Talk:Coefficient of relationship

Step siblings are missing from the diagram
Step siblings are much closer than many of the relatives shown in the diagram, but step siblings are omitted. — Preceding unsigned comment added by 123.3.163.53 (talk) 06:55, 26 May 2015 (UTC)
 * They're not related... — Preceding unsigned comment added by 173.16.124.119 (talk) 22:15, 1 January 2020 (UTC)

Moved from main article
== Sexual organisms == === Parent-offspring === For a sexual organism, the coefficient is 0.5. == Haplodiploidy == In haplodiploidy, things get screwed around a bit. == Philandering == being naughty reduces the coefficient. == Sex chromosomes == Sex chromosomes reduce the relatedness slightly http://www.husdyr.kvl.dk/htm/kc/sexrelat.htm == Finite populations == Finite and small population size mean that it can be negative. hence spite.

It was already commented out, so I put it somewhere more public. Anyone want to fix this? alerante &#x2706; 20:55, 9 September 2005 (UTC)

Wording updates needed
The article addresses the reader directly, for example by posing questions, which WP:TONE says is inappropriate tone for an article. Can someone take a crack at re-writing these sections to use a better tone? --ΨΦorg (talk) 22:27, 6 October 2008 (UTC)

I am a person with a background in biology and biochem, am (obviously) good at math, have a basic understanding of genetics, and am a writer to boot. That being said, I cannot understand what this writer is trying to say. My opinion is that some people are very knowledgeable in their subject area, but make terrible teachers. No matter how hard they try, they simply can't explain something to another person using plain, easily-understood English. Others are very gifted in the art of 'splainin' (Carl Sagan: astronomy; Bill Clinton: politics). This writer belongs to the former group, unfortunately. I was left shaking my head with confusion, due almost exclusively to the wording.

Two examples from the text.

"Each such line connects the two individuals via a common ancestor, passing through no individual who is not a common ancestor more than once."

WT--?

In trying to understand the latter half of this sentence, I jumped through the following logic:
 * If two people share a 'common ancestor,' then that person appears in the family tree of both people.
 * A person who is a 'common ancestor more than once' is someone who appears in the shared family tree in two or more places.
 * A person who is 'NOT a common ancestor more than once' appears in the shared family tree of both people only ONCE.
 * '...passing through NO INDIVIDUAL who is NOT a common ancestor more than once' means passing through NO SHARED ANCESTOR WHO APPEARS ONCE ONCE.

Thus, the line must pass through a common ancestor two or more times. Is this what the writer intended? I don't think so, because first cousins share only a grandparent. Once. This is as clear as mud.

"...where p enumerates all paths connecting B and C with unique common ancestors (i.e. all paths terminate at a common ancestor and may not pass through a common ancestor to a common ancestor's ancestor)..."

First, by definition, a 'common ancestor' is not a person who is unique. So, what is a "unique common ancestor"?

And does the writer simply mean that a path should be traced back to the first ancestor that is common to both people, and that it should STOP THERE? If so, say this.

This wording is just unbelievably poor! There are other small grammar corrections I would probably make ("To given an artificial example"; "fewer generations," not "less generations"; etc.), but I'd start with easy-to-follow wording and a real-life example or two--worked start to finish.

IMHO, this is the kind of article that makes people hate science and math.

Mrs rockefeller (talk) 22:53, 12 April 2015 (UTC)

Not numerology
Someone acting under the cover of a bare IP number categorized the article as "Numerology". I have reverted this as it isn't numerology, but describes a useful quantity that connects to scientific theories. Felsenst (talk) 12:01, 17 July 2010 (UTC)

Coefficients incorporating incestuous ancestry?
Has it been calculated what effect previous incestuous relationships in the parents' ancestries have?

E. g., how would 'r' increase for two first cousins where one is already the result of such relationship and the other has a parent who is also such a result? -- 91.14.188.5 (talk) 22:22, 6 October 2011 (UTC)

2 x the coefficient of inbreeding?
On 2018-12-01, following the suggestions below that the pages should be split, I removed the redirect from coefficient of inbreeding to this page. I do hope someone who understands this better than I can add some content to that page. Emrys2 (talk) 15:21, 1 December 2018 (UTC)

The coefficient of inbreeding of who? One individual? Perhaps you meant the coefficient of kinship of a random copy of the gene in one individual with a random copy of the gene in the other? Felsenst (talk) 12:28, 25 October 2011 (UTC)

The second sentence of the page is


 * The 'Coefficient of Relatedness' (or: coefficient of kinship) is defined as the probability that the alleles at a particular locus chosen at random from two individuals are identical by descent.

