Wardogs: Play with your racist, cheating, griefer mates

Why I'm writing this

Bulkhead's shooter puts a hundred players across three teams, with infantry, vehicles and an economy that gives you something to work towards. There's plenty here I want to keep playing. Game overview.

It's the accumulation that leaves me bitter: the abuse, the losses, the balance complaints and the uncertainty about whether reporting helps. Giving up Wardogs would also mean missing out on playing with my mates. Staying shouldn't be read as a vote of confidence in everything happening around us.

Joe Brammer, Bulkhead's CEO, clearly makes time for public feedback. He responded to JbowenDaMan's critique of balancing and XP. During launch, he also addressed the moderator mistake that saw Summit1g banned instead of the player the streamer was reporting, as Dexerto documented.

Those are useful responses. They also make the gaps elsewhere stand out.

I've raised racism and griefing directly, followed up about support, and posted the clip below. The publicly visible parts of those threads didn't show a substantive reply from Brammer or Wardogs. X restricts access to the full reply view, so that check can't establish that no response exists. My experience is that I'm still waiting for a clear answer about what happens next.

The studio's two-change hotfix explains decisions about thermals and the Havoc. Alongside that detail, I'm looking for clear communication about racism, griefing, overwhelmed support and the many other problems affecting a match. Why do one or two balance decisions receive a specific explanation while the community problems I've raised remain so unclear?

Public posts aren't a timesheet showing where every employee's effort went. An equipment fix can be tested repeatedly; a voice report may require a difficult investigation. Those are plausible explanations for different response times. They don't prevent the studio from acknowledging a problem or giving players a dependable route to help.

I'd like a public issues page that covers community safety alongside balance: what the studio recognises, what it's investigating, who owns the next step and when it will update us. “Still investigating; next update on Friday” would be more useful than having to guess. Ordinary players should be able to find that information without a large following or a personal video from the CEO.

The rest of this article pairs the problems with changes I'd like Bulkhead to consider. Some are safeguards I would have wanted before launch; others need development, testing and ongoing staff. I don't know which are already being built.

Racism needs evidence and consequences

I'm hearing racism and real hatred in most matches. Sometimes it's aimed at somebody; sometimes people apparently just want the whole server to know what they think of others.

In the video I posted, the exchange includes:

00:03.46: “What do you guys think of brown people?”

00:08.72: “Are they even real people?”

00:10.96: “Stinky.”

00:16.00: “Something about eugenics.”

Transcript excerpt; punctuation added. I corrected the final line after the automatic transcription misheard it. Speaker identities haven't been verified, so these words aren't attributed to any visible player name.

This wasn't directed at me personally, as I said in the post. It doesn't need to be for me to find it disgusting.

Hearing this repeatedly has left me questioning more than whether to queue for another match. I've found myself questioning whether I want to live in a country where I keep hearing people speak this way. A game lobby can't tell me how many Australians hold those views. But knowing that doesn't make the experience less wearing.

Other players describe racism and homophobia in voice chat. Muting helps immediately, but the abusive player can carry on with everyone else. Turning off voice entirely also means giving up part of a game built around coordinating with strangers.

Make a report preserve the evidence

On 2 October, I posted that support was saying it had too many requests to receive further ones. I had video evidence. A full queue shouldn't mean the game stops accepting serious abuse reports.

Bulkhead's Season 1 changelog describes report tracking, weighting against bad-faith reporting and in-server cheater-ban announcements when match termination is disabled. Those are useful measures. I'd like the reporting experience to make their effect clearer.

A voice report could preserve a short, clearly disclosed rolling buffer around the incident, with the match, timestamp, voice channel and the account associated with each transmitted stream. Keeping streams distinguishable would reduce reliance on guessing who spoke from a mixed recording. It still wouldn't prove which person was physically speaking into a microphone.

There is a practical precedent. Xbox announced a reporting feature in 2023 that lets players capture a 60-second voice incident and submit it for review. That doesn't establish what Wardogs can implement cheaply, but it shows what a usable report can contain.

Use AI to help review, with the recording beside it

I'd combine AI transcription, player reporting and AI pattern matching in a process like this:

1. Capture and attach. The report carries the relevant audio and match context. The player chooses racism, harassment, griefing or suspected cheating and receives a receipt.
2. Transcribe and flag. Speech recognition produces a timestamped transcript. A separate classifier flags possible hate speech or harassment, with uncertainty and the surrounding exchange available to the reviewer.
3. Connect repeated incidents. Group reports concerning the same event, distinguish independent reports from duplicates, and identify repeated harmful behaviour across matches. A coordinated pile of reports shouldn't count as a pile of independent proof.
4. Prioritise and protect. Urgent or repeated cases reach a staffed queue. Give players immediate mute and block controls; allow trained moderators to apply proportionate temporary restrictions when the evidence supports them.
5. Decide and explain. Serious sanctions need an accountable decision. Tell the reporter that the case was reviewed or action taken, and tell the sanctioned player the rule, duration and route to appeal without exposing the reporter's identity.

