Start with a single roll
Every random event in a game comes down to a number between 0 and 1, or equivalently between 0 percent and 100 percent. A 25 percent critical hit chance means that, on any one attack, the event happens with probability 0.25. A 1 percent drop means 0.01. A fair six-sided die gives each face a probability of 1/6, which is about 0.167.
That number describes one trial. It does not describe what will happen over the next ten, and it does not remember what happened on the last one. In most games each roll is independent: the game does not owe you a success because you have had several failures in a row. The main exception is a deliberately designed system that tracks your history, and those are covered in the discussions of pity timers and pseudo random distribution.
Combining two events
Two rules handle most of what a player needs.
“And” for independent events: multiply. If an attack has a 50 percent chance to hit and a hit has a 20 percent chance to crit, the chance of a critical hit on any given attack is 0.5 times 0.2, which is 0.10, or 10 percent.
“Or” for mutually exclusive events: add. If a die roll of 1 triggers one effect and a roll of 2 triggers another, the chance that one of them happens is 1/6 plus 1/6, which is 1/3. Adding only works when the two outcomes cannot happen together.
For events that can happen together, adding gives the wrong answer. Two separate 30 percent chances do not add up to a 60 percent chance of at least one. The next section shows the safe way to handle that.
The chance of at least one success
The most useful trick for random events is the complement. Instead of counting every way to succeed, work out the chance of never succeeding and subtract it from 1.
If the chance of success on one try is p, the chance of failing every one of n tries is (1 minus p) raised to the power n. So the chance of at least one success is 1 minus that.
Some examples:
- A 20 percent crit chance across five attacks: the chance of no crits is 0.8 to the fifth power, which is 0.32768. The chance of at least one crit is about 67 percent. Roughly one sequence of five attacks in three has no crit at all.
- A 25 percent chance across three attempts: 0.75 cubed is 0.421875, so the chance of at least one success is about 57.8 percent.
- A 1 percent drop across 100 attempts: 0.99 to the 100th power is about 0.366, so the chance of at least one drop is about 63.4 percent. About 37 percent of players doing exactly 100 attempts would still have nothing.
That last result catches many people out. A 1 percent chance does not become certain at 100 tries. The average number of attempts is 100, which is what the idea of expected attempts describes, but the typical spread around that average is wide.
How many tries to reach a given confidence
You can reverse the calculation. To find how many attempts give a 50 percent chance of at least one success at 1 percent per try, solve 0.99 to the power n equals 0.5. The answer is about 69 attempts. For a 95 percent chance, the answer is about 299 attempts, so roughly 300.
Two points follow. The median number of attempts (69) is lower than the mean (100), because a minority of very unlucky runs pull the average up. And going from “likely” to “nearly certain” costs a lot: reaching 95 percent takes more than four times as many attempts as reaching 50 percent.
Reading published odds correctly
When a game states a rate, check what it applies to. A few common ambiguities:
- Per attempt or per item? Some systems roll once per kill, others once per drop table entry. A rate of 1 percent per roll is different from 1 percent per hour of play.
- Independent or adjusted? Many games use plain independent rolls. Others adjust the odds based on recent results, which makes streaks less extreme than the plain formulas suggest.
- Stated odds versus effective odds. A 5 percent chance that depends on a prior 50 percent condition is a 2.5 percent chance overall. Hidden conditions are easy to miss.
Where a game lists several items together, understanding how drop tables share their total probability helps, and a published drop rates list is most useful when you know whether it applies per roll or per session.
Common misreadings
The gambler’s fallacy. After five failures at 20 percent, the next try is still 20 percent. The five failures are part of the past, not a debt. Players farming for items often fall for this, which is discussed in the context of item farming.
Confusing odds with frequencies in small samples. If a 10 percent event happens twice in ten tries, that is a perfectly ordinary result, even though the observed rate is 20 percent. Small samples bounce around the true probability considerably.
Treating a rare event as impossible. An event with a 2 percent chance will occur for one attempt in fifty on average. In a game played by many thousands of people, it will happen to someone every day.
A short checklist
When you meet a new random mechanic:
- Write the chance for a single trial as a decimal.
- Check whether trials are independent or whether the game adjusts odds based on history.
- Decide whether you need the chance of exactly one result, at least one, or all of them.
- For “at least one”, compute 1 minus (1 minus p) to the power n.
- Sanity-check against the average, 1 divided by p, remembering the typical spread is large.
These steps do not predict a particular roll, and nothing can. They tell you what range of results to treat as normal, which is usually what matters when deciding whether an outcome was unlucky or whether something about the system is worth investigating.