Morph odds without the hand-waving

Pairing decisions get made on odds, and the odds get quoted wrong constantly. The categories matter because each one produces a completely different clutch distribution.

Recessive traits

Two copies are needed for the animal to look different. Visual × visual gives all visuals. Visual × het gives half visual, half het. Het × het gives one in four visual — and, importantly, two thirds of the normal-looking siblings are hets, which is where '66% het' comes from.

'100% het' means proven by parentage. '50% het' is a probability, not a description of the animal. Sell it labelled honestly, and note that only breeding or genetic testing converts a possible het into a known one.

Co-dominant and incomplete dominant traits

One copy shows, two copies show differently — the 'super' form. A single-gene animal bred to a normal gives half single-gene, half normal. Two single-gene animals give one super, two single-gene, one normal.

This is where the super form matters most: in some lines the super is a lethal or severely compromised animal, and the pairing is a decision about welfare, not just about odds. Know what the super of your gene does before you set the pairing.

Dominant traits

One copy shows and a second copy adds nothing visible, so a dominant animal bred to a normal produces roughly half showing the trait. Because the super is indistinguishable, you cannot tell a homozygous animal from a heterozygous one by looking.

Polygenic and line-bred traits

Structure, pattern quality, white coverage in many crested gecko lines, colour saturation — these are not single genes. They respond to selection over generations, not to a Punnett square, and they are heavily influenced by husbandry and age.

A common and costly mistake is treating a line-bred appearance as if it were a simple dominant gene. Buying a 'high white' animal does not buy you a gene; it buys you a place in someone's selection programme, which you then have to continue.

Where the maths goes wrong

• •Averaging odds across a clutch. Every egg is an independent roll — a four-egg clutch at 25% odds very often produces zero hits.
• •Stacking traits without multiplying. Two independent 1-in-4 outcomes together is 1 in 16, not 1 in 8.
• •Ignoring known lethal or deleterious super combinations.
• •Compounding relatedness. Odds say nothing about inbreeding; track a coefficient of inbreeding separately and set a ceiling you will not cross.