A paradox about paternal surnames--the 300 years rule--why are there so many surnames?
raylopez99
Statistically, a male descendant cannot survive from a common ancestor
for more than about 12 generations, or roughly 300 years, because
after 12 generations or so there's a good chance no more male
offspring will result (you can do computer simulations to verify this,
as I have done--assume 50% probability of a male heir and 0-5 kids per
generation, per male heir; you'll find after about 12 iterrations the
male offspring will die out).
So how can a name like "Smith" or "Lopez" survive for more than this
time? I surmise there were many people with surnames "Smith" or
"Lopez", genetically 'unrelated' (insofar as possible), so that's why
there's so many of these surnames today. Another possibility is that
surnames are relatively recent, as recent as the middle ages, which
was about 500 to 1000 years ago, and lots of surnames existed then,
and have since gone extinct, so we still have a lot of unique surnames
but they will eventually dwindle down to a handful of names (not
unlike what is present today in Korea--everybody is named "Kim"), as
time progresses, assuming patermonial descent name convention of
course (unlike some tribes in Africa, which use matrimonial surnames,
though that also has the same problem, long term).
In short, if you do the simulations long enough, eventually you should
get (I would imagine, I haven't done this simulation yet), one surname
("Adam"?) for the whole population.
Any thoughts on this paradox? Surely it must have been discussed
before.
RL
the_ver...@comcast.net
Just why should we assume 50% probablity of a male heir and 0-5 kids
per generation?
Once we nail that down we can move on to other parts of your theory;
but I really believe those assumptions to not be representative of
reality.
Charani
> Just why should we assume 50% probablity of a male heir and 0-5 kids
> per generation?
> Once we nail that down we can move on to other parts of your theory;
> but I really believe those assumptions to not be representative of
> reality.
I agree. If there were 0 children in a generation, then the
probability of a male heir drops to - zero and the family line dies
out immediately.
In one family I'm researching there was a male born 1928, the next
male in his direct line wasn't born until 1987. It was all girls in
between with no boys at all.
My great grandfather was the youngest of a family of 11, eight of his
siblings were girls. His father (my great great grandfather) had
three brothers who had 43 children between the four of them and well
over half of them were girls. His grandfather (my 3x great
grandfather) was an only child.
As for the name SMITH: since that derives from the occupation, not
only of a blacksmith, but also of whitesmiths, silversmiths,
goldsmiths, etc, there's bound to be numerous unrelated families with
the same surname.
Keith nuttle
simplify the model for discussion assume four children to a family the
first generation there would be four, two boys, the second 8, four boys,
the third there would 16, eight boys, and so on. Though I have not set
it up in a spread sheet as you, it seems that the only way the name
would die out is in the first couple of generations.
As you say in a direct line from four children to four children to four
children the branch may end
--
Keith Nuttle
3110 Marquette Court
Indianapolis, IN 46268
317-802-0699
Donna
given the surnames of their mothers---in this way surnames are not being
"daughtered out," as I think the model supposes.
Regarding the ubiquitous "Kim" in Korea: In the U.S. we need not worry
about a dwindling of given names----new ones pop up with great regularity as
parents make up names out of (seemingly) random syllables.
In all my family lines up until the 1920s, given names were selected for
traditional reasons----parents named their children for their own parents,
their own brothers and sisters, etc. It really helps me sort who's who and
who probably belongs to whom as I begin to untangle the twisted yarns of
same-surnamed families. For example, one line has carried the name "Mason"
for generations. It was a grand discovery as I went back and back as
discovered that the use of the name in our family is derived from a maternal
line before the Revolution. And now, trying to decipher this extended
family, I do have this Mason to continue as a clue.
It is wonderful to see solid maternal family surnames show up as given names
in the descendants. Again, this helps decipher maternal lines that are so
often forgotten. This using of surnames as given names has been prevalent
in my family lines.
Now, all too often a poor baby is named after a soap opera starlet rather
than her dear 3rd gr grandmother Catherine Patience. So we get the asinine
rather than the solid-----I include my own name of Donna Dell as assine. I
would have made a lovely Catherine Patience---as my gr grandmother was, and
her Catherine and Patience names were derived from the names of her own two
grandmothers.
All too often our maternal lines get swallowed up in the black hole of the
patrilineal surname. I began documenting my family lines for the very
reason that I didn't want my family to get lost from the memory and
awareness of my children. And I have found some of our richest history in
the maternal lines way behind me.
However, I must control my genealogy talk with my own adult children. I see
their eyes glaze over after only a moment or two. The passion is not
shared----but oh, to get into a wonderful dialog with a new cousin found on
the genealogy internet circuit! Then the talks begin. . . .
As yet another aside----and to make genealogy documentation even more a maze
than before .......... I recently read that to avoid confusion in legal
documents, an American 18th century court clerk might make arbitrary
additions to a name. For example, if there were two Mason Colvins in the
county, the clerk might record the elder as "Sr." and the younger "Jr." even
though they were not father and son. It was just a way of keeping track.
Or he might tuck a middle name in ---- John Smith who lived on Waterford
Plantation might thus be called John Waterford Smith to avoid confusion with
the five other John Smiths in the county. And the local grocer might use
yet a different improvized distinction in his account ledgers--and the
doctor yet another, all for the same person.
But I digress, as usual . . . . . .
Warmest Regards,
Donn
cecilia
>[...] I recently read that to avoid confusion in legal
> documents, an American 18th century court clerk might make arbitrary
Surnames arose in order to distinguish people with the same given
name. One would expect something similar to happen if a community
became one-surnamed, particularly if they we not all obviously
related.
I know of a case where two brothers in a "Smith" family adopted
different surnames - one becoming Carrington.
.
granitegriz
Adams that came to the colonies in the first place. Most likely due to Lord
Mayor Thomas Adams obtain land grants for all his relatives etc. Although my
particular lineage does run the risk of vanishing because I have only one
male Adams left for grand child. This runs the since the proven records of
1620.
singhals
> Here's a paradox that's a real puzzler.
>
> Statistically, a male descendant cannot survive from a common ancestor
ER?? Define your terms, please. ANY male descendant or
merely a same-surnamed male descendant? If the first, this
is absurd and we need go no further.
> for more than about 12 generations, or roughly 300 years, because
> after 12 generations or so there's a good chance no more male
Ummmm, I have on my very own home computer a family
genealogy beginning with a marriage in 1690 which is into
the 13th generation by now with males galore.
> offspring will result (you can do computer simulations to verify this,
> as I have done--assume 50% probability of a male heir and 0-5 kids per
> generation, per male heir; you'll find after about 12 iterrations the
> male offspring will die out).
Actually, if memory serves you'd be better off with a 51%
probability of a male from any given pregnancy. More, a
male "heir" need not be offspring of the deceased, a fact
which muddies up your analysis.
[snip]
> In short, if you do the simulations long enough, eventually you should
> get (I would imagine, I haven't done this simulation yet), one surname
> ("Adam"?) for the whole population.
You might want to run that sim before you interpret the results.
Cheryl
wgwhalley
>
> Any thoughts on this paradox? Surely it must have been discussed
> before.
>
> RL
There is quite a literature on this interesting topic. I just found
this web page concerning name frequencies http://homepages.newnet.co.uk/dance/webpjd/index.htm
Check the "Variance" page for some ideas about how new surnames arose.
My reading is that any given unique surname has a pretty good chance
of dying out, but that new ones arise and that a surname shared by
several people can persist for quite a while. A surname could "get
lucky" early in its history by
W. Whalley
raylopez99
> On Jul 28, 3:22 pm, raylopez99 <raylope...@yahoo.com> wrote:
>
>
>
> > Any thoughts on this paradox? Surely it must have been discussed
> > before.
>
> > RL
>
> There is quite a literature on this interesting topic. I just found
>
> Check the "Variance" page for some ideas about how new surnames arose.
>
> My reading is that any given unique surname has a pretty good chance
> of dying out, but that new ones arise and that a surname shared by
> several people can persist for quite a while. A surname could "get
> lucky" early in its history by
>
> W. Whalley
Thanks W. Whalley. I checked out this website and it looks
promising. I agree that a surname shared can persist; also, since
posting my original post, perhaps "SMITH" and "LOPEZ" are the
surnames, like "KIM", that are "taking over the world" (if you follow
my logic).
I am in a process of writing a more sophisticated version of my
program to prove that surnames die out very quickly (well within 12
generations). And, reading a book on the medieval ages, I find this
phenomena of patrimonial descent dying out is commented on (regarding
the French Capetian kings dying out in 1328 for example).
For the rest of you posters (sorry I cannot address each of you
individually, and I thank you for replying), here is my answer, in no
particular order: use your common sense, I'm not suggesting all males
will die out; you can do the simulation yourself, on paper, as
follows:
1/ go to http://www.random.org/integers/ fill in numbers between 1
and X, where X is any number of maximum children you think an average
family can have (I picked five, but feel free to pick a larger number,
keep in mind we're talking about children who end up reproducing, not
dying in infancy)
1.5/ On a piece of paper, do a tree (I trust you know what a tree
is, this being a genealogy forum), and pick a number for your first
level (branches), say the first generation (the first set of parents)
has 6 kids.
2/ go to http://www.random.org/coins/
3/ fill in, for step 2/, from the drop down box enter the same number
as in 1.5/, or, in this example, "6".
4/ "Flip virtual coin(s)" from step 3/. Heads means a boy, tails
means a girl, and eliminate all branches that are girls.
5/ keep doing steps 1/ through 4/ for each generation, for each child
in the generation, picking a different number, consecutively, from 1/,
as the tree expands (I trust you can follow what I'm saying here, use
your smarts people).
You will find that after 10 or 12 generations the boys die out, and
therefore the surname dies out. Even (surprisingly) if one guy has 4
sons, as one of my nodes in the tree did--eventually they peter out.
