Is Dad or Mad Accurate? An Honest Look at What It Measures

Short answer: it gives a structured AI analysis of visible resemblance — which is not the same as an objective measurement or paternity result.

It's a fair question to ask of any tool that hands you a confident-looking percentage. Here's the straight answer about what Dad or Mad does well, what it doesn't, and how much to trust the number.

What it's useful at

Dad or Mad is built for one narrow job: organizing visible family-resemblance observations into a clear report. With clean photos, the AI can call out feature-level similarities, show a directional parent lean, and summarize the photo set with a score. That score is the model's structured visual judgment, not a laboratory measurement or an identity match.

What it is NOT accurate at (because it doesn't do it)

• Paternity or any biological relationship. It reads pixels, not DNA. A high score is not proof of a family link, and a low score doesn't rule one out — see is an online paternity test real?.
• Predicting the future. Faces change a lot with age, especially in kids.
• Settling anything serious. It's entertainment; only an accredited DNA test can answer parentage.

What moves the score

Photo quality is the biggest factor. Clear, front-facing, well-lit images with one face produce the steadiest results; blurry, angled, or multi-face photos lower both the score and the confidence. Matching the angle, lighting, and expression between two photos makes the comparison far more meaningful — full checklist in best photos for a face comparison.

How to read your result

Treat the percentage as a fun, directional signal and lean on the reliability level shown alongside it. If you want the mechanics, see how AI face matching works and our about page for the philosophy behind keeping it honest and private.