Palworld’s breeding farm looks like a slot machine for the first few hours you use it. You put two Pals in, wait for an egg, and something arrives that you did not particularly plan for.

It is not random at all. The result of any pairing is fixed before the egg appears, and it can be worked out in advance with a little arithmetic.

That single realisation changes how the second half of the game plays, because breeding stops being a way to pass the time and becomes a way to get exactly what you need.

Every Pal Carries a Hidden Number

Each species has an internal breeding value that never appears anywhere in the interface. It has nothing to do with level, rarity or how strong the Pal is in a fight.

When two Pals are paired, the game takes the average of their two values and produces whichever species sits closest to that average. Swap either parent and the answer moves accordingly.

Because the calculation is deterministic, the same two parents will always produce the same child. Repeating a pairing does not give you another roll of the dice.

Gender is the only thing that needs to line up beyond that, since a pairing requires one of each. Two Pals of the same species will simply produce more of themselves, which is useful for other reasons entirely.

The Lookup Part

Working the averages out by hand is possible and thoroughly unpleasant, which is why breeding calculators became standard equipment. paldb.gg is one of the more complete Palworld references, built by extracting values directly from the game files rather than by transcribing them from videos.

Its breeding calculator works in both directions. You can pick two parents and see what they produce, or name a target Pal and get back every pairing capable of producing it, which is the version most people actually want.

The wider database sits behind that, covering 288 Pals, 1,880 items, 579 technologies, 330 skills and 43,993 map nodes with exact positions. There is an interactive map carrying roughly 8,000 spawn points and 407 effigies, a tier list computed from real stats rather than opinion, plus capture rate, crafting and base planning tools.

Everything is refreshed automatically when the game updates, which matters more in Palworld than in most games of its type. It is an unofficial project with no connection to Pocketpair, and it does not claim otherwise.

Special Pairs Override the Maths

There is a second layer sitting on top of the averaging rule. A list of fixed combinations produces specific results regardless of what the arithmetic would otherwise say.

These are the pairings that make certain Pals obtainable at all, and they are the reason a calculator beats a spreadsheet. You cannot derive them, you can only look them up.

Most of the genuinely useful breeding routes involve at least one of these steps. Working purely from averages will send you down a much longer path than necessary.

The same applies in reverse when you are trying to avoid a result. A pairing that looks sensible on paper can drop straight into a special combination and hand you something entirely unrelated.

Passives Are the Real Reason to Breed

Producing a particular species is usually the easy half. What most players are actually chasing is a Pal with a specific set of passive skills attached.

Passives inherit from the pool of both parents, with a limited number of slots on the child. That means a breeding project is normally several generations long, gradually concentrating the traits you want while discarding the ones you do not.

The practical approach is to fix the species first and then run repeated pairings for the passives. Trying to solve both problems in a single step wastes an enormous amount of cake.

The Numbers Move Between Patches

None of this is stable across versions, which is where returning players get caught out. Pocketpair has been willing to rebalance systems well after launch, and breeding values and passive effects have both been touched.

The Steam update notes are the primary record of what changed and when, and they are worth a skim before restarting an old save. A guide written two updates ago may be describing routes that no longer exist.

Tools rebuilt from the game files handle that automatically, which is the honest argument for using one. A written guide cannot update itself, and a great many of them never have.

Where to Start

Pick the Pal you want rather than the Pal you have. Working backwards from a target is far quicker than experimenting forwards from whatever happens to be in the box.

Then check whether a special pair exists for it before doing any averaging at all. That single check removes most of the guesswork.

After that it is a question of stocking the farm with cake and being patient. Eggs take real time to hatch, and the incubator is usually the bottleneck rather than the breeding itself.

For anyone who follows the game through its development updates rather than by playing it constantly, breeding is the system worth relearning after a long gap. It has changed more than the surface of the game suggests.