In fact, if you pick "0" kids (meaning some children capable of
reproduction will not reproduce) your surname will die out faster
(that is, I've been alternating between having 0-5 kids and 1-5 kids,
and with "0" your tree dies much faster.
I will try, once I perfect my program, upload an executable or source
file you can compile (I'm using C#/CLI Visual Studio 2005).
Thanks!
And keep in mind your name, unlike the Egyptians, will die out in
about 300 years and will be known no more. BAHAHAHAHAHA!!!
But the good news is that some lucky family line (by chance) will end
up, as time progresses, being the majority surname for the vast
majority of people. Remember, I'm not saying the human race will die
out; to the contrary, population will increase with each passing
generation.
RL (<--Lopez, a good name, like Smith and Wang, to bet on!)
raylopez99
few years ago, it was speculated the reason the Chinese "look the
same" unlike most Europeans (aside from racial insensitivities--BTW,
if you visit China like I have, the Chinese will comment that
Westerners all look the same to them, LOL), is that the Chinese
traditionally practice making sure husband and wife are widely
seperated by consanguity, meaning making sure they're far apart in
terms of being relatives (i.e., to put in more crudely, there's less
incest in China than in Europe). Consequently, you don't get
phenotypes that look radically different when you do this (don't ask
me why), so there's more "blending" if you will, and the people start
"looking the same". For this reason perhaps the "Han" tribe in China
is so pervasive. Now if you take this argument further, along the
lines of this thread, you can suppose that perhaps this is why the
"WANG" name is so popular in China--this 'lucky' surname is quickly
"infecting" or "taking over" all the other male surnames, as time
progresses.
That is, WANG might eventually become one of the only surnames in
China, as well as the rest of the world, unless the chauvinist habit
of adopting only the male surname is dropped.
RL (<--Lopez, a good name to bet on, like Genghis KHAN's!)
http://news.bbc.co.uk/2/hi/asia-pacific/6745259.stm
China re-think on names shortage
Officials say the prevalence of some names causes confusion
Chinese officials are considering measures to expand the number of
surnames in the country in order to prevent confusion, state media
says.
At the moment around 85% of China's 1.3bn residents share around 100
surnames, a survey in April by the Ministry of Public Security found.
The most popular name, Wang, is shared by some 93 million people.
Now the ministry wants to give parents the option of combining both
surnames for their children, China Daily said.
"If a father's family name is Zhou, and the mother, Zhu, the baby
could have four options for the surname: Zhou, Zhu, Zhouzhu or
Zhuzhou," the newspaper said.
This could create around 1.28m new surnames, said Guan Xihua, a
household registration officer with the Beijing public security
bureau.
The prevalence of some surnames caused problems in daily life and more
of them would reduce repetition, she said.
In addition to Wang, some 92 million people are called Li, while
another 88 million are called Zhang, the survey in April found.
More than 100,000 people share China's most popular name, Wang Tao,
another report by the Chinese Academy of Social Sciences found.
Charani
> I am in a process of writing a more sophisticated version of my
> program to prove that surnames die out very quickly (well within 12
> generations).
If "surnames die out very quickly (well within 12 generations)" how
come one surname that I'm researching goes back to 1380 and hasn't
died out at - and that's well over 12 generations. The spelling has
changed but that doesn't mean the name has died out, far from it.
There was no fixed spelling of surnames (or given names come to that)
until the late 19th century at the earliest when literacy levels rose.
I'm afraid your theory and program are flawed.
--
http://home.comcast.net/~webact1/Collingridge/
OPC for Walton, SOM, England
Bruce Remick
"Charani" <SGBNOSPAM@ mail2genes.invalid> wrote in message
news:46adab0c$0$97252$892e...@authen.yellow.readfreenews.net...
With my own genealogy research, I find that my surname "dies out" in the
PAST, with little danger of it disappearing in the present or future. I can
trace twelve generations of my surname in America back to 1650, but I
haven't been able to locate the family or origin of the immigrant. His
hundreds of living male descendants today attest to the shakiness of the
"disappearing surname" theory.
Bruce
Phyllis Nilsson
before then), that is over 300 years of my maiden name not dying out. In
fact, the police chief of Kansas City that caught the "BTK" (?) killer is
bearing our same last name. If I google my maiden name they are still all
over the country.
Another thing to remember is that marrying first cousins was not only legal
in our ancestors' time, but is still legal today. A male cousin with the
family name could marry a female cousin of another name, and their children
would continue the surname (and it is not incest). I've also found, at
least in my family, where the daughter and son of two brothers married which
meant they both had the same last name.
I agree with you that the theory is flawed.
"Charani" <SGBNOSPAM@ mail2genes.invalid> wrote in message
news:46adab0c$0$97252$892e...@authen.yellow.readfreenews.net...
raylopez99
> If "surnames die out very quickly (well within 12 generations)" how
> come one surname that I'm researching goes back to 1380 and hasn't
> died out at - and that's well over 12 generations. The spelling has
> changed but that doesn't mean the name has died out, far from it.
> There was no fixed spelling of surnames (or given names come to that)
> until the late 19th century at the earliest when literacy levels rose.
>
Because my friend there are lies, damn lies, and statistics. The
AVERAGE might be 12 generations (actually I am to confirm this, when I
perfect my program--my preliminary calculations were based on a
hypothetical family having, each generation, zero to four children,
which I don't know if that's historically accurate).
Average does not mean your surname. Your surname could be defying the
averages.
But I doubt your surname will last, say, 30 generations, unless your
ancestors get very lucky with XY chromosome pairings.
Sorry, but your name will die out like the rest of us Charani. Well,
most of us. As a LOPEZ I have a good chance to join WANG, KHAN and
SMITH in the panoply of immortal names.
RL
raylopez99
> With my own genealogy research, I find that my surname "dies out" in the
> PAST, with little danger of it disappearing in the present or future. I can
> trace twelve generations of my surname in America back to 1650, but I
> haven't been able to locate the family or origin of the immigrant. His
> hundreds of living male descendants today attest to the shakiness of the
> "disappearing surname" theory.
>
> Bruce
Well Bruce, you are confusing the size of the tree and the paucity of
finding documents from the past with statistical facts.
Computers don't lie. But, if you don't believe me, read this thread
and do the hand calculations I suggest at random dot org.
Then get back to us with what you find.
RL
raylopez99
express.com> wrote:
> While I've only gone back so far as 1688 in this country (can't verify
> before then), that is over 300 years of my maiden name not dying out. In
> fact, the police chief of Kansas City that caught the "BTK" (?) killer is
> bearing our same last name. If I google my maiden name they are still all
> over the country.
Phyllis your forte is clearly not statistics. 300 years is the
average, and you're saying that you've not surpassed the average.
There's a contradiction here. Get it? Probably not.
>
> Another thing to remember is that marrying first cousins was not only legal
> in our ancestors' time, but is still legal today. A male cousin with the
> family name could marry a female cousin of another name, and their children
> would continue the surname (and it is not incest). I've also found, at
> least in my family, where the daughter and son of two brothers married which
> meant they both had the same last name.
That introduces an interesting phenomena--but I don't think it will
affect the calculation any--we're still talking male surnames.
>
> I agree with you that the theory is flawed.
Your agreement belies the facts. Do the calculations, by hand, as I
outlined in this thread, and get back to us with your findings. It
should take no more than half an hour, and you might just learn
something. Your "gut feeling" is about as useful as those philosopher
poets who said a computer would never beat a human in chess, and if
man was mean to fly or go to the moon--he'd have wings. How many
angels can dance on the head of a pin Phyllis?
RL <--a good name, LOPEZ, that's likely to withstand the test of time
(on average)
Dr. Brian Leverich
> Because my friend there are lies, damn lies, and statistics. The
> AVERAGE might be 12 generations (actually I am to confirm this, when I
> perfect my program--my preliminary calculations were based on a
> hypothetical family having, each generation, zero to four children,
> which I don't know if that's historically accurate).
Hi Ray -
This is old stochastic processes stuff, specifically discrete
branching theory. You can find it discussed, for example, in
"The Theory of Stochastic Processes" by Hilton David Miller
and David Roxbee Cox [1977]. You can buy it from Amazon or
look at the relevant bits on Google.
I think Miller and Cox's bottom line was for American lines
using some old census data, the probability of extinction of
the male lines of a single individual is 0.86. If you have
N males with the same surname as "initial conditions", the
probability of extinction is 0.86 ** N (assuming independent
breeding), which means the probability of extinction is pretty
low for most surnames in existence today (if birth patterns
continued as they were).
Cheers, B. (I am an applied mathematican, among other things.)
--
Dr. Brian Leverich Co-moderator, soc.genealogy.methods/GENMTD-L
Angeles Chapter LTC Admin Chair http://angeles.sierraclub.org/ltc/
P.O. Box 6831, Frazier Park, CA 93222-6831 leve...@mtpinos.com
Donna
their own surnames is being ignored in this discussion. I hope my post did
get through.
In some U.S. cultures this is common, and becoming more so. It decreases
the likelihood of a surname being "daughtered out."
Is this being considered in the OP's statistical model and in other models
mentioned?
Donna
James A. Doemer
donnamcr_...@hotmail.com, typed this:
> I think my earlier comment about unmarried mothers giving their
> children their own surnames is being ignored in this discussion. I
> hope my post did get through.
Additionally, the vast numbers of immigrants whose surnames where anywhere
from mildly to radically changed, either voluntarilly or unvoluntarily by
the immigration check-in authority is probably not being considered. I
have a 3rd Great Grandfather on my dad's side whose name was Johannes
Wüesthoff. He wrote in his diary that the clerk looked him up and down and
said, "From now on, your name is John West. Get used to it. Next!"
Mark Fields
side the family name was Wogenstahl. She had two brothers and one had a son
(my cousin) who now has a son. Counting backward from my cousin's son, I
can trace the name back 10 generations. I have not traced it further.
