What a Pity Timer Is
A pity timer is a counter that tracks how many attempts have passed since a player last received a particular reward. When the counter reaches a set value, the game either guarantees the reward or raises its chance. The idea is to put a ceiling on bad luck. Without it, a pure chance system leaves some players with a waiting time far longer than the average, while a counter makes the worst case a known number.
The same concept goes by other names: bad luck protection, guaranteed drops, mercy counters or dry streak protection. The details differ by game, so the examples below use made-up numbers to illustrate the principle.
Hard Pity: A Fixed Guarantee
Suppose a reward has a 1% chance per attempt and a hard guarantee on the 90th attempt if nothing has dropped before. Without the guarantee, the average number of attempts is 1 / 0.01 = 100, and some players would wait several hundred.
With the guarantee, nobody waits more than 90 attempts. The average falls, because the long tail is cut off. One way to compute it is to add up, for each attempt number from 1 to 90, the probability that you are still trying at that point. The chance of still being in the game at attempt k + 1 is the chance of k straight failures, 0.99^k. Summing 0.99^k for k from 0 to 89 gives (1 - 0.99^90) / 0.01. Since 0.99^90 is about 0.405, the average is (1 - 0.405) / 0.01, about 59.5 attempts.
So the guarantee lowers the average from 100 to roughly 60 under these assumptions. The chance that you reach the guarantee, meaning 89 failures in a row, is about 0.99^89, which is roughly 41%. In other words, a large share of attempts end by the guarantee, not by the base chance.
Notice what the guarantee did not do. The base chance is still 1% on every attempt before the cutoff. It is the cap, not the underlying chance, that changed the average.
Soft Pity: A Rising Chance
A soft system raises the chance gradually instead of jumping to 100%. A simple illustration: start at 10% and add 10 percentage points after each failure. The probabilities of success on successive attempts are 10%, 20%, 30%, and so on, reaching 100% on the tenth.
The probability of still trying after each attempt is the product of the failure chances so far: 0.9, then 0.9 x 0.8 = 0.72, then 0.72 x 0.7 = 0.504, then 0.504 x 0.6 = 0.3024, and so on. Adding these “still trying” probabilities, starting with 1 for the first attempt, gives an average of about 3.66 attempts. A flat 10% chance would average 10. The rising chance cuts the average to roughly a third and also removes the long tail entirely, since success is certain by attempt 10.
Real systems usually raise the chance more gently and later, so they bring smaller reductions, but the logic is the same: the more the chance grows after failures, the more the average falls and the more the outcomes bunch around that average.
What Pity Systems Do and Do Not Do
They reduce the worst case. Variation shrinks because very long waits become impossible or very rare.
They lower the average cost, but only relative to the same base chance. If a designer adds a guarantee and leaves the base rate unchanged, the average falls. Designers often account for this by lowering the base rate, in which case the average may end up near where it started. The comparison to make is the full expected value of the cost with the system in place, not the stated base chance.
They make attempts no longer independent. A pure random system has no memory. A pity system does, because the result of one attempt changes the next attempt’s chance. Some of the intuition players bring from independent events no longer applies, and in a pity system, “I am due” is partly true, which is different from the gamblers fallacy.
They can reset in surprising ways. Counters may reset on any success, on a success of a certain tier, when the banner or event ends, or never. Whether the counter carries over affects the real cost, so check the rules for the specific game.
They can encourage overspending. If a player is close to a guarantee, stopping can feel like wasting the progress already made. That is sunk cost thinking: the earlier attempts are gone either way, and the decision about the next one depends on whether the reward is worth its remaining cost.
Pity Systems Outside Purchases
Bad luck protection also appears in ordinary play. An action game may increase the chance of a rare drop after each kill that does not produce it. A roguelike may adjust offers so a shop or healing item appears after a stretch without one. These work like the soft example above and are closely related to the idea behind pseudo-random distribution. They are usually less contentious because no money is involved, though they still change the average and should be read like any other drop rate.
Reading a Pity System Before You Engage
When a game describes a counter, a short checklist helps.
- What is the base chance, and does it stay fixed until the guarantee?
- Is the guarantee hard or soft, and at what count?
- Does the counter persist between sessions, events or banners?
- Does the guaranteed result cover the exact item you want, or only a tier of items?
- What is the cost per attempt, and what is the average and worst-case cost?
Even with the best protection, outcomes remain uncertain, and the worst case is real money or time in any system that charges for attempts. A guarantee is a tool for understanding limits, not a reason to spend more than you intended. If you find yourself chasing a counter, step back and decide whether the reward is worth the remaining cost.