Learn
RNG Explained: Diablo Drops, Pokemon Pulls, and Gacha Rates
How RNG actually works in games: Diablo loot tables, card pulls, and gacha pity systems explained with real math instead of superstition.
What RNG actually is
RNG is not magic. In games, it usually means the engine is sampling from a probability table. Every possible outcome has a weight, and the game rolls against those weights when a monster dies, a card pack opens, or a summon animation starts. That matters because players often talk about streaks as if the system has intent, when in reality the system is usually just doing repeated weighted draws.
A good mental model is a bag filled with different numbers of colored tokens. If one color has more tokens, it shows up more often over time. That does not mean outcomes are evenly spaced, and it does not mean short sessions feel fair. Random systems create clumps, dry spells, and sudden bursts. Those patterns feel emotional because humans are great at noticing streaks and terrible at accepting that streaks can happen naturally.
Why Diablo drops feel rigged
Loot-heavy games like Diablo stack multiple tables on top of each other. First the game can decide whether an item drops at all, then what slot it belongs to, then what rarity it has, then whether affixes or stat rolls land in the range you wanted. Even if the headline legendary rate sounds reasonable, the exact version you want may be buried behind several independent rolls.
That is why two players can farm the same boss for an hour and come away with completely different stories. One gets a useful legendary in ten minutes. The other gets crafting trash for fifty-five minutes and assumes the system is punishing them. Usually it is just compounded probability. When several low-probability events need to line up, the bad runs get long fast.
Pokemon packs and MTG draws are independent
Card games create the same illusion. If you open a bad Pokemon pack, that does not make the next one better. If you draw lands three turns in a row in Magic, the deck does not owe you gas on turn four. Each draw is its own event, conditioned only on the actual remaining deck state or pack collation rules, not on your frustration level.
Players often remember droughts more vividly than normal variance. A dead hand or a pack with no chase hit feels personal. Mathematically, though, the system has no memory beyond the rules of the container you are drawing from. If the process is independent, past misses do not change future hit rates. If the process draws without replacement, the changing odds come from the remaining pool, not from destiny trying to even things out.
How pity systems change the feel
Gacha games add soft pity and hard pity to make bad variance survivable. Soft pity means the rate starts climbing after a threshold. Hard pity means you are guaranteed something by a fixed pull count. Those mechanics do change the math, but they are still product design features, not fairness gifts. They are there because a totally memoryless banner can feel brutal enough to push players out of the game.
The important point is that pity systems only affect odds according to their published rules. Pull number 12 usually has the base rate. Pull number 78 might have an elevated rate. If you are not inside pity, the previous misses did not make the next pull special. If you are inside pity, the improvement comes from a rule trigger, not from the universe deciding you have suffered enough.
The useful takeaway
Every pull, drop, or draw should be understood in terms of probability tables and rule triggers. The faster you stop reading intention into streaks, the better you get at evaluating whether a system is generous, stingy, or simply noisy. That mindset is useful whether you are building a game economy, opening packs, or deciding whether to keep grinding for one more run.
If you want to reason about outcomes instead of vibes, simulate the odds, count the paths, and work from the actual rules. Random systems do not need to feel fair in the short run to be mathematically fair over many trials.