However, the name changed from Bockstaller about 8 generations ago (in the
mid 1700's). I have not been able to trace further back At some point
surnames did not exist or else persons surnames were related to occupations.
Many unrelated persons took on identical surnames in medival times so a
name like "Smith" will not trace back to a single person.
Mark
"James A. Doemer" <ckdbigt...@earthlink.net> wrote in message
news:7nuri.12637$tj6...@newsread4.news.pas.earthlink.net...
Delbert Stanley
"raylopez99" <raylo...@yahoo.com> wrote in message
news:1185822456.8...@g4g2000hsf.googlegroups.com...
A few years ago I remember that there was so many Johanssons in Sweden
that it created a minor social crisis. The government started giving tax
credits,
and outright payments to those so named to change their names to something
else.
I don't know if the program was successful, or not.
Lopez? Don't flatter yourself yet. I admit that I have used no computer
models,
nor have I done any research on this, but I suspect that Gonzalez(s) and
Garcia
are whipping you, maybe even Sanchez too. Don't raise that Lopez banner just
yet.
There was time in the late 70s and early 80s when there were so many
refugees
from Vietnam named Nguyen, that I actually started thinking that it was a
kind of
title of respect---like Sherpa. Again I suspect (don't know) that this name
as
a percentage of the total population names of Vietnam is perhaps the highest
of any country. For example, if the name Nguyen is 30% (extremely high) of
total Vietnamese names, I suspect the next country's highest total is
probably no more than 10-15 %. Like I said---pure speculation by me.
Del
Delbert Stanley
"cecilia" <my...@ic24.net> wrote in message
news:1185716096.8...@57g2000hsv.googlegroups.com...
> "Donna" wrote:
>>[...] I recently read that to avoid confusion in legal
>> documents, an American 18th century court clerk might make arbitrary
>> additions to a name. [...]
>
> Surnames arose in order to distinguish people with the same given
> name. One would expect something similar to happen if a community
> became one-surnamed, particularly if they we not all obviously
> related.
Sometimes it just added to the confusion. On many large U.S. plantations
there
were slave clusters that took the surname of their masters after the Civil
War.
Most of the people using the master's name were not related to other slaves
using the same name. Some who didn't use the master's name were actually
related to the master, and not related to other slaves named after the
master!
Sally Hemings of Thomas Jefferson fame for example.
Think of all the confusion with new legal filings in a certain county for
all
these new Smiths. Sometimes people would change the name on a dime.
Moving to a new county? Get a new name! Sometimes that name would
be the county name.
Del
Bruce Remick
I believe I already told you what I have found in the real world. I don't
need any hand calculations. Maybe you're right. I AM confused here.
Considering the fact that my surname will continue to survive long after my
grandson is gone, what point does your "paradox" make that can be useful to
us? Can you prove, as you claim, that my particular surname has died out
after twelve generations? The surname has remained the same in America for
13 generations and counting. Perhaps you should step back from your
computer and formulas for a bit and collect some more real world data. As
raw machines, computers may not lie, but the programs and data people feed
into them often produce erroneous and misleading results.
Bruce
catalpa
"James A. Doemer" <ckdbigt...@earthlink.net> wrote in message
news:7nuri.12637$tj6...@newsread4.news.pas.earthlink.net...
> donnamcr_...@hotmail.com, typed this:
>
>> I think my earlier comment about unmarried mothers giving their
>> children their own surnames is being ignored in this discussion. I
>> hope my post did get through.
>
> Additionally, the vast numbers of immigrants whose surnames where anywhere
> from mildly to radically changed, either voluntarilly or unvoluntarily by
> the immigration check-in authority is probably not being considered. I
> have a 3rd Great Grandfather on my dad's side whose name was Johannes
> Wüesthoff. He wrote in his diary that the clerk looked him up and down
> and said, "From now on, your name is John West. Get used to it. Next!"
>
Immigration officials did not change anyone's name. Passengers to America
had to have some sort of ID and the passenger manifests were done on the
foreign side of the ocean, not on the American side. Many immigrants changed
their own names after arrival to more Anglo-American names to fit in. Most
were too embarrassed to tell the truth about the new name and thus blamed
the name change on some faceless and nameless immigration clerk.
Name change myth article at:
http://genealogy.about.com/od/ellis_island/a/name_change.htm
Phyllis Nilsson
wasn't discussing statistics, I was discussing a fact.
I didn't say anything about a "gut feeling", just the facts as they took
place in my own family. Please don't read something into a post that isn't
there.
Get it, Mr. Lopez? Probably not. I hope you have a pleasant day anyway.
Charani
> Because my friend there are lies, damn lies, and statistics.
Exactly, and statistics (like computer programs) can prove anything.
The same set of stats can even be used to prove opposite results!
Incidentally, I'm not your friend for which I'm very grateful!
> The AVERAGE might be 12 generations (actually I am to confirm this,
> when I perfect my program--my preliminary calculations were based
> on a hypothetical family having, each generation, zero to four
> children, which I don't know if that's historically accurate).
Have it your own way since it's quite clear that nothing that anyone
says about your flawed program is going to have any effect at all.
Theories are fine but that's all they are - theories. To be
validated they have to be proved and you cannot prove your theory
because it's flawed. There's nothing to support it in *fact*.
> Average does not mean your surname. Your surname could be defying
> the averages.
That shows just how flawed your theory is. There are hundreds,
thousands of surnames in the world that have been around since
surnames started to be used. Some of the more uncommon ones may
fizzle out eventually, but the more common ones will continue.
> But I doubt your surname will last, say, 30 generations, unless your
> ancestors get very lucky with XY chromosome pairings.
Since you don't know what my surname is, you are in no position to be
able to state that with any certainty at all - and luck has nothing
to do with it either.
> Sorry, but your name will die out like the rest of us Charani.
Sorry, but it won't :))
> Well, most of us. As a LOPEZ I have a good chance to join WANG,
> KHAN and SMITH in the panoply of immortal names.
Ah so *this* is what it's really all about!! The desire to be
different from others. It's a common human failing and completely
ignores the fact that every human being is unique. However, it's the
first time I've come across anyone who wants to be different who
aligns themselves with hundreds of others because those aren't
immortal names at all!! They're *common* names!! ROFL
Well, one of the names I'm researching is PARKER and that's a common
name so, according to you, it won't die out.
ROFL
I think it's time you joined the trolls :))
--
http://home.comcast.net/~webact1/Collingridge/
raylopez99
wrote:
> On 2007-07-30, raylopez99 <raylope...@yahoo.com> wrote:
>
> Hi Ray -
>
> This is old stochastic processes stuff, specifically discrete
> branching theory. You can find it discussed, for example, in
> and David Roxbee Cox [1977]. You can buy it from Amazon or
> look at the relevant bits on Google.
Cool! I'll take a quick look...no such exact title at Amazon by this
author, though a number of pricey books on this subject and those
severe yellow Springer monographs (does anybody understand these
books? Hats off to you mathematicians who can).
>
> I think Miller and Cox's bottom line was for American lines
> using some old census data, the probability of extinction of
> the male lines of a single individual is 0.86. If you have
> N males with the same surname as "initial conditions", the
> probability of extinction is 0.86 ** N (assuming independent
> breeding), which means the probability of extinction is pretty
> low for most surnames in existence today (if birth patterns
> continued as they were).
Excellent. When I perfect my program, which should not take long,
I'll try and confirm this. I'm building an N-ary tree, and will run a
series of simulations with N=1,2,3... N=number of children (max) per
generation, selected at random from 0 to N, and will see how fast, on
average, the tree peters out for every value of N.
>
> Cheers, B. (I am an applied mathematican, among other things.)
>
> --
> Dr. Brian Leverich Co-moderator, soc.genealogy.methods/GENMTD-L
> Angeles Chapter LTC Admin Chair http://angeles.sierraclub.org/ltc/
> P.O. Box 6831, Frazier Park, CA 93222-6831 lever...@mtpinos.com
AWESOME! I have great respect for mathmaticians, as well as anybody
who uses their real name on Usenet, LOL. I fancy myself as
understanding mathematics, even though I really am not that good at it
(I once tried, in my college years, to prove the Mean Value Theorem--
and got as far as, or rather, I still remember, "Dedekind Cuts" (sic),
which are used to prove (methinks) things like '1 + 1 = 2') I am
slowly working my way through the Fontana History of the Mathematical
Sciences by Ivor Grattan-Guinness, where I am learning things like
Euler's formula for complex analysis (http://en.wikipedia.org/wiki/
Euler's_identity) from the 1740s was anticipated in a different form
by Roger Cotes (1714), de Movre (1722) and d'Alembert. The things
they never teach you in school.
Thanks Dr. Brian, and if my tone in this thread was a bit flippant or
harsh, my appologies, but I use this name primary for flaming (my best
work can be found in the alt.global-warming group).
I will, if anybody is interested, post my analysis from my program
later.
Ray LOPEZ <--a good name, if it was really my own, to survive the p(x)
= 0.86 surname dieoff!
Charani
> I think my earlier comment about unmarried mothers giving their
> children their own surnames is being ignored in this discussion.
I think it is.
> I hope my post did get through.
Yes, it did.
> In some U.S. cultures this is common, and becoming more so. It
> decreases the likelihood of a surname being "daughtered out."
I think this is common everywhere.
> Is this being considered in the OP's statistical model and in other
> models mentioned?
I doubt very much that it is any more than the cases below.
I believe there are some societies where the mother's name is taken
in preference to the father's.
There are also cases where, for reasons of inheritance, a beneficiary
has to adopt a surname other than their own.
In Icelandic society, surnames change with every generation and within
a family. An Icelandic friend's surname is THORMODSSON but his
children's surnames are PALLSSON and PALLSDOTTIR. His wife's surname
is JONSDOTTIR but her brother's is JONSSON.
