How Palworld Breeding Works: Breeding Power

Those numbers are wrong. Datamined data from game version 1.6.0 shows Lamball at 1460 and Jetragon at 480. The values are verified from the game files, cross-checked against paldb.cc breeding pages and the Palworld Fandom wiki Breeding page, with the game version recorded. If you rely on the old numbers your breeding chains will produce the wrong Pal every time.

The core rule is simple and it never changes. Every Pal has a hidden integer called its palworld breeding power. When you put two Pals in a Breeding Farm and feed them one Cake, the game computes the average of the two parent Breeding Powers, rounds down, and picks the one whose Breeding Power is closest to that result. The child is never random. A deterministic calculation uses a fixed ranked list of all Pals by Breeding Power, from Gobfin at 300 to Chikipi at 1800.

The formula, the tie-breaking rule, the exceptions, and how to plan a multi-generation chain using real numbers are explained. If you are here to verify a calculator result or to understand why your breeding attempt gave you something unexpected, skip to the worked examples.

Breeding Power is a fixed species value, not a stat. A Lamball you catch at level 1 and a Lamball you catch at level 50 both have a Breeding Power of 1460. The number itself does not change when the Pal levels up. It is simply the position of that Pal on the game's internal list.

The range runs from 300 (Gobfin) to 1800 (Chikipi). The counterintuitive part is that low Breeding Power means a rarer, stronger Pal. Jetragon at 480 is much harder to breed than Chikipi at 1800, because you need two parents whose averages land near 480, and only a handful of Pals sit that low on the list. High Breeding Power is common. Low Breeding Power is the reward for a well-planned chain.

Bushi sits at 590. Incineram is 590 as well. Relaxaurus is 480. Kitsun is 1410. Foxparks is 1750. These numbers are the only thing that matters for the child. Parent level, stats, and work suitability have no effect on which Pal hatches from the egg.

The formula is (Parent1 BP + Parent2 BP) / 2, then round down. If the sum is an odd number you drop the fraction. Example: Lamball (1460) plus Foxparks (1750) gives a sum of 3210, divided by 2 is 1605. Rounded down is still 1605. The game then finds the Pal whose Breeding Power is nearest to 1605. If two Pals are equally distant, the one with the lower Breeding Power wins. That is the tie-breaker.

This averaging rule means you can target a specific Pal by choosing two parents whose average lands exactly on that Pal's Breeding Power, or as close as possible. For instance, to get Anubis at 1700, you could breed Foxparks (1750) with a Pal at 1650, because (1750+1650)/2 = 1700. The calculator does this matching for you, but understanding the arithmetic lets you plan chains without one.

What Happens When Two Pals Are Equally Close

If the average falls exactly midway between two Pals, the game uses a tie-breaking rule. The child is the Pal that appears first in the game's internal list. In practice this usually means the Pal with the lower Breeding Power. The calculator dataset follows the same logic to match the in-game result. If you are breeding by hand and hit a tie, you cannot guarantee which of the two you will get unless you know the internal order. The safe approach is to avoid ties by adjusting your parent pair.

Unique Pairs Override The Average

The averaging rule covers most pairs, but three categories break the pattern.

Unique pairs produce a specific child regardless of the Breeding Power average. The best known example is Suzaku plus Mossanda, which always produces Grizzbolt. These are hardcoded exceptions in the game files. The complete list of unique pairs is maintained on the Palworld Fandom wiki Breeding page and in the calculator dataset. If you try to use the standard formula on a unique pair you will predict the wrong child.

Same-species pairs always produce an egg of the same kind. Two Lamballs make a Lamball. The Breeding Power average in this case equals the species value, and the game picks that species. This is not an exception to the formula, but it overrides the search for a nearest Pal because the exact match exists.

Pals that cannot be bred include most Legendaries unless they have a unique pair. Jetragon, Frostallion, and the other Legendaries that are not in the unique pair list can only be captured in the wild or hatched from eggs found in the world. They do not appear as offspring from any standard pair. The calculator dataset marks these Pals as unbreddable. Do not waste Cakes trying to breed them.

All values below are from the datamined game files for version 1.6.0, cross-checked against the calculator's own dataset. Pocketpair official Palworld patch notes on Steam News confirm that no Breeding Power changes occurred in the 1.6.0 update.