Activision's voice-moderation FAQ describes ToxMod classifying and reporting behaviour, with Activision determining enforcement. It also distinguishes detection from action and says recordings may need further review for context. That is a useful precedent for assistance and review, not a guarantee that another game's approach transfers unchanged.

Our own clip shows why the recording matters. Its automatic transcript initially rendered “Something about eugenics” as “Something about you Jenny”. A word filter could miss the point entirely. A model could also mistake a person condemning a slur for the person using it abusively.

I'd test against accents, noisy microphones, overlapping speakers, supported languages and quoted speech before letting a model affect accounts. Start by comparing its flags with human review, then measure missed abuse and wrongful flags separately. Low confidence should send a case for review, rather than turn into a confident accusation.

Any audio capture needs clear notice, appropriate consent, a short published retention policy, secure access and a defined purpose. Unreported buffer material should expire; evidence needed for a case or appeal can follow a separate, limited schedule. Routine voice chat shouldn't quietly become an indefinite training archive.

Keep the route open and the consequences visible

The report system should continue accepting and preserving evidence when moderators have a backlog. Separate urgent safety cases from ordinary support, merge duplicates and show meaningful states such as received, reviewing and closed. If evidence is missing, explain what would help.

Publish aggregate response times, repeat-offence rates and the proportion of appealed sanctions overturned. Ban totals alone don't tell us whether evenings are improving. Clear warnings, voice restrictions, match removal and escalating suspensions should follow severity and repeated conduct; severe abuse shouldn't need several free attempts first.

None of that removes the need for people. AI can help sort and inspect evidence. Someone still has to run the queue, review difficult cases and correct mistakes.

Why the abuse can feel so pervasive

I don't have data showing that racism in Wardogs is increasing over time. I have repeated encounters that make it feel pervasive. A few mechanisms could explain that experience.

The people who want ordinary conversation may leave public voice first. They mute, move into a private call or stop speaking. The same abusive players then account for more of what everyone else hears, which gives more cooperative players a reason to withdraw.

A survey of 293 women who played online games recorded communication avoidance among responses to harassment and associations with leaving games. It doesn't establish that the same feedback is happening in Wardogs. It gives us a reason to ask who has stopped speaking, as well as who is still shouting.

Exposure also accumulates. Suppose, purely as an illustration, each independent interaction had a 2% chance of including abuse. Across 10 interactions, the chance of hearing it at least once would be about 18%; across 100, about 87%.

Those are invented inputs, not measured Wardogs rates. Real interactions aren't necessarily independent, and a hundred-player match doesn't mean hearing a hundred people. “Most of my sessions” and “most of the players” are very different claims.

If hateful remarks keep passing without a visible response, newcomers may conclude this is accepted. A randomised experiment across 2,190 Reddit science discussions found that displaying community rules increased newcomer participation and rule compliance. A science forum isn't a combat game's voice channel, but clear expectations are worth testing here too.

Short conduct reminders on joining, accessible moderator contact and visible, privacy-preserving confirmation that rules are enforced could help. Reward useful logistics, revives and cooperation as well as kills. Check whether people feel able to keep public voice on; fewer reports could otherwise mean more people have simply given up.

Evidence can be difficult to recover later. Interviews with 25 Discord moderators found that fleeting voice interactions made moderation harder. A backlog could therefore add investigation work as well as waiting, if useful context isn't preserved. That is a possible mechanism, not evidence that Wardogs currently loses recordings.

Separate suspected cheating from connection problems

I've seen players who appear to wipe out entire opposing teams, disappear and then return. A separate player report describes a similar departure and return. Neither account proves an anti-cheat ban was bypassed. A disconnect, kick and permanent ban aren't interchangeable.

I've also noticed that most players I suspect of cheating appear to have around 170–320 ms ping. That's an observation from my matches, not a logged sample, and it isn't evidence of cheating by itself.

Riot's explanation of VALORANT's networking describes how delayed updates, prediction and buffering affect what players see. Wardogs may make different trade-offs. The ping belongs in the report alongside match evidence; it doesn't establish guilt, location or nationality.

Investigate patterns against what the server knows

AI pattern detection could help prioritise unusual sequences across matches, using server-recorded events and the equipment, spotting information and network conditions that applied at the time. Review should distinguish an impossible event from a strong player, coordinated squad or misleading replay.