--
http://home.comcast.net/~webact1/Collingridge/
raylopez99
Thanks Phyllis Diller. I will try.
So, based on Dr. Brian's post in this thread, let's try and make some
sense of the male surname dieoff problem.
Pay attention class. You have 10 people named SMITH. They may or may
not be related, and they generally are big and strong (think
blacksmith) and often lame, perhaps from a hunting accident (think
HEPHAESTUS, the lame Greek blacksmith god; lame not unlike some of the
replies in this thread), because lame people cannot hunt but making
iron is more important to the village anyway.
WHat is the probability, according to Dr. Brian's post, that ALL TEN
of the male surnamed people named SMITH will perish (if they breed
independently, otherwise they will perish even faster)? Pull out your
scientific calculator, which is found in Windows under Start | All
Programs | Accessories | Calculator (choose the Scientific
calculator). enter 0.86 , hit the x^y button, enter 10, hit 'enter',
and you get 0.22 or 22%. Good.
Now let's pretend 100 Smiths exist. What's the chance of dieoff?
Enter 100 instead of 10, and you get a number that's nearly zero.
So, there you have it Ms. Diller. If you are the last SMITH on earth,
the probability of dieoff would be 86% (enter '1' instead of '10'
above). Even if 10 of you SMITHs were around, 22% of the time you all
(or rather, your surname) would die off. But 100 SMITHs is a
different story, and likely your surname will last nearly forever.
See? That wasn't so bad. Math can be fun.
RL
Lesley Robertson
>
> See? That wasn't so bad. Math can be fun.
>
ignoring the simple fact that biological systems rarely take any notice of
statistics. There's families in my database with only 1 child, there's
others with 14 sons, all of whom grew to marry. Families with money are more
likely to raise a number of healthy kids than those who're poor. Many of us
have actual trees going back more than 10 generations, with plenty of
continuous male lines.
This argument is about as useful as the collapsing pedigree one. You can
prove mathematically that the world must have been hopelessly overpopulated
in the past since the number of ancestors doubles every generation, but that
has nothing to do with biology.
If you're serious, learn some biology. If you're trolling, not a bad
attempt.
Lesley Robertson
raylopez99
wrote:
> "raylopez99" <raylope...@yahoo.com> wrote in message
>
> news:1185878105....@b79g2000hse.googlegroups.com...
>
> > See? That wasn't so bad. Math can be fun.
>
> While you're so busy patronising the members of this group, you seem to be
> ignoring the simple fact that biological systems rarely take any notice of
> statistics.
There's an institute in Sante Fe, New Mexico that disagrees with you.
And a host of biologists who have PhDs. But you can cling to your
ignorance if you want to; it's the internet after all.
> There's families in my database with only 1 child, there's
> others with 14 sons, all of whom grew to marry. Families with money are more
> likely to raise a number of healthy kids than those who're poor. Many of us
> have actual trees going back more than 10 generations, with plenty of
> continuous male lines.
OK. How about posting some stats I can use in my analysis? Post
here, strip out the confidential stuff (global replace of any
confidential stuff with a placemarker will do).
>
> This argument is about as useful as the collapsing pedigree one. You can
> prove mathematically that the world must have been hopelessly overpopulated
> in the past since the number of ancestors doubles every generation, but that
> has nothing to do with biology.
Huh? Of course it does. This is called exponential growth, and is
tempered by logarithmic growth, that's why it levels off ("s-shaped
curve"). Biology 101, which you never learned.
> If you're serious, learn some biology. If you're trolling, not a bad
> attempt.
> Lesley Robertson
What are your net credentials "Lesley Robertson"? You can lie, but I
doubt, as a troll, that you can even fabricate a good lie, since you
don't appear that smart.
Remember, post your data here; maybe I can use it.
RL
Robert Melson
Hmmmm.
Lopez' argument is kinda like this: No cat has 8 tails. Every
cat has one more tail than no cat. Therefore, all cats have
9 tails. QED
Statistical Ol' Bob
--
Robert G. Melson | Rio Grande MicroSolutions | El Paso, Texas
-----
"People unfit for freedom---who cannot do much with it---are
hungry for power." ---Eric Hoffer
raylopez99
> Hmmmm.
>
> Lopez' argument is kinda like this: No cat has 8 tails. Every
> cat has one more tail than no cat. Therefore, all cats have
> 9 tails. QED
>
> Statistical Ol' Bob
>
That's not it at all.
Gawd dang Billy Bob, youze stupid!
Ray
Fred McKenzie
"Lesley Robertson" <l.a.ro...@tnw.tudelft.nl> wrote:
> While you're so busy patronising the members of this group, you seem to be
> ignoring the simple fact that biological systems rarely take any notice of
> statistics. There's families in my database with only 1 child, there's
> others with 14 sons, all of whom grew to marry. Families with money are more
> likely to raise a number of healthy kids than those who're poor. Many of us
> have actual trees going back more than 10 generations, with plenty of
> continuous male lines.
Leslie-
You may be correct that the argument is not useful. However, our
tendency is to look up the family tree into the past, not down. (Or is
it the other direction?)
We all have a male parent, whether we know them or not, so ALL genealogy
has a very large number of male ancestors going back to some primitive
first human male. However this argument looks the other way, and
considers the likelihood of one individual's unique name dying out in
some number of generations vs the likelihood of several possibly
unrelated individuals with the same name dying out.
Yes, the real world has other variables. Male children adopting new
names and maternal names would tend to reduce the chances of the name
dying out, but not the chances of the actual male line dying.
How does this apply to Genealogy? I doubt that it does, since Genealogy
is a study of what actually happened, not what might have happened.
Fred
Robert Melson
Hoooooooooo, dogies! No need for name calling,
Ray - and don't take yourself so seriously.
While there MAY be a kernel of truth in your hypothesis,
and I'd certainly encourage you to pursue it, I don't think
the evidence at this point supports it. While you can't, as
some have, argue on the basis of one surname's/bloodline's
continu{ed,ing} existence, seems to me there are far too
many of'em out there that _have_ lasted well beyond your
conjectured point of extinction for that conjecture to be
taken seriously at this point. And, yeah, that's a
subjective appraisal.
Let me suggest this as an experimental design:
(1) state your null hypothesis as clearly and simply as
possible (not your hypothesis but its reverse),
something on the order of "bloodlines/surnames
do NOT extinguish after 300 years" (that certainly
needs refinement, but I think you see what I mean);
(2) from several lists of surnames - maybe those associated
with some of the RootsWeb individual pages - randomly select
some number of surnames, preferably > 1000 in order to assure
a reasonably large statistical universe;
(3) trace those surnames' origins as far back as you can, then
bring them forward, following branch extinctions, rejoinings,
current population numbers, etc;
(4) run your collected, ahhh, durations, through, e.g., an
analysis of variance, and such other statistical tests
as you might think appropriate;
(5) if you've structured things right and haven't otherwise
"cooked" the data, you should be able to accept or reject
your null hypothesis with a reasonable degree of certainty -
I'd suggest a p=0.05, but something in the range of
0.05 - 0.15 would probably be adequate. Anything higher
is really questionable, IMO, and you might just as well
have flipped a coin. I
Report your results, here, wherever, and don't be surprised
if folks ask to see the data and your calculations. I'm
sure we'd all be interested. No hand waving allowed,
however.
Smilin' Ol' Bob
Donna
Donna
Lesley Robertson
> There's an institute in Sante Fe, New Mexico that disagrees with you.
> And a host of biologists who have PhDs. But you can cling to your
> ignorance if you want to; it's the internet after all.
A whole Institute? Wow.
> OK. How about posting some stats I can use in my analysis? Post
> here, strip out the confidential stuff (global replace of any
> confidential stuff with a placemarker will do).
>
Do your own research. You can't afford my time.
>>
>> This argument is about as useful as the collapsing pedigree one. You can
>> prove mathematically that the world must have been hopelessly
>> overpopulated
>> in the past since the number of ancestors doubles every generation, but
>> that
>> has nothing to do with biology.
>
> Huh? Of course it does. This is called exponential growth, and is
> tempered by logarithmic growth, that's why it levels off ("s-shaped
> curve"). Biology 101, which you never learned.
Read what I said again... Statistics show, as you have so admirably
illusrtrated, that there should have been more people alive in the distant
past because we must have had all those ancestors. Pedigree collapse is why
it's not true.
And the log curve you describe describes microbial growth and flattens off
because the population density reaches a point at which nutrients become
inadequate. Try Googling on Monod.
Lesley Robertson
raylopez99
wrote:
>
> Read what I said again... Statistics show, as you have so admirably
> illusrtrated, that there should have been more people alive in the distant
> past because we must have had all those ancestors. Pedigree collapse is why
> it's not true.
You're not thinking straight Lesley. What did they put in your
herringbone?
Read carefully this thread including the reply by Dr. Brian. Then do
the simulation by hand that I suggested for a small tree (0-3) so that
your precious time is not wasted. You'll see, as Dr. Brian's (0.86^x)
stat shows, that with "at least" 86% probability this tree will die
out. (At least because with a small tree of 0-3 children per node it
will die faster I am almost certain).
I am working on a simulation now, as a C# programming exercise (I'm
really interested in building an n-ary tree and traversing it both
recursively and iteratively, rather than this particular genealogy
problem, but I digress). I'll post results here when I finish, maybe
(not that you people, ignorant as sin or humanities majors, will
appreciate it).
> And the log curve you describe describes microbial growth and flattens off
> because the population density reaches a point at which nutrients become
> inadequate. Try Googling on Monod.
Stupid is as stupid does. OK, I'll do that. http://en.wikipedia.org/wiki/Jacques_Monod
That has almost nothing to do with this topic.
