Season 5 Reloaded: AN-94 retakes the long-range rifle tier as MXR-17 loses four stats at once

Two rifles, two directions of wind

In the Season 5 Reloaded balance notes, there is a symmetry worth pausing on longer than most people paused for it. AN-94 received four changes pulling in a favourable direction: increased maximum damage, expanded damage ranges, reduced reticle shake, and 10% less vertical recoil. MXR-17 received four changes pulling the other way: reduced damage, reduced range, reduced bullet velocity, and added recoil. In a single update, two long-range assault rifles stepped into two opposing winds.

That symmetry carries the signature of a familiar balance philosophy: dethrone the reigning king, elevate the challenger. But nodding along at that point skips the hardest part. The note says AN-94 is stronger. It does not say by how much, at what distance, against how many armour plates, or in whose hands. The gap between those two sentences is where every hasty conclusion gets buried.

In my own records of South-East Asian ranked lobbies during the first days of the update, the notable detail was not how many players switched weapons. The notable detail was that the first ones to switch clustered in the middle skill band — players who do not earn their living by taming recoil, but by picking the right gun. The patch handed that group exactly what it needed, and that group is the largest group in every lobby.

A mid-season patch and three layers of questions

Season 5 Reloaded belongs to the mid-season update family. This is the kind of patch developers use to correct weapons that have drifted out of the safe zone after the first weeks of a season, rather than waiting for the next one. The consequence is a very short life cycle for any gun: a few weeks at the top, a few weeks pushed down, then another mention in the following patch.

This update is described as a multi-weapon balance pass. The long-range assault rifle tier was pushed toward mid-to-long-range engagements, where accuracy matters more than fire rate. AN-94 and MK35 ISR are the two names leading the beneficiary group. On the other side, MXR-17 was pushed out of its default-pick position. The submachine gun tier received smaller adjustments: CBRS-3 was slightly reduced as a preventive measure, while MPC-25 and Ryden 45K received support, with VST mentioned for mobility.

There are three layers of questions to separate. The first is the stat layer: which number changed, and by how much. The second is the behavioural layer: how players react, and how long it takes for a weapon to become a common pick. The third is the outcome layer: pick rate, win rate, eliminations per match. The analysis I am working from only has the first layer. The other two have no data yet.

Based on my experience tracking matches, the lag between a patch and the actual shift of the meta usually falls between one and two weeks. During that window, most players stick to old habits while a small group reads the patch notes and swaps early. That lag is a temporary edge, not a permanent truth.

I still remember the lesson from 2026, when I was still typing up numbers for a Vietnamese football club. That team generated an average of 2.1 expected goals per match but scored only 0.8, and I concluded that keeping the coaching staff was the most rational path to survival. The board decided otherwise, sacked the coach, and the club was relegated with 21 points. Data does not lie — the listener was simply not patient enough. That story repeats in every sport, including in a weapon patch.

AN-94: four changes and one arithmetic threshold

AN-94's four changes look good at first glance. Increased maximum damage. Expanded damage ranges. Reduced reticle shake. 10% less vertical recoil. All four push the same weapon forward, and when four stats move in the same direction, their effect does not add — it multiplies.

But the hard part lies elsewhere. A damage change only has practical meaning when it crosses a specific arithmetic threshold: the threshold that reduces the number of bullets required to kill by one. If each bullet deals X damage and the target has a fixed pool of health plus armour, then raising X by a few percent only makes a difference when it pulls the required bullet count from five down to four, or from four down to three. Below that threshold, the damage increase is just a pretty number on paper.

This is the point most readers skip when reading a balance note. They read the words "increased damage" and assume time-to-kill fell. Those two things are not the same. A weapon can gain damage without its time-to-kill changing, if the bullet threshold is not broken. Conversely, a very small change can create a large jump if it lands exactly on the threshold.

The patch note does not publish that threshold. Nor does it publish the armour configuration the developers use as a balancing baseline. Without those two facts, any claim along the lines of "AN-94 is now the king of long range" is a hypothesis presented as a conclusion. One number is an accident. A cluster of numbers is a confession — but a cluster without a threshold is still an unchecked statement.

