Your opponent is a range, not a hand. The villain is dealt real cards from a realistic range for their position and action, and plays them with strength-based logic. But your decisions are graded against their whole range — because at the table you never see their cards, and the right play is the one that wins against everything they can have.
The equity is real. Every grade comes from live Monte Carlo simulation: your hand versus the villain's current range, run hundreds of times in your browser at each decision point, compared against the price the pot is laying.
Honest simplifications: one opponent, fixed bet sizings (2/3 pot bets, 2.5x raises), heuristic villain behavior, and preflop grading based on stack-depth charts rather than a solver. The Lab trains the decisions that matter most — count the equity, respect the price, and know your stack zone.