Learning Curve

A learning curve plots performance against amount of practice, and across an enormous range of skills, that curve takes a strikingly consistent shape. Newell and Rosenbloom's (1981) power law of practice showed that reaction time and error rate on a practiced task decrease as a power function of the number of repetitions: improvement is steep at first and then flattens, so that the hundredth hour of practice yields far less visible gain than the first. The shape holds across skills as different as cigar rolling, typing, and mental arithmetic, which is part of why it is treated as a near-universal property of skill acquisition rather than a quirk of any one task.

Reference: Newell & Rosenbloom, 1981. The NeuroRank implementation holds the canonical form and scales interference via task-irrelevant stimulus density.

1. The steep early-rank climb a new player experiences in their first weeks, followed by a much slower grind through the same rank tier a year later.
2. Large week-to-week aim-trainer score jumps early on that shrink to fractional gains once a player is already mechanically strong.
3. A newly-released hero or agent's mechanics feeling mastered within days, while the last few percent of optimal play takes months.

A single NeuroRank combine captures one point on a player's learning curve, not the curve itself. The platform's retest credits, which let a claimed profile be re-run and compared over time, are the mechanism designed to eventually trace this shape across a player's own history rather than infer it from a single session.