Why your osu! rank drops when you have not played
An osu! global rank is a position in a sorted list, not a number stored on your account. The osu! servers order every ranked player by total performance points, and a rank is simply where a player falls in that order. osu! applies no pp decay: a score set years ago keeps its full pp value and its place in the top 100 indefinitely. What changes during a break is everyone else. Each player who climbs past a pp total pushes the players below down exactly one place, so a rank can slide for weeks while every number on the profile itself stays identical.
Rank is a position, not a possession
Nothing on the osu! servers stores the sentence "this player is rank 42,000". The servers store a pp total for every player, sort those totals, and report the position that falls out of the sort. Global rank is that position across all ranked players; country rank is the same sort filtered to one country.
The consequence is that a rank has two inputs, and only one of them belongs to the player. Your pp is yours. The list you are sorted into is shared with everyone else who plays the game. A queue is the closest everyday comparison: standing perfectly still does not move you backwards, but it does nothing to stop the people behind you walking forward past you. Rank movement with no play activity is not a penalty, a bug, or a stat reset. It is the arithmetic of being sorted against a moving population.
osu! has no pp decay
The weighting that produces a profile pp total contains no time term at all. The osu! wiki gives it as total pp = p × 0.95(n−1), where p is a score's full pp value and n is its placing in the Best Performance list. The exponent is the score's rank within your top 100, not its age. A play that sits third in your list is weighted at 90.25% whether you set it last night or in 2019.
The wiki's own worked example: top five scores of 110pp, 100pp, 100pp, 90pp and 80pp contribute roughly 110, 95, 90, 77 and 65 pp. Nothing in that calculation asks when the plays happened. The only way a top play loses weight is by being pushed further down the list by better plays — which requires you to have played, not to have stopped.
Bonus pp does not decay either
Alongside the weighted top 100, osu! awards a bonus based on how many ranked beatmaps you have set a score on. The wiki gives the formula as 416.6667 × (1 − 0.995min(N, 1000)), where N is the number of ranked maps with a score set, reaching a maximum of 413.894 pp at 1,000 scores. Roughly 137 scores gets you about half of that maximum, so the bonus climbs quickly early and crawls afterwards.
Two things follow. First, this component is also time-free: maps you have already scored on stay counted, so the bonus never shrinks during a break. Second, it is capped at a few hundred pp, which means it cannot rescue a rank either. Players sometimes assume a falling rank means the bonus has drained away. It has not — the play count and map coverage behind it are permanent.
What actually moves: the players around you
If your pp is frozen, every place you lose corresponds to exactly one player whose pp crossed yours. That population is large and permanently in motion: newer accounts climbing through the ranks, returning players rebuilding a top 100, and people mid-way through a good month.
The arithmetic is blunt. If 250 players overtake your total during a month away, you finish the month 250 places lower with an identical profile. The same effect is why players who are playing sometimes still lose ground: gaining 40pp in a week is a real improvement, but if the players immediately above you gained more, your position still slips. Rank measures relative progress, so it answers a different question from pp. Pp asks "how good are your best plays"; rank asks "how good are they compared with everyone playing right now". A break only silences the first half of that comparison.
How much drift to expect depends on how crowded your neighbourhood is
There is no universal drift rate, and any tool quoting one is guessing. Drift is simply the number of players who cross your pp total in a given period, which depends entirely on how tightly packed players are around your total. You can measure your own neighbourhood in about a minute.
Open the osu! performance rankings, find the pp total at your rank, then find the total 100 places above you. Divide the difference by 100 to get the pp cost of one place near you. If the gap is 30pp, one place costs about 0.3pp, and every player sitting within 30pp below you is a single strong top play away from passing you — expect a fast, continuous slide. If the gap is 300pp, one place costs about 3pp, individual plays rarely reorder anything, and the drift is slow. The same break produces very different rank losses at those two densities.
pp reworks change your total without you touching the game
A second, less obvious cause of movement is recalculation. The osu! wiki is explicit that the pp algorithm is occasionally updated and existing scores are recalculated to match, which can surface as a sudden increase or decrease in a total. These are announced by the Performance Points Committee rather than applied quietly.
The updates of 16 January 2021 and 28 October 2024 are the documented recent examples. The 2024 pass removed combo scaling in osu!, adjusted rhythm complexity and fast-flow aim, changed how sliders contribute to accuracy pp, altered stamina and accuracy scaling in osu!taiko, changed the No Fail penalty in osu!catch, and rebalanced long-note scaling in osu!mania. Because every player is recalculated in the same pass, a rework can move a rank in either direction overnight. A rank jump with no plays and no pp loss on your side is often a rework redistributing pp elsewhere in the list.
Long inactivity removes you from the listings entirely
Gradual drift has a separate, sharper cousin. osu! flags accounts that have gone a long time without a ranked score as inactive and drops them out of the ranking listings, which is why an extended break can end with no rank shown rather than a worse one. The commonly reported threshold in community and issue-tracker discussion is around three months without a ranked score; osu! does not publish this as a formal policy, so treat the exact figure as approximate rather than guaranteed.
Two points matter more than the exact number. Being delisted is not the same as losing pp — the pp total and top 100 are untouched, and a single ranked score restores the listing. And because inactive players leave the list, everyone below them moves up by one; batches of delistings can therefore improve a rank for players who never stopped playing. Older third-party tools reading the legacy API may show a delisted player as having no pp at all.
Measuring your own drift with session history
Because passive drift is personal to your pp neighbourhood, the only accurate figure is your own. TrackMyOsu records global rank and country rank in the snapshot it takes when you press Start and again when you press Stop, and keeps every session permanently.
That gives you a clean measurement nobody has to calculate by hand. Compare the rank recorded at the end of one session with the rank recorded at the start of the next. You did not play between those two snapshots, so the entire difference is other players moving. Divide by the days between them and you have your drift rate in places per day. Do it across several gaps and the number stabilises — which tells you how long a break your current rank can absorb, and how much of a session's pp gain is going into recovering ground rather than making it.
What to do with the information
The only lever that acts on drift is playing, but the useful conclusion is narrower than "play more". If your measured drift is a few places a day, a fortnight off costs very little and is not worth planning around. If it is a hundred places a day, your rank is a high-maintenance number and treating it as your main progress metric will be miserable.
For long breaks, the delisting threshold is the thing worth protecting: one ranked score every few weeks keeps a profile listed even if the rank itself slides. And for tracking progress across a gap, prefer metrics that only move when you move them. New top plays, profile pp and accuracy record what you did; rank records what you did relative to everyone who did not take the break.
By carrot12_. Last updated 2026-08-12.