Just Wiki this, Einstein: http://en.wikipedia.org/wiki/Logistic_curve
Ray Lopez
raylopez99
wrote:
> In article <1185897801.342273.127...@19g2000hsx.googlegroups.com>,
> Smilin' Ol' Bob
>
> --
> Robert G. Melson | Rio Grande MicroSolutions | El Paso, Texas
> -----
Hahaha! Good one Bob! I see you have a sense of humor, and like to
post silly stuff with a "straight face", as a comedic foil. I
appreciate good humor and commend you for it. My apologies for
flaming you in an earlier email amigo--for a minute I thought you were
serious.
All kidding aside, as we know, what I am talking about is going
forward, not what "historically" has happened. What you propose is
essentially refining the probability function for males/females born
so that it's something more accurate than p(x) = 50% (which is what my
default is right now). I've thought about this, and if I was writing
a commercial program, I would do what you suggest to find this true
p(x). But it's a minor point (BTW, you also have to correct for p(x)
for different cultures as well, that practice female infantcide, but I
digress).
For now, I'll stick with 50% as a good rule of thumb, confident it
won't change the stats much from the 'true' (historical) stats for
generation male surname dieoff.
RL
William Whalley
Surnames" by C. M. Sturges and B. C. Haggett, 1990, ISBN 0785548475.
It is in a few British and US libraries, but not on Amazon.com or ABE
books.
According to a Scientific American Computer Recreations article (May
1986, p 12-16), the authors were British Ministry of Defence employees
who traced genealogy in their spare time. They were puzzled by the
gradual disappearance of surnames in the records. They wrote a
simulation program such as you describe. They used these factors derived
from a statistical analysis of genealogical records.
Number of males
in a family Probability
who will marry
0 0.317
1 0.364
2 0.209
3 0.080
4 0.023
5 0.005
6 0.001
I.e., they estimated that in any given family, there was a 31.7% chance
there would be no son to marry and (possibly) carry on the surname, a
36.4% chance that one son would marry, etc.
Robert Melson
raylopez99 <raylo...@yahoo.com> writes:
> On Jul 31, 11:44 am, mels...@aragorn.rgmhome.net (Robert Melson)
> wrote:
> post silly stuff with a "straight face", as a comedic foil. I
> appreciate good humor and commend you for it. My apologies for
> flaming you in an earlier email amigo--for a minute I thought you were
> serious.
Well, you can take it as you wish - if you're amused by whativer
it is in my previous ...
>
> All kidding aside, as we know, what I am talking about is going
> forward, not what "historically" has happened. What you propose is
> essentially refining the probability function for males/females born
> so that it's something more accurate than p(x) = 50% (which is what my
> default is right now). I've thought about this, and if I was writing
> a commercial program, I would do what you suggest to find this true
> p(x). But it's a minor point (BTW, you also have to correct for p(x)
> for different cultures as well, that practice female infantcide, but I
> digress).
>
> For now, I'll stick with 50% as a good rule of thumb, confident it
> won't change the stats much from the 'true' (historical) stats for
> generation male surname dieoff.
>
> RL
>
Unfortunaely for your project, though, you do have to forecast
a trend and doing so does involve historical - not hysterical -
data (although I will agree that some of the data out there really
_is_ hysterical). The crucial part of your project is stating and
refining your null-hypothesis: without _that_ you're really just
making meaningless noise. After that, then a properly selected
universe is most critical.
Diverging slightly, you've no doubt read of the concern in some
quarters regarding the, ummmm, arabification of Europe: massive
immigration from the Islamic world, coupled with a declining
birth-rate - in some countries less than the replacement rate -
is leading to a "die-off" of the original population and its
replacement with a new, if not heartier, stock. I mention
this because it suggests you might need to trace historic
trends in birthrates as part of you "study" and not rely on
your back-of-the-envelope guess of 0-5 (male) births per
family.
Y'wanna play games with numbers, fine by me. But if those
numbers are going to be meaningful to anybody but you, you've
got to take the time to design a proper experiment. And,
so far, it doesn't appear that you have.
No need for a reply. I hope you learn some day that folks aren't
near as stupid as you think they are. When you've figured that
out and get over your arrogance, maybe then you'll be worth
talking to.
Ciao and <plonk!>,
Bob Melson
--
Robert G. Melson | Rio Grande MicroSolutions | El Paso, Texas
-----
hlmw
marriage, and there are plenty of us on this list. As usual, we get left
out of the genetic slippery slope! By the way, if I follow my maternal
line re male ancestors, it is unbroken back to 1411.a.d.
Lorna
> -------------------------------
> To unsubscribe from the list, please send an email to ALT-GENEAL...@rootsweb.com with the word 'unsubscribe' without the quotes in the subject and the body of the message
>
>
>
raylopez99
>
> Y'wanna play games with numbers, fine by me. But if those
> numbers are going to be meaningful to anybody but you, you've
> got to take the time to design a proper experiment. And,
> so far, it doesn't appear that you have.
Not so fast you fat lard ash Texan red neck. Hahaha! I'm kidding of
course.
In my program you'll be able to enter any range of integers for
children born per generation. In this thread I picked "0 to 4"
because otherwise the tree takes longer to die, if not forever. But
in the program it will be easy enough to allow the user to enter any
number. As one can appreciate, if you enter a very high (and
unrealistic) number, like "30" (that is, up to 30 offspring children
per husband-wife pairing--and I mean every husband and wife,
individually, meaning the wife can give birth to up to thirty (30!)
kids that reach reproductive age), then since the computer picks a
random number between zero and 30, your tree will likely not die off
ever, and will grow forever, because of the rule of thumb (see this
thread) that the probability of extinction is (0.86)^^N, here
N=30/2(average) = 15, and if you do this calculation then the
probability of the "forefather's" surname going extinct is only 10%
(0.1041), which means nine times out of ten your tree will never die.
What this means is that your program will go into an "infinite loop"
and/or your hand calculations will never cease. In the actual
program, I intend to monitor this, and if the program starts "running
away" I'll abort it. But, like I said in this thread, for a more
realistic ratio of zero to four kids (meaning two kids per husband &
wife, "on average") your surname goes extinct in about 10 to 12
generations I've found, from my rough calculations as outlined in this
thread.
>
> No need for a reply. I hope you learn some day that folks aren't
> near as stupid as you think they are. When you've figured that
> out and get over your arrogance, maybe then you'll be worth
> talking to.
>
> Ciao and <plonk!>,
> Bob Melson
An elitist are you? In another thread, I'll post on why I think
geneologists are racist Luddites, who don't like science and
erroneously believe that homogeneous population societies are best
(actually, homogeneous societies are indeed very slightly better
economically that a heterogenous population, but only very slightly, a
fraction of a point per year or so, and it's much more important that
your economy be open rather than closed. Thus homogeneous Sweden is
slightly faster growing historically than heterogenous USA, as
evidenced by the Swedish stock market long term, which beats even the
US stock market, but any relatively heterogenous country, even Canada
or Spain, is better --grows faster-- than a closed economy homogeneous
population country like North Korea. And don't even think about
"racial purifying" by 'sending them Mexicans back home"--you gringos
will stew in your own miasma if you try that).
See you later you racist Luddites!
Ray LOPEZ <--a name that will live in this newsgroup, in infamy, for a
long, long time!
raylopez99
> (actually, homogeneous societies are indeed very slightly better
> economically that a heterogenous population, but only very slightly, a
> fraction of a point per year or so, and it's much more important that
> your economy be open rather than closed. Thus homogeneous Sweden is
> slightly faster growing historically than heterogenous USA, as
> evidenced by the Swedish stock market long term, which beats even the
> US stock market, but any relatively heterogenous country, even Canada
> or Spain, is better --grows faster-- than a closed economy homogeneous
> population country like North Korea.
Just to clarify this economic point, since I sometimes feel I'm
talking to children in this thread: we're comparing heterogenous open
economies to homogeneous open economies and saying the latter (the
second one) grows slightly faster, based on historical data, but a
closed economy always grows much slower than an open economy,
regardless of whether the closed economy is homogeneous or not. So
the important factor is openness, not homogenuity. Putting all this
together, if China has a truly open economy (and I'm not entirely sure
they do, having visited), with their homogeneous population they will
grow faster economically than anybody else in history.
RL <--Ray LOPEZ, who is thinking of changing his surname to WANG, so
his surname can live forever!
Bruce Remick
Did you ever consider pointing out what *practical use* all your statistical
gibberish can be to any of us researching our ancestors? Without all the
impressive statistical jargon-- for us inferior humans, you know. You seem
to be so into yourself and into whatever you're doing that you have never
bothered to learn how to tactfully market your theories without alienating
much of your audience. You strike me as one of those people who have never
thrown a baseball, lit a campfire, taken a date to the prom, .......... or
enjoyed digging up any family genealogy information.
Bruce
'an institute in Santa Fe, NM? Why would I have guessed that? Actually,
Roswell was my first guess.'
raylopez99
>
> Did you ever consider pointing out what *practical use* all your statistical
> gibberish can be to any of us researching our ancestors? Without all the
> impressive statistical jargon-- for us inferior humans, you know.
OK, I'll try. You are indeed inferior you know, but I won't rub it in.
Practical use: a client walks in the door (let's assume you do
genealology for pay). He wants to know whether he and his brothers
surname will "live forever". Since you have not read this post, you
throw up your hands, say "Hell if I know dude--that's math and I
flunked math!", and the client thanks you and leaves to go to your
competitor. Your competitor, having read this thead, simply asks the
client: "how many brothers and nephews and other relatives do you have
that share your last name?". The client answers "25". Or, even
better, the client doesn't know, and the geneologist researches the
answer (for a fat fee) and finds it is "25". The competitor
geneologist plugs in "25" into the formula y(x) = 0.86^25 = 0.023 =
2.3% and proudly tells the pleased client: "Your surname has a 97.7%
chance of surviving forever!". And gets paid. The satisfied client,
as he's walking out the door, says to your competitor: "your
competition Bruce couldn't help me!", to which your competition says,
with a chuckle, "Well, Bruce is retired and really only does geneology
as a hobby"
> You seem
> to be so into yourself and into whatever you're doing that you have never
> bothered to learn how to tactfully market your theories without alienating
> much of your audience. You strike me as one of those people who have never
> thrown a baseball, lit a campfire, taken a date to the prom, .......... or
> enjoyed digging up any family genealogy information.
