AN-94 and MK35 ISR: Warzone's New Power Map After Season 5 Reloaded

One number is worth pausing on: 10%. That is the vertical recoil reduction the AN-94 received in Call of Duty: Warzone's Season 5 Reloaded update. Alongside it came increased maximum damage, expanded damage ranges, and reduced reticle shake. At the same moment, the MXR-17 — the assault rifle that had become the default pick in the long-range tier — was cut on four fronts at once: damage, damage range, bullet velocity, and increased recoil.

This is the kind of patch I always read slowly. Across 23 years of watching this industry, I have seen hundreds of weapon balance passes go by, and most of them did not produce what the community believed they would. But the architecture of Season 5 Reloaded is clear enough that misreading it takes effort: one weapon dethroned by four simultaneous cuts, and two long-range rifles elevated through deliberate accuracy adjustments.

I do not trust intuition; I trust numbers that speak after being asked the right question. And the right question here is not "which gun is strongest." The right question is: what behaviour is this patch trying to change, and does it have enough force to change that behaviour before the next patch arrives?

Context: why a mid-season patch carries this much weight

Warzone is a battle royale, and that changes entirely how a weapon balance pass should be read compared with round-based shooters.

In a round-based game, players buy weapons with in-match currency and rebuild their loadout each round. If a weapon is nerfed, you lose one option. In Warzone, you carry a loadout prepared in advance — a primary, a secondary, and a set of perks. When a patch changes stats, it does not change an in-match choice. It changes something the player has invested time unlocking, practised to learn the recoil pattern of, and bonded with through muscle memory.

That is why mid-season Warzone patches generate reactions far larger than their technical scale. A 10% vertical recoil change is not a large change in arithmetic terms. It is a large change in behavioural terms, because it touches what I call the reachable skill ceiling of a weapon.

Some foundational terms deserve spelling out, because I know many readers of weapon analysis are never given the basics. TTK — Time To Kill — is the time required to eliminate an opponent with a given weapon. But TTK on paper and TTK in a match are two different things. TTK on paper assumes you hit continuously. TTK in a match depends on whether you can keep your reticle on target while the weapon is kicking upward.

And that is precisely where Season 5 Reloaded lands. It does not primarily change theoretical TTK. It changes the ability to convert theoretical TTK into practical TTK.

Vertical recoil is the upward force on the barrel during firing. Reticle shake is the wobble of the aiming reticle while you are aiming. Both are variables that directly affect hit rate at distance. A 10% vertical recoil reduction does not mean a weapon is 10% stronger. It means that at 60 metres, the number of rounds you place into a target within a burst goes up — and in the long-range rifle tier, that is the entire story.

Core analysis: the architecture of a "nerf the king, elevate the challengers" patch

What Season 5 Reloaded actually does is restructure the long-range rifle tier toward accuracy-based engagement rather than raw firepower — and this is a deliberate balance philosophy, not an accident of numbers.

Let us walk through each group.

AN-94: the weapon designed to become the "safe pick"

The AN-94 received four changes in the same direction: increased maximum damage, expanded damage ranges, reduced reticle shake, and 10% less vertical recoil.

Read individually, none of these is remarkable. Read together, they form a very clear profile. Increased maximum damage widens the window of rounds needed to eliminate. Expanded damage ranges hold that damage further out before it drops off. Reduced reticle shake makes the first shot more trustworthy. Reduced vertical recoil keeps the second and third bursts inside the target zone.

These four changes do not add up to "stronger." They add up to "easier to use at distance." Those are entirely different things.

A weapon with a high skill ceiling rewards good players but punishes average ones. A weapon with a low skill floor does not reward good players much, but does not punish average ones either. After this patch, the AN-94 is being pushed toward the second category. It becomes a weapon that a player at 40% skill can still use to apply pressure at 80 metres.

In my analytical framework, I mark this at medium-to-high confidence. Medium because we have no post-patch pick-rate or win-rate data. High because the direction of the four changes is unmistakable.

MK35 ISR: the weapon for players who already have skill

The MK35 ISR was also elevated, but with a different profile. It is stronger, yet still described as less forgiving than the AN-94.