The expanded damage range needs to be read the same way. A damage range is the set of distances at which a weapon retains its highest damage value. Expanding it only helps if the expansion falls inside the distances where real fights happen. If the expansion sits in the 80-to-95-metre band while most gunfights in the main modes end before 60 metres, it is a gift nobody opens.

The remaining two changes are more practical. Reduced reticle shake directly affects the first shot and the ability to stay on a moving target. A 10% reduction in vertical recoil affects the second and third shots — that is, it affects the hardest part of a long-range exchange: keeping the bullet line steady while the sight drifts upward.

Combined, these four changes do not create a weapon for elite players. They lower the skill threshold required to shoot accurately at distance. Skilled players were already hitting; they gain little from an easier-to-control bullet line. Average players gain a great deal, because most of their error lives in exactly those second and third shots.

The design conclusion: the patch is redistributing power toward the middle of the skill distribution. If that is right, the trace will appear earliest in mid-tier ranked lobbies, not at the top. That is a testable prediction, and I will test it using weapon pick rates across skill bands.

MK35 ISR: a high ceiling and the price of being demanding

MK35 ISR is the second name in the beneficiary group, but it entered this patch carrying a different label: less forgiving than AN-94. In balance language, "less forgiving" means the weapon punishes error more heavily and rewards precision more generously.

That is the profile of a weapon with a high skill ceiling. Players in the upper band can extract far more from it than average players, and the gap between the two groups will widen. Where AN-94 flattens the skill distribution, MK35 ISR steepens it. Two weapons benefit from the same patch while producing opposite social effects.

One thing must be stated plainly about the data: the available analysis describes MK35 ISR as buffed but does not publish full per-stat detail as it does for AN-94. That gap matters. With AN-94, I at least know exactly what changed. With MK35 ISR, I know it sits in the beneficiary group but not why at the stat level. Writing about a weapon without its stat sheet is writing on faith, and faith is not something I put into an article.

A demanding weapon has a noteworthy statistical trait: its pick rate usually rises slowly, while its win rate when picked is high. If that pattern appears, the reasonable conclusion is that MK35 ISR is becoming the tool of a narrow but effective group. If the pick rate rises fast while the win rate does not move, that is the signature of an imitation wave with no skill foundation beneath it.

MXR-17: an execution across four stats

MXR-17 lost damage, range, bullet velocity, and gained recoil. The analysis describes it as no longer the automatic best pick, and that phrasing tells us it used to be exactly that.

Four stats struck at once is a heavy-handed decision, and it has its own logic. In the balance cycles of shooter titles, one reduced stat usually drops a weapon a few places. Two reduced stats can change the weapon's role within a team composition. Four reduced at once means the developers decided to remove that weapon from high-level play, not merely to tune it.

Bullet velocity is the most underrated of the four. Players usually look at damage and recoil. But bullet velocity determines the travel time of the projectile, and travel time determines how far ahead a player must lead a moving target. Lower bullet velocity increases the required lead. Skilled players compensate. Average players do not, and they will feel the gun is "strange" without being able to explain why.

That is why a weapon hit across four stats rarely returns to its old usage level, even when each individual cut is small. Their effects do not sit on the same axis. Damage and range affect the kill window. Bullet velocity affects the skill required. Recoil affects controllability. Four different axes pushed back at once create a gap that practice on a single skill cannot close.

Behind the numbers sits a column that never appears in balance notes: cost. Players invested time building muscle memory for this weapon, fine-tuned attachments to its exact behaviour, and in many cases bought cosmetic blueprints for it. A patch does not refund that investment. That is why large nerfs always generate a stronger backlash than the numbers suggest, and that backlash is data, not idle emotion.

The SMG tier: variety and a hidden selection tax

In the submachine gun tier, the patch did three things. CBRS-3 was slightly reduced. MPC-25 and Ryden 45K were buffed. VST was mentioned for mobility. The result is described as a relatively varied SMG tier.

The conventional reading is positive: variety means options, options mean good balance. But another reading deserves a place on the table. In a shooter, a weapon tier with no king usually means that tier has no sharp enough edge to impose itself on other tiers. Variety at class level can be a symptom of dispersed power rather than evenly shared strength.

The selection tax also needs accounting. A tier with one obvious choice costs the player nothing to decide. A tier with five comparable choices costs attachment tuning time, testing across distances, and adaptation to different recoil patterns. That cost never appears in a stat sheet, but it appears in a player's match history.