>
I've done all of the above except taken a date to the prom. I skipped
my prom since it was a waste of time IMO. Come to think of it I
skipped my unofficial high school graduation, my professional degree
graduation, my high school 10 yr reunion, and the like. Waste of
time. Too busy making money.
> Bruce
>
> 'an institute in Santa Fe, NM? Why would I have guessed that? Actually,
> Roswell was my first guess.'
If you were intelligent Bruce I would have gotten a chuckle out of
that remark, but I'm afraid you're being literal.
RL
Bruce Remick
> On Jul 31, 5:38 pm, "Bruce Remick" <rem...@cox.net> wrote:
>>
>> Did you ever consider pointing out what *practical use* all your
>> statistical
>> gibberish can be to any of us researching our ancestors? Without all the
>> impressive statistical jargon-- for us inferior humans, you know.
>
> OK, I'll try. You are indeed inferior you know, but I won't rub it in.
>
> Practical use: a client walks in the door (let's assume you do
> genealology for pay). He wants to know whether he and his brothers
> surname will "live forever".
Excuse me. This is your best example of a *practical use*?
Since you have not read this post, you
> throw up your hands, say "Hell if I know dude--that's math and I
> flunked math!", and the client thanks you and leaves to go to your
> competitor.
Actually, I would have simply said said "yes". Math plays absolutely no
part in it. Happy client.
Your competitor, having read this thead, simply asks the
> client: "how many brothers and nephews and other relatives do you have
> that share your last name?". The client answers "25". Or, even
> better, the client doesn't know, and the geneologist researches the
> answer (for a fat fee) and finds it is "25". The competitor
> geneologist plugs in "25" into the formula y(x) = 0.86^25 = 0.023 =
> 2.3% and proudly tells the pleased client: "Your surname has a 97.7%
> chance of surviving forever!". And gets paid. The satisfied client,
> as he's walking out the door, says to your competitor: "your
> competition Bruce couldn't help me!", to which your competition says,
> with a chuckle, "Well, Bruce is retired and really only does geneology
> as a hobby"
(Wait a minute. You earlier had me doing genealogy for pay! Stick to your
math.)
And then the client reminds my competitor that half of the males in his
surname line are currently patrolling in the Middle East, and 75% of the
past two generations of males had died of cancer before age 50. And then
the client gets hit by a car on the way out, before he even had time wonder
what in the hell he can do with that number 25 and before he could realize
how he had been ripped off.
I'd be curious as to how many professional genealogists have ever handled a
question like your classic example.
>
>> You seem
>> to be so into yourself and into whatever you're doing that you have never
>> bothered to learn how to tactfully market your theories without
>> alienating
>> much of your audience. You strike me as one of those people who have
>> never
>> thrown a baseball, lit a campfire, taken a date to the prom, ..........
>> or
>> enjoyed digging up any family genealogy information.
>>
>
> I've done all of the above except taken a date to the prom. I skipped
> my prom since it was a waste of time IMO. Come to think of it I
> skipped my unofficial high school graduation, my professional degree
> graduation, my high school 10 yr reunion, and the like. Waste of
> time. Too busy making money.
Hmmm. I suspected right again. No math needed either.
>>
>> 'an institute in Santa Fe, NM? Why would I have guessed that?
>> Actually,
>> Roswell was my first guess.'
>
> If you were intelligent Bruce I would have gotten a chuckle out of
> that remark, but I'm afraid you're being literal.
I am. You should have. And I am.
Phyllis Nilsson
whether one's surname will die out 300 years after one is dead or even
sooner. A classic case of lights on but nobody 's home.
Charani
> My head hurts.
LOL
I think the OP's will as well - from banging his head against a brick
wall with his flawed program <VBG>
--
http://home.comcast.net/~webact1/Collingridge/
raylopez99
The Wall Street Journal had an article last year about a newly made
rich person who wanted to make sure his family name survived in
perpetuity. The gist of the article was that the person found it was
hard to do this--not because of his last name dying out--but because
his teenage kids were lazy and really didn't want to carry on any
family tradition.
RL
Bruce Remick
Too "lazy" to marry and have families? Tradition and surname continuation
then becomes automatic with little or no extra work. It appears your
formula wasn't needed here.
Bruce
Allen
> Here's a paradox that's a real puzzler.
>
> Statistically, a male descendant cannot survive from a common ancestor
> for more than about 12 generations, or roughly 300 years, because
> after 12 generations or so there's a good chance no more male
> offspring will result (you can do computer simulations to verify this,
> as I have done--assume 50% probability of a male heir and 0-5 kids per
> generation, per male heir; you'll find after about 12 iterrations the
> male offspring will die out).
>
> So how can a name like "Smith" or "Lopez" survive for more than this
> time? I surmise there were many people with surnames "Smith" or
> "Lopez", genetically 'unrelated' (insofar as possible), so that's why
> there's so many of these surnames today. Another possibility is that
> surnames are relatively recent, as recent as the middle ages, which
> was about 500 to 1000 years ago, and lots of surnames existed then,
> and have since gone extinct, so we still have a lot of unique surnames
> but they will eventually dwindle down to a handful of names (not
> unlike what is present today in Korea--everybody is named "Kim"), as
> time progresses, assuming patermonial descent name convention of
> course (unlike some tribes in Africa, which use matrimonial surnames,
> though that also has the same problem, long term).
>
> In short, if you do the simulations long enough, eventually you should
> get (I would imagine, I haven't done this simulation yet), one surname
> ("Adam"?) for the whole population.
>
> Any thoughts on this paradox? Surely it must have been discussed
> before.
>
> RL
>
Perhaps this has been mentioned before in this to-me silly thread, but
there is a major problem in the fifth word in the second paragraph"
"cannot". Probability can hardly express an absolute, but only a degree
of likelihood. "Statistically very unlikey" would have been a better
expression. I know that later the OP uses "a good chance", but that is
buried farther down. This a tabloid headline quality to the posting.
As far as "who would want to know", never underestimate the degree of
vanity that exists in some people. Does "Egyptian pyramids" ring a bell?
Allen
the_ver...@comcast.net
It is really rather easy to insure that.
Just set up a trust with all the money and they only get the income of
the trust if they use the surname. The original funds in the trust are
never dispursed. Somewhat akin to a "Name and Arms" clause.
singhals
> On Tue, 31 Jul 2007 14:47:52 -0500, Donna wrote:
>
>
>>My head hurts.
>
>
> LOL
>
> I think the OP's will as well - from banging his head against a brick
> wall with his flawed program <VBG>
Hey, man, if it keeps off the streets ...?
Cheryl
Dennis
wrote:
>There is quite a literature on this interesting topic. I just found
>this web page concerning name frequencies http://homepages.newnet.co.uk/dance/webpjd/index.htm
Yes. Even ignoring the OP's lack of social skills, the topic is an
interesting one.
I wrote a quick simulation program in C++ just to see what would happen.
I've posted the source at http://tinyurl.com/26n73t for anyone who is
interested. I've probably run it a couple of hundred times while testing
and tweaking it. I've found that many lines die out very quickly.
Relatively few lines die out once you get above 15 generations or so. By
the time you get to 50 generations, there are usually so many male
descendants left (in the thousands) that it would seem improbable that
the line would go extinct.
Of course, my simulation assumes a perfect world (every male passes his
surname on to the next generation unchanged) ... which we all agree is
not the case.
--
Dennis K.
raylopez99
> On Sun, 29 Jul 2007 16:21:16 -0000, wgwhalley <bwhal...@comcast.net>
> wrote:
>
> >There is quite a literature on this interesting topic. I just found
>
> Yes. Even ignoring the OP's lack of social skills, the topic is an
> interesting one.
Lack of social skills says a PROGRAMMER? That's harsh!
>
> I wrote a quick simulation program in C++ just to see what would happen.
> I've posted the source athttp://tinyurl.com/26n73tfor anyone who is
> interested. I've probably run it a couple of hundred times while testing
> and tweaking it. I've found that many lines die out very quickly.
Nice work; took a quick look but I see you are cheating--you are using
a formula (Sturges and Hackett) to work backwards and come up with the
number of males who marry. In intend to write a program in C# that
will actually build and traverse an n-ary tree, and assume any number
of male children, then see which lines die off and which run away.
Should be done in a few weeks (I really am just learning the nuances
of C#, which doesn't use pointers like C++ but only references).
RL
raylopez99
>
> I wrote a quick simulation program in C++ just to see what would happen.
> I've posted the source athttp://tinyurl.com/26n73tfor anyone who is
> interested. I
Besides your program not being object oriented, I think I've spotted a
weakness: your rand() is local, and srand() is not reseeding it.
Therefore you're not really doing different simulations everytime you
instantiate your program--just running it as before it appears. I
could be wrong, but that's how it looks to me.
RL
Mark Fields
> Practical use: a client walks in the door (let's assume you do
> genealology for pay). He wants to know whether he and his brothers
> surname will "live forever".
ROTFLMAO
I don't need a program or theroem. If they are willing to pay for an answer
I will take their money and tell them there's a 100% chance.
If this proves to be untrue, they can get there money back from one of my
ancestors but they won't find them since my surname will be gone.
This thread is good for a new laugh every day.
Mark
James A. Doemer
typed this:
> "James A. Doemer" <ckdbigt...@earthlink.net> wrote in message
> news:7nuri.12637$tj6...@newsread4.news.pas.earthlink.net...