This is totally unclear. The coefficient of kinship between individuals A and B is the probability that a random one of the two copies from A is identical by descent to a random one of the two copies chosen from individual B. That is also the inbreeding coefficient of the offspring that would result from mating individual A and individual B.  (Although we can calculate the coefficient of kinship even for two indviduals who cannot be mated, as when they are of the same sex). The coefficient of relationship used in (say) kin selection calculations is different. Felsenst (talk) 15:29, 26 October 2011 (UTC)

Well, the case is much more complicated than this article suggests. These coefficients were defined for use in actual breeding (of animals, I should say, even though they are now also used to examine human genealogies). Their definition is correspondingly complicated. To calculate the coefficient, you would need to know the complete family tree of the two individuals down to all of their common ancestors. This is usually the case in breeding situations, where you start with a given ancestor population and then keep accurate records of each pairing, but it is clearly impossible in most human scenarios. The point is that if the number of generations separating the two individuals from their common ancestors increases, the coefficient approaches zero.

So, the "dumbed down" definition of this coefficient is that you assume that all common ancestors except the ones under explicit consideration are assumed to be arbitrarily far removed. In this case, you get the simple powers-of-two rules of "siblings 2^-1, cousins 2^-3, second cousins 2^-5" etc. But obviously this article should give the full definition, and then treat the simplified "cousins table" as a special case. --dab (𒁳) 15:03, 5 January 2012 (UTC)

ok, I should admit that I am struggling here, as I am reading up on this definition for the first time. Help is appreciated. What is clear is that the article was completely mistaken. The coefficient isn't even defined in terms of genetics, but in terms of genealogy. It can be calculated precisely if the full genealogy is known, never mind genetics. Of course it is intended to still make a statement about genetic relatedness, but that's not part of its strict definition. It is intended to describe breeding processes of mammals, so I am not sure it can even be meaningfully applied to hymenoptera genealogies. I have so far looked up the 1922 definition of r. It involves summing over all paths in the full genealogy. The thing being summed are path coefficients, and these path coefficients are in turn defined in terms of the inbreeding coefficient. As the inbreeding coefficient isn't defined in the 1922 paper, the reader just being referred to the 1921 one, I have not so far been able to supply the full definition. --dab (𒁳) 15:30, 5 January 2012 (UTC)

I have not studied dab's comments in any detail, but I am disheartened to note that he called attention to the gross error in 2012, and it has never been fixed.

COEFFICIENT OF RELATIONSHIP AND COEFFICIENT OF INBREEDING ARE NOT THE SAME THING. The coefficient of inbreeding for an individual may be one-half of the coefficient of relationship between the individual's parents, yet the article misses BOTH of the key points here. And this error persisted for FIVE YEARS after dab called attention to it? I knew Wikipedia had become a pale imitation of its former self, but this is unbelievable.

I'd fix the article myself, but it babbles on and on and on from its totally misconceived origin.Jamesdowallen (talk) 09:20, 13 September 2017 (UTC)

So someone removed the note about the faulty redirection. ::whack:: It is the faulty redirection that needs to be removed. Let me try that with All-Caps: IT IS THE FAULTY REDIRECTION THAT NEEDS TO BE REMOVED. I'd do it myslef, but have better ways to spend my time than learning Wiki macros. Hope this helps.Jamesdowallen (talk) 06:18, 30 October 2017 (UTC)


 * I agree with Dab and James on this one. Coefficient of Relationship and Coefficient of Inbreeding, while related (no pun intended ;-) are not the same things. The Coefficient of Relationship (R) can be stated much more simply than it is here; the equation shown uses the Coefficient of Inbreeding (F) as a required variable, but never adequately defines how it is obtained. Determining the Coefficient of Inbreeding is more complex because, as Dab noted, the chain of meiosis events has to be known not just to an MRCA, but to all (significantly contributing) ancestors who donated to the inheritance chain, even the ones pre-dating the MRCA. I have no clue how to do mathematical markup for Wiki, so I can't even attempt an edit. For someone who might want to dive into it, good sources of information that I've used in the past are: http://www.genetic-genealogy.co.uk/Toc115570144.html and http://www.genetic-genealogy.co.uk/Toc115570135.html. Williams-Texas (talk) 14:15, 7 November 2018 (UTC)


 * See: Help:Math. Hyacinth (talk) 13:09, 21 December 2018 (UTC)

This discussion is out of order. Please place newer comments after earlier comments. Hyacinth (talk) 14:08, 21 December 2018 (UTC)

