Post by RebeccaMonkey on Sept 28, 2015 3:25:19 GMT
Hopefully I can actually finish this haha but I want to see what the probabilities are for getting each eye color. I have seen stuff like this and would like to see if there does seem to be genes and probabilities involved (or more like what they are). Same with getting spots and stuff on coats.
But hopefully I'll remember to post pics here! If anyone knows how many eye colors there are, that would be helpful I read 104 ish?
I'm not great with the coding end of things and such but I'm decent with the breeding of things. I just had a spree of turquoise eyed alleyz. I've noticed some breeds are more likely to support a colour over another breed. Also most of my petz likely have hexed lineage somewhere along the lines. Dunno how that might factor in to your calculations.
Your friendly neighborhood hacker here. Lately I've been thinking about delving into Petz's genetics and how a puppy/kitten is generated. Right now I can tell you that there are a couple different...things...in Petz's code that appear to deal with genetics. They're called things like Allele, Chromosome, Phenotype... I can also tell you that Petz makes heavy use of randomness and semi-randomness for many things; by semi-randomness I mean there are functions called 'PickFrom2', 'PickFrom3', etc, all the way up to 6. This seems to indicate, in the context of genetics, that the game may randomly select 3 eye colors from all available colors, and then randomly select one of those three. Or something like that. That's just speculation on my part, however. Once I finish writing my P5 conversion tool I'll probably delve further into this.
intronexon.host56.com/ this website is my mecca. Reflet if you make any headway I'd be really, really keen to know!
Ah yes I was thinking of that site last night when I was writing my answer! Couldn't remember what it was though.
I will definitely post a thread with information I find out!
In the meantime, I've compiled screenshots of all the functions in Petz's code that LOOK like they deal with genetics/offspring generation. Pretty interesting stuff. There may be more, or some of these might be irrelevant. Not sure yet. What we can learn, however, is that pets have things called alleles, chromosomes, etc.
Check it out. Note that "arguments" are simply parameters one can set when calling a function.
as someone into stuff like differential cell signalling in a big way, this excites me
Are you into genetics? I'm not so good with this stuff. Haven't studied it since... idk, middle school? 10th grade? (I graduated in 2015)
I'm researching these terms to try and map out exactly what everything is, in order to get a better idea of what a pet's genes actually contain. Definitely welcoming any info on such terms (Genotypes...Phenotypes...Ksfkljfdvnotypes), particularly how they relate to one another. ...As in, how it works irl. Not petz lol.
Looks like PFM did their homework, though. If Petz goes down to the allele level, I'm guessing they structured Petz genetics after irl genetics.
I think that your studies into petz genetics are really fascinating, but as someone who has a bit of experience in both real-world genetics and petz genetics, i'd like to weigh in my two cents, if that's okay!
I have bred, for a good period of time, animals that carry recessive, dominant, co-dominant, and polygenic genetics in the real world. I've selectively bred polygenic animals for a couple of generations and I've made two different selective breeds in Petz that took 8+ hours to make. There are some similarities, but petz genetics are by no means the same as real genetics, and I think it starts and ends in the fact that there are definitely random mutations, and there are also phenotypes and genotypes. I'm going to describe these both in Layman's terms for my sake, and your sake, and everyone else's sake lmao.
Phenotype is, to put very simply, the genetics and traits visibly expressed in a living being, both the ones given at birth (the genes you express thanks to your genetics) and the ones developed by life (by infection, illness, injury, etc). The latter isn't relevant in petz, except I guess you could mention hexed traits such as addball ears chucked on a bulldog, though some hexed traits (eye and coat color) will pass on, so that's a slippery slope. But surely PF. Magic had no idea that hexed traits would ever become a thing, so I'm sure they didn't intend for that half of the definition of phenotype.
Genotype is, again, very simple, the genes you carry. You might carry ten toes and blue eyes, but you've got a perfectly normal amount of toes with brown eyes. That's your phenotype vs your genotype.
I do believe that petz, in a very basic sense, have a phenotype and a genotype. There are obviously genes that they show, and those that they pass on that they recieve from their relatives, and they're given a "genotype" at birth. But I think that the relation to real genetics stops there, because the rest of it appears to be a randomly generated free-for-all, because there are no true recessive, dominant, or co-dominant traits in Petz. There may be slots in the game code where certain genes for certain petz take these roles, but they don't exist outside of randomization. In real genetics, if you know the exact lineage of whatever you're raising, you can come up with a percentage of the probability of creating such outcome.
So let's say I have a fish. It's a normal, wildtype variety fish. Wildtypes of all species (their natural, unaltered coloration) are dominant. If I breed it to a visibly albino fish, albino is recessive (meaning you need two copies of the gene for it to appear), their offspring will have a normal phenotype, but carry albino in their genotype. If you breed two normal fish heterozygous for albino together, there is a 25% chance on every offspring that it will be albino. If you breed a carrier of albino to a visible albino, there is a 50% chance that offspring will be albino, and if you breed two albinos together, 100% of all offspring will be albino, period. Nothing works like this in Petz. And if recessives don't work, then everything can't just..be dominant. And there's no petz traits that express themselves anything like a co-dominant (there's no "super" traits).
