Trust & Methodology

There's got to be a better way to earn trust.

Trust isn't claimed. It's built — through transparency, accountability, and showing your work. Here's ours.

Scoring code is open-source on GitHub for anyone who wants to inspect it. See also: FAQ →

How each pillar works.

Who built this.

Me — Matt Blumberg. Vibe coded with Claude.

No political party. No political donors. No institutional backer with an agenda. A technology entrepreneur who got frustrated enough to build something. Read the full story →

Who funds this.

Right now: me personally. Maybe someday if the platform grows: small user donations, nonpartisan civic foundation grants, and potentially a nonprofit structure with full financial transparency. Before then — and if and when it does — no political parties, PACs, political donors, advertising, or any organization with a stake in where you land will ever fund this.

The core design decision.

Most civic quizzes map you onto a single left-right spectrum. We think that's wrong — not just imprecise, but actively misleading. Real political identity is multidimensional. Flattening it into one axis loses almost everything that matters.

So we built an eight-dimension model instead. Each dimension captures a genuine tension that every thoughtful voter navigates — not a proxy for party affiliation, not a coded version of left vs. right. Eight real spectrums, each with honorable positions at both ends.

The eight dimensions — deep dive.

Before you answer a single question, we show you all eight. No black boxes.

1

Stability ↔ Change

Steady vs. Bold

How much should existing systems, institutions, and norms change — and how fast? This isn't about being timid or reckless. It's about whether you trust gradual improvement more than structural transformation, or vice versa.

Example: The federal tax code runs to tens of thousands of pages, and almost no one defends its complexity. Do you reform it from within — close loopholes, simplify the brackets, tighten oversight — or does the structure itself need to be scrapped and rebuilt from scratch?

2

Local ↔ Federal

Close to Home vs. Bigger Stage

When a problem needs solving, where should the decision-making power live — closest to the people affected, or at the scale needed to make it stick? Local control can mean responsiveness and accountability. Federal power can mean consistency and reach.

Example: Should school curriculum be set by local school boards who know their communities, or by national standards that ensure every child gets the same foundation regardless of zip code?

3

National ↔ Global

Home First vs. Bigger Picture

When American interests and global cooperation pull in different directions, which wins — and how often? It's not isolationism vs. globalism. It's about where you draw the circle of concern, and why.

Example: If a trade agreement would create jobs abroad and lower prices for American consumers, but cost American manufacturing jobs — is that a win, a loss, or something more complicated?

4

Rules ↔ Outcomes

Fair Process vs. Fair Result

If a process is fair but produces unequal outcomes, is that acceptable? If an outcome seems just but the process was messy, does that matter? This is one of the deepest tensions in democratic life.

Example: A judge follows sentencing guidelines precisely and gives two people identical sentences for the same crime. One grew up with resources and opportunity, the other didn't. Is that justice?

5

Markets ↔ Governance

Let It Compete vs. Set the Rules

When something important is broken — housing, healthcare, education, the environment — is the better lever competition and private incentives, or regulation and public investment? This isn't capitalism vs. socialism. It's a practical question about which tools work better, and when.

Example: Prescription drug prices are high. Do you want more competition between pharmaceutical companies to drive prices down, or a government body that negotiates or sets prices directly?

6

Pragmatism ↔ Idealism

What Works vs. What's Right

Are your positions anchored to a vision of what should be — a principle you won't compromise — or are they constantly negotiated against what's actually achievable? Neither is naive. Both are honorable.

Example: A bipartisan immigration bill would meaningfully reduce illegal crossings and create a path to legal status for long-term residents — but it requires compromises that neither side loves. Do you support passing it?

7

Individual ↔ Collective

Personal vs. Shared Responsibility

Where does responsibility primarily live — with the individual, or with the community? This isn't about laziness vs. generosity, or freedom vs. control. It's about how you understand the relationship between personal agency and the systems people are born into.

Example: Someone is struggling financially after a job loss. How much of their path forward is on them — and how much should a social safety net absorb?

8

Trust ↔ Skepticism

Trust the System vs. Question It

Do you believe existing institutions — courts, agencies, elections, media — are basically legitimate and worth working within? Or are they captured or structurally flawed in ways that demand challenge? Healthy skepticism has a long and honorable tradition across the entire political spectrum.

Example: A court rules in a way you believe is deeply wrong. Do you accept it as legitimate even if you'll fight to change it, or do you think the institution itself has lost its claim to authority?

How we tested for bias.

Every dimension was designed so that both poles have defensible, honorable positions. Neither end should feel like the obviously correct answer or the obviously wrong one. We ran five stress tests before finalizing the model:

The partisan smell test

Does either pole secretly smell like Democrat or Republican? If yes, we rewrote it.

The independence test

Are the dimensions actually measuring different things, or are some just proxies for each other?

The real person test

We mapped eight distinct political archetypes across all eight dimensions and confirmed they produce genuinely different profiles.

The questionability test

Can we write at least two or three good, non-leading questions for each dimension?

The output test

Would two candidates with genuinely different governing philosophies score differently across these dimensions? We tested against real political figures. They do.

How we bias-checked every question.

The five tests above are about the model. But a neutral model can still be undermined by a single leading question, a loaded answer option, or a “fun fact” that quietly takes a side. So before launch, every question, every answer option, every micro-reaction, and every historical easter egg went through a separate, item-by-item review.

The method is adversarial, and we run it twice. For each item we ask: what would a sharp, good-faith critic from the left say — does any option read as the obviously wrong answer for a progressive, does the framing assume a center-right baseline, does a reaction reward one direction over the other? Then we ask the identical question from the right. An item only passes when it survives both.

Three things we specifically check:

Every option has a home.

On every question, all of the options — not just the middle one — have to be a position a thoughtful person actually holds, stated in its own strongest terms. If one side's real view is missing, or shows up only as a straw man, that's a defect we fix — even when fixing it means adding an option or rewriting one we liked.

The easter eggs don't argue.

The historical asides are there to be interesting, not to make a point. A true fact selected because it flatters one side is still a thumb on the scale, so the eggs are held to history and Americana that illuminate a tension rather than resolve it.

Structure does some of the work.

Answer order is randomized for every user, so no position gets a permanent advantage from sitting first. Internal scoring keys to the position of an option's content, never its letter. And the Layer 4 dealbreakers are deliberately paired left and right so the balance is visible at a glance — including, where the honest answer required it, items that filter candidates on our own founder's positions.

We're not going to claim the result is perfect. Bias-checking is a practice, not a one-time certificate — language drifts, the news cycle re-codes old words, and a sharp critic will always find something we missed. When that happens, the feedback button on every question comes straight to us, and the commitment below holds: if a question is biased, we fix it and say so.

Three accountability commitments.

Published methodology

Open to scrutiny, updated when we learn something. You're reading it now.

Open-source scoring logic

You can see exactly how your profile is built. GitHub →

No political donors

Ever. The independence of this platform is non-negotiable.

Still skeptical?

Good. You should be. Skepticism is healthy — we built an entire dimension around it.

hello@bedrock.guide. A human reads it.