Combine that work with checks on actions the server can validate, reliable match identifiers and evidence from player reports. Preserve enough context to review the reported encounter without turning every kill into another player's private surveillance feed. A model's suspicion score or a high kill count shouldn't be the whole case for a ban.

Bulkhead could also distinguish connection removal from a confirmed enforcement action in the messages players see. If a match is ended because cheating is confirmed, a published rule for restoring eligible loadout costs or progression would stop innocent players carrying the whole loss.

The studio has already published cash and XP exploit fixes, progression resets and possible permanent bans. I'd like regular reports on outcomes and repeat problems. Those can explain progress without publishing detection thresholds or giving people instructions to evade checks.

Use ping limits to improve match quality

I'd support clear, configurable ping limits, paired with enough regional capacity that players have somewhere suitable to go. Show the limit and measured latency before joining. Prefer nearby servers, and let players choose whether to accept a longer queue or a higher-latency match.

Use sustained latency over a window, alongside packet loss and jitter, rather than ejecting someone after one spike. Warn them, allow a grace period and explain the removal. Community admins could choose published limits within sensible bounds, while friend groups could deliberately choose a more permissive server.

The actual threshold needs testing against Wardogs' networking and regional player population. An arbitrary low ceiling could exclude legitimate OCE players without catching a single cheat. Ping limits should make encounters more consistent; anti-cheat decisions still need their own evidence.

Make moderated OCE servers available

Bulkhead's pre-launch anti-cheat announcement directs players towards community-hosted servers, where admins can remove bad actors immediately. That only helps when players can access suitable servers or run them themselves.

On 1 October, I reported that I couldn't purchase an Oceania community server from the providers. Other players report using OCE servers, so I'm not claiming none exist. QONZER's hosting page listed Sydney as out of stock when checked for this article; stock can change.

I'd like Bulkhead to work with hosts on more regional community-server capacity, publish availability and give waitlisted communities a credible update. Additional approved providers, reservable peak-time capacity or controlled self-hosting could widen access. Each needs operational and security support; simply adding a provider logo doesn't create a server people can buy.

A server browser should show region, latency, rules and whether administrators are available. Favourites, squad joining and a shared queue would make moving with mates easier. That matters when the alternative is staying in an unpleasant lobby because leaving splits the group.

Admins need recent incident evidence, temporary mute and kick controls, a clear escalation route to Bulkhead, and a record of their own moderation actions. Shared information about confirmed sanctions could help, but a community admin's accusation shouldn't silently become a game-wide ban. Players need an appeal route and protection from abusive admins too.

Regular communities could also change what people expect. If you meet the same players tomorrow, a reputation and a moderator's standards have somewhere to persist. Limited local choice could instead keep someone encountering the same hostile group. I haven't tracked that overlap, so it remains an explanation worth testing.

Community hosting should complement staffed official servers. Players shouldn't have to buy and administer their own refuge before they can expect the game's rules to mean something.

Make team killing and sabotage costly

I've paid for goods to sell, only for a griefer to blow them up. I lost 140,000 in-game cash in a single match this way, with no refund. Initially, I thought it was a glitch with the game.

That's time earning the cash, followed by someone else destroying what I bought. Before I could even report the sabotage, I had to work out whether another player caused it or the game had failed to work. Other players describe supply crates deliberately destroyed at spawn.

Stronger consequences for deliberate team killing

Track repeated friendly damage, team kills, vehicle sabotage and destruction of team resources together. A player shouldn't escape scrutiny by spreading the harm across different mechanics. Look at the sequence and circumstances, with forgiveness for genuine accidents and review for repeated patterns.

Consequences could escalate from a warning to damage restrictions, losing access to heavy weapons or vehicles, removal from the match and longer suspensions for repeated deliberate conduct. Restoring the victim's eligible loadout and spawn cost would address some of the damage immediately. Any financial penalty on the offender should be bounded so baiting an accident doesn't become a new way to bankrupt people.

Reverse friendly fire is another option to test. Ubisoft introduced it in Rainbow Six Siege so certain further attacks hurt the offending player instead of their teammates. Its documented exceptions and later adjustments also show why explosives and gadgets need careful treatment.

Wardogs would need to account for crowded vehicles, blast damage and players deliberately stepping into friendly fire. Blocking further friendly damage or restricting the offending equipment may be safer than reflecting every explosion. The aim is a consequence the offender feels, while protecting people who made an ordinary mistake.

