WARDOGS Ban Wave: 185,000 Accounts Explained

Guide status checked against the live build on 2026-11-12 (Update 0.3.0, Season 03).

Quick answer

This wave removed 185,000 accounts. The developers describe the number as close to 5% of purchases, and compare it with Season 1, when about 10,000 accounts were banned from a player base of roughly 4 million. The scale jump is the story: the same team, a much larger action, and a public explanation of method that did not exist in Season 1.

Two things are worth separating. The numbers are official. The interpretation, including claims that the wave caught innocent players, is not.

The numbers as published

The 5% figure is not a measured cheat rate for the whole player base. It is the developers' description of this action relative to copies sold. Those are different claims, and the second one is the only one being made.

How a ban is actually assembled

The most useful part of the official explanation is that no single signal decides an outcome. Enforcement is described as a profile, built from several independent sources:

• Session playback. Available since Season 2, this lets reviewers replay a flagged session rather than infer from a scoreboard.
• Raw mouse input against server-side hits. Input is compared with what the server registered, which is where aim behaviour becomes visible as a pattern rather than a suspicion.
• Hit-rate history. A curve over time, not a single match. A player whose headshot rate collapses after a detection update looks different from a consistently strong player.
• Player reports. These flag an account for review. The developers state plainly that a report alone does not ban anyone.
• Behaviour patterns. Reloading while facing a wall, pre-firing the ground before a slope, and similar routines that make sense only with information a player should not have.

Before a rule goes live, the team runs a simulated ban pass and reviews the recordings. That review step is the part most players never see, and it is also the reason the official position is that a single bad match cannot decide an outcome.

What the confidence figure does and does not mean

The developers say their confidence reaches the 99.8 to 99.9 percentile. Read that as a claim about the model's separation between flagged and normal play, not as a guarantee about any individual account.

It does not mean 99.8% of banned players were cheating, because percentile and error rate are not the same quantity. It is also not independently audited: no third party has published a review of the method, the training data or the outcomes. The figure is the team's own, and it should be labelled that way.

Hardware IDs and the digital bloodline

The alt-account policy is the part with the widest side effects.

According to the official explanation, a banned player's new account can be linked after playing for a short while, because hardware identifiers tie the machine to the earlier account. The link is described as working in both directions: a flagged alt can also lead back to a main account. The developers used the phrase "digital bloodline" for this.

What is not answered is the case that matters most to ordinary players: a shared PC. Households, student flats, internet cafes and borrowed machines all put multiple accounts on one set of hardware. The official material does not describe how those cases are separated from evasion, and no exception process has been published. Treat that as an open question, not as a known safe harbour.

Regional data and the bias correction

The regional breakdown was published together with an explicit correction. The developers stated that the claim "WARDOGS cheaters are mostly Chinese" is wrong, and gave the rate by share of copies sold:

• Africa: about 12%, the highest rate.
• South America: about 8%.
• Europe and China: lower rates, but the largest absolute numbers because of sales volume.
• North America: below expectations.

The distinction between rate and absolute count is the whole point. A large market produces a large count even at a lower rate, and reading the raw count as a statement about a region's players is a mistake the developers called out directly.

Gold farming servers

A separate action targeted gold farming. The team says it infiltrated farming servers, described as carrying Chinese server names while hosted in an EU datacentre. Participants had cash, unlocks, gold bars and XP wiped, and were permanently banned.

That detail matters for appeals: buying progress, participating in organised farming, or joining a server whose purpose is currency generation is treated as a separate category from cheating in a match, with its own consequence.

The false-ban videos

In Chinese-language communities there are videos describing a "false ban wave". In the comments, some players say they were permanently banned after only playing knife or melee servers.

This is a community self-report, not a verified fact. There is no published list of affected accounts, no third-party adjudication, and no official response to those specific claims. It is also in tension with the official description, in which reports only flag an account and enforcement requires a reviewed profile. Both statements can coexist, but only one of them is currently documented.

Presenting those videos as proof of a broken system would go beyond the record. Presenting the 185,000 figure as proof that every ban was correct would go beyond it in the other direction.

If you believe your ban is wrong

Do not open a second account to test the system. Hardware linking is exactly the mechanism described above, and a test account can associate the new one with the original.

Collect your SteamID64, the ban date and time with timezone, and the sessions you think were reviewed. Then follow the appeal routes in the account banned guide. No outcome should be promised, and no response time has been published.

Official silence and unverified points

• Shared computers. How a household or cafe machine is handled has not been explained.
• Regional split of the 185,000. Only rates and rankings were given, not a full table.
• How much of the wave was manual review versus model output.
• Whether hardware-ID links can be appealed separately from the account action.
• Any independent audit of the confidence figure or the method.
• The specific false-ban claims. No official response to the videos, and no verified list of cases.

Related guides

• WARDOGS graphics settings and anti-cheat
• WARDOGS account banned: appeal routes
• WARDOGS report a player
• WARDOGS Season 3
• WARDOGS patch notes