Statistical analysis of month-dates in the Book of Mormon
Duwayne Anderson
You might want to check out the article at
Here is what the abstract says:
Abstract
The Book of Mormon claims to be a literal history of the ancient
Americans. However, a statistical analysis of month-dates in the Book
of Mormon strongly suggests that the dates are none random, and thus
probably fabricated. This document explains in detail the process of
arriving at the relevant probabilities, and the rational behind the
resulting conclusion that the Book of Mormon is a fake.
Read the entire article to see the details.
Duwayne Anderson
American Quarter Horse: The ultimate all-terrain vehicle.
Roy Stogner
> You might want to check out the article at
>
> http://64.4.16.250/cgi-bin/linkrd?_lang=EN&lah=d6673340c4196977ac52c36e8c36f42b&lat=1014429772&hm___action=http%3a%2f%2fwww%2elds%2dmormon%2ecom%2fnumbersinthebookofmormon%2eshtml
I can't seem to reach the link; the IP address appears to be a
hotmail.com server, but I get the message "There was an error displaying
this link" from the webserver when I try the URL.
A google search for phrases in the abstract comes up blank, too.
Could you find a more permanent URL or email me a copy?
---
Roy Stogner
Duwayne Anderson
My mistake. Sorry. You should be able to see it at
Buzzard
Andersons little analysis.
1. A sample of 8 is definitely on the small side. While small samples can
be used in statistics, the margin of error is quite large.
2. Along the same lines. Because three of the dates were essentially at the
start of the year, the sample is severely skewed. I know that Mr. Anderson
would see this as supporting his argument, but it illustrates how quickly a
small sample can be thrown off.
3. While 3.4 standard deviations is way out on the bell curve, a tiny
percentage of results will be found out there. Can you say with certainty
that this one isn't that one? Yes, I saw your analyis that the chance is one
in 2,000, and I took stat too long ago to run any different numbers, but I
suspect a more competent mathemetician could look at the numbers in some
other way and come up with other figures. Besides, one random sample in
2,000 *would* be located at this point on the curve, even according to you.
4. You assume that the events in the BOM had to happen on random dates.
That assumption may be incorrect. In the Old Testament, the story of David
and Bathseba begins by referring to a time of year when the kings went out
to make war. Why could not the Nephites have a similar situation for say,
sending a formal letter or going to war? I'm not making the claim, I'm
saying that you don't have enough information to make the statement that
dates in the BOM should be distributed randomly. You don't even know for
sure if the Nephites were using the Hebrew calendar after 500 years in the
Americas. What if their month's had 15 days instead of 29 or 30? If memory
serves, the Aztec calendar, which we know to have been in use in America at
some point after BOM times, had shorter months. Heck, (LDS cuss word there),
the fact that only 8 dates made it into Mormon's abridgement might very well
mean that those days had some particular significance to a current or future
audience. I'm speculating, I know, but frankly, your analysis contains some
speculation as well.
Feel free to flame away in return.
Lorin John
aka Buzzard
"Duwayne Anderson" <duwa...@hotmail.com> wrote in message
news:a42139e3.02022...@posting.google.com...
Roy Stogner
> What if their month's had 15 days instead of 29 or 30?
Then "why are all the dates listed (and a couple other instances where a
month but not a day is mentioned) during the first half of the year"
becomes a question.
---
Roy Stogner
Jim Allison
"However, since the Book of Mormon presents the month-dates
matter-of-factly, as being part of a literal history, we conclude that the
Book of Mormon itself is a fabrication. "
And if the days would have added up to 124, you would have concluded that
the Book of Mormon is not a fabrication?
I'm probably guessing that if the book of mormon mentioned 8 days of the
week, and half of those days of the week were the same day, you'd conclude
that the Book of Mormon was a fabrication too =)
Using your samples:
Arizona: 24th of Feb, 1863 Tuesday
Arkansas: 2nd of march, 1819 Tuesday
Colorado: 28th of Feb, 1861 Thursday
Idaho: 4th of March, 1863 Wednesday
Iowa: 12th of June, 1838 Tuesday
Kansas: 30th of May, 1854 Tuesday
Lousisiana: 26th of March, 1804 Monday
Minnesota: 3rd of march, 1849 Saturday
If you add the number of each month...you get 27...when you would expect it
to be 6.5 * 8 = 52
(Weird...no months in the last half of the year...doesn't mean that these
states are fabrications, it just means its weird/unusualy/unlikely).
Assuming that the events happen on random days of the week, you wouldn't
expect half of your samples to fall on the same day of the week either.
Duwayne Anderson
> I am not an expert in stats, but I see several obvious flaws in Mr.
> Andersons little analysis.
> 1. A sample of 8 is definitely on the small side.
The sample size was included in the calculation of probabilities. Did
you read the article? Do you understand how the probabilities were
calculated? There is no mathematical error.
> While small samples can
> be used in statistics, the margin of error is quite large.
So you assert. Calculate the "margin of error," and show your work.
As I said, the sample size was included in the calculation of the
probabilities. There is no mathematical error.
> 2. Along the same lines. Because three of the dates were essentially at the
> start of the year, the sample is severely skewed.
That's the point. The dates are not random. That is why I said the
distribution supports the conclusion that the Book of Mormon is a
fake.
> I know that Mr. Anderson
> would see this as supporting his argument,
Do you understand that you are using circular logic?
> but it illustrates how quickly a
> small sample can be thrown off.
The sample size was part of the calculation of probabilities. There
is no mathematical error.
> 3. While 3.4 standard deviations is way out on the bell curve, a tiny
> percentage of results will be found out there. Can you say with certainty
> that this one isn't that one?
You obviously did not read the article. This was stated clearly in
the conclusion section. Yes. There is a finite possibility that the
dates are from a real history. The probability is about 1/200. There
is a much higher probability that the dates are from a concocted
history. This is also consistent with the Book of Mormon's failure to
properly describe ancient American life.
> Yes, I saw your analyis that the chance is one
> in 2,000, and I took stat too long ago to run any different numbers, but I
> suspect a more competent mathemetician could look at the numbers in some
> other way and come up with other figures.
Well, instead of "suspecting" this and "suspecting" that, why don't
you show how to do it?
> Besides, one random sample in
> 2,000 *would* be located at this point on the curve, even according to you.
You clearly have no idea how probabilities are calculated.
> 4. You assume that the events in the BOM had to happen on random dates.
I listed the events used in the analysis. Are you saying they should
have their month-dates correlated? Why?
> That assumption may be incorrect.
Why?
> In the Old Testament, the story of David
> and Bathseba begins by referring to a time of year when the kings went out
> to make war. Why could not the Nephites have a similar situation for say,
> sending a formal letter or going to war?
Did you read the article? Did you pay any attention? Not all dates
are associated with war. There is nothing in the list of activities
associated with the dates to suggest in any way – even with the
most strained ad hoc arguments – that the events are at all
correlated through days of the month.
> I'm not making the claim,
Oh, no. You wouldn't stick your neck out like that, would you?
> I'm
> saying that you don't have enough information to make the statement that
> dates in the BOM should be distributed randomly.
Real dates are distributed randomly over the days of the month.
Didn't you see the examples? Even you admit you can provide no
rational reason why the events in table four would not be distributed
randomly over the days of the month.
> You don't even know for
> sure if the Nephites were using the Hebrew calendar after 500 years in the
> Americas.
They followed the law of Moses.
> What if their month's had 15 days instead of 29 or 30?
Why would transplant Hebrews change their calendar? How would they
continue to live the law of Moses if they did?
> If memory
> serves, the Aztec calendar, which we know to have been in use in America at
> some point after BOM times, had shorter months.
Which is another good reason to understand the Book of Mormon as a
fraud. There are no ancient American civilizations that used the
Hebrew calendar, yet the Book of Mormon says the ancient Americans
were Hebrews who followed the law of Moses.
> Heck, (LDS cuss word there),
> the fact that only 8 dates made it into Mormon's abridgement might very well
> mean that those days had some particular significance to a current or future
> audience.
You really need to read the article and understand the mathematics
before making foolish statements like this. The probabilities were
based on the sample size. Your intuition notwithstanding, the
probability of 8 days of the month being distributed like those in the
Book of Mormon is roughly 1/2000
> I'm speculating, I know,
Oh, you are doing worse than that. You are also demonstrating a
deplorable lack of understanding of statistics.
> but frankly, your analysis contains some
> speculation as well.
Which are all carefully called out and identified in the article.
You'd know that, of course, if you had bothered to actually read it.
> Feel free to flame away in return.
No flames necessary. There was nothing of substance in what you said.
Duwayne Anderson
> Duwayne Anderson <duwa...@hotmail.com> wrote in message
> news:a42139e3.02022...@posting.google.com...
>
> "However, since the Book of Mormon presents the month-dates
> matter-of-factly, as being part of a literal history, we conclude that the
> Book of Mormon itself is a fabrication. "
>
> And if the days would have added up to 124, you would have concluded that
> the Book of Mormon is not a fabrication?
The probability of finding a certain sequence was evaluated by looking
at the total probability of finding any sequence a certain distance
from the mean. This is a common technique. If the days of the months
listed in the Book of Mormon were in a region relatively close to the
mean, one would be unable to make a strong statement either way (for
or against its authenticity). But the sum of the days of the months
in the Book of Mormon are far from the mean. The probability of a
random sum of 8 days of the month being where the Book of Mormon dates
are is roughly 1/2000.
As the article says, this does not prove the dates are non-random. It
simply says it is very unlikely (probability = about 1/2000) that the
dates are random. If we assume the more likely conclusion that the
dates are not random, that is consistent with the conclusion that the
Book of Mormon is a fake. This conclusion is also consistent with the
fact that the Book of Mormon so poorly describes the conditions in
ancient America.
> I'm probably guessing that if the book of mormon mentioned 8 days of the
> week, and half of those days of the week were the same day, you'd conclude
> that the Book of Mormon was a fabrication too =)
The study I did was for days of the month. Not days of the week.
> Using your samples:
> Arizona: 24th of Feb, 1863 Tuesday
> Arkansas: 2nd of march, 1819 Tuesday
> Colorado: 28th of Feb, 1861 Thursday
> Idaho: 4th of March, 1863 Wednesday
> Iowa: 12th of June, 1838 Tuesday
> Kansas: 30th of May, 1854 Tuesday
> Lousisiana: 26th of March, 1804 Monday
> Minnesota: 3rd of march, 1849 Saturday
>
> If you add the number of each month...you get 27
Correct. Add the numbers of the days of the month and you get 27.
> ...when you would expect it
> to be 6.5 * 8 = 52
Very good, Jim. The expected value (mean) of the distribution is 52.
> (Weird...no months in the last half of the year...doesn't mean that these
> states are fabrications, it just means its weird/unusualy/unlikely).
Well, you forgot to calculate the standard deviation. Let's finish
the work you started and calculate the probability. Shall we?
The standard deviation for summing 8 months is 9.764. This means
that, for the example you picked above, the sum of the days of the
month is (52-27)/9.764 = 2.56 standard deviations from the mean. The
probability of a sum being this far from the mean (using the months
with 8 tries, as in your example above) is about 1/100 (determined
using Monte Carlo analysis with 1 million samples). That's a pretty
small number, but not as small as the figure of 1/2000 for the Book of
Mormon.
But you do make a good point. Just because the probability is small
does not mean it *cannot* happen. In fact, if you were to look at a
large enough sample of random dates, you'd expect to find some that
have a probability of 1/2000.
But the objective for rational people is not one of finding any
possibility for hope, but of logically examining the probabilities and
all the data. So, while the dates in the Book of Mormon could be real
dates, it's more likely that they are fake dates.
> Assuming that the events happen on random days of the week, you wouldn't
> expect half of your samples to fall on the same day of the week either.
Of course, the article makes clear the nature of the conclusions.
Since folks are apparently having difficulty reading the whole thing,
I've reprinted the conclusion section below:
"Throughout all of this analysis, we should be careful to remember
that conclusions based on statistical inference do not constitute
proof. Indeed, given a sufficiently large sampling of statistical
data from real historical events, there is a finite probability that
some will have the same distribution found in the Book of Mormon.
These, however, will be very rare; it is far more likely that the
dates in the Book of Mormon were concocted as part of a fabricated
story. This conclusion is buttressed by other important information,
such as the Book of Mormon’s almost universal failure to
properly describe the realities of ancient-American life."
So, you see, there is a finite probability that the month-dates in the
Book of Mormon are randomly distributed. The probability is about
1/2000. Does this prove the Book of Mormon is a fake? No. Does it
prove the dates are concocted? No. Does it say it's very unlikely
that the dates are true, random, historical dates? Yes. Is this all
the evidence we have for consideration? No. We also have
archeological evidence from ancient America. Is this other data
consistent with the conclusion that the dates are fabricated? Yes.
Jackie Chan
What if their month's had 15 days instead of 29 or 30?
JC comments ===
the chances improve dramatically to about 1 in 12
I though the ancient Americans used a 20-day calendar, however, which
would yield a probability of only 1/100 for the Book of Mormon 8 point
days of the month data set
Duwayne Anderson
There are a couple of things wrong with that. First, the Book of
Mormon needs to be evaluated according to its specific claims. As
I've pointed out several times, now, the Book of Mormon claims the
ancient Americans were Hebrews, and as Hebrews they practiced the Law
of Moses. An important part of the Law of Moses are the feasts and
Holy Days, which are an integral part of the Hebrew calendar (see the
reference in the article). So, it involves somewhat circular logic to
claim that a 20-day month should be used because the ancient Americans
used one, when the 20-day month of the Maya has no similarity
whatsoever to the Hebrew calendar which is implied by circumstances
described in the Book of Mormon. Instead, the fact that the ancient
Americans did not use a Hebrew calendar is strong evidence against the
Book of Mormon.
That said, let's look at what happens if we use the 20-day calendar.
For 20-day months the mean is 10.5*8 = 84 and the standard deviation
is root 8*5.766 = 16.31. Since the sum of days for the Book of Mormon
month dates is 41, this sum lies about (84-41)/16.31 = 2.636 standard
deviations from the mean. The probability of being this far, or
further, from the mean is about 1/110. Quite a bit better than the
1/2000 calculated for the situation using the Hebrew calendar.
But with the assumption of 20 months the sum of the month is now also
very improbable (it isn't if you assume, as I did, 13 months with 30
days each). For 20-day months there are 18 months in a year. The
mean is 9.5*8 = 76 and the standard deviation is (root 8)*5.993 =
14.674. The sum of the months in the Book of Mormon dates is 40 so
this is (76-40)/14.674 = 2.453 standard deviations from the mean. The
probability of a sum of months being this far or further from the mean
is also about 1/110.
In the original analysis I did not include the distribution of months
because it was not dramatically removed from the mean. Doing so would
have changed the probability from 1/2000 to about 1/3000. With the 18
20-day months, the distribution of months is now *also* highly
unlikely. So one must not only rationalize away the probability of
1/110 for the days, but the equally improbably distribtion of months.
So trying to use the 20-day month does not really help things much.
The distribution is still very unlikely, and (more importantly) the
assumption violates one of the key claims of the Book of Mormon --
namely that the ancient Americans were transplanted Hebrews who
supposedly lived the Law of Moses.
Jackie Chan
Correct
> But with the assumption of 20 months the sum of the month is now also
> very improbable (it isn't if you assume, as I did, 13 months with 30
> days each). For 20-day months there are 18 months in a year. The
> mean is 9.5*8 = 76 and the standard deviation is (root 8)*5.993 =
> 14.674. The sum of the months in the Book of Mormon dates is 40 so
> this is (76-40)/14.674 = 2.453 standard deviations from the mean. The
> probability of a sum of months being this far or further from the mean
> is also about 1/110.
you're very sharp, Duwayne, I buzzed right by this
I found two other months referenced in the BoM, without any day
attached: the second month, in (Alma 56:27), and the sixth (Third Nephi
4:7, 17) - if you add them to the mix, you then get a sum of 51 for 10
data points, and the probability is less than 1/300, where (105-51)/18 =
3 SD's out from the population Mean
> In the original analysis I did not include the distribution of months
> because it was not dramatically removed from the mean. Doing so would
> have changed the probability from 1/2000 to about 1/3000. With the 18
> 20-day months, the distribution of months is now *also* highly
> unlikely. So one must not only rationalize away the probability of
> 1/110 for the days, but the equally improbably distribtion of months.
got it - the problem with decreasing the number of days in a month
bites the apologist in the butt when he has to increase the number of
months per year
> So trying to use the 20-day month does not really help things much.
> The distribution is still very unlikely, and (more importantly) the
> assumption violates one of the key claims of the Book of Mormon --
> namely that the ancient Americans were transplanted Hebrews who
> supposedly lived the Law of Moses.
yes - toot! toot! - the 20 month calendar must have been invented
later, by accursed loin-clothed dark Lamanite savages
Jim Allison
> simply says it is very unlikely (probability = about 1/2000) that the
> dates are random. If we assume the more likely conclusion that the
> dates are not random, that is consistent with the conclusion that the
> Book of Mormon is a fake.
But it is by no means the only conclusion that can reached. So you're
basically end up where you started from.
>
> Well, you forgot to calculate the standard deviation. Let's finish
> the work you started and calculate the probability. Shall we?
pssst, this is your work. You should have finished it to begin with silly!
> The standard deviation for summing 8 months is 9.764. This means
> that, for the example you picked above, the sum of the days of the
> month is (52-27)/9.764 = 2.56 standard deviations from the mean. The
> probability of a sum being this far from the mean (using the months
> with 8 tries, as in your example above) is about 1/100 (determined
> using Monte Carlo analysis with 1 million samples). That's a pretty
> small number, but not as small as the figure of 1/2000 for the Book of
> Mormon.
So at what probability can you conclude that it must a fabrication? 1/100?
1/1000? 1/2000?
> But you do make a good point. Just because the probability is small
> does not mean it *cannot* happen. In fact, if you were to look at a
> large enough sample of random dates, you'd expect to find some that
> have a probability of 1/2000.
>
> But the objective for rational people is not one of finding any
> possibility for hope, but of logically examining the probabilities and
> all the data. So, while the dates in the Book of Mormon could be real
> dates, it's more likely that they are fake dates.
Do you believe that the dates you used for the States being admitted into
the union are fake dates because they lie far away from the mean?
> These, however, will be very rare; it is far more likely that the
> dates in the Book of Mormon were concocted as part of a fabricated
> story.
As shown above, real stories have dates that stray far from the mean as
well. At what point can you conclude that the dates in a story come from
reality or a fabrication? 1/100? 1/300? 1/2000?
> So, you see, there is a finite probability that the month-dates in the
> Book of Mormon are randomly distributed. The probability is about
> 1/2000. Does this prove the Book of Mormon is a fake? No. Does it
> prove the dates are concocted? No. Does it say it's very unlikely
> that the dates are true, random, historical dates? Yes. Is this all
> the evidence we have for consideration? No. We also have
> archeological evidence from ancient America. Is this other data
> consistent with the conclusion that the dates are fabricated? Yes.
So close, yet so far...
Jim Allison
of months.
>
> got it - the problem with decreasing the number of days in a month
> bites the apologist in the butt when he has to increase the number of
> months per year
What about throwing in what decade and what year they happened in?
Doesn't it seem like a conspiracy that of all the day/month combinations
mentioned in the BoM, all but one of them appear in the book of Alma. What
is the probability of that happening?
=P
atwood
news:a42139e3.02022...@posting.google.com...
>
> Book of Mormon are randomly distributed. The probability is about
> 1/2000. Does this prove the Book of Mormon is a fake? No. Does it
> prove the dates are concocted? No. Does it say it's very unlikely
> that the dates are true, random, historical dates? Yes. Is this all
> the evidence we have for consideration? No. We also have
> archeological evidence from ancient America. Is this other data
> consistent with the conclusion that the dates are fabricated? Yes.
You've used statistical analysis carefully and appropriately. As you say, it
doesn't "prove" anything here, but rather shows the improbability that the
BOM has any truth. When your analysis is combined with other evidence, the
result is devastating. Now, we all ought to understand that none of this
will make any differerence to believing Mormon cultists. There is already
massive evidence that everything about Mormon history, doctrine and
scripture, such as it is, is fraudulent. That said, your work is great, and
thanks for doing it.
Jackie Chan
compared to what?
it is clear, however - given the nature of counting time, and the set
of dates actually found in the Book of Mormon - that in all probability
the author of the Book of Mormon has about twelve months per annum in
mind, not 20
Duwayne Anderson
> Duwayne Anderson <duwa...@hotmail.com> wrote in message
> news:a42139e3.02022...@posting.google.com...
> > As the article says, this does not prove the dates are non-random. It
> > simply says it is very unlikely (probability = about 1/2000) that the
> > dates are random. If we assume the more likely conclusion that the
> > dates are not random, that is consistent with the conclusion that the
> > Book of Mormon is a fake.
>
> But it is by no means the only conclusion that can reached. So you're
> basically end up where you started from.
What part about statistics and probability don't you understand, Jim?
The nature of probability means that that some things are more likely
than other things. It's not about what is possible and what is
impossible, but what is more likely than other stuff.
>
> >
> > Well, you forgot to calculate the standard deviation. Let's finish
> > the work you started and calculate the probability. Shall we?
>
> pssst, this is your work.
It was your example. I finished it and showed that your example does
not at all invalidate the conclusion that it is more likely that the
Book of Mormon dates are fabricated than that they are from a real
history.
> You should have finished it to begin with silly!
I did. In the article to which you responded. Would you like me to
repeat the calculations for you?
>
> > The standard deviation for summing 8 months is 9.764. This means
> > that, for the example you picked above, the sum of the days of the
> > month is (52-27)/9.764 = 2.56 standard deviations from the mean. The
> > probability of a sum being this far from the mean (using the months
> > with 8 tries, as in your example above) is about 1/100 (determined
> > using Monte Carlo analysis with 1 million samples). That's a pretty
> > small number, but not as small as the figure of 1/2000 for the Book of
> > Mormon.
>
> So at what probability can you conclude that it must a fabrication? 1/100?
> 1/1000? 1/2000?
I see you are still having difficulty with the concept of probability.
At a probability of 1/100 there is a chance of 1/100 that it is is a
real history. At a probability of 1/1000 there is a chance of 1/1000
that it is a real history. At a probability of 1/2000 there is a
chance of 1/2000 that it is a real history.
The *REAL* question is "at what probability -- combined with all the
other evidence -- will Mormons finally admit that the Book of Mormon
is a fake." The answer to that question is "never."
>
>
> > But you do make a good point. Just because the probability is small
> > does not mean it *cannot* happen. In fact, if you were to look at a
> > large enough sample of random dates, you'd expect to find some that
> > have a probability of 1/2000.
> >
> > But the objective for rational people is not one of finding any
> > possibility for hope, but of logically examining the probabilities and
> > all the data. So, while the dates in the Book of Mormon could be real
> > dates, it's more likely that they are fake dates.
>
> Do you believe that the dates you used for the States being admitted into
> the union are fake dates because they lie far away from the mean?
Okay, Jim. Let's try this one more time. I see the concept of
probability is a difficult one for you. I'll repeat the last
paragraph in the Internet article. You might want to consider reading
the entire article again:
"Throughout all of this analysis, we should be careful to remember
that conclusions based on statistical inference do not constitute
proof. Indeed, given a sufficiently large sampling of statistical
data from real historical events, there is a finite probability that
some will have the same distribution found in the Book of Mormon.
These, however, will be very rare; it is far more likely that the
dates in the Book of Mormon were concocted as part of a fabricated
story. This conclusion is buttressed by other important information,
such as the Book of Mormon’s almost universal failure to
properly describe the realities of ancient-American life."
>
> > These, however, will be very rare; it is far more likely that the
> > dates in the Book of Mormon were concocted as part of a fabricated
> > story.
>
> As shown above, real stories have dates that stray far from the mean as
> well. At what point can you conclude that the dates in a story come from
> reality or a fabrication? 1/100? 1/300? 1/2000?
>
> > So, you see, there is a finite probability that the month-dates in the
> > Book of Mormon are randomly distributed. The probability is about
> > 1/2000. Does this prove the Book of Mormon is a fake? No. Does it
> > prove the dates are concocted? No. Does it say it's very unlikely
> > that the dates are true, random, historical dates? Yes. Is this all
> > the evidence we have for consideration? No. We also have
> > archeological evidence from ancient America. Is this other data
> > consistent with the conclusion that the dates are fabricated? Yes.
>
> So close, yet so far...
No, the Book of Mormon is not even close.
Duwayne Anderson
> Jackie Chan <nos...@donotreply.net> wrote in message
> news:3C79AE5B...@donotreply.net...
> of months.
> >
> > got it - the problem with decreasing the number of days in a month
> > bites the apologist in the butt when he has to increase the number of
> > months per year
>
> What about throwing in what decade and what year they happened in?
Decades and years are sequential, so they are correlated (the 21st
year, for example, always follows the 20th). If you had a long enough
history you might look at decades mentioned in each century. If only
the first and second decade, for example, were mentioned in 8
different centuries, you'd have something that's improbable.
The other point I'd like to make is that the evidence found in
distribution of dates is asymmetric. It's easy enough to fake a
random distribution. Just get a table of random numbers, or toss a
dice. So the existence of a random set of dates does not provide
compelling evidence for a legitimate history.
On the other hand, if the distribution is sufficiently unlikely the
evidence weighs heavily in favor of a fabricated history. This is
buttressed, in the case of the Book of Mormon, by a nearly universal
failure to properly describe the ancient Americans that supposedly
wrote the Book.
>
> Doesn't it seem like a conspiracy that of all the day/month combinations
> mentioned in the BoM, all but one of them appear in the book of Alma.
Are you suggesting a "conspiracy" in the writing of the Book of
Mormon? In the analysis I performed, I used *all* the month dates
mentioned in the book. I did not pick and choose. That said, you
should be aware that Alma is the most history-specific book in the
Book of Mormon. So finding all the month-dates there is not too
surprising. But if you have another explanation, why not offer it?
> What
> is the probability of that happening?
Well, that's a good question, Jim. Are you suggesting the Book of
Mormon is a fabricated history because of it?
Roy Stogner
> I see you are still having difficulty with the concept of probability.
> At a probability of 1/100 there is a chance of 1/100 that it is is a
> real history. At a probability of 1/1000 there is a chance of 1/1000
> that it is a real history. At a probability of 1/2000 there is a chance
> of 1/2000 that it is a real history.
I don't think this is quite correct. The statement that "1/2000 of real
histories have this property" is not the same as the statement that "there
is a 1/2000 chance that it is a real history". Arguments such as yours
can strongly weight the probability that an observation is true, but
cannot set that probability from scratch.
Suppose I draw from a deck of cards, half of which have "true" written on
the back, and half of which have "false". If you tell me that only one in
a hundred "true" cards are blue on the front, and that one in two of the
"false" cards are blue, then there is an easily calculable (under 2%)
probability that my card will say "false" when I flip it over. Note,
though, that to calculate this probability you need to know how many of
the "false" cards fit the improbable condition, as well as what the
initial "true"/"false" probabilities were in the first place! The final
calculation depends on these conditions.
Example: if I begin with the belief that there is a 1/4 chance that the
Book of Mormon is true and the assumption that there is a 1/2 chance that
a man-made lists of dates will be as statistically deviant as the Book of
Mormon dates, then I am left with a weight of (1/4*1/2000) that it is true
and a weight of (3/4 * 1/2) that is false, giving (1/8000)/(1/8000+3/8),
i.e. approximately a 1/3000 chance that it is true.
If I begin with the belief that there is a 3/4 chance that the Book of
Mormon is true, then adding your argument leaves me with a
(3/8000)/(3/8000+1/8) =~ 3/1000 chance that it is true. Similarly, "99%
sure" turns to "9% sure" (would that last pattern be a meta-coincidence?).
If you really want to go all the way and help people change their
estimations of probability, you need to find some estimate of what the
distribution of artificial groups of dates is, so you can measure the Book
of Mormon's difference from that standard as well. I just pulled the
"1/2" number out of thin air.
> The *REAL* question is "at what probability -- combined with all the
> other evidence -- will Mormons finally admit that the Book of Mormon is
> a fake." The answer to that question is "never."
Unfortunately, if you begin with the belief that there is a 100% chance
that the Book of Mormon is true, this sort of adjustment doesn't change
that probability at all, it just leads to the conclusion that "even 1/2000
improbabilities happen".
Don't drop the calculations or anything, though. Those people merely at
the "99.9%" level may still find them interesting.
---
Roy Stogner
Duwayne Anderson
> On Sun, 24 Feb 2002 13:40:09 -0600, Duwayne Anderson wrote:
>
> > I see you are still having difficulty with the concept of probability.
> > At a probability of 1/100 there is a chance of 1/100 that it is is a
> > real history. At a probability of 1/1000 there is a chance of 1/1000
> > that it is a real history. At a probability of 1/2000 there is a chance
> > of 1/2000 that it is a real history.
>
> I don't think this is quite correct. The statement that "1/2000 of real
> histories have this property" is not the same as the statement that "there
> is a 1/2000 chance that it is a real history".
I think I have a tendency to agree with you, but I'll have to think
about this some more.
> Arguments such as yours
> can strongly weight the probability that an observation is true, but
> cannot set that probability from scratch.
See above.
>
> Suppose I draw from a deck of cards, half of which have "true" written on
> the back, and half of which have "false". If you tell me that only one in
> a hundred "true" cards are blue on the front, and that one in two of the
> "false" cards are blue, then there is an easily calculable (under 2%)
> probability that my card will say "false" when I flip it over. Note,
> though, that to calculate this probability you need to know how many of
> the "false" cards fit the improbable condition, as well as what the
> initial "true"/"false" probabilities were in the first place! The final
> calculation depends on these conditions.
Well, in the test that I applied I calculated the probability that a
random selection of dates would fall a certain distance from the mean.
The conclusion I drew was that the sum of dates has a low probability
of being drawn from a random distribution. The inference is that, if
the distribution of dates is unlikely to have been drawn from a random
distribution, it is more likely to have been part of a concocted
story.
I assume you agree with that?
>
> Example: if I begin with the belief that there is a 1/4 chance that the
> Book of Mormon is true and the assumption that there is a 1/2 chance that
> a man-made lists of dates will be as statistically deviant as the Book of
> Mormon dates, then I am left with a weight of (1/4*1/2000) that it is true
> and a weight of (3/4 * 1/2) that is false, giving (1/8000)/(1/8000+3/8),
> i.e. approximately a 1/3000 chance that it is true.
>
Note that my analysis did not start from any assumptions about the
Book of Mormon. The assumption was that the month-dates of a real
history are randomly distributed over the days of the month. But, as
you can see in the analysis, the month-dates in the Book of Mormon are
many standard deviations from the expected value if the dates are
randomly distributed.
The conclusion is that the month-dates in the Book of Mormon are not
from a real history because the sum of the dates is so far from the
expected value one gets with a real (that is, randomly distributed)
set of dates.
> If I begin with the belief that there is a 3/4 chance that the Book of
> Mormon is true, then adding your argument leaves me with a
> (3/8000)/(3/8000+1/8) =~ 3/1000 chance that it is true. Similarly, "99%
> sure" turns to "9% sure" (would that last pattern be a meta-coincidence?).
>
> If you really want to go all the way and help people change their
> estimations of probability, you need to find some estimate of what the
> distribution of artificial groups of dates is, so you can measure the Book
> of Mormon's difference from that standard as well. I just pulled the
> "1/2" number out of thin air.
Well, this I definitely disagree with. The test was for the
probability that the dates were chosen at random -- the way
month-dates in real histories are. So there is absolutely no need to
make any estimate of what the distribution of artificial group dates
is. The test is sufficient to state that the sum of the month dates
is very unlikely given a random distribution. And, if the
distribution was not random, it was probably not from a real history
(because the month dates in a real history are random).
>
> > The *REAL* question is "at what probability -- combined with all the
> > other evidence -- will Mormons finally admit that the Book of Mormon is
> > a fake." The answer to that question is "never."
>
> Unfortunately, if you begin with the belief that there is a 100% chance
> that the Book of Mormon is true, this sort of adjustment doesn't change
> that probability at all, it just leads to the conclusion that "even 1/2000
> improbabilities happen".
True. Things with probability of 1/2000 do happen. But the question
for rational people is about what's most probable. In the case of the
Book of Mormon, we have a book that fails to properly describe ancient
American at all, and it has a distribution of dates that is very
unlikely for a real (random) history. Superstitious people can always
ignore those facts and hold on to hope. No matter the odds. But
that's not what rational people do.
>
> Don't drop the calculations or anything, though. Those people merely at
> the "99.9%" level may still find them interesting.
> ---
> Roy Stogner
Thanks for your comments.
Jim Allison
> > mentioned in the BoM, all but one of them appear in the book of Alma.
>
> Are you suggesting a "conspiracy" in the writing of the Book of
> Mormon? In the analysis I performed, I used *all* the month dates
> mentioned in the book. I did not pick and choose. That said, you
> should be aware that Alma is the most history-specific book in the
> Book of Mormon. So finding all the month-dates there is not too
> surprising. But if you have another explanation, why not offer it?
Wow, you surprise me Duwayne, seriously =)
Jim Allison
Duwayne Anderson <duwa...@hotmail.com> wrote in message
news:a42139e3.02022...@posting.google.com...
> "Jim Allison" <alli...@home.com> wrote in message
news:<GL5e8.111756$Zc.32...@news2.rdc1.mi.home.com>...
> > Duwayne Anderson <duwa...@hotmail.com> wrote in message
> > news:a42139e3.02022...@posting.google.com...
> > > As the article says, this does not prove the dates are non-random. It
> > > simply says it is very unlikely (probability = about 1/2000) that the
> > > dates are random. If we assume the more likely conclusion that the
> > > dates are not random, that is consistent with the conclusion that the
> > > Book of Mormon is a fake.
> >
> > But it is by no means the only conclusion that can reached. So you're
> > basically end up where you started from.
>
> What part about statistics and probability don't you understand, Jim?
> The nature of probability means that that some things are more likely
> than other things. It's not about what is possible and what is
> impossible, but what is more likely than other stuff.
Exactly, you seem to not understand that when you make it seem that the only
conclusion is that the Book of Mormon is a fake because the probably of the
dates being random is 1/2000.
> > > Well, you forgot to calculate the standard deviation. Let's finish
> > > the work you started and calculate the probability. Shall we?
> >
> > pssst, this is your work.
>
> It was your example. I finished it and showed that your example does
> not at all invalidate the conclusion that it is more likely that the
> Book of Mormon dates are fabricated than that they are from a real
> history.
Are you under the impression that dates which are unlikely to be random
_must_ be fake?
That is the error which I think you are making.
> > You should have finished it to begin with silly!
>
> I did. In the article to which you responded. Would you like me to
> repeat the calculations for you?
You should have finished it on your website Duwayne =P
Incomplete work makes it look shobby.
> I see you are still having difficulty with the concept of probability.
> At a probability of 1/100 there is a chance of 1/100 that it is is a
> real history. At a probability of 1/1000 there is a chance of 1/1000
> that it is a real history. At a probability of 1/2000 there is a
> chance of 1/2000 that it is a real history.
Duwayne, please try to follow...
You show that there is a 1/2000 chance that the dates in the BoM are
completely random. Because of the low probably, you conclude that the BoM
is a fake. The dates for the states have a roughly 1/100 chance of being
truly random, but you conclude that those dates are real.
The probably of dates being randomly generated has nothing to do with
whether the dates are real or fake.
You had a really interesting website, but when you started making
unwarranted conclusions it made the rest of it look bad.
Its foolish to conclude that if something is unlikely to be randomly
generated that it cannot be real, and that is the conclusion which you
reached.
Jim Allison
Duwayne Anderson <duwa...@hotmail.com> wrote in message
news:a42139e3.02022...@posting.google.com...
>
> evidence weighs heavily in favor of a fabricated history. This is
> buttressed, in the case of the Book of Mormon, by a nearly universal
> failure to properly describe the ancient Americans that supposedly
> wrote the Book.
How about taking a bunch of other stories (that are known to be made-up) and
finding out the mean value of fabricated dates, and than comparing it to the
BoM's dates.
1/2000 chance of being randomly generated does mean a 1999/2000 chance of
being made up.
Joel Rees
duwa...@hotmail.com (Duwayne Anderson) wrote:
> All:
>
> You might want to check out the article at
>
[snipped URL]
>
> Here is what the abstract says:
>
> Abstract
> The Book of Mormon claims to be a literal history of the ancient
> Americans. However, a statistical analysis of month-dates in the Book
> of Mormon strongly suggests that the dates are none random, and thus
> probably fabricated. This document explains in detail the process of
> arriving at the relevant probabilities, and the rational behind the
> resulting conclusion that the Book of Mormon is a fake.
Two lists of dates, Duwayne:
(I'll tell you later which I cribbed from where and which I cheated at.)
List 1:
6/27/1850
4/19/1850
8/14/1850
4/13/1834
10/29/1850
5/20/1832
1/8/1799
10/28/1830
1/12/1835
2/14/1838
9/6/1846
List 2:
2 Oct 1788
13 Mar 1800
14 May 1800
16 Aug 1807
16 Dec 1824
15 Jun 1827
1 Dec 1827
29 Jul 1829
20 May 1830
6 Feb 1848
17 Jan 1852
Have fun.
Joel
cdowis
One example,
3 Nephi 8
[5] And it came to pass in the thirty and fourth year, in the first
month, on the fourth day of the month, there arose a great storm, such
an one as never had been known in all the land.
"Buzzard" <Buzz...@prodigy.net> wrote in message news:<HWHd8.10108$Lv1.146...@newssvr17.news.prodigy.com>...
snip
Duwayne Anderson
Joel, I see you didn't read the article, or it's conclusions. Let me
help you out:
Throughout all of this analysis we should be careful to remember that
conclusions based on statistical inference do not constitute proof.
Indeed, given a sufficiently large sampling of statistical data from
real historical events, there is a finite probability that some will
have the same distribution found in the Book of Mormon. These,
however, will be very rare; it is far more likely that the dates in
the Book of Mormon were concocted as part of a fabricated story. This
conclusion is buttressed by other important information, such as the
Book of Mormon’s almost universal failure to properly describe
the realities of ancient-American life.
Duwayne Anderson
Jim, you didn't answer or even acknowledge any of the questions.
Duwayne Anderson
> Duwayne Anderson <duwa...@hotmail.com> wrote in message
> news:a42139e3.02022...@posting.google.com...
> >
> > On the other hand, if the distribution is sufficiently unlikely the
> > evidence weighs heavily in favor of a fabricated history. This is
> > buttressed, in the case of the Book of Mormon, by a nearly universal
> > failure to properly describe the ancient Americans that supposedly
> > wrote the Book.
>
> How about taking a bunch of other stories (that are known to be made-up) and
> finding out the mean value of fabricated dates, and than comparing it to the
> BoM's dates.
Had you read the article at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
you would know that no claim was made regarding someone's ability to
fabricate a list of dates that looks random. Indeed, it would be an
easy thing to do. In fact, I've already described in this thread how
to do it, by using a table of random numbers, or something as simple
as a set of dice.
So finding a set of random dates in a book of questionable origin
provides no information one way or the other regarding the book's
authenticity.
See, that's the interesting thing about statistical tests. They are
often very asymmetrical in the level of proof they provide. If the
list of dates is very unlikely for a random distribution (as the dates
in the Book of Mormon are) then we can use the following line of logic
(with a little summary at the start):
1) The month-days in the Book of Mormon are very unlikely given a
random distribution (the probability is less than about 1/2000).
2) The month-days in real history are random, or unbiased (that is,
they don't tend to favor any particular day of the month).
3) The conclusion that the Book of Mormon is a fake is consistent
with the distribution of month-days found in it.
4) The conclusion that the Book of Mormon is a fake is consistent
with the lack of archeological evidence for the Book of Mormon.
5) Both the distribution of month-days and the lack of archeological
evidence are inconsistent with the Book of Mormon's claim to be a real
history.
>
> 1/2000 chance of being randomly generated does mean a 1999/2000 chance of
> being made up.
The article did not say "1/2000 chance of being randomly generated.."
It said the probability of finding the sort of month-day distribution
found in the Book of Mormon, in a real (i.e., random or non-biased)
set of dates is less than 1/2000.
This is consistent with the hypothesis that the Book of Mormon is a
fake, particularly when taken with all other data, such as the Book of
Mormons failure to properly describ Ancient America.
Duwayne Anderson
<snip>
> Exactly, you seem to not understand that when you make it seem that the only
> conclusion is that the Book of Mormon is a fake because the probably of the
> dates being random is 1/2000.
This staetment by Jim is false. The article clearly states that other
data should be considered. From the article:
"Throughout all of this analysis we should be careful to remember that
conclusions based on statistical inference do not constitute proof.
Indeed, given a sufficiently large sampling of statistical data from
real historical events, there is a finite probability that some will
have the same distribution found in the Book of Mormon. These,
however, will be very rare; it is far more likely that the dates in
the Book of Mormon were concocted as part of a fabricated story. This
conclusion is buttressed by other important information, such as the
Book of Mormon’s almost universal failure to properly describe
the realities of ancient-American life."
<snip>
> Are you under the impression that dates which are unlikely to be random
> _must_ be fake?
Once again, for Jim (who seems to enjoy commenting on articles he has
not read):
"Throughout all of this analysis we should be careful to remember that
conclusions based on statistical inference do not constitute proof.
Indeed, given a sufficiently large sampling of statistical data from
real historical events, there is a finite probability that some will
have the same distribution found in the Book of Mormon. These,
however, will be very rare; it is far more likely that the dates in
the Book of Mormon were concocted as part of a fabricated story. This
conclusion is buttressed by other important information, such as the
Book of Mormon’s almost universal failure to properly describe
the realities of ancient-American life."
> That is the error which I think you are making.
No, Jim. The error is yours in commenting on an article you have not
read, and don't understand.
<snip>
> Duwayne, please try to follow...
>
> You show that there is a 1/2000 chance that the dates in the BoM are
> completely random.
Wrong, Jim. What I showed mathematically is that the probability of
the dates in the Book of Mormon arising from a random set (as in real
histories) is less than 1/2000.
> Because of the low probably, you conclude that the BoM
> is a fake.
Wrong again, Jim. Let's quote the conclusion of the article you
didn't read one more time:
Throughout all of this analysis we should be careful to remember that
conclusions based on statistical inference do not constitute proof.
Indeed, given a sufficiently large sampling of statistical data from
real historical events, there is a finite probability that some will
have the same distribution found in the Book of Mormon. These,
however, will be very rare; it is far more likely that the dates in
the Book of Mormon were concocted as part of a fabricated story. This
conclusion is buttressed by other important information, such as the
Book of Mormon’s almost universal failure to properly describe
the realities of ancient-American life.
> The dates for the states have a roughly 1/100 chance of being
> truly random, but you conclude that those dates are real.
There are two big differences between the dates for the states and
those found in the Book of Mormon:
1) One has a probability of 1/100 associated with it, and the other
has a probability of 1/2000 associated with it.
2) The dates for the States have other corroborating evidence
supporting them. The Book of Mormon, on the other hand, is already
virtually disproven by the fact that it's description of ancient
America is totally at odds with what we know from science.
> The probably of dates being randomly generated has nothing to do with
> whether the dates are real or fake.
Real historical dates don't favor one day of the month over any other.
The Book of Mormon month-days do. The probability of a real history
having the month-day distribution found in the Book of Mormon is less
than 1/2000. Again, from the conclusion of the article Jim did not
read:
Throughout all of this analysis we should be careful to remember that
conclusions based on statistical inference do not constitute proof.
Indeed, given a sufficiently large sampling of statistical data from
real historical events, there is a finite probability that some will
have the same distribution found in the Book of Mormon. These,
however, will be very rare; it is far more likely that the dates in
the Book of Mormon were concocted as part of a fabricated story. This
conclusion is buttressed by other important information, such as the
Book of Mormon’s almost universal failure to properly describe
the realities of ancient-American life.
> You had a really interesting website,
It isn't my website, Jim.
> but when you started making
> unwarranted conclusions it made the rest of it look bad.
Oh, someone is looking pretty silly all right.
> Its foolish to conclude that if something is unlikely to be randomly
> generated that it cannot be real, and that is the conclusion which you
> reached.
Well, to the point of being repetitious, it seems I need to repeat
this one more time:
Throughout all of this analysis we should be careful to remember that
conclusions based on statistical inference do not constitute proof.
Indeed, given a sufficiently large sampling of statistical data from
real historical events, there is a finite probability that some will
have the same distribution found in the Book of Mormon. These,
however, will be very rare; it is far more likely that the dates in
the Book of Mormon were concocted as part of a fabricated story. This
conclusion is buttressed by other important information, such as the
Book of Mormon’s almost universal failure to properly describe
the realities of ancient-American life.
Anyone wanting to see the full article that Jim did not read can find
it at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
Kevin Thurston
I wondered what that series of dates was supposed to be. I thought it might
be dates of statehood, until I got to the listing for my state, Colorado.
Since you got the date wrong, I dismissed that idea. Colorado became a
territory in 1861, but that was on November 6th, not February 28th, as you
listed. Colorado became a state in 1876, which is why it is known as the
centennial state. Perhaps you wanted Duwayne to apply his statistics to your
dates, to discover that they were fictional?
Kevin Thurston
--
"Let's see now,,,, my horoscope: Today will be just like every other day.
D'oh, it just gets worse and worse."... Homer Simpson
Duwayne Anderson
<snip>
> I wondered what that series of dates was supposed to be. I thought it might
> be dates of statehood, until I got to the listing for my state, Colorado.
> Since you got the date wrong, I dismissed that idea. Colorado became a
> territory in 1861, but that was on November 6th, not February 28th, as you
> listed. Colorado became a state in 1876, which is why it is known as the
> centennial state. Perhaps you wanted Duwayne to apply his statistics to your
> dates, to discover that they were fictional?
<snip>
I think the problem here is that the LDS on ARM are still smarting
from their inability to take my "challenge" regarding the "wordprint"
analysis that Charles posted.
There are, however, some serious difference. Let's review:
1) The statistical analysis of dates rests primarily on a single
assumption: that dates from real history don't favor one day of the
month over any other. That is, an important date in history is as
likely to fall on the first day of the month as the last day, or any
day inbetween. This is a relatively easy assumption to validate.
2) The fundamental assumption in "wordprint" analysis is that people
have distinct "wordprint" signatures that they cannot alter or change,
even when writing a book of fiction. This is a very difficult
assumption to validate because it relates directly to human abilities.
So, there is significant asymmetry in the assumptions of the two
approaches. In addition, the force and direction of the conclusions
is very different:
1) The statistical analysis of the dates in the Book of Mormon shows
that it is highly unlikely that a real history of dates would be as
biased as those in the Book of Mormon. In other words, out of several
thousand real histories, only a few would have dates as biased as
those found in the Book of Mormon. When taken along with other
evidence (such as the virtual total lack of corroborating
archeological evidence) this suggests that the Book of Mormon does not
represent a real history.
2) The statistical analysis in the "wordprint" article purports to
show that more than one author was involved in its writing. While
this conclusion is consistent with the Book of Mormon's internal
account, it is also at least equally consistent with the idea that
more than one 19'th century person was involved in writing the Book of
Mormon. Taken with the fact that the Book of Mormon so poorly
describes conditions in ancient America, the proper conclusion to draw
from the "wordprint" analysis -- assuming it's fundamental assumption
is ever validated -- is that more than one 19'th century person was
involved in writing it.
So, you see, I've made no claims to be able to determine if a set of
dates is from a true history or not. Indeed, I state in the
conclusion section of my paper that such dates can happen -- but
infrequently. And I calculated the probability. At the end of the
day, the basic facts remain: The Book of Mormon describes virtually
nothing correctly about ancient America, and the distribution of dates
found in the Book are highly unlikely for a real (unbiased, random)
set of historical month-days.
Duwayne Anderson
> May I also point out that he perhaps missed dates.
>
> One example,
>
> 3 Nephi 8
> [5] And it came to pass in the thirty and fourth year, in the first
> month, on the fourth day of the month, there arose a great storm, such
> an one as never had been known in all the land.
No, Charles. That day was included. It's in table 4, third line
down. Since you obviously have not read the article allow me to point
you to an Internet site where you can view it.
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
Also, Charles, note that (unlike yourself) I've made all the
calcuations, tables, etc. totally available. Nothing to hide. All
the details are out in the open.
Duwayne Anderson
>
> > I see you are still having difficulty with the concept of probability.
> > At a probability of 1/100 there is a chance of 1/100 that it is is a
> > real history. At a probability of 1/1000 there is a chance of 1/1000
> > that it is a real history. At a probability of 1/2000 there is a chance
> > of 1/2000 that it is a real history.
>
> I don't think this is quite correct. The statement that "1/2000 of real
> histories have this property" is not the same as the statement that "there
> is a 1/2000 chance that it is a real history". Arguments such as yours
> can strongly weight the probability that an observation is true, but
> cannot set that probability from scratch.
I've given this some additional thought.
There are two possibilities in the case (as I analyzed it):
1) The dates were generated without bias (randomly, the way real
historical dates are generated).
2) Not #1. That is, there was some bias involved.
The analysis shows there is a probability of less than 1/2000 that the
dates were generated without bias. If we start with no other
information (this is an important) then the alternative set -- that
bias was somehow involved, even though how the bias was applied may be
unknown -- is 1-(1/2000.
If you *add* additional information the probabilities change (they
almost always do for any statistical analysis). For example, if the
LDS could add information to the effect that a city in Central America
has been discovered with the name "Zarahemla" at the city gate, that
changes the probability that bias was not involved. On the other
hand, if there is a great deal of information about ancient America
that is inconsistent with the Book of Mormon, and if the Book of
Mormon is virtually never right about the way ancient Americans lived,
where they came from, what the eat, etc., this also changes the
probabilities.
> Suppose I draw from a deck of cards, half of which have "true" written on
> the back, and half of which have "false". If you tell me that only one in
> a hundred "true" cards are blue on the front, and that one in two of the
> "false" cards are blue, then there is an easily calculable (under 2%)
> probability that my card will say "false" when I flip it over.
Actually, since you never said whether you knew if the card in your
had was blue or not, the probability that it said "false" when flipped
over is 50/50. In your example, you are pointing to the addition of
other information (and that was your intention, if I read you
correctly). See comments above.
<snip>
> Example: if I begin with the belief that there is a 1/4 chance that the
> Book of Mormon is true and the assumption that there is a 1/2 chance that
> a man-made lists of dates will be as statistically deviant as the Book of
> Mormon dates, then I am left with a weight of (1/4*1/2000) that it is true
> and a weight of (3/4 * 1/2) that is false, giving (1/8000)/(1/8000+3/8),
> i.e. approximately a 1/3000 chance that it is true.
Simple belief does not alter the probabilities. But other evidence
can and does. It won't change the fact that the Book of Mormon's
distribution of dates is statistically very unlikely, but it will
change how we interpret them.
This is why I specifically described the statistical analysis of dates
within the larger context of a Book that is, by virtually all other
accounts, a rather clumsy fraud. If, as I said, there were specific
Book of Mormon cities and people identified through the archeological
record, one would be justified somewhat in dismissing the highly
unlikely distribution of month dates in the Book of Mormon. But this
is not the case. Instead, the Book of Mormon is found to violate
virtually everything science tells us about the way ancient Americans
lived. Taken in *that* context, the unlikely distribution of dates in
the Book of Mormon looks exactly like what it probably is: a
fabricated history.
<snip to end>
Kevin Thurston
"Duwayne Anderson" <duwa...@hotmail.com> wrote in message
news:a42139e3.02022...@posting.google.com...
news:<a5dmss$m...@dispatch.concentric.net>...
>
> <snip>
> > I wondered what that series of dates was supposed to be. I thought it
might
> > be dates of statehood, until I got to the listing for my state,
Colorado.
> > Since you got the date wrong, I dismissed that idea. Colorado became a
> > territory in 1861, but that was on November 6th, not February 28th, as
you
> > listed. Colorado became a state in 1876, which is why it is known as the
> > centennial state. Perhaps you wanted Duwayne to apply his statistics to
your
> > dates, to discover that they were fictional?
> <snip>
>
> I think the problem here is that the LDS on ARM are still smarting
> from their inability to take my "challenge" regarding the "wordprint"
> analysis that Charles posted.
<snip review>
> So, you see, I've made no claims to be able to determine if a set of
> dates is from a true history or not.
It appeared to me that it might be a trap to try to get you to make such a
claim.
I still don't understand why he would get the date so wrong though unless he
was deliberately attempting to produce a fictionalized set of dates that
were supposed to look like the statehood dates.
Kevin Thurston
"You're not allowed to say sex on the internet!"... Marge Simpson
Roy Stogner
> There are two possibilities in the case (as I analyzed it): 1) The
> dates were generated without bias (randomly, the way real historical
> dates are generated).
> 2) Not #1. That is, there was some bias involved.
>
> The analysis shows there is a probability of less than 1/2000 that the
> dates were generated without bias.
It shows that there is a probability of 1/2000 that dates generated
without bias would resemble these. That's not the same thing.
> If we start with no other
> information (this is an important) then the alternative set -- that bias
> was somehow involved, even though how the bias was applied may be
> unknown -- is 1-(1/2000.
The probability that dates generated without bias would be more centered
than Book of Mormon dates is 1-(1/2000). Again, that's not the same
thing.
Let me phrase this with set theory? I'm a bit more comfortable there.
Call the sets of "historical dates" and "artificial dates" H and H'. Call
the sets of "dates more randomly distributed than the Book of Mormon" and
"dates equally or less randomly distributed" R and R'. Then:
The probability (x in H => x in R') = 1/2000.
The probability (x in H => x in R) = 1999/2000.
The probability (x in R' => x in H) is not calculable without more
information. This is what we're expressing in English as "the probability
that the Book of Mormon is a real history".
>> Suppose I draw from a deck of cards, half of which have "true" written
>> on the back, and half of which have "false". If you tell me that only
>> one in a hundred "true" cards are blue on the front, and that one in
>> two of the "false" cards are blue, then there is an easily calculable
>> (under 2%) probability that my card will say "false" when I flip it
>> over.
>
> Actually, since you never said whether you knew if the card in your had
> was blue or not, the probability that it said "false" when flipped over
> is 50/50. In your example, you are pointing to the addition of other
> information (and that was your intention, if I read you correctly). See
> comments above.
I intended to say "my blue card", of course. Note that without the
addition of this additional information, the probability that my blue card
says "false" cannot be calculated at all, even though we know the
probability that "true" cards are blue.
> <snip>
>> Example: if I begin with the belief that there is a 1/4 chance that the
>> Book of Mormon is true and the assumption that there is a 1/2 chance
>> that a man-made lists of dates will be as statistically deviant as the
>> Book of Mormon dates, then I am left with a weight of (1/4*1/2000) that
>> it is true and a weight of (3/4 * 1/2) that is false, giving
>> (1/8000)/(1/8000+3/8), i.e. approximately a 1/3000 chance that it is
>> true.
>
> Simple belief does not alter the probabilities. But other evidence can
> and does.
I suppose "belief" was a poor choice of word in this context; "preexisting
evidence" or "starting probability" might have been better. My main point
is simply that mathematically, your analysis can be used to alter such a
probability estimate, but not to generate one from scratch.
(by the way, I'm snipping a lot of your text here, in the belief that the
passages I cut are either answered by what I've said or are something I
agree with. If you're unsure which category a deletion falls in, let me
know)
---
Roy Stogner
cdowis
> cdo...@my-dejanews.com (cdowis) wrote in message news:<93c36e92.02022...@posting.google.com>...
> > May I also point out that he perhaps missed dates.
> >
> > One example,
> >
> > 3 Nephi 8
> > [5] And it came to pass in the thirty and fourth year, in the first
> > month, on the fourth day of the month, there arose a great storm, such
> > an one as never had been known in all the land.
> <snip to end>
>
> No, Charles. That day was included. It's in table 4, third line
> down.
You're right. Sorry. I missed it.
Duwayne Anderson
> On Mon, 25 Feb 2002 15:24:26 -0600, Duwayne Anderson wrote:
>
> > There are two possibilities in the case (as I analyzed it): 1) The
> > dates were generated without bias (randomly, the way real historical
> > dates are generated).
> > 2) Not #1. That is, there was some bias involved.
> >
> > The analysis shows there is a probability of less than 1/2000 that the
> > dates were generated without bias.
>
> It shows that there is a probability of 1/2000 that dates generated
> without bias would resemble these. That's not the same thing.
Okay. Let's see if I can be as accurate as possible. The analysis
shows that if we select dates without bias, there is a probability of
less than 1/2000 of finding a series of dates that is as far, or
further, from the mean than those dates in the Book of Mormon.
>
> > If we start with no other
> > information (this is an important) then the alternative set -- that bias
> > was somehow involved, even though how the bias was applied may be
> > unknown -- is 1-(1/2000.
>
> The probability that dates generated without bias would be more centered
> than Book of Mormon dates is 1-(1/2000). Again, that's not the same
> thing.
Okay. Let's see if I can be as accurate as possible. The analysis
shows that if we select dates without bias, there is a probability of
less than 1/2000 of finding a series of dates that is as far, or
further, from the mean than those dates in the Book of Mormon.
>
> Let me phrase this with set theory? I'm a bit more comfortable there.
> Call the sets of "historical dates" and "artificial dates" H and H'.
How can you do this? The set of artificial dates is a partial subset
of the historical dates. There is no way to separate the sets. The
thing that distinguishes them is the probability of finding a string
of dates.
This seems to be a common point of confusion on the analysis I
presented. It's easy to fake (create artificially) a set of data that
looks random. So finding such a set in the Book of Mormon would have
told us nothing about it. But that's not what happened. The author
of the Book of Mormon was bad at faking historical data, and this
mistake is obvious by looking at the statistically very improbably
list of dates there. The test is only good if the author is bad at
faking dates. The same goes for fingerprints. The test is only good
if the burgler forgets to wear his gloves.
Call
> the sets of "dates more randomly distributed than the Book of Mormon" and
> "dates equally or less randomly distributed" R and R'. Then:
>
> The probability (x in H => x in R') = 1/2000.
>
> The probability (x in H => x in R) = 1999/2000.
>
> The probability (x in R' => x in H) is not calculable without more
> information. This is what we're expressing in English as "the probability
> that the Book of Mormon is a real history".
>
> >> Suppose I draw from a deck of cards, half of which have "true" written
> >> on the back, and half of which have "false". If you tell me that only
> >> one in a hundred "true" cards are blue on the front, and that one in
> >> two of the "false" cards are blue, then there is an easily calculable
> >> (under 2%) probability that my card will say "false" when I flip it
> >> over.
> >
> > Actually, since you never said whether you knew if the card in your had
> > was blue or not, the probability that it said "false" when flipped over
> > is 50/50. In your example, you are pointing to the addition of other
> > information (and that was your intention, if I read you correctly). See
> > comments above.
>
> I intended to say "my blue card", of course. Note that without the
> addition of this additional information, the probability that my blue card
> says "false" cannot be calculated at all, even though we know the
> probability that "true" cards are blue.
Adding additional information usually changes the probabilities. That
was my point in framing the distribution of dates in the Book of
Mormon within the larger context of a book that is also wrong about
virtually every aspect of life in ancient America.
>
> > <snip>
> >> Example: if I begin with the belief that there is a 1/4 chance that the
> >> Book of Mormon is true and the assumption that there is a 1/2 chance
> >> that a man-made lists of dates will be as statistically deviant as the
> >> Book of Mormon dates, then I am left with a weight of (1/4*1/2000) that
> >> it is true and a weight of (3/4 * 1/2) that is false, giving
> >> (1/8000)/(1/8000+3/8), i.e. approximately a 1/3000 chance that it is
> >> true.
> >
> > Simple belief does not alter the probabilities. But other evidence can
> > and does.
>
> I suppose "belief" was a poor choice of word in this context; "preexisting
> evidence" or "starting probability" might have been better.
Yes.
> My main point
> is simply that mathematically, your analysis can be used to alter such a
> probability estimate, but not to generate one from scratch.
As I said, "Adding additional information usually changes the
probabilities. That was my point in framing the distribution of dates
in the Book of Mormon within the larger context of a book that is also
wrong about virtually every aspect of life in ancient America."
Woody Brison
news:<HRKd8.109891$Zc.31...@news2.rdc1.mi.home.com>...
> Duwayne Anderson <duwa...@hotmail.com> wrote...
>
> "However, since the Book of Mormon presents the month-dates
> matter-of-factly, as being part of a literal history, we conclude that the
> Book of Mormon itself is a fabrication. "
>
> And if the days would have added up to 124, you would have concluded that
> the Book of Mormon is not a fabrication?
Come now, he would conclude that it WAS a fabrication,
because the dates were scattered around -- the forger
was clever, and tossed the dice when he wanted a date.
If they are biased toward the front of the month, then
the forger was not clever. But he was still a forger.
In real life, if they are distributed randomly, then it's
either a forgery or real. If they are skewed, then it's
a forgery or there is a reason why they are skewed. But
here, genuine is ruled out. Why? It might be because, as
I seem to remember Duwayne noting at some point, statistics
can be assymetric in what they show or don't show. Or, it
might be because the fast talker has directed your
attention away from something that might occur to you
otherwise.
> Arizona: 24th of Feb, 1863 Tuesday
> Arkansas: 2nd of march, 1819 Tuesday
> Colorado: 28th of Feb, 1861 Thursday
> Idaho: 4th of March, 1863 Wednesday
> Iowa: 12th of June, 1838 Tuesday
> Kansas: 30th of May, 1854 Tuesday
> Lousisiana: 26th of March, 1804 Monday
> Minnesota: 3rd of march, 1849 Saturday
> ...
> (Weird...no months in the last half of the year...doesn't mean that these
> states are fabrications, it just means its weird/unusualy/unlikely).
>
> Assuming that the events happen on random days of the week, you wouldn't
> expect half of your samples to fall on the same day of the week either.
I was sort of wondering why there are only 8 states listed
and why those particular 8 were in the set. Maybe Duwayne
cooked the list of states until the days of the month looked
random, and didn't notice that he'd skewed the months. (Keep
the expression "probably invented in the mind of an
unscrupulous man" in mind; this kind of mind tends to imagine
their target to have their own character...)
Or, there might be a real reason why states are born in the
first half of a year. When did Congress meet? When did
they adjourn? When did they tend to begin work on important
historic business, January? How long did it tend to take?
Did they tend to do sign big stuff on a particular day after
the Sabbath?
I would expect that if we listed the start date of many
military campaigns we'd see a correlation with the time of
year the mud dries. Real events often have patterns in them.
Thus, there might be a reason why dates in the Book of Mormon
tend to the first week in the month.
The fact that all these are from Alma seems odd. Maybe Alma
found meaning in certain dates of great portent, and maybe it
was because they were in the first week of the year for some
reason. There is also the strong possibility of intermediate
scribes standing as it were between the orginal authors such
as Alma and the historian Mormon. See
http://www.jefflindsay.com/chiasmus.shtml
It would seem from this that the stack of plates was not
quite a simple journal of day to day events but some kind of
amalgamated collection, worked over by scribes with poetic
tendencies. Maybe they introduced date biases, who knows.
Certain it is that numerous scribes contributed to the whole.
Absolutely certain it is that Joseph couldn't have made this
kind of stuff up.
The danger in any kind of analysis like Duwayne's is that
there can be included a subtle bias, or more subtly, excluded
some fairness. The Bible Codes flap was a recent example of
this. Let's see, I remember writing something about them,
http://groups.google.com/groups?selm=87ant6%24eo3%241%40nnrp1.deja.com
Any time you find that statistics is asymmetric in what it is
able to show, ask why the asymmetry.
Wood
Joel Rees
Duwayne, I am just offering two control groups. I thought you might
like to analyze them, since you seem to have the time.
One is computer generated, something like generating a list by
throwing dice. The other is the dates from the first 11 individuals
that a quick search on a common name in the IGI produced. I have not
taken the time to perform your analysis on them, I have more important
things to do, begging your pardon.
11 is too few, of course, to do much with. Do you want the program? I
assume you know where the IGI can be found on-line?
BTW, I disagree with your assumption that there should be no
legitimate source of skew in a list of this sort. A uniform calendar
system of the sort we enjoy now has by no means been the rule during
recorded history in general. Ours has only been really uniform since
the time of one Gregory.
Internal evidence in the BoM includes a claim that the _general_
accounting of years changed at least every time there was a new king,
down to the period called the reign of the judges. Such a change
cannot be discounted, IMO, as a source for this sort of skew.
And you will accuse me of rationalizing, and of having no idea of what
science is for, I suppose. But I do find other evidence more moving,
such as Mosiah 28 and 29, particularly the verses about the inequality
of making the king carry the burden of rule.
Joel
Duwayne Anderson
<snip>
> Duwayne, I am just offering two control groups.
No, Joel. You are demonstrating that you didn't read the article and
that you don't understand the statistical test that was involved.
> I thought you might
> like to analyze them, since you seem to have the time.
The equations in the article will allow you to calculate all the
probabilities yourself. Did you miss the link to the article? Here
it is again:
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
> One is computer generated, something like generating a list by
> throwing dice. The other is the dates from the first 11 individuals
> that a quick search on a common name in the IGI produced. I have not
> taken the time to perform your analysis on them, I have more important
> things to do, begging your pardon.
Well, see here you illustrate nicely your failure to understand the
test that was conducted. Let's go over it again. Shall we?
If we select 8 days of the month, at random, each day of the month has
an equal probability of being selected. No day will be selected
preferentially over any other day.
If we select 8 days of the month, and add those days, there will be an
expected (mean) value and standard deviation associated with the
resulting probability distribution. The further a series of dates
lies from the mean of the distribution the more unlikely it is to
occur.
When we look at the 8 month dates in the Book of Mormon, and add them
up, we find that they are several standard deviations from the mean.
The probability of finding a string of dates this far from the mean
(or further) is less than 1/2000.
The conclusion from this is that the dates in the Book of Mormon are
very unlikely to occur in a real history of random dates. Yet that is
exactly the way the dates are presented -- as a real history of
unbiased dates.
We then look at other data, such as the Book of Mormon's total failure
to properly describe ancient America. Taken with this other data, and
the statistically improbably string of dates found in the Book of
Mormon, we conclude that the Book of Mormon is a fake. The summary in
the article pretty much says it all:
"Throughout all of this analysis we should be careful to remember that
conclusions based on statistical inference do not constitute proof.
Indeed, given a sufficiently large sampling of statistical data from
real historical events, there is a finite probability that some will
have the same distribution found in the Book of Mormon. These,
however, will be very rare; it is far more likely that the dates in
the Book of Mormon were concocted as part of a fabricated story. This
conclusion is buttressed by other important information, such as the
Book of Mormon’s almost universal failure to properly describe
the realities of ancient-American life."
Note that this analysis is based on probability and statistics. It
makes no claims about whether or not a person or computer could invent
a string of dates that looks random. In fact, that is a distinct
possibility. Unfortunately for the author of the Book of Mormon,
however, they didn't do that. Those dates are very unlikely in a real
history of random dates.
<snip to end>
Woody Brison
13th of the month. He usually has a birthday party, but it's not often
possible to schedule the party on the exact day. Over the last 8
years, he's managed to hit the following days: 13, 16, 15, 16, 17, 15,
17, 15. Applying Duwayne's Math, these numbers add to 124 -- right on
the money, a totally random scatter over the month, altho his dates
have really only spanned 5 days.
His brother does the same thing. He was born on the 5th, and over the
same period he's hit these days: 2, 22, 1, 6, 3, 9, 6, 4. Duwayne's
Math detects that the brother has a very high marksmanship. The sum
is 53, almost 3 sigma out from the mean, which means that the
probability of it being random is only a half a percent. The reality
is, however, that he's managed to randomize it out over a span of
three weeks across the month!
But funnier yet, their sister, organized little sweetheart that she
is, mails out invitations 3 months in advance, so she can pretty much
schedule when it's convenient for her and everyone can reserve that
evening. Her birthday is on the 30th of the month, and over the last 8
years, she's hit the days 30, 1, 30, 1, 30, 1, 30, 1. The sum is 124.
We can tell using Duwayne's math that her scheduling is really pretty
close to totally random.
Wood
Woody Brison
> No, Joel. You are demonstrating that you didn't read the article and
> that you don't understand the statistical test that was involved.
I was wondering, Duwayne, did you read the article yourself?
Wood
Woody Brison
> May I also point out that he perhaps missed dates.
A casual glance at his list suggests that he's gotten it
by doing a computer search on the word "month". However,
actually reading the book is a better way of knowing what
it says.
As an example, his first date in the list is from Alma
52:1 -- Year of the Judges 26, month 1, day 1. Boy, that
weights the dates to the front end of the month, like a fat
lady stepping into the side of the boat! You can check the
passage, it's about how the Lamanites woke up in the
morning and found their guy had died in his tent in the
night. No orders from him today about how to fight the
Nephites, who are waving their spears. But how'd he get
that way? Ho, ho, for that we have to rewind to the
previous chapter and actually read more than a quick verse.
We find that on the LAST DAY of the previous year, Teancum
had basically been the cause of the guy dying, altho it
could be ascribed to heart failure. It doesn't take a genius
to infer that this day, the activities of which are described,
including the setting of the sun, had a date, altho it isn't
stated explicitly by number.
If we add that one teeny date to the chart, we'll have to
guess at how many days in a month, but say it's 30, that
brings the average to 53, a trimming of the bias by a full
sigma (and this strictly according to Duwayne Math, no
cheating).
He's missed lots of other dates similarly. 3 Ne 2:1 talks
about the last day of 95 Judges, v. 4 names the last day
of 4 different years. There are many others. But, if you
think about it, where you've got the last day of one year,
you've got the first day of the next year. You're really
being told about the year rolling over, a lot of dates are
implied in there -- if you are trying to comb it for
information instead of pretending to be as dumb as possible
in order to manufacture accusations.
A real analysis, one having the intention of finding out
what the book really shows, should start with a simple
reading of the book, with a notebook beside to note down
all the days found.
The analysis we're given here, in contrast, comes from a
guy who was certain the book's a fake before he started,
but can't seem to demonstrate that he's ever read it; before
any numbers are even entered, his analysis is only going to
be able to show that the book is fake, not genuine; and it
depends on some pretty bizarre statistical methods to get
its conclusions. Other than that, however, it's a nice bit
of expository writing. It would probably get an A in an
English class. I don't think it would fare so well from a
competent stats teacher or even a history teacher.
Wood
Markg91359
analysis one may well come up with a probability that certain dates would fall
within a certain range are only 1 out of 2000.
However, I think true believers would tell you that religion by its nature is
*exceptional*. The averages aren't likely to apply and the event is likely to
be the 1 out of 2000 precisely because it is sacred, ordained by God, etc.
Mark
Roy Stogner
Thank you for sharing; this really puts into the perspective the time I've
wasted nitpicking Duwayne's trivial errors.
---
Roy Stogner
Duwayne Anderson
> cdo...@my-dejanews.com (cdowis) wrote in message news:<93c36e92.02022...@posting.google.com>...
>
> > May I also point out that he perhaps missed dates.
>
> A casual glance at his list suggests that he's gotten it
> by doing a computer search on the word "month".
No kidding.
> However,
> actually reading the book is a better way of knowing what
> it says.
This was a statistical test that looked at the distribution of dates
in the Book of Mormon. So computer searches are not only appropriate,
they are necessary for completeness.
By the way, I've read the Book of Mormon cover to cover over 17 times.
How many times have you read it cover to cover?
> As an example, his first date in the list is from Alma
> 52:1 -- Year of the Judges 26, month 1, day 1. Boy, that
> weights the dates to the front end of the month, like a fat
> lady stepping into the side of the boat!
The order of the dates in the table has nothing to do with the way the
statistical analysis was carried out.
<snip>
> He's missed lots of other dates similarly.
This statement by Woody is false. The study included all month-dates
in the Book of Mormon.
> 3 Ne 2:1 talks
> about the last day of 95 Judges, v. 4 names the last day
> of 4 different years.
Here is 3 Ne 2:1
3 Nephi 2
[1] And it came to pass that thus passed away the ninety and fifth
year also, and the people began to forget those signs and wonders
which they had heard, and began to be less and less astonished at a
sign or a wonder from heaven, insomuch that they began to be hard in
their hearts, and blind in their minds, and began to disbelieve all
which they had heard and seen --
and here is verse 4:
[4] And thus did pass away the ninety and sixth year; and also the
ninety and seventh year; and also the ninety and eighth year; and also
the ninety and ninth year;
Clearly, for anyone who bothers to actually read the Book of Mormon,
neither verse contains any month dates that could be included in the
study.
> There are many others
This statement by Woody is false. The study used all the month-dates
found in the Book of Mormon.
<snip>
> The analysis we're given here, in contrast, comes from a
> guy who was certain the book's a fake before he started,
I was an active member of the LDS Church for over 35 years, and held
many positions in the Church, including Elder's Quorum President in
two wards. I've been married in the temple, gone on a mission, and
graduated from BYU.
What Woody's doing is using his old standby ad hominem argument.
Trying to dodge the issues by focusing on the person.
> but can't seem to demonstrate that he's ever read it;
For other readers just joining, Woody is still smarting from a bad
loss a few years ago in which he claimed that the Book of Mormon's
Arabian river had been found. His "river" was trounced soundly by
reference to descriptions found in the Book of Mormon -- descriptions
that Woody was apparently unfamiliar with.
Care to have this discussion again, Woody?
> before
> any numbers are even entered, his analysis is only going to
> be able to show that the book is fake, not genuine;
The statistical distribution was wrong because a critic did the math?
This is really lame, Woody.
> and it
> depends on some pretty bizarre statistical methods to get
> its conclusions.
If you think that calculating means and standard deviations, and
deriving probabilities based on location in the distribution is
"bizarre" I suggest a remedial class in mathematics and statistics.
> Other than that, however, it's a nice bit
> of expository writing. It would probably get an A in an
> English class. I don't think it would fare so well from a
> competent stats teacher or even a history teacher.
And this is typical of how the LDS tend to simply brush off things
that don't fit their world view.
Duwayne Anderson
Duwayne Anderson
> 13th of the month. He usually has a birthday party, but it's not often
> possible to schedule the party on the exact day. Over the last 8
> years, he's managed to hit the following days: 13, 16, 15, 16, 17, 15,
> 17, 15. Applying Duwayne's Math, these numbers add to 124 -- right on
> the money, a totally random scatter over the month, altho his dates
> have really only spanned 5 days.
Faking a random distribution is easy. Anyone can do it. The author
of the Book of Mormon *could* have done it, had he been smart. The
problem is, he *didn't* do it.
The test in the article at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
Is a negative test. It looks at the probability that something did
*NOT* happen -- namely that the Book of Mormon dates probably did
*NOT* come from a random distribution; i.e., real history.
As Woody has just shown, it's easy to come up with a string of numbers
that is very near the mean of the random distribution. Because of
this, it's not possible to say that the dates in the Book of Mormon
are *valid* no matter how close they are to the mean.
But that's not what the test results showed. The test results showed
that the Book of Mormon dates are *far* from the mean. The results
show that the Book of Mormon dates are *unlikely* to happen in a
random (real history) distribution of uncorrelated dates.
See the difference, Woody? Should I explain it again?
<snip rest of the same type of misunderstanding about statistical
tests for probability>
Duwayne Anderson
> news:<HRKd8.109891$Zc.31...@news2.rdc1.mi.home.com>...
> > Duwayne Anderson <duwa...@hotmail.com> wrote...
> >
> > "However, since the Book of Mormon presents the month-dates
> > matter-of-factly, as being part of a literal history, we conclude that the
> > Book of Mormon itself is a fabrication. "
> >
> > And if the days would have added up to 124, you would have concluded that
> > the Book of Mormon is not a fabrication?
>
> Come now, he would conclude that it WAS a fabrication,
> because the dates were scattered around -- the forger
> was clever, and tossed the dice when he wanted a date.
The above statement by Woody is false, as he would know if he had read
the article at http://www.lds-mormon.com/numbersinthebookofmormon.shtml
which had several examples of real historical dates, drawn at random,
with sums close to the mean that's expected of random dates.
<snip to end>
Duwayne Anderson
Duwayne Anderson
Well, I think you are right. There will be lots of BS excuses and
hand wringing, ad hoc arguments, bad statistical arguments, and (as
Woody has demonstrated) ad hominem arguments.
But the fact remains that the Book of Mormon *itself* presents these
dates as being non-religious in nature, and part of an allegedly real
history of the ancient Americans. However, the distribution of these
dates is very unlikely to occur in a real history of random
month-dates, as the analysis at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml shows. When
taken together with the fact that the Book of Mormon is wrong about
virtually every testable, non-trivial statement about ancient America
that it makes, these data lead to the conclusion that the Book is a
fake.
Of course, devout LDS will never accept that conclusion. That's the
nature of religious faith -- belief in spite of evidence to the
contrary.
Joel Rees
> No, Joel. You are demonstrating that you didn't read the article and
> that you don't understand the statistical test that was involved.
No, Duwayne, I am telling your that
your arguments fail to convince me!
People are not _random_, we are _arbitrary_. Your test, in fact, is
seeking to find non-randomness in a situation that should _not_ be
random. I told you one reason before, and I just generalized the
reason. If you can't see why, that's your problem. It's your
assertion.
I am giving you two sets of data which you may use as controls to
check your test. I told you where they came from. You can get your own
data sets from the same or similar places.
You've made an assertion. I don't buy it. I'm offering you a test to
check your assertion against. You may choose to test or not, that is
your business, not mine.
Oh. Here's my program. I didn't even bother to use the secure or
statistically good random number generators. I'm not telling you the
seed constant I used, just to be ornery. If you want to use dice
instead, because dice would have been available 170+ years ago, that
should also give useful results.
/**
* This is a really cheap program to generate arbitrary dates,
* doesn't even bother with command line parameters.
*
* @author Joel Rees,
* Cast upon the public domain, 2002.02.25.
* @version 0.0
*
* @deprecated It was written in a hurry, everything hard-coded.
* nextGaussian() method and java.security.SecureRandom would
* provide much more interesting results.
*
*/
import java.util.Random;
public class TrivialApplication {
public static void main(String args[]) {
Random rand = new Random();
rand.setSeed( YOUR_CONSTANT ); // constant for reproducibility
System.out.println( "Here, we go!" );
for ( int i = 0; i < 11; ++i )
{ int month = rand.nextInt( 12 ) + 1;
int year = rand.nextInt( 64 ) + 1788;
int day = rand.nextInt( MaxDays( month, year ) ) + 1;
System.out.print( Integer.toString( month ) + "/" );
System.out.print( Integer.toString( day ) + "/" );
System.out.println( Integer.toString( year ) );
}
}
// could've used the date classes, but why bother?
static final int[][] DAYS_OF_MONTH =
{ { 31, 28, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 },
{ 31, 29, 31, 30, 31, 30, 31, 31, 30, 31, 30, 31 }
};
static int MaxDays( int month, int year )
{ if ( ( month < 1 ) || ( month > 12 ) )
return 0;
--month;
return ( ( year % 4 != 0 ) || ( year % 100 == 0 ) )
? DAYS_OF_MONTH[ 0 ][ month ]
: DAYS_OF_MONTH[ 1 ][ month ];
}
}
Roy Stogner
> I am giving you two sets of data which you may use as controls to check
> your test. I told you where they came from. You can get your own data
> sets from the same or similar places.
Using Duwayne's simplification of a 30 day month, your first set of dates
falls .68 standard deviations from the mean, an event of 49% probability.
Your second set of dates falls .749 deviations from the mean, an event of
45% probability.
Since those sets of dates come from essentially random sources, this is
also what I expected to see. Was there a point to this or did you just
want Duwayne to do more math?
---
Roy Stogner
Duwayne Anderson
<snip>
> There are many others. But, if you
> think about it, where you've got the last day of one year,
> you've got the first day of the next year. You're really
> being told about the year rolling over, a lot of dates are
> implied in there
Historical dates are specific, named, historical events in which the
month and day are both mentioned. The end of the year is a calendar
date. Calendar dates are not randomly distributed. So "the first day
of the year" and the "last day of the year," (or similar phrases)
don't constitute historical dates unless associated with a specific
non-correlated historical event.
One of the reasons for listing the specific events mentioned in the
Book of Mormon (and used in the study) was so readers could tell for
themselves that the events were actual historical dates (not calendar
events) that were uncorrelated. You can see the study at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
> -- if you are trying to comb it for
> information instead of pretending to be as dumb as possible
> in order to manufacture accusations.
Biasing the results is a clear possibility, as you've just illustrated
by trying to argue for the inclusion of non-random calendar dates
(wherever the Book of Mormon mentions the end of another year)into the
analysis.
<snip to end>
Joel Rees
> On Wed, 27 Feb 2002 21:35:16 -0600, Joel Rees wrote:
>
> > I am giving you two sets of data which you may use as controls to check
> > your test. I told you where they came from. You can get your own data
> > sets from the same or similar places.
>
> Using Duwayne's simplification of a 30 day month, your first set of dates
> falls .68 standard deviations from the mean, an event of 49% probability.
> Your second set of dates falls .749 deviations from the mean, an event of
> 45% probability.
>
> Since those sets of dates come from essentially random sources,
This is the point. Duwayne says he expects randomness. I ask why?
About half of the events he shows would have, not random, but
arbitrary dates. A good statistician knows the difference and will
refrain from setting up a model that expects randomness when he should
be expecting biases of an arbitrary nature to show up. And Duwayne
only considers one possible cause for the arbitrariness. He is also
mixing the set of dates, as well.
Given the list of events, the events which I would expect not to be
biased are not biased, and I _don't_ need to play statistical games to
observe the patterns.
Also, eight samples is not enough to draw any decent conclusions from,
no matter how far off the normal the distribution appears.
> this is
> also what I expected to see. Was there a point to this or did you just
> want Duwayne to do more math?
This is irrelevant, but can you set up a model that can tell which is
which?
Prophecy, Roy, prophecy for us.
Joel
cdowis
In my personal opinion, Dwayne has shown us that the bom people used
the mayan calendar of 20 days. His own statistical analysis shows
that such a calendar would fall into the expected range.
The earliest date cited is in Alma, which is centuries after the first
landing, and after the integration of the Mulekites into their
culture.
I would wildly speculate that the Mulekites, no longer keeping the
Mosiac laws and having no scriptural records, began using the 20 day
month calendar. The Nephites were the smaller group, and still
probably had the Jewish calendar.
Here is direct evidence that they were using the 20 day calendar of
the Mulekites, and, perhaps, also the Jewish calendar as well to keep
track of the Mosaic rituals. A dual calendar system.
Duwayne Anderson
>
> > No, Joel. You are demonstrating that you didn't read the article and
> > that you don't understand the statistical test that was involved.
>
> No, Duwayne, I am telling your that
You are telling me I didn't read the article I wrote?
> your arguments fail to convince me!
I'm not surprised. But, then, you made up your mind a long time ago
to believe in the Book of Mormon no matter the facts. So no argument
would convince you.
> People are not _random_,
Nobody said people are random, Joel. You would know that if you read
the article at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
> we are _arbitrary_.
I agree. You are arbitrary. You arbitrarily arrive at conclusions
with no regard for evidence.
> Your test, in fact, is
> seeking to find non-randomness in a situation that should _not_ be
> random.
The test assumes that *historical* events don't favor one day of the
month over any other. You clearly disagree. So, tell us, Joel, which
days of the month are favored mostly by historical dates? According
to you, most historical dates must happen during the first week of the
month -- after all, that's where they show up in the Book of Mormon.
Duwayne Anderson
>
> > On Wed, 27 Feb 2002 21:35:16 -0600, Joel Rees wrote:
> >
> > > I am giving you two sets of data which you may use as controls to check
> > > your test. I told you where they came from. You can get your own data
> > > sets from the same or similar places.
> >
> > Using Duwayne's simplification of a 30 day month, your first set of dates
> > falls .68 standard deviations from the mean, an event of 49% probability.
> > Your second set of dates falls .749 deviations from the mean, an event of
> > 45% probability.
> >
> > Since those sets of dates come from essentially random sources,
>
> This is the point. Duwayne says he expects randomness.
The assumption is that historical dates don't favor any particular day
of the month. You apparently think they do. So, Joel, tell us which
days of the month historical dates favor. You must think they favor
the days during the first week. Right? Why should historical dates
favor days during the first week of the month, Joel?
> I ask why?
Well, the reason you think that (I'm sure) is that you want so badly
to believe in the Book of Mormon. I don't think you can give us a
logical/rational reason why historical dates should favor the first
week of the month.
> About half of the events he shows would have, not random, but
> arbitrary dates.
Joel, you are embarassing yourself. If the dates are "arbitrarily"
distributed they don't favor any particular day of the month.
Anyone wanting to see the article can find it at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
Duwayne Anderson
Roy, you've been sold a bill of goods.
The fact remains, the string of dates in the Book of Mormon (see the
article at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml) is
statistically unlikely. Had the string of dates been close to the
mean -- as in Woody's made-up example, no conclusion could be drawn.
But that's not what happened, and that's not what was tested. In
fact, the string of dates was very *FAR* from the mean. So far, in
fact, that it's probability of arising from a non-biased historical
account is very unlikely.
At the end of the day, after all the strained arguments offered by
Mormons, the result is still the same: they are trying to argue that
there is nothing suspicious about something as improbable as rolling
snake eyes 8 times in a row with a pair of unloaded dice.
Lee Paulson
>
> > On Wed, 27 Feb 2002 21:35:16 -0600, Joel Rees wrote:
> >
>
> Also, eight samples is not enough to draw any decent conclusions from,
> no matter how far off the normal the distribution appears.
Have you ever looked at veterinary trials?
Woody Brison
> > My thoughts on this subject are this. If one applies Duwayne's statistical
> > analysis one may well come up with a probability that certain dates would fall
> > within a certain range are only 1 out of 2000.
> >
> > However, I think true believers would tell you that religion by its nature is
> > *exceptional*. The averages aren't likely to apply and the event is likely to
> > be the 1 out of 2000 precisely because it is sacred, ordained by God, etc.
>
> Well, I think you are right. There will be lots of BS excuses and
> hand wringing, ad hoc arguments, bad statistical arguments,
No duh!
>...and (as
> Woody has demonstrated) ad hominem arguments.
It's actually a fallacy that ad hominem is a fallacy. It's good to
look at the messenger as well as the message. If the messenger is
a liar, insane, biased, mentally retarded, it casts doubt on the
validity of the message. If the message shows signs of bias, bad
logic, etc. then this also casts doubt on the message.
> But the fact remains that the Book of Mormon *itself* presents these
> dates as being non-religious in nature, and part of an allegedly real
> history of the ancient Americans. However, the distribution of these
> dates is very unlikely to occur in a real history of random
> month-dates, as the analysis at
> http://www.lds-mormon.com/numbersinthebookofmormon.shtml shows.
That analysis is deeply flawed, as I have shown. You have not
addressed those flaws.
>... When
> taken together with the fact that the Book of Mormon is wrong about
> virtually every testable, non-trivial statement about ancient America
> that it makes,
That's not a fact; that's your wishful thinking. There are
hundreds of things in the book that test positive.
>... these data lead to the conclusion that the Book is a
> fake.
No; you are positing the data to be conclusive in the direction
you wish they were, starting from your incorrect paradigm, and
there are serious questions about your logical abilities. This
is just to clarify my position. I have no wish to argue further
about your problems.
Wood
Woody Brison
> > cdo...@my-dejanews.com (cdowis) wrote in message news:<93c36e92.02022...@posting.google.com>...
> >
> > > May I also point out that he perhaps missed dates.
> >
> > A casual glance at his list suggests that he's gotten it
> > by doing a computer search on the word "month".
>
> No kidding.
So you did confine your data to the results of a computer search.
> > However,
> > actually reading the book is a better way of knowing what
> > it says.
>
> This was a statistical test that looked at the distribution of dates
> in the Book of Mormon. So computer searches are not only appropriate,
> they are necessary for completeness.
That's hogwash. It's fine to use a computer search, but to rely
on only that is to confine yourself to a partial set of data, and
to put a blinder on yourself as to the real contents of the book.
The computer search does not understand themes. It can only find
instances of specific words. There are many passages in books that
talk about something but don't use just one word to talk about it.
> By the way, I've read the Book of Mormon cover to cover over 17 times.
> How many times have you read it cover to cover?
583,410. My assertion here is just as good as yours, anyone
can claim anything on the internet. But, your claim of 17 is
not well supported by your obvious lack of knowledge of the book's
contents. We've been over this before, so I leave you with it.
> > As an example, his first date in the list is from Alma
> > 52:1 -- Year of the Judges 26, month 1, day 1. Boy, that
> > weights the dates to the front end of the month, like a fat
> > lady stepping into the side of the boat!
>
> The order of the dates in the table has nothing to do with the way the
> statistical analysis was carried out.
My point had nothing to do with the order of dates.
> <snip>
> > He's missed lots of other dates similarly.
>
> This statement by Woody is false. The study included all month-dates
> in the Book of Mormon.
I've given you several examples of dates you left out. Why don't
you go ride your horse? You're not much good at debating.
> I was an active member of the LDS Church for over 35 years, and held
> many positions in the Church, including Elder's Quorum President in
> two wards. I've been married in the temple, gone on a mission, and
> graduated from BYU.
Those plus /enduring to the end/ will get you eternal life. Minus
the enduring to the end they buy you nothing. And, they do not
give you knowledge of statistics, or of the book. I've known lots
of missionaries who didn't actually study the scriptures, oodles of
married people who don't, and a few EQ prezzes who don't.
> What Woody's doing is using his old standby ad hominem argument.
> Trying to dodge the issues by focusing on the person.
No. Your analysis is flawed. I pointed out the flaws. My points
have nothing to do with your Church involvement in the past, and in
fact did not mention it. It's a nice try at claiming some kind of
omniscience, or something, but it does nothing for your flawed
stats presentation.
> > but can't seem to demonstrate that he's ever read it;
>
> For other readers just joining, Woody is still smarting from a bad
> loss a few years ago in which he claimed that the Book of Mormon's
> Arabian river had been found.
Your assessment of that battle is as bad as the rest of your debating
skills.
> Care to have this discussion again, Woody?
No, you are useless as a source of information, insight, or method.
> > before
> > any numbers are even entered, his analysis is only going to
> > be able to show that the book is fake, not genuine;
>
> The statistical distribution was wrong because a critic did the math?
> This is really lame, Woody.
My point was that your test was devised to fail the book. It
was not devised to test the book. It is not a real test; it is
a deception aimed at persuasion away from something wonderful.
It is the fox that lost his tail all over again.
> > and it
> > depends on some pretty bizarre statistical methods to get
> > its conclusions.
>
> If you think that calculating means and standard deviations, and
> deriving probabilities based on location in the distribution is
> "bizarre" I suggest a remedial class in mathematics and statistics.
You took a mean of the data but did not calculate any s.dev over
any of it. You did not account for the circular nature of the days
in the month -- 30/31 is followed by 1,2,3... Your calculations
are cute but the rationale behind them is almost blank.
Tell you what. Go over to sci.math or any of the sci. groups
and try out your essay on them. I will watch but not say a word.
See if you get a consensus that your methods and expertise in
stats are valid.
> And this is typical of how the LDS tend to simply brush off things
> that don't fit their world view.
And this is typical of how Duwayne Anderson addresses problems
with his output -- it's always right, it could never have a flaw,
it's always the LDS who are blind. Poor, blind fools!
Wood
Woody Brison
>
> <snip>
> > There are many others. But, if you
> > think about it, where you've got the last day of one year,
> > you've got the first day of the next year. You're really
> > being told about the year rolling over, a lot of dates are
> > implied in there
>
> Historical dates are specific, named, historical events in which the
> month and day are both mentioned. The end of the year is a calendar
> date. Calendar dates are not randomly distributed. So "the first day
> of the year" and the "last day of the year," (or similar phrases)
> don't constitute historical dates unless associated with a specific
> non-correlated historical event.
"The great enemy of clear language is insincerity. When there is a gap
between one's real and one's declared aims, one turns as it were
instinctively to long words and exhausted idioms, like a cuttlefish
squirting out ink." -- George Orwell
If you wish to exclude certain recorded dates from the sample set, you
need to explain why those, and not the ones you want to include. And,
the fact that the ones you want to include make the book look
contrived is not a good reason. It would be an obvious case of
cooking the data to urge a predetermined conclusion.
> Biasing the results is a clear possibility, as you've just illustrated
> by trying to argue for the inclusion of non-random calendar dates
> (wherever the Book of Mormon mentions the end of another year)into the
> analysis.
You missed my argument. I would prefer to see ALL dates included,
in a real study, not just a lot of blather about stats competence
and fancy charts, but the real analysis ADMITTEDLY confined to one
quick computer search, yielding only eight dates out of the
potentially hundreds that are in the text.
And, would you believe it, this kind of "analysis" from a person
who actively campaigns against his former religion night and day.
It's fairly clear why you don't seem to want to discuss the flaws
in your essay. You're not trying to discover the truth of the
matter; you are trying to push a sour interpretation on the matter.
And, I'd want explanation of what is attempted. Launching off into
math calculations is all very fine, but we need to know what you are
trying to show by your calculations. This is done by providing
explanations in clear, simple language prior to showing the math.
You rush off to take means of the data. Well, you failed to explain
why you are doing that. The mean of a sample set doesn't begin to
compute any measure of its distribution, or random nature, etc.
That was basically handwaved past in your essay. But it's central.
If you're going to try to claim expertise in stats, I guess we could
all just vote you an honorary degree from Internet University if
that's what you need to booster your ego, but it's rather obvious
you don't really know too much about stats. So, don't try waving
your expertise about in lieu of evidence or logic -- "I'm an expert
in stats, because I can copy a graph out of a book, and I say it's
this way, so don't try to question what I say, just abandon the
Book of Mormon; trrrrrrrruuuuuust me".
Wood
Woody Brison
I wonder if there is a simple reason why the dates seem to be skewed
toward the front of their months. Possible reasons:
1. Duwayne has cooked the data, excluded dates from the back half
of months.
2. Historians who contributed to the Book of Mormon tended to note
dates in the front of the month because it was cumbersome in their
notation to write larger numbers. An unconscious bias perhaps.
3. People tended to do things on or soon after the beginning of the
month -- kicking off military campaigns, assassinations connected
to elections and festivals, things that are tied to the calendar.
In a sample size of 8, it only takes 1 or 2 such to seriously bias
the outcome, and there are easily four such things on Duwayne's
list (three military attacks and an assassination). The great
storm and cataclysms at the Savior's crucifixion were tied to the
calendar, since ancient calendars were tied to the day of his
future resurrection -- the beginning of their year, so that's five
of the eight.
In simulations of submarine warfare, I noticed the hilarious tendency
of ship/sub captains to set courses like 90, 180, 270, etc. People
tend to shoot for round numbers. If they show a course of 173 or
something odd like that, you could project that course straight to
some destination, and intercept them there, plus infer things going
on at that place. That in fact is how the Battle of Midway was won.
American aviators found a Japanese destroyer going lickety-split in
a certain direction, so they flew ahead along that vector and found
the Japanese fleet. Destroyed 75% of their carrier force in 15
minutes.
Wood
Woody Brison
> > I had to laugh at something. Say this one guy's birthday is on the
> > 13th of the month. He usually has a birthday party, but it's not often
> > possible to schedule the party on the exact day. Over the last 8
> > years, he's managed to hit the following days: 13, 16, 15, 16, 17, 15,
> > 17, 15. Applying Duwayne's Math, these numbers add to 124 -- right on
> > the money, a totally random scatter over the month, altho his dates
> > have really only spanned 5 days.
>
> Faking a random distribution is easy. Anyone can do it. The author
> of the Book of Mormon *could* have done it, had he been smart. The
> problem is, he *didn't* do it.
That's true; Joseph Smith didn't fake anything, except sometimes he
would play jokes on people, like pretending to be drunk and see if
they would react to what they knew by revelation or what they saw
with their eyes as seemingly certain evidence. When he translated
the Book of Mormon, he just rendered it simply as it stood.
> The test in the article at
> http://www.lds-mormon.com/numbersinthebookofmormon.shtml
> Is a negative test. It looks at the probability that something did
> *NOT* happen -- namely that the Book of Mormon dates probably did
> *NOT* come from a random distribution; i.e., real history.
>
> As Woody has just shown, it's easy to come up with a string of numbers
> that is very near the mean of the random distribution.
So, you are now on record that you saw them. We will see how soon
you address their implications to your flawed statistical methods
in your essay. Remember, Duwayne, "the whole wide world's a watchin'".
>... Because of
> this, it's not possible to say that the dates in the Book of Mormon
> are *valid* no matter how close they are to the mean.
>
> But that's not what the test results showed. The test results showed
> that the Book of Mormon dates are *far* from the mean. The results
> show that the Book of Mormon dates are *unlikely* to happen in a
> random (real history) distribution of uncorrelated dates.
>
> See the difference, Woody? Should I explain it again?
I would rather see you address the flaws in your analysis. Sure,
the data you show exhibits a bias toward the front of the month.
You're a long way from demonstrating that your dates are a fair
sample; and you can't run "stats" on a "random sample" (deliberately
chosen) and bleat about how competent you are in the field of
statistics. Not when your competence is so obviously so thin as
demonstrated in your statistical calculations.
Tell, us, Duwayne, how does averaging the data show anything about
its distribution? This ought to be good.
Wood
Woody Brison
> The equations in the article will allow you to calculate all the
> probabilities yourself. Did you miss the link to the article? Here
> it is again:
>
> http://www.lds-mormon.com/numbersinthebookofmormon.shtml
I think that Mr. Duwayne is more than a little too impressed with
his own work. He has posted this link what, now, a dozen times?
Wood
Roy Stogner
> Roy Stogner <royst...@SPAMiname.com> wrote:
>
>> Using Duwayne's simplification of a 30 day month, your first set of
>> dates falls .68 standard deviations from the mean, an event of 49%
>> probability. Your second set of dates falls .749 deviations from the
>> mean, an event of 45% probability.
>>
>> Since those sets of dates come from essentially random sources,
>
> This is the point. Duwayne says he expects randomness. I ask why? About
> half of the events he shows would have, not random, but arbitrary dates.
> A good statistician knows the difference
I'm clearly not a good statistician. Would you explain the difference?
---
Roy Stogner
Roy Stogner
> In my personal opinion, Dwayne has shown us that the bom people used the
> mayan calendar of 20 days. His own statistical analysis shows that such
> a calendar would fall into the expected range.
Actually, the days in such a calendar would fall 2.6 standard deviations
away from the mean. This is only a 1/100 event, not 1/2000, but it's not
a complete improvement.
Do you have any theory for why the month dates in the calendar only
fall into the 1-11 range?
It's a shame there aren't a few more dates to look at; it's sad that we
apparantly can't decide whether Alma 49 takes 50 days or 150.
---
Roy Stogner
Roy Stogner
You sound like you are misinterpreting the meaning of my statement; I
suppose I shouldn't have made it so short. I'm annoyed that I wasted time
arguing over the differences in English phrasings of statements about
probability with you, when there are apparantly posters here who
completely fail to understand what a test like yours can say about
randomness, or even what the word "random" means. It's as if my home was
on fire and I spent time fiddling with the thermostat.
---
Roy Stogner
Roy Stogner
> markg...@aol.com (Markg91359) wrote in message
> news:<20020227185803...@mb-fp.aol.com>...
>> My thoughts on this subject are this. If one applies Duwayne's
>> statistical analysis one may well come up with a probability that
>> certain dates would fall within a certain range are only 1 out of 2000.
>>
>> However, I think true believers would tell you that religion by its
>> nature is *exceptional*. The averages aren't likely to apply and the
>> event is likely to be the 1 out of 2000 precisely because it is sacred,
>> ordained by God, etc.
>
> I wonder if there is a simple reason why the dates seem to be skewed
> toward the front of their months. Possible reasons:
>
> 1. Duwayne has cooked the data, excluded dates from the back half of
> months.
Which dates has he excluded?
Note: the reason why cdowis cannot be sure whether the back half of month
"dates" should be day number 30 or number 20 is because *they aren't dates*.
> 2. Historians who contributed to the Book of Mormon tended to note dates
> in the front of the month because it was cumbersome in their notation to
> write larger numbers. An unconscious bias perhaps.
Boy, I do love concordance.com:
"Word searched is: THIRTY, 37 occurrences"
And of course, several of those occurances are in the middle of "two hundred
and thirty", "five hundred thirty and two", "thirty thousand"...
> 3. People tended to do things on or soon after the beginning of the
> month -- kicking off military campaigns, assassinations connected to
> elections and festivals, things that are tied to the calendar.
Quite possibly. Can you find an example of this effect outside of the
Book of Mormon? That would be a lot more convincing than the bare
assertion.
> In a
> sample size of 8, it only takes 1 or 2 such to seriously bias the
> outcome,
It takes 5 or 6 such to bias the outcome as much as in the Book of Mormon.
> In simulations of submarine warfare, I noticed the hilarious tendency of
> ship/sub captains to set courses like 90, 180, 270, etc. People tend to
> shoot for round numbers.
Like 1, 2, 3, 4, and 5?
---
Roy Stogner
Duwayne Anderson
<snip>
> In my personal opinion, Dwayne has shown us that the bom people used
> the mayan calendar of 20 days. His own statistical analysis shows
> that such a calendar would fall into the expected range.
<snip>
I see Charles is still failing to read the threads on which he
responds. The claims made by Charles are, of course, false. Read on:
duwa...@hotmail.com (Duwayne Anderson) wrote in message news:<a42139e3.02022...@posting.google.com>...
> Jackie Chan <nos...@donotreply.net> wrote in message news:<3C793AF2...@donotreply.net>...
> > Buzzard wrote ===
> > What if their month's had 15 days instead of 29 or 30?
> >
> > JC comments ===
> > the chances improve dramatically to about 1 in 12
> >
> > I though the ancient Americans used a 20-day calendar, however, which
> > would yield a probability of only 1/100 for the Book of Mormon 8 point
> > days of the month data set
>
> There are a couple of things wrong with that. First, the Book of
> Mormon needs to be evaluated according to its specific claims. As
> I've pointed out several times, now, the Book of Mormon claims the
> ancient Americans were Hebrews, and as Hebrews they practiced the Law
> of Moses. An important part of the Law of Moses are the feasts and
> Holy Days, which are an integral part of the Hebrew calendar (see the
> reference in the article). So, it involves somewhat circular logic to
> claim that a 20-day month should be used because the ancient Americans
> used one, when the 20-day month of the Maya has no similarity
> whatsoever to the Hebrew calendar which is implied by circumstances
> described in the Book of Mormon. Instead, the fact that the ancient
> Americans did not use a Hebrew calendar is strong evidence against the
> Book of Mormon.
>
> That said, let's look at what happens if we use the 20-day calendar.
> For 20-day months the mean is 10.5*8 = 84 and the standard deviation
> is root 8*5.766 = 16.31. Since the sum of days for the Book of Mormon
> month dates is 41, this sum lies about (84-41)/16.31 = 2.636 standard
> deviations from the mean. The probability of being this far, or
> further, from the mean is about 1/110. Quite a bit better than the
> 1/2000 calculated for the situation using the Hebrew calendar.
>
> But with the assumption of 20 months the sum of the month is now also
> very improbable (it isn't if you assume, as I did, 13 months with 30
> days each). For 20-day months there are 18 months in a year. The
> mean is 9.5*8 = 76 and the standard deviation is (root 8)*5.993 =
> 14.674. The sum of the months in the Book of Mormon dates is 40 so
> this is (76-40)/14.674 = 2.453 standard deviations from the mean. The
> probability of a sum of months being this far or further from the mean
> is also about 1/110.
>
> In the original analysis I did not include the distribution of months
> because it was not dramatically removed from the mean. Doing so would
> have changed the probability from 1/2000 to about 1/3000. With the 18
> 20-day months, the distribution of months is now *also* highly
> unlikely. So one must not only rationalize away the probability of
> 1/110 for the days, but the equally improbably distribtion of months.
>
> So trying to use the 20-day month does not really help things much.
> The distribution is still very unlikely, and (more importantly) the
> assumption violates one of the key claims of the Book of Mormon --
> namely that the ancient Americans were transplanted Hebrews who
> supposedly lived the Law of Moses.
cdowis
> > Roy Stogner <royst...@SPAMiname.com> wrote:
> >
> > > On Wed, 27 Feb 2002 21:35:16 -0600, Joel Rees wrote:
> > >
> > > > I am giving you two sets of data which you may use as controls to check
> > > > your test. I told you where they came from. You can get your own data
> > > > sets from the same or similar places.
> > >
> > > Using Duwayne's simplification of a 30 day month, your first set of dates
> > > falls .68 standard deviations from the mean, an event of 49% probability.
> > > Your second set of dates falls .749 deviations from the mean, an event of
> > > 45% probability.
> > >
> > > Since those sets of dates come from essentially random sources,
> >
> > This is the point. Duwayne says he expects randomness.
>
> The assumption is that historical dates don't favor any particular day
> of the month. You apparently think they do. So, Joel, tell us which
> days of the month historical dates favor. You must think they favor
> the days during the first week. Right? Why should historical dates
> favor days during the first week of the month, Joel?
Simple. If the month has 20 days instead of 30, as you yourself aptly pointed out.
Woody Brison
> In my personal opinion, Dwayne has shown us that the bom people used
> the mayan calendar of 20 days. His own statistical analysis shows
> that such a calendar would fall into the expected range.
At the least, you've uncovered yet another hidden assumption
in Duwayne's essay: his standard of judgement is based on a
month of 28/29/30/31 days, applied to his sample of BofM dates.
There's no certainty that BofM dates are based on a 30 day
month, nor any discussion of that little item. One more
thing flipped away by the left hand while the right hand
dazzles us with graphs and references to books. Which, by
inference, he claims to have read. You wouldn't have skipped
that too, would you D.?
> The earliest date cited is in Alma, which is centuries after the first
> landing, and after the integration of the Mulekites into their
> culture.
>
> I would wildly speculate that the Mulekites, no longer keeping the
> Mosiac laws and having no scriptural records, began using the 20 day
> month calendar. The Nephites were the smaller group, and still
> probably had the Jewish calendar.
>
> Here is direct evidence that they were using the 20 day calendar of
> the Mulekites,
Yes, except that it's still subject to Duwayne's artificial
exclusion of dates which his computer search did not find. If
we do a real study and it still bears out this bias, then I
will credit you with a transcendant insight.
>... and, perhaps, also the Jewish calendar as well to keep
> track of the Mosaic rituals. A dual calendar system.
If they had a dual calendar system, and the idea is hard to
dismiss, and if the reported/implied dates are some from one
system and some from another, then it becomes a worthy puzzle.
Wood
Woody Brison
> > duwa...@hotmail.com (Duwayne Anderson) wrote...
> >
> > > No, Joel. You are demonstrating that you didn't read the article and
> > > that you don't understand the statistical test that was involved.
> >
>
> I wrote it. You can too, at
> http://www.lds-mormon.com/numbersinthebookofmormon.shtml
I could never write something that wacky.
But did you read it was my question.
If you can't be bothered to read it yourself, it wouldn't
seem too hard to run it thru a spellchecker...
Wood
Duwayne Anderson
> > wwbr...@lds.net (Woody Brison) wrote...
> > > duwa...@hotmail.com (Duwayne Anderson) wrote...
> > >
> > > > No, Joel. You are demonstrating that you didn't read the article and
> > > > that you don't understand the statistical test that was involved.
> > >
> > > I was wondering, Duwayne, did you read the article yourself?
> >
> > I wrote it. You can too, at
> > http://www.lds-mormon.com/numbersinthebookofmormon.shtml
>
> I could never write something that wacky.
But you can call it "wacky" without giving any logical/rational reason why.
Duwayne Anderson
<snip>
> That's true; Joseph Smith didn't fake anything,
That's not what I said, Woody. I said he didn't fake a random
distribution. Instead, he put in a very unlikely distribution, as you
can see in the article at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
<snip>
> I would rather see you address the flaws in your analysis. Sure,
> the data you show exhibits a bias toward the front of the month.
The bias is so strong that the probability of finding this
distribution in a real set of non-correlated historical dates is less
than 1/2000.
> You're a long way from demonstrating that your dates are a fair
> sample;
Really? Which dates in table 4 in the article at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml are not
"fair.?"
> and you can't run "stats" on a "random sample" (deliberately
> chosen) and bleat about how competent you are in the field of
> statistics.
This makes no sense. You need to explain yourself.
> Not when your competence is so obviously so thin as
> demonstrated in your statistical calculations.
Deal with the issues, Woody. Your ad hominem arguments won't save the
Book of Mormon.
> Tell, us, Duwayne, how does averaging the data show anything about
> its distribution? This ought to be good.
Woody, what "averaging" are you talking about?
Duwayne Anderson
Okay. Understood. Sorry.
Duwayne Anderson
How many times do you think I should post it, Woody, before you read
it and deal rationally/logically with what it says?
Duwayne Anderson
> cdo...@my-dejanews.com (cdowis) wrote in message news:<93c36e92.02022...@posting.google.com>...
>
> > In my personal opinion, Dwayne has shown us that the bom people used
> > the mayan calendar of 20 days. His own statistical analysis shows
> > that such a calendar would fall into the expected range.
>
> At the least, you've uncovered yet another hidden assumption
> in Duwayne's essay: his standard of judgement is based on a
> month of 28/29/30/31 days, applied to his sample of BofM dates.
> There's no certainty that BofM dates are based on a 30 day
> month, nor any discussion of that little item.
I see Woody has joined Charles in failing to read up on what's already
been covered in discussion on this thread. Read the following, Woody.
It might help:
duwa...@hotmail.com (Duwayne Anderson) wrote in message news:<a42139e3.02022...@posting.google.com>...
<snip>
Duwayne Anderson
<snip>
> It's actually a fallacy that ad hominem is a fallacy.
This must be why Mormons us the ad hominem argument so often, instead
of dealing rationally/logically with the issues.
<snip>
> That analysis is deeply flawed, as I have shown. You have not
> addressed those flaws.
And now Woody, in Dowis fashion, resorts to making naked assertions.
<snip to end>
Duwayne Anderson
<snip>
> 1. Duwayne has cooked the data, excluded dates from the back half
> of months.
This statement by Woody is false. The study used all month-dates
describing historical events. Woody's in a tizzy because I did not
include calendar events, namely the passing phrase that a certain year
had passed.
The funny thing is, that would not help Woody because the phrase is
used so often that it would stack the dates to the other extreme.
And, as I've already explained for Woody, calendar dates are not
randomly distributed.
> 2. Historians who contributed to the Book of Mormon tended to note
> dates in the front of the month because it was cumbersome in their
> notation to write larger numbers.
Now this is truly hilarious! Woody, you've outdone yourself.
> An unconscious bias perhaps.
Oh, we can see lots of bias here. It's almost fun watching you squirm
with these sorts of absurd excuses for the bias in the dates in the
Book of Mormon.
<snip to end>
Duwayne Anderson
> duwa...@hotmail.com (Duwayne Anderson) wrote in message news:<a42139e3.02022...@posting.google.com>...
> > wwbr...@lds.net (Woody Brison) wrote in message news:<f36171a3.02022...@posting.google.com>...
> > > cdo...@my-dejanews.com (cdowis) wrote in message news:<93c36e92.02022...@posting.google.com>...
> > >
> > > > May I also point out that he perhaps missed dates.
> > >
> > > A casual glance at his list suggests that he's gotten it
> > > by doing a computer search on the word "month".
> >
> > No kidding.
>
> So you did confine your data to the results of a computer search.
I used a computer search to make sure no events were excluded. And
none were.
>
> > > However,
> > > actually reading the book is a better way of knowing what
> > > it says.
> >
> > This was a statistical test that looked at the distribution of dates
> > in the Book of Mormon. So computer searches are not only appropriate,
> > they are necessary for completeness.
>
> That's hogwash.
No, it's statistical analysis.
> It's fine to use a computer search, but to rely
> on only that
Where did you get this "only" stuff? Did I say I relied *only* on
that?
> is to confine yourself to a partial set of data, and
> to put a blinder on yourself as to the real contents of the book.
No dates were excluded.
> The computer search does not understand themes.
Nobody said it did.
> It can only find
> instances of specific words.
No month dates were excluded. You have provided NO such reference to
any historical month dates that were excluded.
> There are many passages in books that
> talk about something but don't use just one word to talk about it.
>
> > By the way, I've read the Book of Mormon cover to cover over 17 times.
> > How many times have you read it cover to cover?
>
> 583,410.
Okay, so you are afraid to answer the question. Is anyone surprised
by this?
> My assertion here is just as good as yours, anyone
> can claim anything on the internet.
As you keep demonstrating with your assertions about the study.
> But, your claim of 17 is
> not well supported by your obvious lack of knowledge of the book's
> contents.
And here is yet another unsubstantiated assertion from Woody.
> We've been over this before, so I leave you with it.
>
> > > As an example, his first date in the list is from Alma
> > > 52:1 -- Year of the Judges 26, month 1, day 1. Boy, that
> > > weights the dates to the front end of the month, like a fat
> > > lady stepping into the side of the boat!
> >
> > The order of the dates in the table has nothing to do with the way the
> > statistical analysis was carried out.
>
> My point had nothing to do with the order of dates.
Then what did you mean when you complained that the first date was
weighting the results? Don't you like the first date? Did Joseph
Smith make a mistake in using it?
>
> > <snip>
> > > He's missed lots of other dates similarly.
> >
> > This statement by Woody is false. The study included all month-dates
> > in the Book of Mormon.
>
> I've given you several examples of dates you left out.
I quoted the verses you "gave." There were no dates in them.
> Why don't
> you go ride your horse?
Feeling in a tight spot, Woody?
> You're not much good at debating.
Well, why don't you give me some lessons. You can start by dealing
rationally with the article on dates in the Book of Mormon.
>
> > I was an active member of the LDS Church for over 35 years, and held
> > many positions in the Church, including Elder's Quorum President in
> > two wards. I've been married in the temple, gone on a mission, and
> > graduated from BYU.
>
> Those plus /enduring to the end/ will get you eternal life.
Is that what you are doing with your ad hominem arguments, Woody?
> Minus
> the enduring to the end they buy you nothing.
I see. So you are actually arguing for your eternal salvation.
Right?
> And, they do not
> give you knowledge of statistics, or of the book.
God does not know about statistics?
> I've known lots
> of missionaries who didn't actually study the scriptures, oodles of
> married people who don't, and a few EQ prezzes who don't.
And yourself, too, of course.
>
> > What Woody's doing is using his old standby ad hominem argument.
> > Trying to dodge the issues by focusing on the person.
>
> No. Your analysis is flawed.
Well, you need to stick to the issues and show that, instead of using
your ad hominem approach of trying to focus on me.
> I pointed out the flaws.
So you assert, in Charles Dowis style.
> My points
> have nothing to do with your Church involvement in the past, and in
> fact did not mention it.
Oh, but they do. It was part of your ad hominem argument.
> It's a nice try at claiming some kind of
> omniscience, or something, but it does nothing for your flawed
> stats presentation.
Well, deal with it then. Rationally.
>
> > > but can't seem to demonstrate that he's ever read it;
> >
> > For other readers just joining, Woody is still smarting from a bad
> > loss a few years ago in which he claimed that the Book of Mormon's
> > Arabian river had been found.
>
> Your assessment of that battle is as bad as the rest of your debating
> skills.
Well, why don't you try again, Woody? Tell us all about how the
Arabian river has been found. Make sure to demonstrate its existence
with verifiable evidence that's consistent with the Book of Mormon.
>
> > Care to have this discussion again, Woody?
>
> No, you are useless as a source of information, insight, or method.
Well, why don't you show us the truth, then?
>
> > > before
> > > any numbers are even entered, his analysis is only going to
> > > be able to show that the book is fake, not genuine;
> >
> > The statistical distribution was wrong because a critic did the math?
> > This is really lame, Woody.
>
> My point was that your test was devised to fail the book.
How would you know, and why should it make any difference?
> It
> was not devised to test the book.
Another assertion, with no explanation.
> It is not a real test;
Why?
> it is
> a deception aimed at persuasion away from something wonderful.
Oh. I see. It gave a result you don't like. So you are duty bound
to reject it. Is that how it works?
> It is the fox that lost his tail all over again.
What is this supposed to mean?
>
> > > and it
> > > depends on some pretty bizarre statistical methods to get
> > > its conclusions.
> >
> > If you think that calculating means and standard deviations, and
> > deriving probabilities based on location in the distribution is
> > "bizarre" I suggest a remedial class in mathematics and statistics.
>
> You took a mean of the data but did not calculate any s.dev over
> any of it.
This statement by Wood is false, as anyone can verify by actually
reading the article at
http://www.lds-mormon.com/numbersinthebookofmormon.shtml
> You did not account for the circular nature of the days
> in the month -- 30/31 is followed by 1,2,3...
How does this circular nature cause historical dates to favor the
first week of the month, Woody?
> Your calculations
> are cute but the rationale behind them is almost blank.
Says the guy who never read the article.
Duwayne Anderson
> duwa...@hotmail.com (Duwayne Anderson) wrote in message news:<a42139e3.02022...@posting.google.com>...
> > wwbr...@lds.net (Woody Brison) wrote in message news:<f36171a3.02022...@posting.google.com>...
> >
> > <snip>
> > > There are many others. But, if you
> > > think about it, where you've got the last day of one year,
> > > you've got the first day of the next year. You're really
> > > being told about the year rolling over, a lot of dates are
> > > implied in there
> >
> > Historical dates are specific, named, historical events in which the
> > month and day are both mentioned. The end of the year is a calendar
> > date. Calendar dates are not randomly distributed. So "the first day
> > of the year" and the "last day of the year," (or similar phrases)
> > don't constitute historical dates unless associated with a specific
> > non-correlated historical event.
>
> "The great enemy of clear language is insincerity. When there is a gap
> between one's real and one's declared aims, one turns as it were
> instinctively to long words and exhausted idioms, like a cuttlefish
> squirting out ink." -- George Orwell
Woody proves his point by acting the roll. I described in simple,
rational language why calendar dates (like the end of the year) are
not randomly distributed and Woody ignores it all and quotes George
Orwell instead.
> If you wish to exclude certain recorded dates from the sample set, you
> need to explain why those, and not the ones you want to include.
I did. I said that the study was based on historical dates. I showed
that historical dates are distributed randomly over the days of the
month. Calendar dates fall on the same day of the month. They are not
randomly distributed.
Having said that, let me point out that, in his mad rush to sweep away
the statistically unexpected distribution of dates in the Book of
Mormon, Woody is actually setting himself up. That's because, if you
include all the times that the Book of Mormon mentions the last day of
the year (a calendar date) the distribution switches. It becomes even
*MORE* unlikely, but for a different reason -- all those mentions of
the last day of the year.
Of course using calendar dates is not proper for the reasons that I
just described. But this illustrates the almost frantic manner in
which the LDS try any ad hoc excuse available to save their book from
critical analysis.
> And,
> the fact that the ones you want to include make the book look
> contrived is not a good reason.
As I said, Woody, I restricted the dates I used to dates of historical
events. I excluded only calendar events. I included all historical
dates where a day and month were given.
> It would be an obvious case of
> cooking the data to urge a predetermined conclusion.
Well, it's clear who is trying to cook the data to support their
predetermined conclusion.
Nice try.
Duwayne Anderson
<snip>
> Simple. If the month has 20 days instead of 30, as you yourself aptly pointed out.
Charles, I did not say that 20-day months make the problem go away.
In fact, it makes the problem even worse. Here is how I answered this
objection before. You need to deal with these issues, instead of
simply repeating yourself:
Duwayne Anderson
Joel Rees
> Have you ever looked at veterinary trials?
And we wonder how all those killer wonder drugs get on the market.
This one _especially_ fails to pursuade me.
Joel
Joel Rees
[stuff snipped about the difference between arbitrary and random]
> I'm clearly not a good statistician. Would you explain the difference?
If the shoe fits, _wear_it_with_pride_.
Roy, if you don't understand the difference, or have chosen to ignore
it, you'd better not be working professionally in statistics. If you
are in charge of approving drugs, keep me away from them. Or, I would
hope your professional work is more rigorous than your hobby work.
For those who may not be aware, arbitrary processes are those which an
experimenter can't control, or which he refrains from controlling in
an experiment. Random processes are processes whose results are not
predictable in the individual case but when collected show a specific
sort of arbitrariness that can be described by the mathematics of
statistics.
Proper analysis has to not simply assert that an experiment should
show a particular kind of randomness, but should show why.
Joel
cdowis
> cdo...@my-dejanews.com (cdowis) wrote in message news:<93c36e92.0202...@posting.google.com>...
>
> <snip>
> > Simple. If the month has 20 days instead of 30, as you yourself aptly pointed out.
>
> Charles, I did not say that 20-day months make the problem go away.
> In fact, it makes the problem even worse. Here is how I answered this
> objection before. You need to deal with these issues, instead of
> simply repeating yourself:
>
> First, the Book of Mormon needs to be evaluated according to its
> specific claims. As
> I've pointed out se
> veral times, now, the Book of Mormon claims the
> ancient Americans were Hebrews, and as Hebrews they practiced the Law
> of Moses.
European Jews practice the Law of Moses, to a certain extent.
An important part of the Law of Moses are the feasts and
> Holy Days, which are an integral part of the Hebrew calendar (see the
> reference in the article).
An important cultural aspect for the European Jews are their
festivals.
So, it involves somewhat circular logic to
> claim that a 20-day month should be used because the ancient Americans
> used one, when the 20-day month of the Maya has no similarity
> whatsoever to the Hebrew calendar which is implied by circumstances
> described in the Book of Mormon.
European Jews use the Julian calendar, in case you were not aware.
That does not prevent them from practicing their religion. Indeed
American Jews also use that calendar, but can still follow their
religion
Instead, the fact that the ancient
> Americans did not use a Hebrew calendar is strong evidence against the
> Book of Mormon.
How sad that your mind is so closed and your eyes are so blind. Look
around you. There are Jews practicing their religion everywhere,
Soviet Union, America, Africa, even China. Now you tell me what
calendar they use.
Open your eyes and your mind. They use two calendars.
>
> That said, let's look at what happens if we use the 20-day calendar.
> For 20-day months the mean is 10.5*8 = 84 and the standard deviation
> is root 8*5.766 = 16.31. Since the sum of days for the Book of Mormon
> month dates is 41, this sum lies about (84-41)/16.31 = 2.636 standard
> deviations from the mean. The probability of being this far, or
> further, from the mean is about 1/110. Quite a bit better than the
> 1/2000 calculated for the situation using the Hebrew calendar.
Which proves they were using the mayan calendar, as the european Jews
use the modern calendar.
>
> But with the assumption of 20 months the sum of the month is now also
> very improbable (it isn't if you assume, as I did, 13 months with 30
> days each). For 20-day months there are 18 months in a year. The
> mean is 9.5*8 = 76 and the standard deviation is (root 8)*5.993 =
> 14.674. The sum of the months in the Book of Mormon dates is 40 so
> this is (76-40)/14.674 = 2.453 standard deviations from the mean. The
> probability of a sum of months being this far or further from the mean
> is also about 1/110.
>
> In the original analysis I did not include the distribution of months
> because it was not dramatically removed from the mean. Doing so would
> have changed the probability from 1/2000 to about 1/3000. With the 18
> 20-day months, the distribution of months is now *also* highly
> unlikely.
You will note that several of the dates involve wars. Read the paper
on the **seasonality** of war in the bom. If you will read it, I will
get you the reference.
So one must not only rationalize away the probability of
> 1/110 for the days, but the equally improbably distribtion of months.
This has already been handled.
>
> So trying to use the 20-day month does not really help things much.
> The distribution is still very unlikely, and (more importantly) the
> assumption violates one of the key claims of the Book of Mormon --
> namely that the ancient Americans were transplanted Hebrews who
> supposedly lived the Law of Moses.
Open your eyes.
Roy Stogner
> Roy Stogner <royst...@SPAMiname.com> wrote:
>> I'm clearly not a good statistician. Would you explain the difference?
>
> If the shoe fits, _wear_it_with_pride_.
>
> Roy, if you don't understand the difference, or have chosen to ignore
> it, you'd better not be working professionally in statistics.
Did I say I was? I'm sorry if I left that impression; there are CAM
researchers here using statistics for economics simulations, but most
students are just interested in various PDE apps. The closest I've come
to stat is combinatorics years ago.
Your valuable explanation makes up for the rude delivery; thank you.
So we've established that the Book of Mormon dates appear to have "biases
of an arbitrary nature", and that Duwayne is only explicitly mentioning
one explanation for those biases. The alternative explanations Woody and
Charles Dowis have offered sounded somewhat weak to me; do you have any to
add?
---
Roy Stogner
Duwayne Anderson
>
> [stuff snipped about the difference between arbitrary and random]
>
> > I'm clearly not a good statistician. Would you explain the difference?
>
> If the shoe fits, _wear_it_with_pride_.
>
> Roy, if you don't understand the difference, or have chosen to ignore
> it, you'd better not be working professionally in statistics. If you
> are in charge of approving drugs, keep me away from them. Or, I would
> hope your professional work is more rigorous than your hobby work.
Joel, Roy has demonstrated far more understanding about statistics
than you have. Your rather comical attempts to describe why
historical dates should be biased toward the first week of the month
are a case study in how superstition can fry a person's ability to
think rationally.
> For those who may not be aware, arbitrary processes are those which an
> experimenter can't control, or which he refrains from controlling in
> an experiment. Random processes are processes whose results are not
> predictable in the individual case but when collected show a specific
> sort of arbitrariness that can be described by the mathematics of
> statistics.
All of which illustrates Joel's total confusion on the matter. The
study at http://www.lds-mormon.com/numbersinthebookofmormon.shtml uses
historical dates in the Book of Mormon. Real historical dates don't
show a preference for any particular day of the month. They are as
likely to fall on the first day as the last, or any day in between.
The Book of Mormon, though, shows something different. The historical
month-dates in the Book of Mormon have a distinct tendency to fall
during the first week of the month. The probability of this happening
with real (unbiased) historical dates is less than 1/2000.
> Proper analysis has to not simply assert that an experiment should
> show a particular kind of randomness, but should show why.
Here Joel illustrates that he's not even bothered to read the article
he's trying to argue against. Several examples were used to
illustrate the fact that real historical dates don't favor the first
week of the month, the way they do in the Book of Mormon. But Joel is
so determined to continue his faith in the Book of Mormon that no
evidence will ever convince him the book is a fake.
Duwayne Anderson
Nobody ever claimed that logic or reason would "pursuade" someone who
thinks the Book of Mormon is the product of an "unseen world" where
the "scientific method does not work," Joel.
Duwayne Anderson
>
> > In my personal opinion, Dwayne has shown us that the bom people used the
> > mayan calendar of 20 days. His own statistical analysis shows that such
> > a calendar would fall into the expected range.
>
> Actually, the days in such a calendar would fall 2.6 standard deviations
> away from the mean. This is only a 1/100 event, not 1/2000, but it's not
> a complete improvement.
Not only that, if you go to a 20-month year, the distribution of
months has a probability of just 1/100. So instead of making the
problem go away, it only makes it worse, because now, instead of one
distribution with odds of 1/2000 you have two distributions of 1/100.
> Do you have any theory for why the month dates in the calendar only
> fall into the 1-11 range?
Yes. He does. Charles' "theory" is that the Book of Mormon is true,
and that any ad hoc argument that Charles thinks supports the Book of
Mormon is thus valid.
<snip to end>
Jackie Chan
In the original analysis I did not include the distribution of months
because it was not dramatically removed from the mean. Doing so would
have changed the probability from 1/2000 to about 1/3000. With the 18
20-day months, the distribution of months is now *also* highlyunlikely.
cdowis ===
You will note that several of the dates involve wars. Read the paper on
the **seasonality** of war in the bom.
JC comments ===
seasonality doesn't affect dates within a month - the
days-of-the-month notices present the chief problem with Book of Mormon
(p < 1/2000)
further, the distribution of months fits a 12 month model very nicely
(as Duwaynea notes), but not 20 (p < 1/110) - there is no basis here to
propose a 20 month calendar for the Book of Mormon
finally, European Jewry uses a 12 month calendar, as does the Law of
Moses, AND as did the 19th century author(s) of the Book of Mormon - why
would you expect the White American "Nephites" to try something
different?
cdowis
> Duwaynea ===
> In the original analysis I did not include the distribution of months
> because it was not dramatically removed from the mean. Doing so would
> have changed the probability from 1/2000 to about 1/3000. With the 18
> 20-day months, the distribution of months is now *also* highlyunlikely.
>
> cdowis ===
> You will note that several of the dates involve wars. Read the paper on
> the **seasonality** of war in the bom.
>
> JC comments ===
> seasonality doesn't affect dates within a month - the
> days-of-the-month notices present the chief problem with Book of Mormon
> (p < 1/2000)
There are *two* separate issues -- the dates within the month and the
months themselves.
I dealt with both.
1. Dates within the month -- the nephites used the mayan calendar.
Dwayne's analysis shows that this would be within the expected range.
2. Months -- five of the eight dates are related to war-time
activities. As documented by FARMS, the wars in the bom are
seasonally related. Thus the months are skewed by seasonality.
>
> further, the distribution of months fits a 12 month model very nicely
> (as Duwaynea notes), but not 20 (p < 1/110) - there is no basis here to
> propose a 20 month calendar for the Book of Mormon
I do not understand your point. I said a 20 day month.
>
> finally, European Jewry uses a 12 month calendar,
You missed my point. They use a **dual** calendar. Dwayne insists
upon the jewish calendar ONLY, but I have proven that the Jews can
happily use two different calendar systems.
as does the Law of
> Moses, AND as did the 19th century author(s) of the Book of Mormon - why
> would you expect the White American "Nephites" to try something
> different?
Re-read the post. You are confusing the issues.
Duwayne Anderson
<snip Charles Dowis' ad hoc excuses about the Hebrews in the Book of
Mormon using the Mayan calendar>
> > But with the assumption of 20 months the sum of the month is now also
> > very improbable (it isn't if you assume, as I did, 13 months with 30
> > days each). For 20-day months there are 18 months in a year. The
> > mean is 9.5*8 = 76 and the standard deviation is (root 8)*5.993 =
> > 14.674. The sum of the months in the Book of Mormon dates is 40 so
> > this is (76-40)/14.674 = 2.453 standard deviations from the mean. The
> > probability of a sum of months being this far or further from the mean
> > is also about 1/110.
> >
> > In the original analysis I did not include the distribution of months
> > because it was not dramatically removed from the mean. Doing so would
> > have changed the probability from 1/2000 to about 1/3000. With the 18
> > 20-day months, the distribution of months is now *also* highly
> > unlikely.
>
> You will note that several of the dates involve wars. Read the paper
> on the **seasonality** of war in the bom. If you will read it, I will
> get you the reference.
<snip to end>
And so we see how the LDS apologetic mind operates. Charles, without
doing any actual thinking about it, and without any direct evidence
tries to save the Book of Mormon by claiming ad-hoc fashion that they
used a 20-month calendar. But the problem is, the historical dates
are even *more* unlikely with the 20-month calendar because now both
the month-dates *and* the months themselves have very unlikely
distributions.
To save himself and the Book of Mormon from this particular
embarrassment, Charles tries to argue that the historical dates
included wars, and that wars were seasonal.
Then, in typical Dowis circular style, he says that if I read the
paper on seasonality of wars he will get me the reference for the
paper I'm supposed to read.
Jackie Chan
Dates within the month -- the nephites used the mayan calendar. Dwayne's
analysis shows that this would be within the expected range.
JC comments ===
no, using Duwayne's method, an 18 month Mayan calendar yields months
of 20 days each, with a probability < 1/110 for the complete data set,
Sum(12,4,5,10,1,3,2,4) = 41
this is 2.6 Standard Deviations from the population mean for randomly
generated dates (84.1, SD=16.4)
do the math!
cdowis ===
five of the eight dates are related to war-time activities. As
documented by FARMS, the wars in the bom are seasonally related. Thus
the months are skewed by seasonality
JC comments ===
no, in fact several of the dates have nothing at all to do with war
(Alma 10:6, 14:23, Third Nephi 8:5), and the wars that are mentioned
began, BEFORE the beginning of the year (read Alma 52:1), yet no month
is mentioned past number Eleven, strongly suggesting a 12 month year is
in the author's mind
JC wrote ===
finally, European Jewry uses a 12 month calendar,
cdowis ===
You missed my point. They use a **dual** calendar. Dwayne insists upon
the jewish calendar ONLY, but I have proven that the Jews can happily
use two different calendar systems.
JC comments ===
and the dual calendars resembled each other very closely, not only in
number of months per year, but in days per month
the main reason Jews used both calendars was pressure from the
Gentile society they lived within, but this is obviously not a factor in
the Book of Mormon story
Duwayne Anderson
<snip>
> There are *two* separate issues -- the dates within the month and the
> months themselves.
>
> I dealt with both.
Actually, Charles. No. You have not dealt with them. You have, for
example, ignored the point that I made, to the effect that the
distribution of months under your imaginary 20-month year is as
unlikely as the distribution of days, with each distribution having a
probability of less than 1/100.
So what you did, in your attempt to massage the data and save the Book
of Mormon from critical analysis, was like squeezing on a water
balloon. You reduced the probability of the day distribution from
1/2000 to about 1/100 (still not great odds, by any means) but you
*also* change the distribution of months by assuming 20-month years.
Now you have to come up with a credible explanation how there can be
20 months in the year, yet 8 randomly drawn months are all closely
weighted toward the front end. In a nutshell, what you've managed to
do in your attempt to bias the numbers and save the Book of Mormon is
to jump out of the proverbial frying pan and into the proverbial fire.
> 1. Dates within the month -- the nephites used the mayan calendar.
Well, as I've shown above, your ad hoc argument only makes the
probabilities worse. This reminds me of the kids I knew long ago who
thought they could solve simultaneous equations by guessing. But as
soon as they got an answer for one equation, they'd find out it no
longer worked in the next equation.
In your case, you have blissfully gone off thinking that all you had
to do was figure a way to massage the numbers and fix the distribution
of days. So you dream up this 20-month year. But you forgot about
the distribution of months, and the 20-month year makes the
distribution of months all wrong.
Besides that, your claim about the Mayan calendar is totally
inconsistent with the Book of Mormon's narrative. The Book of Mormon
specifically states that the Promised Land had been kept from other
nations as an inheritance for Lehi and his family. There were no
people there when Lehi arrived, by the Book of Mormon's *OWN*
statement. Yet the Maya are known to have lived in the Americas long
before Lehi is said to have arrived.
> Dwayne's analysis shows that this would be within the expected range.
Charles, before you stick both feet and both hands in your mouth, why
don't you read the replies I've been making. I've explained all this
to you, now, at least three times. When you say my analysis supports
your 20-month year you are lying.
> 2. Months -- five of the eight dates are related to war-time
> activities.
So?
> As documented by FARMS, the wars in the bom are
> seasonally related.
This illustrates nicely the circlar nature of LDS apologetics. "Why
should wars be seasonal? Why, because the Book of Mormon's dates are
skewed, of course, and we know it's right, so there must have been
seasonal wars."
> Thus the months are skewed by seasonality.
What possible, logical explanation can you give for wars being
seasonal? There is nothing in the Book of Mormon that suggests this
-- beyond the circular logic that you and FARMS bring to the debate.
<snip to end>
cdowis
> cdo...@my-dejanews.com (cdowis) wrote in message news:<93c36e92.02030...@posting.google.com>...
>
> <snip Charles Dowis' ad hoc excuses about the Hebrews in the Book of
> Mormon using the Mayan calendar>
>
> > > But with the assumption of 20 months the sum of the month is now also
> > > very improbable (it isn't if you assume, as I did, 13 months with 30
> > > days each). For 20-day months there are 18 months in a year. The
> > > mean is 9.5*8 = 76 and the standard deviation is (root 8)*5.993 =
> > > 14.674. The sum of the months in the Book of Mormon dates is 40 so
> > > this is (76-40)/14.674 = 2.453 standard deviations from the mean. The
> > > probability of a sum of months being this far or further from the mean
> > > is also about 1/110.
> > >
> > > In the original analysis I did not include the distribution of months
> > > because it was not dramatically removed from the mean. Doing so would
> > > have changed the probability from 1/2000 to about 1/3000. With the 18
> > > 20-day months, the distribution of months is now *also* highly
> > > unlikely.
> >
> > You will note that several of the dates involve wars. Read the paper
> > on the **seasonality** of war in the bom. If you will read it, I will
> > get you the reference.
> <snip to end>
>
> And so we see how the LDS apologetic mind operates.
I make a proposal and you respond. That is how discussions are
conducted.
Charles, without
> doing any actual thinking about it, and without any direct evidence
> tries to save the Book of Mormon by claiming ad-hoc fashion that they
> used a 20-month calendar. But the problem is, the historical dates
> are even *more* unlikely with the 20-month calendar because now both
> the month-dates *and* the months themselves have very unlikely
> distributions.
Could you be specific.
>
> To save himself and the Book of Mormon from this particular
> embarrassment, Charles tries to argue that the historical dates
> included wars, and that wars were seasonal.
If you will look at the dates, five of the eight dates have to do with
wars. If you read the references you give us you will discover that
for yourself.
>
> Then, in typical Dowis circular style, he says that if I read the
> paper on seasonality of wars he will get me the reference for the
> paper I'm supposed to read.
I suspected that you would not be intested, but I tried.
Woody Brison
> wwbr...@lds.net (Woody Brison) wrote...
>
>
> I used a computer search to make sure no events were excluded. And
> none were.
How do you know that? Because the computer didn't turn up any,
I gather. Or did you perhap not even author this rubbish, but
just copied it off some website somewhere?
> No month dates were excluded. You have provided NO such reference to
> any historical month dates that were excluded.
You are right, I believe; I have not studied the entire book
targeting this specific thing. On the other hand, I don't
think you have either.
> > > > As an example, his first date in the list is from Alma
> > > > 52:1 -- Year of the Judges 26, month 1, day 1. Boy, that
> > > > weights the dates to the front end of the month, like a fat
> > > > lady stepping into the side of the boat!
> > >
> > > The order of the dates in the table has nothing to do with the way the
> > > statistical analysis was carried out.
> >
> > My point had nothing to do with the order of dates.
>
> Then what did you mean when you complained that the first date was
> weighting the results? Don't you like the first date? Did Joseph
> Smith make a mistake in using it?
I'm going to cease and desist these little explanations about
straws. You obviously have all day and night to niggle over
every single one, but I don't and doubt that I ever will, in
all of eternity. Most of them you could get with a little
thought.
Wood
Duwayne Anderson
<snip>
> How do you know that [no month-dates were excluded in the study described
> at http://www.lds-mormon.com/numbersinthebookofmormon.shtml?
> Because the computer didn't turn up any,
> I gather.
Well, anyone can do their own search using engines like the one at
http://www.hti.umich.edu/m/mormon/prox.html, and easily verify the
dates I used. And then there's also the fact that nobody's posted any
month-dates that are not included in the study.
> Or did you perhap not even author this rubbish, but
> just copied it off some website somewhere?
I wrote it, and take full credit for the article and all the
consternation it's causing you.
Duwayne Anderson
> duwa...@hotmail.com (Duwayne Anderson) wrote in message news:<a42139e3.02030...@posting.google.com>...
> > cdo...@my-dejanews.com (cdowis) wrote in message news:<93c36e92.02030...@posting.google.com>...
> >
> > <snip Charles Dowis' ad hoc excuses about the Hebrews in the Book of
> > Mormon using the Mayan calendar>
> >
> > > > But with the assumption of 20 months the sum of the month is now also
> > > > very improbable (it isn't if you assume, as I did, 13 months with 30
> > > > days each). For 20-day months there are 18 months in a year. The
> > > > mean is 9.5*8 = 76 and the standard deviation is (root 8)*5.993 =
> > > > 14.674. The sum of the months in the Book of Mormon dates is 40 so
> > > > this is (76-40)/14.674 = 2.453 standard deviations from the mean. The
> > > > probability of a sum of months being this far or further from the mean
> > > > is also about 1/110.
> > > >
> > > > In the original analysis I did not include the distribution of months
> > > > because it was not dramatically removed from the mean. Doing so would
> > > > have changed the probability from 1/2000 to about 1/3000. With the 18
> > > > 20-day months, the distribution of months is now *also* highly
> > > > unlikely.
> > >
> > > You will note that several of the dates involve wars. Read the paper
> > > on the **seasonality** of war in the bom. If you will read it, I will
> > > get you the reference.
> > <snip to end>
> >
> > And so we see how the LDS apologetic mind operates.
>
> I make a proposal and you respond.
That's right. I showed that your proposal of a 20-month year is
inconsistent with what the Book of Mormon describes, and not only
that, it does not even get rid of the problem. The distribution of
dates looks even more unlikely to have come from a real, unbiased
history because now the months are all skewed.
> That is how discussions are
> conducted.
But you don't hold up your end of it, Charles. You don't deal with
the issues I raise. You don't respond with verifiable evidence,
logic, or reason. You just invent another assertion and/or ad hoc
excuse.
> Charles, without
> > doing any actual thinking about it, and without any direct evidence
> > tries to save the Book of Mormon by claiming ad-hoc fashion that they
> > used a 20-month calendar. But the problem is, the historical dates
> > are even *more* unlikely with the 20-month calendar because now both
> > the month-dates *and* the months themselves have very unlikely
> > distributions.
>
> Could you be specific.
Yes, I can, Charles. It's right up there in the post you responded
to. The post that you responded to without even bothering to read.
Here, let me repeat it for you (this is about the sixth time, now):
But with the assumption of 20 months the sum of the month is now also
very improbable (it isn't if you assume, as I did, 13 months with 30
days each). For 20-day months there are 18 months in a year. The
mean is 9.5*8 = 76 and the standard deviation is (root 8)*5.993 =
14.674. The sum of the months in the Book of Mormon dates is 40 so
this is (76-40)/14.674 = 2.453 standard deviations from the mean. The
probability of a sum of months being this far or further from the mean
is also about 1/110.
See, Charles? By asserting a 20-month year you only made the problem
worse. Now you have a distribution of days that is found in only
1/100 non-biased histories *AND* a distribution of months that is
*also* found in only 1/100 non-biased histories. You now have twice
as much explaining to do.
> >
> > To save himself and the Book of Mormon from this particular
> > embarrassment, Charles tries to argue that the historical dates
> > included wars, and that wars were seasonal.
>
> If you will look at the dates, five of the eight dates have to do with
> wars. If you read the references you give us you will discover that
> for yourself.
And this is how Charles biases the dates. He willy-nilly decides that
anything to do with wars is not part of a biased history.
>
> >
> > Then, in typical Dowis circular style, he says that if I read the
> > paper on seasonality of wars he will get me the reference for the
> > paper I'm supposed to read.
>
> I suspected that you would not be intested, but I tried.
And once again, Charles shows he cannot hold up his end of a
conversation. In every case, Charles, I've addressed each and every
one of your issues head on with verifiable evidence. You, in turn,
tell me to read a book, and that if I read the book you will tell me
what book to read. And when I ask you what book to read, you tell me
I "would not be interested," but that your "tried."
Such is the biased, circular, nature of LDS apologetics.
Woody Brison
> Actually, Charles. No. You have not dealt with them. You have, for
> example, ignored the point that I made, to the effect that the
> distribution of months under your imaginary 20-month year is as
This is why it's useless to try to examine anything in ARM,
or around D. Anderson, there's no connection to reality; just
a giant effort to obfusticate, at any cost.
Goodby.
Wood
Roy Stogner
>> Actually, the days in such a calendar would fall 2.6 standard
>> deviations away from the mean. This is only a 1/100 event, not 1/2000,
>> but it's not a complete improvement.
>
> Not only that, if you go to a 20-month year, the distribution of months
> has a probability of just 1/100. So instead of making the problem go
> away, it only makes it worse, because now, instead of one distribution
> with odds of 1/2000 you have two distributions of 1/100.
I thought of that, but the obvious response of "historical months may be
biased around one season" sounds at least plausible. People less likely
to start wars or assassinate each other during the rainy season, maybe.
---
Roy Stogner