TL;DR
Each of a kitten's seven stats is an independent roll, and the room's Stimulation tilts every roll toward the higher-statted parent. The skew tops out at 2/3 per stat (66.67%) near Stimulation 100, which makes a fully perfect kitten land about 5.85% of the time, roughly 1 in 17. That ceiling holds for clean pairings. Patch 1.1 weakens Stimulation's benefit the more inbred a kitten is, so keep your breeding pairs at third cousins or further to keep the full skew. The rest of this guide walks the formula and runs three worked examples.
The Seven-Stat System
Every cat carries seven core stats:
- STR scales melee damage at +1 per point
- DEX scales ranged damage at +1 per two points
- CON sets max HP at CON x 4
- INT sets mana regen per turn equal to the stat value
- SPD controls grid movement and turn order
- CHA sets max mana at CHA x 3 and feeds the breeding compatibility check
- LCK sets crit chance at 10% baseline plus 2% per point above or below 5
Bred kittens roll their base values in the 3 to 7 range. Class modifiers stack on top after birth, so a Fighter rolled at STR 7 stays ahead of a Fighter rolled at STR 4 forever. Base rolls are permanent. You get one shot per kitten. For a roundup of which collars amplify which stat rolls, see the Mewgenics Best Starter Classes guide.
How Each Stat Rolls
When two parents breed, the game runs seven independent rolls, one per stat. For each roll it tags the higher-statted parent as "better" and the lower as "worse," then checks the room's Stimulation value and applies this formula:
P(better) = (1 + 0.01 * Stimulation) / (2 + 0.01 * Stimulation)
At Stimulation 0, that resolves to 1/2, a coin flip. At Stimulation 100, it resolves to 2/3, or 66.67%. The function is asymptotic. P(better) approaches but never reaches 100%, no matter how high Stimulation climbs. Even at Stimulation 500 the skew sits around 85.7%.
Each of the seven rolls is independent. STR has no influence on DEX, and DEX has no influence on CON. The same Stimulation value feeds all seven rolls, but the rolls themselves are separate events. That independence produces the 5.85% figure.
The Perfect-Roll Math
A "perfect" kitten inherits the higher of each parent's stats across all seven rolls. Because each roll is independent, you multiply the probabilities:
P(perfect) = P(better)^7
At Stimulation 100, P(better) is 2/3. So:
P(perfect) = (2/3)^7 = 128 / 2187 = 0.0585 = 5.85%
That is the number. About 1 in 17 kittens hits the upper-bound roll on every stat at Stimulation 100. The community-cited "around 6% perfect" figure on mewgenicswiki.org matches this exactly.
Lower Stimulation collapses the rate fast. At Stim 50, P(better) is 60%, so P(perfect) is 0.60^7, about 2.8%. At Stim 0, you are betting on seven coin flips landing heads, which is 0.5^7, about 0.78%.
Stimulation 95 is the practical sweet spot. It guarantees passive ability inheritance and pushes P(better) to 65.5%, so the perfect-roll rate sits at 5.2%, just shy of the Stim 100 ceiling.
Inbreeding and the 5.85% Ceiling
The 5.85% rate assumes a clean pairing. Patch 1.1 changed how Stimulation interacts with related parents: the more inbred a kitten is, the more its inbreeding penalty stacks scale Stimulation's benefit downward. On an inbred line the realized per-stat skew drops below the formula value, so the perfect rate sinks under 5.85%. Breed clean and the full skew stays intact.
Relatedness follows r = 2^(-C), where C counts the generations of separation. Parent-child and full siblings share r = 0.5, half-siblings and grandparent pairs r = 0.25, first cousins r = 0.125, and second cousins r = 0.063. Once a pairing reaches third cousins or further, the game treats the two cats as unrelated, which means you can run a closed pool indefinitely as long as you keep three generations between mates.
Patch 1.1 also flipped how defects pass down. Stimulation now decreases the chance of inheriting a parent's birth defect, where it used to treat inherited defects like positive mutations and pass them more often. The catch: inherited-defect chance rises sharply in inbred kittens. So on a clean line high Stimulation works twice for you, skewing stats up while suppressing inherited defects. On an inbred line both benefits erode together. Pairing across distinct bloodlines is the single move that protects everything good about Stimulation.
Worked Example 1: Similar Parents
Setup:
- Father: STR 6, DEX 6, CON 5, INT 5, SPD 6, CHA 5, LCK 6
- Mother: STR 5, DEX 5, CON 6, INT 6, SPD 5, CHA 6, LCK 5
- Stimulation: 100
The father wins STR, DEX, SPD, and LCK. The mother wins CON, INT, and CHA. Each stat has a 2/3 chance of going to the winner.