Title of article
The phrase "coefficient of inbreeding" shows up in Google results over 5 million times, while "coefficient of relationship" shows up only 3.5 million times. Since "coefficient of inbreeding" is by far the more commonly used phrase, shouldn't that be the article's title (with "coefficient of relationship" redirecting to it? 128.148.231.12 (talk) 15:17, 22 October 2014 (UTC)
 * Well I don't think this article has anything to do specifically with inbreeding. Only if the people with high Coefficients of relationship started copulating would if have anything to do with inbreeding.   Perhaps a "See also."  Cloudswrest (talk) 17:07, 22 October 2014 (UTC)

NO. WRONG. Perhaps there should be an article titled coefficient of inbreeding but that article is not this article. It's too bad that these closely related terms ended up with one twice the other, but that's the way it is and I can attest that some Wiki-readers have ended up with faulty definitions by following the redirection.

(That there are two different terms shouldn't surprise. My sister and I have a consanguinal relationship but to call that an inbreeding coefficient would assume  that our sibling relationship is also sexual.  ::whack:: ) Jamesdowallen (talk) 06:25, 30 October 2017 (UTC)

Double first cousins
In what way is marriage between double first cousins a "rare case"? There are cultures where it is quite common, driven by a combination of arranged marriages and a dowry culture where the large dowry one family is expected to provide but cannot is offset by the expectation of an equal and opposite dowry going the other way. Philip Trueman (talk) 17:07, 22 December 2014 (UTC)

Degree of relation
Why is uncle-neice listed as a 3rd degree relationship? This is a 2nd degree relationship ...

128.147.28.65 (talk) 21:48, 13 February 2015 (UTC)


 * I believe it is the minimum number of links on the graph connecting two people, niece -> parent -> grandparent -> uncle.  Three links.  Cloudswrest (talk) 22:31, 13 February 2015 (UTC)

128.147.28.65 is correct. The consanguinity is twice what Cloudswrest implies because the linkage is DOUBLED!

niece -> parent -> grandparent -> uncle is actually

niece -> parent -> grandmother -> uncle PLUS

niece -> parent -> grandfather -> uncle

The article needs a lot of work, but don't expect a lot of help from me. Anal-retentive dolts have taken over Wikipedia and it's too hard to fight them unless you're "in with the in-crowd." IF some in-crowd Wikier deletes the inappropriate Inbreeding redirection, THEN I may try to help.Jamesdowallen (talk) 06:33, 30 October 2017 (UTC)


 * True, it is doubled and that affects the coefficient of relationship (r), and while I didn't create the table I don't believe the definitions of "degree of relationship" and "coefficient of relationship" are the same, else the columns would be redundant. The miniumum number of links to another person makes more sense. By your argument full siblings would also be "1" because:


 * sibling -> father PLUS
 * sibling -> mother
 * Cloudswrest (talk) 16:44, 30 October 2017 (UTC)

You are correct, Cloudswrest. The article needs a little simplification and clarification BUT THERE'S NO POINT IN MY WORKING ON IT UNTIL THE FAULTY REDIRECTION FROM Coefficient of Inbreeding is REMOVED.Jamesdowallen (talk) 01:42, 1 November 2017 (UTC)

BTW, my own pages https://fabpedigree.com/consang.htm and https://fabpedigree.com/wrigco.htm might be good external links for this article. They have detailed explanations and show calculations for complex cases. I won't add them, or I'd be cited for self-citing!Jamesdowallen (talk) 01:45, 1 November 2017 (UTC)

Proposed merge with Kinship coefficient
To my eye these topics seem identical. I would suggest Kinship coefficient is merged into a section of the larger Coefficient of relationship article and expanded to clearly explain why/how the two concepts differ. WaggersTALK  11:08, 1 December 2017 (UTC)

Apparently the "kinship coefficient" is always equal to 50% of the "coefficient of relationship". We do not need a separate article just for the sake of the 50% factor. Agree: merge. Ehrenkater (talk) 17:32, 1 December 2017 (UTC)
 * They do appear to be much the same thing. Unless there is a non-obvious difference, merge and redirect.&middot; &middot; &middot; Peter (Southwood) (talk): 05:40, 10 October 2018 (UTC)

✅. Hyacinth (talk) 02:23, 10 November 2018 (UTC)

weird long winded question
Is it possible for there to be a relative that corresponds with 7/16 (43.75%) shared DNA?