This is the detail I think many readers skim past. In the same patch, the developers raised two long-range rifles while preserving a usability gap between them. That is deliberate design: two options for two different player groups, rather than one option for everyone.

If the AN-94 is the weapon for players who want stable results, the MK35 ISR is the weapon for players who believe stable results are not enough. A higher ceiling, paid for with more misses while you are still learning it.

In our analytical trade, this is called a two-tier structure: a general tier and a specialist tier. That structure preserves variety while keeping a safe anchor point for the majority.

MXR-17: four cuts and what they mean

The MXR-17 lost damage, lost damage range, lost bullet velocity, and gained recoil. Four stats, one direction.

In industry balance practice, people rarely cut four stats at once for a weapon that is only slightly ahead. Cutting one stat is fine-tuning. Cutting two is adjustment. Cutting four is a statement.

The treatment of the MXR-17 tells us something the original analysis also confirms: before Season 5 Reloaded, this weapon was the default pick in its tier. In a battle royale, a weapon being the default pick means it is no longer a choice. It is a condition.

And when a weapon becomes a condition, the entire ecosystem behind it distorts. Players stop asking "how should I play" and start asking "do I have that gun yet." Creator content collapses into a single topic. Alternative options become statistically meaningless because their sample sizes are too small.

I once bet on the wrong dataset and received the right lesson. That lesson was: when one variable accounts for too much variance in a system, you cannot read the rest of the system until that variable is removed or demoted. This patch is doing exactly that to the MXR-17.

CBRS-3: the preventive cut and the lesson about internal data

The CBRS-3 was reduced slightly. The original analysis describes this as a preventive measure to stop it pulling too far ahead.

This detail is small but important, and I want to dwell on it a moment longer.

A preventive cut only makes sense if the decision-makers saw a trend before it appeared in public data. That means the developers are looking at a dataset the community does not have: pick rates, win rates, elimination rates by weapon, by map, by mode, by skill bracket.

That is the core information asymmetry of every esport. Players see outcomes. Developers see causes. And when they act preventively, they are giving us an indirect signal about what they see in that data.

I mark confidence for this inference as low, and I state why: a preventive cut may come from data, or it may come from a content release schedule. We cannot distinguish those two possibilities from patch notes alone. But we can log it as a signal to watch.

The SMG tier: a managed system, not a free-floating one

The MPC-25 and Ryden 45K were buffed. The CBRS-3 was reduced slightly. And the original analysis describes the SMG category as relatively varied.

Those three facts, placed side by side, paint a different picture from the long-range rifle tier. At the rifle tier, the developers are doing major work: demoting one, elevating two. At the SMG tier, they are doing minor work: modest buffs to two, a modest nerf to one, and leaving the rest alone.

That is the signature of a healthy tier. You only need major surgery when a system is broken. When a system functions, you need small adjustments to stop it drifting in one direction.

For players, this means the current SMG tier has no single correct answer. And in sports analysis, that is usually the best state a system can reach — not a state with one optimal option, but a state with several equivalent ones.

VST: the mobility variable

The VST is mentioned as a mobility-leaning option. This is a detail I want to separate from the rest, because it represents an entirely different axis.

In a battle royale, mobility is not a secondary stat. It is a survival stat. Mobility determines whether you make the circle, whether you can reposition under pressure, whether you can cross an open field without being eliminated.

A mobility-leaning weapon is a weapon for players who understand that matches are not decided solely by hitting shots. They are decided by being in the right place with the right people.

That is why I never rank weapons on a single axis. A weapon can be the best in a head-on duel and the worst in a retreat scenario. And in a battle royale, retreat scenarios occur more often than head-on duels.

The contrarian angle: patch notes are not performance data

This is the section I want to give the most space to, because it is the section most meta analyses skip.

Everything above rests on a single data type: patch notes. And patch notes are a statement of intent, not a statement of outcome.

Patch notes tell us what the developers want. They do not tell us what players will do. And in a competitive system, player behaviour is what actually decides the meta.

Consider this concretely. The AN-94 had its vertical recoil reduced by 10%. But vertical recoil is not an absolute constant in player experience. It is a perceived variable. A player accustomed to the MXR-17's recoil path will not automatically switch to the AN-94 just because the AN-94 is more controllable. Muscle memory is a switching cost, and it appears in no stat table anywhere.

