Choose the curve that turns XP into a long-term progression system.

Linear growth adds requirement at a steady rate and is easy to explain. Quadratic growth increases the cost more aggressively as levels rise; Atom includes its original Quadratic V1 model and a Quadratic V2 model compatible with Lurkr-style progression. Exponential growth compounds through a base and can make high levels exceptionally distant.

No model is universally correct. A short seasonal server may need a flatter curve, while a long-lived community may want late levels to remain rare. Choose based on expected earning volume and the lifetime of the progression system, not on which formula sounds most sophisticated.

The selected model determines which coefficient, constant, base, or exponent controls appear. The dashboard validates those values and provides a level input for quick reference. Use it repeatedly: check an onboarding level, a meaningful middle reward, and the highest level a committed member might reach.

A single attractive example is not enough. Two curves can be nearly identical at level 10 and diverge dramatically by level 60. Record several projected requirements and translate them into expected days or weeks using the server’s real earning rate and cooldown.

Experience decides how eligible messages, voice, and reactions award XP. Cooldown constrains how often some awards can occur. Multipliers alter context, and Formula turns the resulting total into level thresholds. Changing one layer can make the others feel unexpectedly generous or severe.

Place meaningful rewards at points the chosen curve makes attainable. If a desired role sits behind months of ordinary participation, that may be intentional prestige—or it may be a mismatch between the formula and the reward ladder. Projection lets administrators distinguish the two before members invest the time.

Members already have XP, levels, expectations, and earned roles. Before changing the formula, determine how Atom will recalculate future progress and whether displayed levels can move. Communicate the reason, effective time, and treatment of existing rewards before the save occurs.

After saving, inspect several representative members, the leaderboard, rank cards, and milestone timing. If the result is wrong, use the recorded previous parameters to restore the model rather than guessing. A formula deserves versioned operational notes even when the dashboard itself is simple.