Protect goods, or refund verified destruction

I'd start by disabling friendly destruction of purchased goods and supply crates in HQ and other designated safe areas. Keep an owner or admin removal tool for misplaced items so indestructible crates can't become barricades. If friendly crate destruction serves no useful gameplay purpose elsewhere, remove it there too.

Where destruction is part of legitimate combat, distinguish enemy action from friendly sabotage. Every purchased asset should have an owner, an original paid cost and a server-recorded outcome. A clear loss notification could tell the owner whether it was sold, destroyed by an enemy, destroyed by a teammate or lost to a server fault.

For verified friendly sabotage or a confirmed game failure, refund the remaining paid cost or replace the unused goods. Keep normal enemy interception governed by the game's published rules. If reliable attribution isn't ready, temporarily protecting the goods would be preferable to promising refunds the system can't assess.

Refunds need basic checks: pay the original purchaser, account for goods already used or sold, issue compensation only once, and never pay more than the verified loss. Review repeated suspicious claims and coordinated patterns. The innocent player's refund shouldn't depend on whether the offender still has enough cash to pay a penalty.

The same principles could support reconnect recovery and server-failure compensation: preserve ownership, explain the loss and make restoration consistent. Players shouldn't need a viral clip to recover a purchase the game can already verify.

Balance the whole match, and explain priorities

The IR Rangefinder hotfix disabled the item in vendors until Season 2, when it's due to require batteries. It also strengthened CIWS against the Havoc, correcting what Bulkhead described as an earlier overcorrection. Those were the two balance changes announced in that update, not the only changes ever made.

What I don't understand is why the response was that narrow when so many other parts of a match still feel wrong.

Little Birds seem to find everyone and rack up extraordinary kill counts, sometimes around a hundred. I've also been frustrated by apparent two-shot pistol kills to the legs with hollow-point ammunition, 60-round AMG magazines and mortar rounds that seem to land on someone in the middle of the map immediately.

Those are gameplay complaints, not controlled damage tests or proof that a particular pilot or mortar operator is cheating. Good spotting and coordination can explain a lot. But players need to understand the counterplay and have a reasonable chance to use it.

Brammer has already acknowledged the mortar problem, including ammunition cost, in the balance discussion covered by PC Gamer. If the studio recognises that the problems go further, why were thermals and the Havoc the immediate priorities, and what is holding up the rest?

Give players ways to respond

These are levers I'd want tested against actual match data:

• Thermals and spotting: range, battery use, heat limits, occlusion and ways for an exposed player to break detection. Explain the spotting rules so ordinary information sharing doesn't look like an unexplained tracking advantage.
• Havoc and Little Birds: survivability against available counters, repair and rearm downtime, spawn availability and warning time. Check whether infantry can afford and use the counter before the aircraft has decided the fight.
• Mortars: ammunition cost, firing cadence, accuracy, setup time, spotting requirements and warnings or cover that let targets respond. Increasing cost alone may leave rich squads dominating in exactly the same way.
• Pistols, hollow-point ammunition and AMG magazines: publish meaningful damage and range information, test body-part multipliers, and assess reload time, handling, recoil and magazine size together.
• Match economy and XP: reward logistics and objectives fairly, inspect whether early advantages compound too quickly, and measure whether players have a reasonable route back into a match.

A test environment, staged changes and a clear way to reverse a bad patch would help the studio learn without making every player carry the full experiment. Publish the intended result and whether it happened, including effects across skill levels and regions.

If an expensive item is disabled or substantially reworked, explain what happens to the people who bought it. Temporary substitutes, purchase credits or a stated transition policy would make necessary changes easier to accept. Narrow hotfixes can be reasonable. Communication should still cover the rest of the known problems, including racism and griefing.

The money is fictional. The time isn't.

A player's gold-price complaint describes saving roughly 1.6 million cash for a bar on 28 September, then seeing the price at 166,000 on 3 October. Five days, rather than two.

On those reported figures, the purchase price fell about 89.6%. That doesn't establish that an existing gold balance lost 89.6% of its spending power, or that Bulkhead manually caused the change. It means someone buying later could get the same bar for roughly a tenth of the cash.

The frustration is easy to understand. Reaching a goal feels different when a few days later somebody else can reach it with a fraction of the same in-game cash. An economy needs rules players can understand before converting hours of play.

Give gold an understandable reference price

One option I'd like considered is tying gold to an external real-world gold benchmark through a published conversion rule. The LBMA Gold Price, for example, is an established benchmark with data-licensing requirements. Naming a benchmark would only be the start of the design.

Bulkhead would need to define the fictional bar's unit, the baseline in-game price, which market series it follows, when it updates and how it handles missing data. A simple proposed rule could make the in-game price follow the benchmark's percentage change from a published starting date. Show the calculation, timestamp and recent price history in the game.