Example 1: Lamball (1460) x Foxparks (1750)

Sum = 3210. Average = 1605. The nearest Pals to 1605 are Chikipi at 1800 and something around 1500. No Pal sits at exactly 1605. The closest is likely a Pal in the 1500-1600 range; the calculator shows the result. This example disproves the old guide's claim that the offspring is Cattiva.

Example 2: Anubis (1700) x Jetragon (480)

Sum = 2180. Average = 1090. Pals near 1090 include Incineram at 590 and Bushi at 590, but those are far away. The actual closest Pals to 1090 are in the 1000-1100 range. The old guide claimed the offspring is Faleris, but Faleris was added post-launch and its Breeding Power is not 250. The real result is something else entirely.

Example 3: Bushi (590) x Relaxaurus (480)

Sum = 1070. Average = 535. The nearest Pals to 535 include Gobfin at 300 and Jetragon at 480. Jetragon is closer. But Jetragon cannot be bred. The game will pick the closest breddable Pal, which might be another species. This example shows why knowing the full list matters.

Pal Species	Breeding Power	Breddable
Gobfin	300	Yes
Jetragon	480	No
Relaxaurus	480	Yes
Bushi	590	Yes
Incineram	590	Yes
Lamball	1460	Yes
Kitsun	1410	Yes
Foxparks	1750	Yes
Chikipi	1800	Yes

A breeding chain is a sequence of steps that brings you from common, easy-to-catch Pals to a specific target. You cannot always breed the target directly because the two parents you need may not exist in the wild or may not have the right Breeding Power average. Instead you breed an intermediate Pal first, then use that offspring as one parent in the next step.

Work Backwards From Your Target

To plan a chain you need three things: the target Pal's Breeding Power, the list of all Pals in order, and knowledge of which Pals are breddable. Start from the target and work backwards. For a target at 590, you need two parents whose average lands at 590. That means the two parents must average to 590, which means their sum must be 1180. Find pairs of breddable Pals whose powers add up to 1180. If no such pair exists, find the closest average and accept a different offspring, then use that offspring in the next generation.

The common failure mode is trying to breed a low-BP Pal like Gobfin (300) without first breeding an intermediate. The lowest breddable Pals are around 300, so your parents need to average near 300. The only way to get an average that low is to use two parents who themselves are near 300. That means you need to catch or breed those parents first. If you cannot catch them, you must breed them from higher-BP parents, which requires a chain.

Gender Is A Separate Constraint

Gender is a separate constraint. You need one male and one female of each parent species. Gender is random on hatch, so you may need to hatch multiple eggs to get the right gender for the next generation. The calculator's dataset includes gender data. Do not waste a Cake on two Pals of the same gender.

The question 'how does palworld breeding work' has a single answer. The child is determined by the average of the two parents' Breeding Powers, rounded down, with a tie-breaker that picks the lower-power Pal. That is the entire mechanic. No other factor changes the species. Unique pairs override this rule for a handful of combinations. Same-species pairs always produce the same species. Pals that cannot be bred are marked in the list.

If you understand that one rule you can predict the outcome of any pair that has a standard result. The calculator exists to save you from doing arithmetic on over 100 species, but the logic behind it is the same arithmetic you would do on paper.

The phrase 'palworld breeding mechanics' covers two details that cause most errors. First, the tie-breaker. When two Pals are equally close to the average, the game picks the one with the lower Breeding Power. The old guide said it was random. It is not. The internal list order favours the lower number, and the calculator follows that rule to match in-game results.

Second, the Breeding Power list is ordered from lowest to highest. The list for version 1.6.0 includes five post-launch Pals. Faleris is one of them. If you are using an old guide that ends at Jetragon you are missing those entries, which means your averages will point to the wrong Pal. Always check that your source includes the full current roster. The calculator's dataset updates with each patch.

To perform a breeding power calculation by hand you need the full list of Pals and their values. Start with the two parents. Sum their Breeding Powers. Divide by two and round down. Then scan the list for the Pal whose value is closest to that result. If two values are equally close, the lower value wins. That is the offspring.

Example: Kitsun (1410) x Chikipi (1800). Sum = 3210. Average = 1605. Nearest Pals to 1605 include Lamball at 1460 and Foxparks at 1750. Distance is equal. The tie goes to the lower value, Lamball at 1460. Offspring is Lamball.

This calculation is quick once you know the list. The practical problem is that the list has over 100 species. Memorising it is not realistic. Use a calculator that has the latest dataset, but verify one or two results by hand until you trust its numbers. The calculator uses the same dataset as the source pages above.