CBRS-3 being reduced as a preventive measure is a telling signal about how the developers read internal data. They did not wait for the weapon to explode before cutting it. They cut it when internal data showed it was on a rising path. That is proactive practice, and it has a downside: a weapon that never dominated still got hit, and players will ask why a healthy option was worn down before it could prove its worth.

Crisis does not create phenomena. It only exposes data that was forgotten. If CBRS-3 really was trending toward dominance in internal data, the light reduction is justified. If it was not, the patch erased a diverse option that nobody asked to erase.

The contrarian pass: reading a patch note is not reading a match

Most of the coverage around this patch was written within days of the notes going live. That is understandable: readers want to know which gun to use, and writers want to answer before anyone else. Precisely for that reason, a large gap opens between the question and the answer.

No pick rates were published. No win rates were published. No post-patch time-to-kill table was published. All that exists is the balance note, and a balance note is a document of design intent, not a document of competitive outcomes. Reading it to learn what the developers want is reasonable. Reading it to learn what has happened is reading the wrong source.

The prevailing style offers very strong advice, framed as the safest choice for most players. "Safest" and "strongest" are two different questions. A safe pick is one that reduces variance: it will not produce peak results, but it rarely disappoints. A strongest pick is one that maximises results when wielded by a player of the right skill. Readers need to know which question they are being answered before trusting the answer.

At a deeper level, one methodological limit must be named: correlation is not causation. The patch note changing stats is clear causation. A weapon dominating afterwards is the result of a long chain involving community reaction, the spread speed of content creators, old player habits, and the tolerance threshold of each lobby. A weapon can top a stat sheet and still see little use, for reasons entirely outside the stat sheet.

There is also a paradox inside the balance cycle itself. Elevating the challenger to dethrone the king only resets the clock; it does not solve concentration. If one weapon simply replaces another in the same slot, pick rates in lobbies will still converge on a single point. The real question is not who takes the throne. The real question is whether a patch reduces concentration or merely relocates it from one name to another.

I do not write to be agreed with. I write to be verified. With a patch that has no outcome data yet, the only honest conclusion is a set of falsifiable hypotheses, each accompanied by its falsification condition.

Temporal risk and the front-running window

Every weapon recommendation in this period has an expiry date. A mid-season update means at least one more adjustment round before the season closes. If AN-94's pick rate spikes, the MXR-17 precedent will repeat on AN-94 itself, and the time from coronation to cut could be shorter than a single content cycle.

The front-running window also needs to be described honestly. Readers who study patch notes early and swap weapons in the first week hold an advantage, but that advantage comes from most players not having swapped yet, not from the weapon being stronger. Once the majority follows, the edge disappears and the balance sheet settles into a new equilibrium. That is why I call it a window, not an opportunity.

The most notable risk sits inside the most repeated recommendation. A pick described as safe for most players will pull most players toward it, and when the majority uses one weapon, that weapon becomes the target of the next patch. This loop is not the fault of writers or players. It is the structural consequence of a game balanced around pick rate.

Signals for the next cycle

Three signals should be tracked over the next two weeks, and all three are observable from the outside.

The first is AN-94's pick rate broken down by skill band. If the lowered-skill-threshold hypothesis is right, the rise will be clearest in the middle band and fainter at the top. An even rise across every band would refute that hypothesis.

The second is loadout composition in top lobbies. If MK35 ISR appears there with a high win rate while its pick rate stays low, the high-ceiling profile is confirmed. If it becomes common without winning more, it is an imitation wave.

The third is the timing of the next patch and its content. An adjustment aimed at AN-94 is indirect evidence that its pick rate crossed an internal threshold. An adjustment aimed elsewhere is evidence that this patch achieved its dispersal goal.

The crowd watches the scoreline; I watch the rest of the scoreboard. In this case, the rest of the scoreboard is the bullet threshold, the lead time required by bullet velocity, and the skill distribution of players across lobbies. Those columns never appear in a patch note, and that is exactly why they deserve a scroll.

The first thing to verify is very concrete: whether AN-94's damage increase breaks the threshold of bullets required to kill a sufficiently armoured target, or not. If it does, the story of a new king has a foundation. If it does not, we have just spent a week discussing a gift that sits beyond the reach of the very person it was given to.