Real gold also moves. An external link would make the reference more inspectable, but wouldn't guarantee stable purchasing power inside Wardogs. Cash earnings, equipment prices and the amount of gold entering the game still matter. A market headline shouldn't randomly decide how much grinding a new player faces that evening.

If the goal is predictable progression, a fixed seasonal conversion or a published, bounded adjustment rule may work better. A smoothed external index is another option, provided the game clearly says it doesn't track every market movement. Any of these should remain an in-game pricing mechanism, without implying that a fictional bar is redeemable for real gold or cash.

What I want is a rule people can inspect, notice before major changes and a record of why those changes happened. An opaque internal market replaced by an unexplained external peg would leave too much of the same problem.

Review compensation for the earlier grind

I'd like a compensation review that covers the 1.6 million and 2.4 million price points wherever account records confirm the purchases. The linked player report supports the 1.6 million complaint; the 2.4 million figure is an additional price point I'm asking to have checked, not one independently verified here.

For a substantial repricing or a confirmed economy error, a one-off purchase credit could return the difference between an eligible bar's actual paid cost and the correction's published reference price. It should account for prior compensation and any relevant sales or realised gains, rather than paying the same loss twice.

For illustration only, if that reference were 166,000 in-game cash:

Those are subtraction examples, not verified losses or an entitlement already offered by Bulkhead. The studio would need to establish what changed, the affected dates and what happened to each purchase. A routine, clearly disclosed market fluctuation is different from an error or an abrupt change in progression rules.

Players who were still grinding towards those prices need consideration too. They haven't necessarily bought a bar, so purchase refunds won't reach them. A separate, bounded progress credit based on verified eligible earnings during the affected period could recognise their time. The rules should make clear how buyers and savers are treated, with an appeal route for missing records.

Compensation can distort the economy as well. Model the total payout and its effect on progression before issuing it; if necessary, consider usable, non-tradeable purchase credits with clearly explained limits. A token gesture unrelated to the grind wouldn't address the complaint, but an unchecked currency injection could create another one.

What could have been in place, and what should happen next

Basic conduct rules, an always-available reporting route, useful incident records, safeguards against friendly sabotage and a published economy-transition policy are things I would have wanted designed before launch. Regional hosting demand and moderation coverage should have been part of that preparation too.

That doesn't mean every AI system or balance change could have been finished in advance. It means predictable forms of player harm deserved a plan before the busiest period arrived.

I'd put the next steps in this order, while allowing different teams to work in parallel:

1. Acknowledge and reduce immediate harm. Publish the known community issues, keep evidence submission open, staff urgent reports, and test temporary protections for goods and repeat team damage.
2. Make reports and losses traceable. Add clear receipts, status updates, reliable voice and match context, and transaction records that support fair refunds and compensation.
3. Expand the useful choices. Work with hosts on OCE capacity, improve admin tools and squad migration, and test network limits alongside regional availability.
4. Trial the more complex systems. Evaluate AI-assisted review against human decisions, test wider balance changes and publish a gold-pricing and compensation proposal before switching the economy again.

These are priorities I'd ask the studio to explain, not delivery estimates from someone who has seen its code or staffing plan.

Give staff cover, and players an update

In his post about the launch workload, Brammer described exhaustion, cheating and DDoS attacks, with a substantial part of the team required to take holiday. He also said many still worked during that leave.

People should be able to take time off properly. I'd like leadership to arrange moderation and incident cover across active hours, train backups, use outside support where appropriate and publish when an incident is affecting service. Network providers can help with DDoS protection and capacity; support staff need a route to preserve reports during outages. None of that requires publishing sensitive defensive details.

Then give players a regular update covering the whole experience: exposure to abuse, report turnaround, repeat sabotage, connection quality, regional hosting availability, equipment balance and progression costs. State what changed and what remains unresolved. Counting fewer complaints while more players mute voice or leave would miss the point.

Final note

I'm writing because I want Joe Brammer and Bulkhead to communicate clearly about the community issues I've raised, and make it easier to keep playing without accepting them as part of the price.

I don't expect a small hotfix to repair everything. I do expect racism and deliberate sabotage to be treated as problems that deserve a public response, practical safeguards and follow-through. There are options worth testing beyond asking the people affected to mute, lose their purchases or leave.

If the studio publishes clearer answers or evidence that changes any of this account, I'll update the article. The same applies to mistakes in the facts I've set out here. The proposals above are choices to investigate, with trade-offs to resolve, rather than claims that one tool would fix the game overnight.

My mates are why I'm still here. I'd like the game to give me more reasons to stay too.