That is why I always separate two questions: which weapon is stronger, and which weapon will be used more. The answers usually differ, and the gap between them is where bad predictions are born.

The mistake I made years ago taught me that data never lies — only the reading of it does. In 2026 I built a model on expected goals and progressive passes and concluded a national team should play possession football. My model was arithmetically right. It was humanly wrong: the coach did not have the players to play that way, and no time to build it before a knockout match.

Season 5 Reloaded sits in exactly that position. The numbers have changed. The player base has not changed with them, and will take weeks to do so — if it does at all.

There are three specific blind spots worth naming.

The first is sample size. An analysis published immediately after a patch, like the source analysis I am working from, cannot contain post-patch performance data. It has pre-patch data and patch notes. That does not make it wrong. It makes it a forecast, and a forecast should be labelled as one.

The second is temporal validity. A battle royale meta can shift mid-season. A further mid-season patch can invalidate every recommendation here. In my practice I assign each recommendation an effective window. For Season 5 Reloaded, that window runs from the patch to the next patch.

The third is the "safe choice" framing. The source analysis prioritises the safest choices for the broad player base. That is a reasonable editorial decision. But it carries a rarely stated consequence: it optimises for expected average value, not peak value. For an average player, the safe choice is optimal. For a high-skill player, the safe choice is usually a bad trade.

And this is where I want to talk about markets. The betting market is not wrong; it merely reflects a truth you have not yet seen. When the community calls a weapon "the strongest," what the market is actually pricing is not the weapon's strength. It is pricing consensus about the weapon's strength. Those two can coincide, and usually do in the short run. They separate the moment an external factor — a patch, a tournament, a prominent player — changes the consensus without changing the weapon itself.

I have watched enough cycles to recognise a pattern. After every major patch there is a window of two to three weeks in which the community still plays the old meta while the tools have already changed. Players who notice that gap and act early gain a temporary edge. But that edge has an expiry date, and it disappears the moment the rest of the player base catches up.

Esports does not need luck; it needs people who read the meta faster than the servers do. In this case, "faster than the servers" means reading the patch notes before the community reads them, and understanding them in a way the community does not yet.

Signals to track in the next cycle

I do not issue weapon recommendations. That is the job of guide writers. My job is to name the signals that will tell us whether this forecast holds.

The first signal is the AN-94's pick rate. If it spikes over the next two to three weeks, we will have a very clear precedent: the MXR-17 occupied exactly that position and was cut on four stats. The developers have demonstrated they are willing to do it. A weapon that becomes too popular becomes the target of the next patch, regardless of whether it is genuinely strong or merely perceived as strong.

The second signal is the arrival of a further mid-season patch. If that happens, this entire analytical frame needs resetting. In my practice I stamp every meta recommendation with an assumed expiry date, and I never read weapon analysis without checking the current patch version.

The third signal is competitive uptake. In Warzone events, team weapon compositions typically run either a few weeks ahead of or a few weeks behind the community meta, depending on whether competitors are experimenting or optimising. If the AN-94 and MK35 ISR appear heavily in professional matches, that is a confirmation. If they do not, that is also a signal — but a signal about the limits of the patch, not about the quality of the weapons.

One more thing about reading these signals. They are not independent of each other. A rising pick rate can trigger a new patch, and a new patch can push the pick rate down. Competitive uptake can lift the pick rate, or it cannot, depending on whether viewers imitate competitors. This is a system with feedback, not a linear chain of causation.

And that is what I want to stress at the close: correlation is not causation. A weapon with a high win rate is not necessarily a strong weapon. It may be a weapon that good players choose. The difference between those two readings is the entire foundation of this profession, and it is the thing that online weapon tier lists almost always ignore.

Every season is a ritual, and the analyst is only the person who records the omens. Season 5 Reloaded left three clear omens: the long-range rifle tier has been restructured toward accuracy, the SMG tier sits in a managed state of variety, and the developers have proven they will dethrone an over-popular weapon with multiple simultaneous cuts.

What we do not yet know is whether the community will read those three omens before the next patch arrives. That is the only question the data has not answered.