Why Politics is World of Warcraft
The feeling of changing things is the mechanism by which things stay the same.
This essay explains why most political action produces nothing. Protests, national voting, posting, donating—these are how the system maintains equilibrium. The simulation of transformation prevents actual transformation.
If you've ever felt like you're doing everything right and nothing ever changes, you're playing a game designed to never end. (Local politics is different—more on that later.)
Let's start with the version everyone admits is fake.
World of Warcraft launched in 2004 with two factions: Horde and Alliance. Twenty years later, they're still at war. Players have invested billions of hours, billions of dollars, and genuine emotional energy into this conflict. People actually hate the other faction. They feel real tribal loyalty.
And nothing has ever changed. The war never ends. It was never meant to.
The design is explicit: both factions are necessary for the game to exist. Resolution would end the product. Blizzard profits from the conflict continuing. The "war" is the content—peace would be game over.
Everyone knows WoW is a game. No one thinks they're actually changing Azeroth. The tribal emotions are real, but players understand the structure.
Now ask: What if the same structure existed, but people didn't know they were playing?
| World of Warcraft | Politics |
|---|---|
| Horde vs Alliance | Left vs Right |
| Blizzard profits from conflict | Media/politicians profit from conflict |
| War never resolves | Issues never resolve |
| Both factions necessary | Both parties necessary |
| Players feel real emotions | Voters feel real emotions |
| Everyone knows it's a game | Players maintain the fiction it's not a game |
The analogy isn't perfect—WoW is centrally designed, politics is emergent; WoW has clear factions, political identity is fuzzier. But the structural parallel holds: tribal identity, perpetual conflict, zero resolution, profitable for operators, real emotional investment. WoW players know they're playing. Voters think they're governing.
Real change is expensive. Simulation is cheap. Systems drift toward cheap.
Humans neurologically overvalue immediate rewards relative to delayed ones—hyperbolic discounting. The simulation exploits this directly:
This is the marshmallow test at civilizational scale. Simulation is eating the marshmallow now. Building something real is waiting for two marshmallows later.
Simulation wins because it's cheap, riskless, and socially rewarded. Real change requires sustained effort, coordination, resources, and risk of failure—your project might not work, you might be publicly wrong. Simulation requires only symbolic action. You always successfully voted. You always successfully posted. And you get tribal belonging, moral validation, and social proof of virtue without opposing your tribe or waiting years for results.
The simulation is a local minimum—a comfortable valley that most people never leave.
You are not the player. You are the content.
The player has agency over outcomes. The content has agency only over engagement. You choose which faction to support, how much emotion to invest, which posts to share. You have no power to end the war.
The agency drive is a real biological need. Humans who feel they can't affect their environment become depressed and passive. The simulation satisfies this drive without actual effect—the feeling of action dissipates the pressure, leaving no energy for actual action.
Characteristic simulation: Protests. They feel like rebellion but function as safe venting. The test: what happens if the protest is completely ignored? Usually nothing. Everyone goes home. The policy is unchanged.
Protests that work have "or else": credible threat of sustained disruption, electoral consequences, economic damage, or escalation. The Civil Rights movement had all four—sit-ins disrupted business, voter registration changed elections, boycotts caused economic damage, and the implicit threat of escalation was credible. Most modern protests have none of these. They're scheduled, permitted, polite, and forgotten.
Real change requires coordinated action. The simulation atomizes it. Everyone acts individually and symbolically. You vote alone in a booth. You post alone on your phone.
Characteristic simulation: Social media activism. Cost: zero. Risk: zero. Coordination: zero. Effect on the world: zero.
"The people have spoken" (by voting/protesting/posting). The system claims mandate from your participation. Your attempt to change the system is converted into endorsement of the system.
Characteristic simulation: National voting. At national scale, your vote is 1/150,000,000. You choose between Person A and Person B. But you wanted to choose policies. The candidates were selected by processes you don't control. The policies they'll actually implement are constrained by forces that don't appear on the ballot.
The feeling of democratic participation substitutes for actual influence over policy. You chose! The people spoke! But the menu was fixed before you sat down.
The local exception: This critique applies to national elections. Local elections differ on both dimensions that matter: N-size and causal traceability. Your vote on city council is 1/10,000. You can attend meetings, know candidates personally, trace how your vote affected outcomes. Primaries have turnouts so low that organized groups dominate.
