Random Pokemon GeneratorAsk anyone who has lost a long Nuzlocke how the run ended and you will hear about a hole. Four members weak to Ice. Three that fall to Rock. The run does not usually die to one bad fight — it dies to a weakness nobody counted until it got counted twice. Encounters are the other half of that story, because what the game hands you decides how many holes you can patch. This guide measures both halves, using simulated runs over the same 1,025-species dataset that powers the tools on this site.
What Encounter Math Actually Measures
The method first, so the rest of the numbers can be trusted. These simulations draw Pokemon uniformly from the full dataset — 1,025 species plus their alternate forms — and track one thing: what each draw does to a team's type coverage. Coverage here means exactly one thing, how many of the 18 types your team's STABs can hit for double damage or better. It says nothing about stats, movesets, or level curves, and it is not a route-tuned encounter table, which turns out to matter enormously.
One limitation deserves to be stated up front: uniform draws over the whole dex are far kinder than any real route. Real games deal from small local tables, and that smallness — not bad luck — creates most of the duplicate pain players actually feel. So read the first set of numbers as the best case, what National-Dex-level variety would buy you, and the duplicate section as the correction that reality applies.
The Coverage Curve: What Each Catch Adds
Start with the ceiling: no single Pokemon covers more than 9 of the 18 types by itself. Fighting/Ice (Crabominable) and Ice/Ground (the Mamoswine line) tie for the widest offensive reach in the entire dex, and everything below them falls off quickly. A lone random catch averages about 4 covered types — enough to matter, not enough to lean on. The interesting part starts when catches stack.
| Catches | Types covered (average) | Added by the latest catch |
|---|---|---|
| 1 | 4.1 | — |
| 2 | 7.0 | +2.9 |
| 3 | 9.3 | +2.3 |
| 4 | 11.0 | +1.7 |
| 6 | 13.4 | +1.1 |
| 9 | 15.4 | +0.6 |
| 12 | 16.4 | +0.3 |
The curve bends in a way that rewards front-loading: your first three catches drag coverage past half the type chart (9.3 of 18), and by catch six the squad sits around 13-14 — the point where most accidental holes have names. After that, every new member buys less than one new type. Catch twelve adds about a third of a type on average, and the last two blank spots are stubborn in a way worth understanding.
Which spots stay stubborn says something about how the type chart was built. Simulated teams of twelve still lack Electric coverage 41% of the time and Normal coverage 35% of the time, because hitting Electric for double damage requires a Ground attack and hitting Normal requires Fighting — and reliable Ground and Fighting STABs are not exactly everywhere. If your run has been rolling for ten routes and those two holes are still open, that is not your luck failing. That is the chart being the chart.
Full coverage also stays rare — rarer than most players assume. Only about one simulated twelve in five covers all 18 types (18%), and it takes twenty catches before full coverage becomes a coin flip (53%). Chasing the missing two types with encounters is a losing campaign after the early game. The curve says the war is won or lost in the first six catches.
A slower way to read the same table: coverage arrives lumpy. Four of the eighteen types come in with the average first catch, and the last two can take a dozen more catches to show up. The classic early-game spine of Fire, Water and Grass exists because it buys the steep part of this curve on purpose — three broad attackers, half the chart, before the first gym.
Dupe Clause Economics
The dupes clause — you may skip an encounter that is a species you have already caught — is the most argued-about house rule in the format, and the math explains why both sides feel right. At National Dex scale, duplicates are basically impossible: after six catches, a uniformly random seventh encounter repeats a species only 0.44% of the time. One in two hundred. If routes worked like that, nobody would ever need the clause.
But routes do not deal from the National Dex. They deal from small local tables — call a mid-game route ten to twelve options, older games often fewer. From a twelve-option route, the chance that your first three catches include a duplicate is already 24%, and by your sixth catch it is 78%. From an eight-option route (think of a small early-game cave), it is 34% by the third and 92% by the sixth. The clause exists because route tables are tiny, not because the dice are cruel.
Two practical reads from that. First, write your dupes rule down before Route 1, not in the moment it first bites, because with those numbers it will bite. Second, the clause is worth the most exactly where it feels the most annoying: a small table is a pool that can genuinely run dry, and replacing a repeat with a fresh option is a pacing fix, not a difficulty setting.
The clause has a second-order benefit too: it protects the curve. A repeat that eats an encounter slot adds nothing new by definition — same species, same types — so skipping it converts a dead slot into a fresh roll, which is exactly what the curve wants in its steep early stretch.
The Replacement Roll
When a teammate falls, the replacement roll feels like a fresh start. The math says temper that feeling. Simulate twenty thousand squads — take five random members, add a random sixth, then measure what actually changed. On average the newcomer adds 1.03 new covered types. About one. And in 42.5% of cases it adds none at all: two replacements in five change nothing about the team's offensive shape.
The defensive side is blunter still. In 78.5% of simulations, the new arrival is weak to a type that at least two of the existing five already share a weakness to. Random replacements do not patch holes; they tend to join the crowd around the nearest one. The distribution of what they add says the same thing: 27% of replacements contribute exactly one new type, 19% add two, and the 10.6% that add three or more are the ones worth building around when they land.
None of this argues against taking the roll. It argues for aiming it. Decide which hole you need filled before you go catch the replacement, and if the roll misses, the box is one deeper and the run continues. To feel out combinations before committing a run to them, the Nuzlocke Team Builder deals squads and prints the same coverage report this math is built on.
Five Rules the Math Supports
Distilled to decisions you make mid-run:
- Front-load your coverage. The first three catches do more for your chart than the next nine combined. Treat early routes as the foundation of the team, not its filler.
- Stop chasing eighteen. Full coverage is a coin flip at twenty catches. The twelfth catch is usually better spent on bulk, utility, or a type you are actually afraid of.
- Write the dupes clause down early. On a typical mid-game route, six catches without a repeat is a one-in-five event. Plan for the repeat before it arrives.
- Aim replacements at holes. Two in five random replacements add zero coverage, and four in five join an existing shared weakness. Pick for the gap or expect to shrug.
- Count shared weaknesses like a resource. An average six shares a double-weakness pattern on 7.5 of the 18 attacking types — every new member either narrows that number or feeds it.