OR Half Siblings (share 1 parent and thus 2 grandparents)
 * 1/16 - Half Cousins (one set of parents are half siblings) (share 1 grandparent)
 * 2/16 (1/8) - First Cousins (one set of parents are siblings) or Double Half Cousins (both sets of parents are half siblings) (either way share 2 grandparents)
 * 3/16 - not really an agreed upon term for it but the article mentions "Sesqui Cousins" (one set of parents are siblings other set are half siblings) (share 3 grandparents)
 * 4/16 (2/8)(1/4) - Double Cousins (both sets of parents are siblings) (share 4 grandparents)
 * 5/16 - Half Siblings + Half Cousins (a person had children with 2 half siblings) (share 1 parent and 3 grandparents)
 * 6/16 (3/8) - Half Siblings + Cousins (aka "3/4 siblings") (a person had children with 2 siblings) (share 1 parent and 4 grandparents)
 * 7/16 - ?
 * 8/16 (4/8)(2/4)(1/2) - Siblings (share 2 parents and thus 4 grandparents)

Why would there be a cousin sized gap (for lack of a better term) between 6/16ths and 8/16ths? — Preceding unsigned comment added by 173.16.124.119 (talk) 20:29, February 4, 2020 (UTC)


 * A shared parent (4/16), great grand parent (2/16), and great great grand parent (1/16) in common (4+2+1/16 = 7/16); four generations. Hyacinth (talk) 05:19, 10 June 2020 (UTC)

To the Original Poster: your question is neither weird nor long winded. You have summarized the ways people can be related by sixteenths very well. And to answer your question, I do not believe you can have 7/16 without inbreeding, that is, people procreating with a blood relative. I can think of at least 2 ways...

First, two individuals have the same father, mothers who are sisters, and those sisters themselves have parents who are siblings. Second, two individuals have the same father, mothers who are sisters, and their father is first cousin to their mothers. Perhaps better to think of livestock than people in these cases.

In fact, I would go so far as to say that without inbreeding you cannot get any fraction between 3/4 siblings and full siblings...not 32nds, 64ths, 128ths, nothing.

As to the answer from Hyacinth, I have no idea what they’re talking about, but whatever it is, it’s wrong. For example, if you and I shared one (but not two, or more) great grandparent, then we would be half-second cousins. Yes, we would each be related to our great grandparent by 1/8 or 2/16, but we would be related to EACH OTHER by 1/64. 74.104.189.176 (talk) 21:58, 19 May 2021 (UTC)

Should be careful to not talk in terms of shared DNA. About 99.9% of the DNA sequence is the same between sequences I have and the corresponding sequence that you have. It is how often variations in the DNA are shared that we are talking about. If I have at some site a rare variant, then my sibling will have it 50% of the time. But that does not mean that my sibling has the same base at a random site only 50% of the time. Felsenst (talk) 03:52, 3 June 2021 (UTC)

Well, that's true enough. But when we talk about sharing genes in a genealogical sense, we mean "by common descent"...that is, getting the same thing from the same ultimate source. Thus genetics is what you have, genealogy is where it came from. Genealogy and kinship is not your DNA profile so much as a record of the parent/child connections you have to other people. 74.104.189.176 (talk) 14:46, 4 June 2021 (UTC)

Table Removal
Instead of getting into an edit war, can you walk me through what your concern is with the table? I moved it from the page into a template, and I'd like to understand what the problem is with the table. Mason (talk) 16:19, 8 July 2023 (UTC)


 * To expand @MrOllie, I'm happy to edit the table, as I agree that it is extremely long. However, I wouldn't go as far as calling it vandalism. Mason (talk) 16:25, 8 July 2023 (UTC)
 * There's a long term vandal that hits pages like this and adds nonsense family relationships, often growing the tables out to the Nth degree. They hit this page here. Here are some examples from other pages:, , . For some reason their vandalized table was moved into a template that is used nowhere but on this page. I don't really think any table is needed at all, but it if it is it should feature only common relations (preferably with sourcing) and it should not reside in a separate template. - MrOllie (talk) 16:32, 8 July 2023 (UTC)
 * Thanks for the help/clarifity. I was the one who moved everything to a template, as at the time my primary goal was to make the page easier to navigate, and to make it toggleable. One advantage to the template is that it's probably harder for vandals to edit.
 * But regardless, I think you have good points about many of the relationships being extremely uncommon. However, I actually think that a slimmed down version with some more distant relatives is still useful. I know that it was helpful when I was working on my latest talk on modeling distant cousins. (I'll go dig up that talk to get a better sense of which ones I used...)
 * Would you be ok with keeping it on the page, once I tidy it up? Mason (talk) 17:30, 8 July 2023 (UTC)