A model can be wrong in two completely different ways — and "more accurate" and "more precise" are not the same fix. Fire some shots at four target boards to feel the difference, then drag two sliders to find your own spot on the range, and finish with a quick game that tests whether you can spot overfitting and underfitting on sight.
The bullseye is the truth — the value a model is actually trying to hit, on repeated tries with slightly different training data. Click "Fire" on each board to see ten shots land. Where they cluster relative to the center is accuracy; how tight they cluster with each other is precision.
Drag the sliders to set the model's accuracy and precision independently, then fire a round. Watch how the same "good model" feeling can come from very different combinations — and how low precision alone can hide a model that's accurate on average but unreliable on any single prediction.
A fresh pattern of shots appears with no labels. Decide which of the four failure modes it shows, then check your answer. Ten rounds, see how many you get right.
Every one of these four patterns has a direct counterpart in how a model learns from data — and the fix for one is often the wrong fix for another.
Teams often chase a single accuracy number and miss that the real problem is variance, or chase tighter predictions and miss that they're now consistently wrong. That's the kind of evaluation question we help teams get right before they ship.
Talk to us about your AI roadmap