>> In News 46ae66de$0$4649$4c36...@roadrunner.com,, Donna at
>> donnamcr_...@hotmail.com, typed this:
>>
>>> I think my earlier comment about unmarried mothers giving their
>>> children their own surnames is being ignored in this discussion. I
>>> hope my post did get through.
>>
>> Additionally, the vast numbers of immigrants whose surnames where
>> anywhere from mildly to radically changed, either voluntarilly or
>> unvoluntarily by the immigration check-in authority is probably not
>> being considered. I have a 3rd Great Grandfather on my dad's side
>> whose name was Johannes Wüesthoff. He wrote in his diary that the
>> clerk looked him up and down and said, "From now on, your name is
>> John West. Get used to it. Next!"
>
> Immigration officials did not change anyone's name.
Odd that his papers say Johannes Wüsthoff and his immigration letter says
John West then, huh?
Passengers to
> America had to have some sort of ID and the passenger manifests were
> done on the foreign side of the ocean, not on the American side. Many
> immigrants changed their own names after arrival to more
> Anglo-American names to fit in. Most were too embarrassed to tell the
> truth about the new name and thus blamed the name change on some
> faceless and nameless immigration clerk.
> Name change myth article at:
> http://genealogy.about.com/od/ellis_island/a/name_change.htm
Mary_...@tvo.org
relatively recent construct - prior to that, as I know you know....a
person was known as John the Smith, or John from Balmy Beach, or John
the son of Isaac, or John with the Big Nose or John from Clan X or
some other variation on similar themes that could result in new
"surnames" for every generation. Or maybe you are from a culture like
Iceland where mom, dad, son and daughter all end up with different
"surnames".
And then you layer on top of that illiteracy. In one arm of my
Scottish family, the first literate generation was born in 1875, so
you flash back earlier...and originating from a single household you
find modern lines with surnames Bremner, Brember, Brymner, Brebner
etc. The variation that stuck was a function of where specific
indiviuals ended up, and the whim of the local clerics or census
taker.
Then you throw in changes in names at the time of immigration, such as
Zolotnicov transmogrifying into Weinberg. I have one batch whose names
can be in Hebrew, German, Polish etc.
The vast majority of us(even the very lucky ones) have our genealogy
peter out in tracable form 300-400 years back due to lack or loss of
written records, illiteracy, surname instability etc. The few who can
go back further generally can do so in a single line or two because
they have some ancestor who was a noble muckety muck and thus some
written records survive. I've got a couple of those, and they are the
only ones that push much earlier than 1580, and even their "surnames"
aren't stable.
I'd be pretty much gobsmacked if you could show me a family tree where
the majority of the surnames in the web were all stable and tracable
back 300 years (lets see....15 generations or so.....got all the
blanks filled in back that far? Me neither, and I don't expect to be
able to, for all the reasons discussed.) Your paradox ain't happened
yet. Naming is a crappy system for tracing lines. Maybe going forward
from here, but not from here into the distant past.
If I had to view genealogy as a straight line back tied to my birth
surname and strictly patrilineal, I'd be out of steam entirely at my
2G grandfather, who was a Scot who moved around a lot and was named
John Gordon...which is a whole lot like being a Brit named John Smith.
Fortunately, I'm more interested in the web of relationships and the
history of migration.
M
Texas Gen
asked to spell his last name. The recruiter told him that was not the
correct way to spell that name, and the recruiter recorded another spelling
instead. The uncle became career army.
So ever more, that branch of the family spells their surname the way the
army recruiter dictated. The rest of the family either copied them or
remained with the original spelling (which was a perfectly good variant
spelling of the name).
Donna
catalpa
> In News Ppxri.7542$8u1.4109@trnddc07,, catalpa at cat...@entertab.org,
> typed this:
>
>> "James A. Doemer" <ckdbigt...@earthlink.net> wrote in message
>> news:7nuri.12637$tj6...@newsread4.news.pas.earthlink.net...
>>> In News 46ae66de$0$4649$4c36...@roadrunner.com,, Donna at
>>> donnamcr_...@hotmail.com, typed this:
>>>
>>>> I think my earlier comment about unmarried mothers giving their
>>>> children their own surnames is being ignored in this discussion. I
>>>> hope my post did get through.
>>>
>>> Additionally, the vast numbers of immigrants whose surnames where
>>> anywhere from mildly to radically changed, either voluntarilly or
>>> unvoluntarily by the immigration check-in authority is probably not
>>> being considered. I have a 3rd Great Grandfather on my dad's side
>>> whose name was Johannes Wüesthoff. He wrote in his diary that the
>>> clerk looked him up and down and said, "From now on, your name is
>>> John West. Get used to it. Next!"
>>
>> Immigration officials did not change anyone's name.
>
> Odd that his papers say Johannes Wüsthoff and his immigration letter says
> John West then, huh?
>
You can believe what you want, but immigration officials did not care what
peoples' names were and did not change them. If Johannes Wüsthoff had his
name changed to John West, he did it himself and has conviently blamed some
unknown immigration clerk to remove his responsibilty for the change. In
researching my ancestry I have found that people tell all sorts of untruths
about themselves.
I have no idea what an "immigration letter" is.
James A. Doemer
typed this:
> "James A. Doemer" <ckdbigt...@earthlink.net> wrote in message
> news:lCuti.14875$zA4....@newsread3.news.pas.earthlink.net...
>> In News Ppxri.7542$8u1.4109@trnddc07,, catalpa at
>> cat...@entertab.org, typed this:
>>
>>> "James A. Doemer" <ckdbigt...@earthlink.net> wrote in message
>>> news:7nuri.12637$tj6...@newsread4.news.pas.earthlink.net...
>>>> In News 46ae66de$0$4649$4c36...@roadrunner.com,, Donna at
>>>> donnamcr_...@hotmail.com, typed this:
>>>>
>>>>> I think my earlier comment about unmarried mothers giving their
>>>>> children their own surnames is being ignored in this discussion. I
>>>>> hope my post did get through.
>>>>
>>>> Additionally, the vast numbers of immigrants whose surnames where
>>>> anywhere from mildly to radically changed, either voluntarilly or
>>>> unvoluntarily by the immigration check-in authority is probably not
>>>> being considered. I have a 3rd Great Grandfather on my dad's side
>>>> whose name was Johannes Wüesthoff. He wrote in his diary that the
>>>> clerk looked him up and down and said, "From now on, your name is
>>>> John West. Get used to it. Next!"
>>>
>>> Immigration officials did not change anyone's name.
>>
>> Odd that his papers say Johannes Wüsthoff and his immigration letter
>> says John West then, huh?
>>
>
> You can believe what you want,
I definitely believe the documentation, and his diary, not someone claiming
that something NEVER happened.
Dennis
>I wrote a quick simulation program in C++ just to see what would happen.
>I've posted the source at http://tinyurl.com/26n73t for anyone who is
>interested. I've probably run it a couple of hundred times while testing
>and tweaking it. I've found that many lines die out very quickly.
>Relatively few lines die out once you get above 15 generations or so. By
>the time you get to 50 generations, there are usually so many male
>descendants left (in the thousands) that it would seem improbable that
>the line would go extinct.
OK. I decided to wrap my original code in a 1000 iteration loop. Here
are some typical results...
Total male lines which become extinct after 1 generations = 317
Total male lines which become extinct after 2 generations = 136
Total male lines which become extinct after 3 generations = 66
Total male lines which become extinct after 4 generations = 48
Total male lines which become extinct after 5 generations = 31
Total male lines which become extinct after 6 generations = 28
Total male lines which become extinct after 7 generations = 22
Total male lines which become extinct after 8 generations = 15
Total male lines which become extinct after 9 generations = 8
Total male lines which become extinct after 10 generations = 10
Total male lines which become extinct after 11 generations = 10
Total male lines which become extinct after 12 generations = 7
Total male lines which become extinct after 13 generations = 4
Total male lines which become extinct after 14 generations = 2
Total male lines which become extinct after 15 generations = 5
Total male lines which become extinct after 16 generations = 4
Total male lines which become extinct after 17 generations = 1
Total male lines which become extinct after 18 generations = 6
Total male lines which become extinct after 19 generations = 3
Total male lines which become extinct after 20 generations = 4
Total male lines which become extinct after 21 generations = 4
Total male lines which become extinct after 22 generations = 0
Total male lines which become extinct after 23 generations = 1
Total male lines which become extinct after 24 generations = 2
Total male lines which become extinct after 25 generations = 1
Total male lines which become extinct after 26 generations = 3
Total male lines which become extinct after 27 generations = 0
Total male lines which become extinct after 28 generations = 0
Total male lines which become extinct after 29 generations = 0
Total male lines which become extinct after 30 generations = 0
Total male lines which become extinct after 31 generations = 0
Total male lines which become extinct after 32 generations = 1
Total male lines which become extinct after 33 generations = 0
Total male lines which become extinct after 34 generations = 1
Total male lines which become extinct after 35 generations = 0
Total male lines which become extinct after 36 generations = 0
Total male lines which become extinct after 37 generations = 0
Total male lines which become extinct after 38 generations = 0
Total male lines which become extinct after 39 generations = 0
Total male lines which become extinct after 40 generations = 0
Total male lines which become extinct after 41 generations = 0
Total male lines which become extinct after 42 generations = 0
Total male lines which become extinct after 43 generations = 0
Total male lines which become extinct after 44 generations = 0
Total male lines which become extinct after 45 generations = 0
Total male lines which become extinct after 46 generations = 0
Total male lines which become extinct after 47 generations = 0
Total male lines which become extinct after 48 generations = 0
Total male lines which become extinct after 49 generations = 0
Total male lines which become extinct after 50 generations = 260
After running it several times I find that the number of surnames that
survive 50 generations generally varies from 20%-30%.
I used the Sturges and Hackett data (see
http://homepages.newnet.co.uk/dance/webpjd/intro/big.htm). Feel free to
substitute your own reality.
Dennis
>Total male lines which become extinct after 50 generations = 260
Maybe I should have changed that last display to read something like...
"Total male lines surviving to generation 50 and beyond = 260"
James A. Doemer
nob...@nowhere.com, typed this:
Not being a statistician I have to ask. With all these data models I've
seen mentioned here, couldn't a person just chose whichever data model gave
results that fit into whatever preconceived notion a person might have?
Allen
>
> Not being a statistician I have to ask. With all these data models I've
> seen mentioned here, couldn't a person just chose whichever data model gave
> results that fit into whatever preconceived notion a person might have?
>
>
Ah, you have discovered the first principle of using statistics to prove
a point. See "How to Lie With Statistics", written by Darrel Huff in
1954 and still in print.
Allen
Christopher Jahn
news:46c4fce6$0$15336$4c36...@roadrunner.com:
As Twain said: "There are lies, there are damned lies, and then
there are statistics."
--
}:-) Christopher Jahn
{:-( http://soflatheatre.blogspot.com/
OF COURSE IT HURTS. YOU'RE GETTING SCREWED BY AN ELEPHANT
Graeme Wall
Allen <al...@nothere.net> wrote:
Wasn't it George Bernard Shaw who referred to: Lies, Damned Lies and
Statistics.
--
Graeme Wall
My genealogy website:
<http://www.greywall.demon.co.uk/genealogy/index.html>
Nigel Bufton
"Graeme Wall" <Gra...@greywall.demon.co.uk> wrote in message
news:d8e0a9134f%Gra...@greywall.demon.co.uk...
> In message <46c4fce6$0$15336$4c36...@roadrunner.com>
> Allen <al...@nothere.net> wrote:
>
>> James A. Doemer wrote:
>>
>> >
>> > Not being a statistician I have to ask. With all these data models
>> > I've
>> > seen mentioned here, couldn't a person just chose whichever data model
>> > gave
>> > results that fit into whatever preconceived notion a person might have?
>> >
>> >
>>
>> Ah, you have discovered the first principle of using statistics to prove
>> a point. See "How to Lie With Statistics", written by Darrel Huff in
>> 1954 and still in print.
>> Allen
>
> Wasn't it George Bernard Shaw who referred to: Lies, Damned Lies and
> Statistics.
>
> --
> Graeme Wall
>
Although sometimes attributed to Mark Twain – because it appears in his
posthumously-published Autobiography (1924) – this should more properly be
ascribed to Disraeli, as indeed Twain took trouble to do: his exact words
being, ‘The remark attributed to Disraeli would often apply with justice and
force: “There are three kinds of lies: lies, damned lies, and statistics”.
Nigel
Gene Y.
"Statistics don't lie! But liars use statistics."
--
Gene Y.
Bob F.
>>>> Ah, you have discovered the first principle of using statistics to
>>>> prove
>>>> a point. See "How to Lie With Statistics", written by Darrel Huff in
>>>> 1954 and still in print.
Yes...and I have noticed that a very high percentage of people in my
database have died within one year of their last birthday...very high! ;-)
Bob
Lesley Robertson
news:UOqdnV3t3LQgOljb...@comcast.com...
related?
Lesley Robertson
James A. Doemer
al...@nothere.net, typed this:
I see that it's listed at my local library's website. I'll have to stop and
get it. Thanks.
James A. Doemer
Gra...@greywall.demon.co.uk, typed this:
> In message <46c4fce6$0$15336$4c36...@roadrunner.com>
> Allen <al...@nothere.net> wrote:
>
>> James A. Doemer wrote:
>>
>>>
>>> Not being a statistician I have to ask. With all these data models
>>> I've seen mentioned here, couldn't a person just chose whichever
>>> data model gave results that fit into whatever preconceived notion
>>> a person might have?
>>>
>>>
>>
>> Ah, you have discovered the first principle of using statistics to
>> prove a point. See "How to Lie With Statistics", written by Darrel
>> Huff in 1954 and still in print.
>> Allen
>
> Wasn't it George Bernard Shaw who referred to: Lies, Damned Lies and
> Statistics.
I believe that Shaw was quoting Mark Twain.
James A. Doemer
ni...@bufton.org, typed this:
Ah, thanks for the info.
Graeme Wall
Who was quoting Disraeli...
And Oscar Wilde probably wished he'd said it.
Allen
lying. Most professional statisticians are honest people, but there are
those, mostly employed in marketing or politics, who know how to play
games--and do.My son-in-law teaches advance placement calculus (in high
school)and one year also added a class in statistics. I was very happy
to find out that now students can get some exposure to statistics in
high school.
Allen
John
John
John
Having studied a small amount of statistics (Statistical Process
Control) I decided that statistics are simply a means of creating and
maintaining jobs. LOTS of jobs.
singhals
"Oh, you will, you will." As somebody told him about
another bon-mot...
Bob Jones
"Allen" <al...@nothere.net> wrote in message
news:46c6fe32$0$28797$4c36...@roadrunner.com...
> James A. Doemer wrote:
> I should have said in my post that not all statistics are used for lying.
> Most professional statisticians are honest people, but there are those,
> mostly employed in marketing or politics, who know how to play games--and
> do.My son-in-law teaches advance placement calculus (in high school)and
> one year also added a class in statistics. I was very happy to find out
> that now students can get some exposure to statistics in high school.
> Allen
You should also be aware that approx. 87% of all statistics quoted in
general conversations and in newsgroups are made up on the spot!
--
Bob JONES
How come my family coat-of-arms ties at the back?
John Hill
And I always thought it was 82.67% :-)
It's surprising how many peole confuse precision and accuracy...
John.
--
Please reply to john at yclept dot wanadoo dot co dot uk.
--
Posted via a free Usenet account from http://www.teranews.com
James A. Doemer
rjo2...@bigpond.net.au, typed this:
Yes, I an 98.65% aware of that, +or- a 3% margin for error. :)
James A. Doemer
al...@nothere.net, typed this:
No, certainly not all statistics are used for lying. And not even all
statistics that are erroneous are products of conscious effort to lie.
Outside of open deceit, there are many that enter their data collection and
analysis with the results they want to see fixed in their minds. Whether
they mean it to or not, bias is almost inevitable under such conditions.
Lesley Robertson
news:13cglar...@corp.supernews.com...
>
> No, certainly not all statistics are used for lying. And not even all
> statistics that are erroneous are products of conscious effort to lie.
> Outside of open deceit, there are many that enter their data collection
> and analysis with the results they want to see fixed in their minds.
> Whether they mean it to or not, bias is almost inevitable under such
> conditions.
There's a (possibly mythical) student's report around that claimed that
33.33% of the mice died, 33.33% of the mice lived, and the third mouse
escaped....
Lesley Robertson
James A. Doemer
l.a.ro...@tnw.tudelft.nl, typed this:
LOL! I wonder what his margin for error was? :)
clifto
> No, certainly not all statistics are used for lying. And not even all
> statistics that are erroneous are products of conscious effort to lie.
> Outside of open deceit, there are many that enter their data collection and
> analysis with the results they want to see fixed in their minds. Whether
> they mean it to or not, bias is almost inevitable under such conditions.
Never underestimate the possibility of outright stupidity.
--
"You know the difference between cannibals and liberals?
Cannibals only eat their enemies."
-- Lyndon Baines Johnson
Graeme Wall
Gene Y.
and therefor it underscores that statistics are only as good as the user. <G>
--
Gene Y.
Researching Young, Zies, Harer & Cox.
http://h1.ripway.com/egptech/
Gene Y.
> James A. Doemer wrote:
>> No, certainly not all statistics are used for lying. And not even all
>> statistics that are erroneous are products of conscious effort to lie.
>> Outside of open deceit, there are many that enter their data collection and
>> analysis with the results they want to see fixed in their minds. Whether
>> they mean it to or not, bias is almost inevitable under such conditions.
>
> Never underestimate the possibility of outright stupidity.
>
"Never attribute to malice that which is more easily explained by stupidity!"
Robert Melson
The clock struck one but the rest escaped
with minor injuries.
Bob
--
Robert G. Melson | Rio Grande MicroSolutions | El Paso, Texas
-----
"People unfit for freedom---who cannot do much with it---are
hungry for power." ---Eric Hoffer
James A. Doemer
this:
> James A. Doemer wrote:
>> No, certainly not all statistics are used for lying. And not even
>> all statistics that are erroneous are products of conscious effort
>> to lie. Outside of open deceit, there are many that enter their data
>> collection and analysis with the results they want to see fixed in
>> their minds. Whether they mean it to or not, bias is almost
>> inevitable under such conditions.
>
> Never underestimate the possibility of outright stupidity.
LOL! That too.
Mary_...@tvo.org
My father's paternal line can be traced back to 1280
> and he has several strapping sons and grandsons carrying his name. His
> mother's paternal line can be traced back to 1485 and I know of at
> least a hundred living male descendents of that line today. I can also
> find numerous other male lines in my database going back 400 years.
Mary responds:
Modern DNA testing is showing that something like 10%-30% of children
are not sired by the fathers of record, and those results seem to be
consistent around the world. So....whole other wrinkle on the survival
of "lines" as defined by surnames...and lets not pretend human beings
were different creatures in the past.
Sir Creep
> > a point. See "How to Lie With Statistics", written by Darrel Huff in
> > 1954 and still in print.
> > Allen
>
> Statistics.
Sir Walter Scott: "I think it was DICKENS!"
Charles Dickens: "You BASTARD!"
Phyllis Nilsson
giggles, and laughs than I've had in a very long while. Thank you!
"raylopez99" <raylo...@yahoo.com> wrote in message
news:1185661369.4...@57g2000hsv.googlegroups.com...
> Here's a paradox that's a real puzzler.
>
>