per-round number is the check that it doesn't.
Scope honesty: our weapon data is per-match aggregates, so this measures kill output, not pick rate. A round where an AWP was bought and never killed is invisible to us, and per-round weapon presence needs demo parsing, which our pipeline does not do. If the nerf reduced how often teams buy the gun without reducing what it kills, these numbers would not catch it.
The placebo test
A before/after comparison is only as good as its baseline, so we ran the identical calendar split on the previous season, which had no March patch.
In the patch-free year, AWP share "rose" +0.29 points across that split. In the patch year it rose +0.32. The post-nerf change is indistinguishable from doing nothing, which is the honest headline: not that the nerf buffed the AWP, but that four months of professional play show no measurable change in what the weapon produces.
Fifteen shots was never the constraint
The maths of the ammo cut explains why. Professional AWPers average about 2.3 shots per kill, so a 15-shot budget funds roughly six and a half kills of ammunition in a round. Nobody needs that many.
If anything moved, it is trigger discipline. May and June 2026 are the two most bullet-efficient AWP months in our 18-month series, and the equivalent 2025 months never dropped below 2.30. Two months is a hint rather than a proven trend, but the direction is the opposite of the panic: scarcity made the snipers tidier. The SSG 08, the budget scout some predicted would replace the big rifle, stayed at 0.3-0.5% of kills all year.
The nerf that landed hit the M4A1-S
While the AWP shrugged, the rifle meta flipped. The M4A1-S lost a fifth of its total ammunition (80 bullets across four magazines) and, worse, the discard-on-reload rule taxes exactly the habits the silenced rifle rewards: holding an angle, tapping twice, topping up. Do that with 20-round magazines and you run dry.
Each square is one of the 70 teams with at least 12 depth-covered matches on both sides of the patch; 68 of them were M4A1-S-first before it. By June, 24 of those 68 had re-armed around the M4A4, and only two moved against the tide. At S+A tier the M4A1-S went from roughly 78% of M4-platform kills to 49% in three months; tier-B and C teams followed the same curve a few weeks behind, which is what you would expect when a change propagates from bootcamps outward.
donk is the cleanest single example. Before the patch: 231 M4A1-S kills to 15 on the M4A4. After it: 16 against 175, a documented switch he made within days. The player who liked the update was also the fastest to act on what it really changed.
The elite AWPers didn't blink
If the nerf had teeth, the ten players who live on the weapon would show it first. They show the opposite.
Across the ten dedicated AWPers with at least a dozen matches in both windows, the median change in AWP reliance is +0.6 points. molodoy's falling line is the only real cut, and more of the group leaned in than pulled back. The loudest complaint on record belongs to one of the steepest risers: m0NESY put 45.8% of his kills through the AWP before the patch and 55.4% after it, while his shots-per-kill improved from 2.42 to 2.19.
At IEM Cologne in June he put up 186 AWP kills across the seven matches our depth data covers, 48.7% of everything he killed, and left with his first career Major MVP. Whatever the 15-shot budget took from him, it does not show up anywhere we can measure.
The patch's winners and losers
Usage held, but did anyone's level change? Ranking the twelve stars by the change in their average match rating from the pre-patch window to the post-patch window:
Read this as a scoreboard of form, not a verdict on the patch: rating swings bundle roster moves, opposition strength and plain form together, and none of these deltas is large by rating standards. The class-level numbers carry the real conclusion. Across every qualifying player, the 55 dedicated AWPers averaged a rating change of −0.0015 and the 287 riflers −0.0027: snipers did not underperform riflers after their weapon was nerfed, by any margin worth reporting. The churn is individual. dev1ce's +0.213 is the biggest rise in the field, donk improved while re-learning a rifle, and the two most-discussed AWPers split: molodoy cut his AWP usage and rose, while m0NESY doubled down and dipped in regular play even as he won the Major. torzsi's −0.113 is the largest fall, a number that will not surprise readers of the next chart.
The sniper duels: who actually beats whom
Since the argument is really about who the best AWPer is, here is the head-to-head record among the elite, accumulated across the whole CS2 era. Rows kill columns; the fill marks the dominant side of each pair; the printed number is the raw count.
Some of these cells will start arguments, which is what head-to-head records are for. m0NESY holds a clear edge on ZywOo, 179 to 147, yet molodoy, the man the community keeps calling underrated, owns the Major MVP 111 to 91. sh1ro beats m0NESY too. ZywOo's overall +93 is the best net in the field even without a winning record against his loudest rival, and torzsi absorbs the most elite fire of anyone at -95.
Read them as records, not ratings. These are all-weapon kills between the pair, the sample sizes depend on how often the teams actually met, and a duel matrix says nothing about who carried the harder role in those games. It is a scoreboard, and scoreboards are honest about exactly one thing.
The verdict
Four findings survive every control we could throw at them. The AWP's output is unchanged since the nerf, at every tier, on LAN and online, in share and per round, against a placebo baseline. The players most exposed to the change mostly increased their reliance on the weapon. Their results held too: AWPers' rating changes are indistinguishable from riflers' across 342 qualifying players. And the M4A1-S, the gun that lost its headlines to the sniper, lost the meta instead: 82% of M4 kills to 56%, 24 teams re-armed, one quarter.
What we cannot rule out is a change in buy behaviour that never reaches a kill feed, since per-match weapon aggregates cannot see purchases. That caveat cuts both ways, and it is the first thing we would measure with round-level demo data.
For how the weapon economy shapes rounds more broadly, see the CS2 economy explained and our breakdown of AWP versus rifle impact; for what scoreboard states are actually worth, start with the comeback curve. All of it is queryable through the same API these charts came from.