Levels organize engagement into a visible ladder of tiers, each one dangled just ahead to sustain effort. The core engine is the goal-gradient effect — first shown in Hull’s (1932) animal studies and demonstrated in human reward programs by Kivetz, Urminsky and Zheng (2006), who found people accelerate their effort as they near a goal. A progression bar kept perpetually in view exploits this: the closer the next tier appears, the harder people push, and the moment one tier is reached a new one resets the distance to run.
Level systems can legitimately mark skill development or genuine mastery. They become manipulative when the tiers are an endless grind whose cost rises with each level while the real payoff stays flat or nonexistent — advancement that mostly benefits the platform’s engagement metrics. Tells include escalating time or spend per tier, levels that decay unless you keep feeding them (retention disguised as progression), and “almost there” nudges timed to pull you back. Points (T18.16) typically fuel the bar, and badges (T18.14) mark the tiers, making levels a load-bearing part of the broader gamification layer.
Because the damage is grind and lost time rather than acute harm, levels sit in the Moderate band. The defense is to price the next tier honestly: if advancing yields nothing you value outside the app, the pull you feel is the goal-gradient effect operating, not a real reason to continue. Setting your own outcome-based milestones and letting decaying ranks lapse returns the choice of what counts as progress to you.