How a Scorecard Avoids Being Rigged

The problem with most scorecards

Most political scorecards arrive with a thumb already on the scale. They are built by a group with a position, they reward votes that group likes, and they punish the ones it doesn't. The number at the top tells you less about the politician than about whoever did the scoring.

That is the exact failure we set out to avoid. A scorecard is a powerful object — people share it, argue with it, and remember the number long after they forget the bills behind it. If the number is bent, the bend travels everywhere it goes.

A scorecard is only trustworthy if a conservative and a progressive both believe the math wasn't cooked for the other side.

So we designed for that test from the start. Here is how the math stays honest.

The score measures your values, not ours

The first move is the most important one: we do not have a "correct" set of values that the score is measured against. The score is measured against yours.

When you tell us where you stand on an issue, we compare a politician's record to your stance, not to ours. Two people looking at the same politician can see two different numbers — and both are right, because they are answering two different questions. The method that produces those numbers is identical for everyone. We run the same calculation, the same way, for a self-described progressive and a self-described conservative. Nothing about the engine knows or cares which one you are.

That means there is no house position for the score to drift toward. The only opinion in the room is the reader's own.

A balanced panel checks the inputs

A fair method still depends on fair inputs. Before any analysis of a bill is published, it passes through review by a panel built to be balanced on purpose.

We recruit reviewers in an even split of left, center, and right. Each reviewer discloses their own political lean, and that disclosure is public. But during the work itself, the review is blind: a reviewer is checking whether a claim about a bill is accurate and fairly stated, without being told which side it might help or hurt. The question on the table is never "do I like this?" It is "is this true, and is it neutral?"

We hold to a simple bar for neutrality: a claim is only fair if someone who disagrees with the politician would still concede it is accurate. If a reviewer from one side and a reviewer from another can both look at the same statement and agree it is correct, it can publish. If they can't, it goes back.

No single side sets a number

The last safeguard is in the averaging. Where the panel's judgment feeds a published value, we average across the groups rather than letting any one of them set the figure alone. A reviewer who leans one way cannot move a number without reviewers who lean the other ways landing in the same place.

The effect is structural, not aspirational. We are not asking you to trust that our team is fair-minded. We are showing you a process where no single faction — including ours — has the power to tip the result.

What you can check

None of this asks for blind faith, because every piece of it is inspectable. The reviewer panel's composition and the reviewers' disclosed leans are published on the reviewer cohort page. Every claim a politician's record rests on traces back to a primary source you can open yourself. And the score you see is built from your stated values, which you can change at any time and watch the number move.

A scorecard earns trust the same way a referee does — not by claiming to have no opinions, but by building a process where its opinions can't decide the game.