In Meta Model terms, each pattern points to a different recovery move: deletions call for "what/who specifically?", generalizations for testing the absolute ("always? never? what would happen if you did?"), and distortions for questioning the claimed link ("how do you know?"). The challenge questions you saw after each answer are the practical tool — the classification is just how you choose the right question.
The task: Each round shows you statements like the ones people say every day — in arguments, at work, in their own heads. Your job is to classify which Meta Model pattern the statement shows. In the NLP Meta Model, developed by Richard Bandler and John Grinder from the language patterns of therapists Virginia Satir and Fritz Perls, surface language falls short of full experience in three broad ways:
Feedback: After every answer you see the precise subtype (there are about a dozen across the three categories) and the classic challenge question that recovers the deleted, distorted, or generalized information. Reading these is where the actual skill-building happens.
Scoring: Your accuracy, best streak, and per-category breakdown are shown at the end.
Subtype drills (advanced): The Category Focus setting switches the task from the three broad categories to precise subtype classification within one category — for example, telling a Comparative Deletion from a Nominalization, or a Universal Quantifier from a Modal Operator. This is the level practitioners actually work at, and it's considerably harder than the mixed round.
A note on ambiguity: Real speech often violates several patterns at once. Every statement in this challenge was written so one category is clearly primary, but developing an ear for the overlaps is part of the skill.