The science behind deterministic scoring

Why the same answers should always give you the same result, and how AI guesswork quietly breaks that promise.

If you take a test twice and answer identically, you should get an identical result. That sounds obvious, and yet a surprising number of modern quizzes run your answers through generative models that add noise on every pass.

Reliability is the whole game

Psychometrics lives or dies on test-retest reliability. A measurement that changes when nothing changed isn't measuring anything. Our scoring is pure arithmetic from your item responses: transparent, repeatable, and auditable.

Where AI does belong

Language models are great at explaining results, suggesting next steps, and making reports readable. They have no business deciding your score. We keep the boundary exactly there. ## What to do with this

Treat your result as a well-researched hypothesis about yourself, one worth testing against your real life. The people who get the most from these tools hold their type loosely and their self-awareness tightly.