My theory on petz genetics is:
There's a set number of slots of genes that a pet can carry. I assume there's a set number of different things that can be determined per trait slot. For example: Eyes: Light green, black, gray Head: Poodle, Chi, Bulldog Body: Bulldog, Poodle, Chi
So on and so forth. It's probably more than 3, likely 6 or so, but the point is made. I imagine there's also traits that are prioritized based on their spot in the lineup, and that traits can have multiple slots (and that's how we're able to selectively breed, and how purebreds work). So for example, a purebred's six traits:
And so forth. Things get more complicated for sure with things like dali spots, alley spots, calico patches, calico coloring, and mutt patches, which seem to act like fake polygenics. When you make a selective breed that breeds true, their slots are likely similar with a few wildcards in the very back, but I don't think there's any exact percentage that can be calculated. Even if we knew exactly where each trait stood in the lineup and how many were really assigned per pet, there's still another huge problem. The second parent.
Because we have no idea how genes from each parent are prioritized, we'd never really know. However, my favorite way to breed petz, be it for PUGS or for selective breeding or for whatever, I take a single male and breed them to the entirety of the rest of what I have. This usually shows me, save 4-5 out of 30 at least one-two of his genes in the offspring, and the other day I bred a dane who I knew for certain did not have white lids anywhere in his most recent 6 generations to 40 females and all but 3 babies had white lids. So, we might be able to assume that the genetics from the male petz hold higher priority. I know that sometimes, if I struggle breeding a pair, I will also gender swap and immediately have much better luck, which is also something to consider.
I also want to touch up on my mention of "fake polygenics". Polygenic traits are traits that do not have anything to do with recessive, dominant, or co-dominant traits. They are minor differences that show up on their own and vary from animal to animal, but through selective breeding, you can strengthen the trait.
Let's say I have a gecko who normally, in it's wildtype form, has a lot of spots, and I have a gecko who oddly carries a handful less spots, enough to be noticable. With lots of years of patience and understanding, I could produce a spotless animal. The same can be done in Petz with the dalmatian. However, in real genetics, I know that in my first generation of my project, i'm going to get some animals with normal spots, and some with less. It's questionable as to whether or not I might get an animal with lesser spots than the original lesser spotted animal, but I will at least get a good partner to go with it. (Disclaimer- I don't actually encourage inbreeding. Line breeding is fine, but direct inbreeding causes a lot of serious problems in reptiles which is something that is overlooked by that community in particular. Even when it stares them right in the face with animals who are incredibly weak or seriously deformed, they will blame it on something other than the inbreeding. It's the quickest way to achieve polygenic goals, but it's not safe. I digress.) Then, by breeding two lesser spotted animals together, you will eventually get some that show less and less spots the farther down you selectively breed.
In Petz genetics, you're not going to have that. Ask anyone who's gone on a spotless dali breeding spree, spent 16+ hours invested and now has carpal tunnel and cramping fingers. You are playing with randoms until you get the desired trait, and if you breed them together, you might end up back at square one. Breeding a lesser spotted dali to a lesser spotted dali might take you over 300 offspring to get one that shows less spots from that pairing, unless you're lucky and whatever's in their "genotype" takes pity on you and tosses them relatively easy. But more than likely, if they do give you one, you'll still have to fight through another 100 offspring to get another one. It's a long, painful process and does not, in the slightest, simulate real breeding.
I've also noticed in Petz that, unlike real genetics, a petz can have genes in their phenotype that they don't even carry. You might have a dog who shows chihuahua ears, but he can't reproduce them on his own no matter what. If he's bred to a dog who easily throws them, he might, but not because of his side, because of the other dog's side. I've gotten petz who will also only throw ears different from their own. I have a lab who has dali ears but throws consistent chi ears and will never throw her own ears.
Petz genetics are complicated and interesting, and I know we'd love to know more about them, but I don't think it'll ever be so cut and dry. I hope you understand my ramblings! I do really enjoy the discussion of genetics and I hope that my side of things can be added to your own hypothesis on how Petz genetics work. I do think that your findings back up my theory, and that perhaps the alleles mentioned were PF Magic's way of wording the different genes/slots that I mentioned. 'Course, you never know with PF Magic. There's code for petz humping each other and all kinds of weird messages left in breed files.
Mika that's super interesting and a really good insight for me! I'm far from truly experienced in Petz breeding, considering I was never a part of the community until 2016.
You saw the link to the imgur album in one of my above posts, right? That's what's in the code (give or take), I just have to figure out what some of those functions do. Yes, Petz definitely makes use of randomness, and everything ultimately comes down to numbers - bits, bytes, etc.
Once I start going through this in my debugger, I'll get a better idea of what's happening and how this stuff is laid out. Should be fun!