The mechanism is identical—pulling a lever for a candidate—but the physics differ. The simulation critique applies to treating national elections as your primary agency while neglecting local elections where the same action has 10,000x more effect.
The simulation defines "enough." Vote and your civic duty is done. Protest and your care is demonstrated. Donate and your contribution is made. Now conscience is clear. Return to normal life.
Characteristic simulation: "Caring about issues." If you "care about" twenty issues, you have perhaps two hours per week per issue. Enough to follow the news. Enough to have opinions. Enough to feel informed.
You cannot become effective. You cannot develop the expertise, relationships, or leverage to actually affect outcomes. Caring about everything guarantees affecting nothing.
Media optimizes for engagement. Politicians optimize for re-election. Platforms optimize for time-on-site. Activists optimize for donations. Each optimizes locally. No one decided to create the simulation—it's the equilibrium that survives when many agents locally optimize. The collective result is a system that converts transformative energy into noise.
This is why "raising awareness" never fixes it. The simulation isn't a conspiracy to be exposed—it's a multi-polar trap where each participant is locally rational but the collective outcome is global irrationality. The system maintains equilibrium precisely because people feel like they're changing it. The simulation is the stability mechanism.
The obvious counterargument: Civil Rights, labor movement, suffrage—these involved protests and produced real change.
Examine what they actually did:
The Civil Rights movement combined simulation elements (marches, speeches) with real agency elements: economic disruption (Montgomery bus boycott cost the city $3,000/day), legal strategy (NAACP's decades-long litigation campaign), organizational infrastructure (churches, SNCC, CORE), and credible escalation threat. The marches were advertising for the real mechanisms.
Labor movements worked when strikes actually stopped production. The threat was economic, not symbolic. When unions lost the ability to halt production (automation, globalization, replacement workers), their power collapsed despite continued symbolic action.
The pattern: successful movements used simulation as coordination signal and recruitment tool, but the actual mechanism of change was economic, legal, or electoral—with local feedback loops and traceable causation. The simulation elements were instrumentally valuable: they coordinated action, recruited participants, and signaled credibility. But they were never sufficient alone.
This clarifies the critique. The pathology isn't simulation-presence but simulation-only. Protests with economic disruption and legal strategy produce change. Protests alone are pressure valves. The modern failure mode is treating the advertising as the product.
The exit exists. It's just expensive.
1. Causal Traceability
Can you trace your action to a specific outcome in the world?
2. Skin in the Game
What do you lose if it fails?
3. Falsifiability
What would failure look like?
4. Narrow Depth
Deep expertise in one domain beats shallow engagement in twenty.
5. Local Feedback Loops
Start where you can see effects and adjust.
Consider someone who "cared about housing affordability" and spent five years in the simulation: attending protests, posting about zoning reform, voting for candidates who promised change. Zero measurable effect on housing supply.
Then they exited. Spent two years learning zoning law. Ran for planning commission in a mid-sized city. Won (local elections have small N—their campaign mattered). Changed one specific regulation: eliminated parking minimums for buildings near transit. Three apartment buildings got built that wouldn't have under the old code. 180 units.
That person affected more housing than a million marchers. Not because they're a hero—because they found a causal mechanism with small N and short feedback loops. Their action was traceable, had skin in the game, was falsifiable, required narrow depth, and operated locally. Your path will be different—building a company, teaching, writing software, whatever matches your capabilities. The specific mechanism matters less than finding one.
The war in Azeroth will never end because it was never meant to. The war in politics will never end for the same reason.
You now see the structure. Seeing it changes nothing alone—simulation is still cheap, exit is still expensive, your tribe will still punish defection. But now it's a choice.
You can keep playing. Pull the lever, feel the feeling, perform the virtue, return to normal life. Your attention will be harvested. Your participation will legitimize the system. Nothing will change.
Or you can log off. Find a causal mechanism to outcomes you care about. Probably narrow. Probably local. Probably obscure. Probably slow.
This doesn't fix the system—individual exit from a multi-polar trap doesn't collapse the trap. But the alternative is converting your finite energy into noise. At least this way you might build something that exists.
Related essays: