Calculator · Results · The three traits · Can’t be predicted · References

Goldfish cross calculator

Pick Parent A and Parent B for each trait, and this shows what share of the fry you should expect for each genotype and each look. I only put in traits that breeding studies and the genetics both agree are simple: scale type, telescope eye, and albino. Most of what makes a good show fish can’t be calculated, and I explain those further down.

These are expected ratios, not a forecast for your spawn. They’re the odds for each egg. A real spawn can land well off them, especially when only a few dozen fry make it. Some fry die before you can tell them apart, and once you start culling, the fish left in the tub won’t match the ratios anymore. Every number also assumes you’ve got the parents’ genes right.

Calculator

Tick the traits you want. For each parent, choose what it looks like, or a genotype you know or want to assume. If you’re not sure, pick a “don’t know” option and you’ll get a range.

Results

The three traits

Scale type: metallic, nacreous, matte

One gene (written T) does this. A metallic fish has two copies of the normal form (tt), a matte fish has two copies of the transparent form (TT), and a nacreous fish has one of each (Tt). Because the in-between fish looks different from both, picking the look is the same as picking the genotype.

Nacreous × nacreous: about 25% metallic, 50% nacreous, and 25% matte.
Nacreous × metallic: about half nacreous and half metallic.
Matte × metallic: all nacreous.
Matte × matte: all matte. Metallic × metallic: all metallic.

Things that can throw this off:

Nacreous and matte fry are hard to tell apart by eye, especially under about ¾ inch. If you can’t tell what a parent is, use “nacreous or matte” and you’ll get a range.
Selection can blur the classes. Some lines have been bred so matte fish keep a lot of color and look calico, and a nacreous fish can be bred to look almost matte.
When calico fish are crossed with wild crucian carp, the in-between fish don’t always look in-between.
There are other, rarer genes that also give clear or patchy-looking scales. The calculator only covers this common one.
“Nacreous” only describes the scales. Whether a nacreous fish turns into a good calico is a different question, and it’s one the calculator can’t answer (see calico color quality).

Telescope eye

Telescope eye is recessive: a fish needs two copies (dd) to show it. A normal-eyed fish can carry one copy (Dd) and never show it. A normal-eyed fish with a telescope parent is a carrier, and so is one that has thrown telescopes. The gene behind it has been found.

Telescope × telescope: all telescope.
Carrier × carrier: about 25% telescope.
Telescope × carrier: about 50% telescope.
Telescope × non-carrier: no telescopes, but every fry is a carrier.

Things that can throw this off:

The eyes come out late. Some fry show them at about two months, and others not until they’re a year old. Count after grow-out, or you’ll undercount.
These numbers are for crosses within fancy goldfish. In crosses with wild-type goldfish, far fewer telescopes came out than one gene predicts, so something more than one gene is going on there.
Celestial eyes and bubble eyes aren’t covered. See Can’t be predicted.

Albino

A true albino has pink or red eyes and no black pigment anywhere on it. It takes two genes. I write them p and c, and a fish has to have two recessive copies of both (pp cc) to be albino. An albino crossed to a normal fish only throws albinos if the normal fish carries at least one copy of both genes.

Albino × albino: all albino.
Albino × a fish carrying one copy of both genes (Pp Cc): about 25% albino.
Two fish that each carry one copy of both genes (Pp Cc × Pp Cc): about 1 in 16 albino (6.25%).
Albino × a fish with no albino genes (PP CC): no albinos.

Things that can throw this off:

Count after the fry color up. Some fry look albino in the eye at first and then darken as skin pigment comes in. They aren’t albinos.
A white or pale fish with black eyes isn’t an albino. That’s usually the normal loss of baby color, which the calculator doesn’t cover.
Knowing a normal-looking fish carries albino doesn’t tell you which of the two genes it carries. If its parent was albino, use that option and you’ll get a range.

Can’t be predicted

Goldfish come from a fish whose whole set of genes was doubled long ago, so many genes come in two copies that can cover for each other. On top of that, most of what breeders select for is built from many small genes, plus how the fish was raised. You can’t turn those into percentages, and I won’t make up numbers for them. For all of these, the only way I know to improve a line is to pick the best parents, look at what they throw, and keep going.

Trait	Why it can’t be calculated
Twin tail vs. single tail	A gene for twin tail has been found, but the trait doesn’t always show: twin-tail parents throw single-tails, and breeders report more single-tails when the eggs develop cold. Researchers don’t agree on a simple model, so there’s no honest ratio.
Tail shape and quality	Spread, split, curl, and how the tail is held all come from many genes. Even the anal fins and the tail can be at different stages in the same fish.
Fin length	Fins run from short to long with no clean classes in between, and long fins keep growing for years. One study tied long fins to a gene, but fry don’t sort into neat groups.
Dorsal fin (dorsal-less fish)	There are two models, and I’m not picking one. Single recessive gene: a 2020 genome study of 27 goldfish strains tied dorsal loss to one broken gene (lrp6S). If that’s the whole story, two dorsal-less parents give all dorsal-less fry, and two carriers give about 1 in 4. Many genes: the standard breeder’s genetics handbook treats dorsal loss as suppression spread across several genes, not one. That fits what breeders actually see: dorsal-less lines still throw spikes, partial dorsals, and saw-backs. Nobody has published test crosses that settle it, so the calculator doesn’t put a ratio on dorsal-less.
Head growth (wen, hood)	No single gene for it has been found. It looks like many genes, and diet and environment change how much grows.
Body shape and depth	Short, deep, round bodies vary smoothly from fish to fish. That points to many genes, not one.
Celestial eye and bubble eye	The step from telescope to celestial hasn’t been worked out, and bubble size varies so much that it looks like many genes.
Narial bouquets (pompoms)	Size and texture vary smoothly and look like many genes.
Pearlscale	Often said to be recessive, but I couldn’t find a breeding study that shows it, so it isn’t in the calculator.
Color intensity, and when black fades	How deep the red or orange gets moves with diet and even tank color. The change from dark baby color to red, white, or yellow is controlled by duplicated genes, and its timing depends on the rest of the fish’s genes and on temperature. Some fish change in months, some take years, and some never do.
Calico color quality	The calculator tells you how many fry should have nacreous scales, the scale type calico color sits on. It can’t tell you how many will be good calicos. The blue base, how much black and red there is, where it lands, and whether it holds as the fish ages all come from many genes, and they change with age and with how the fish is kept. A calico × calico spawn full of poorly colored fish doesn’t mean the ratios are wrong. Good calico color comes from picking the best-colored parents, generation after generation.
Blue, brown, lavender, chocolate, and other colors	An old study names one gene for blue vs. brown, but I couldn’t find a modern recheck, and “blue” on a lot of fish is just sparse black. I couldn’t find worked-out genetics for lavender and similar colors.