The expected stat total per kitten is 0.667 * higher + 0.333 * lower summed across all seven stats. Run that out and you get roughly 39.7, above both parents' totals of 39 and 38 because each axis skews toward the better value.
The probability of the kitten inheriting all four of the father's higher stats AND all three of the mother's higher stats is (2/3)^7 = 5.85%. A near-perfect kitten with six of seven hitting the upper roll lands at about 20.5%.
Evenly matched parents do not produce massive single-roll upside. They produce a steady stream of kittens close to the maximum total, which stabilizes a line without dragging it back.
Worked Example 2: Asymmetric Parents
Setup:
- Father: STR 7, DEX 7, CON 7, INT 7, SPD 7, CHA 7, LCK 7 (a max-stat parent)
- Mother: STR 3, DEX 3, CON 3, INT 3, SPD 3, CHA 3, LCK 3 (a min-stat parent)
- Stimulation: 100
The father is the better parent on every stat. Each roll has a 2/3 chance of inheriting his 7, and a 1/3 chance of inheriting her 3.
The expected value per stat: 0.667 * 7 + 0.333 * 3 = 5.67. Across all seven stats, expected total is about 39.7.
Probability of a perfect kitten (all sevens): 5.85%. Probability of an all-threes kitten, every roll picking the worse parent: (1/3)^7 = 0.046%, or about 1 in 2187.
The asymmetric pairing carries the same perfect-roll rate as the symmetric one. What changes is the floor: a bad roll here gives you a 3 instead of a 5. You get plenty of mediocre kittens with mixed totals.
This is why seasoned breeders pair two strong parents instead of dragging a weak parent up through repeated breedings. The math does not reward hauling weak genes forward.
Worked Example 3: The Perfect Roll
Setup:
- Both parents have all sevens (or you have already bred up to a near-max pair), and they are a clean pairing at third cousins or further
- Stimulation: 100
- Comfort: 50 (boosts breeding attempt frequency)
- Twin probability per litter: 17.36% (average fertility)
You run breedings until a perfect kitten lands. Each kitten has a 5.85% chance of being perfect, since the pair is unrelated and the full Stimulation skew applies. The expected number of kittens to produce one perfect roll is 1 / 0.0585, about 17.
Twins boost throughput. About 1 in 6 litters delivers two kittens, so the expected litters per perfect kitten drops to roughly 14.5.
If your breeding pair breeds once per night for 30 nights, you produce around 35 kittens. The probability of zero perfect rolls across that span is (1 - 0.0585)^35, about 12%. So the probability of at least one perfect kitten is 88%.
That is what a month of disciplined breeding buys you. A room running 30 nights at high Stimulation on a clean pair reliably produces at least one perfect kitten, often two or three.
Practical Implications
Three rules follow from the math:
Run many breedings. A single breeding tops out at a 5.85% chance of perfection. The reliable strategy is volume. Set up a high-Comfort, high-Stimulation room and let the pair breed for as many nights as you can spare.
Retire parents and pair across bloodlines. Once both parents have all sevens, additional breedings offer no upside. Move the perfect kitten into a new pairing with another strong cat from a different line. Crossing bloodlines does double duty under 1.1: it keeps relatedness low, which keeps Stimulation's stat skew at full strength and holds inherited-defect odds down. A dedicated carrier cat built to ferry max genes across pairings is the cleanest way to run this rotation.
Build: a632788df0beStack Comfort, Stimulation, and Fertility. Comfort scales breeding attempts per night by sqrt(1 + 0.1 * Comfort), a 2.45x multiplier at Comfort 50. Stimulation shifts per-roll odds toward the better parent. Fertility, a hidden coefficient between 1.0 and 1.25, sets the twin rate. All three stack.
The math is fixed. Your job is to keep the pair clean, maximize attempts, and let the 5.85% land on its own. For room layouts, lover bonuses, and the full compatibility formula, the Mewgenics Breeding Guide covers the upstream mechanics that feed this math.
Where to Go Next
The Mewgenics Breeding Guide carries the full compatibility formula, lover coefficients, and the birth defect math that pairs with everything above.
Once you have a perfect kitten, the next call is which collar to put on it. The Mewgenics Class Tier List ranks every collar against the current meta so you can match the stat spread to a role.
Bring the parents. The math will handle the rest.