How The Book of Mormon’s Dates Were Chosen: A Statistical Approach

to BOM Origins
Posted “in the eleventh month of the [2014th] year, on the tenth day of the month” (Alma 49:1) by TruthIsReason

Summary
A statistical analysis of the dates reported for events in the Book of Mormon has found that, even under the most
generous assumptions, it is virtually impossible for the dates to be genuinely historical. Persistent trends running
through them are so anomalous that their combined odds of occurring by chance are 1 in 1.5 billion. The nature of
these trends strongly suggests that the dates were selected by an author whose choices were heavily biased by the
numbers he had been exposed to most recently and most frequently.

Abstract
A series of rigorous Monte Carlo experiments was performed to determine whether the dates that appear in the
Book of Mormon are consistent with those of a plausible historical record or whether they are indicative of human
selection. A comparison of the dates’ months with their corresponding days found consistent proximity between
them to a degree that is significant at odds against chance of 440,000 to 1. A comparison of these months with the
months mentioned in the King James Bible then revealed a .90 correlation, which would arise by chance only once in
3,400 instances. The likelihood of these two independent phenomena both occurring by chance is 1 in 1.5 billion.
Per Bayes’ theorem, they arguably imply that the probability that the dates reported in the Book of Mormon are
historical is in the vicinity of 1 in 375 million – less than one’s odds of winning the Powerball Lottery. These
findings strongly suggest that the dates were selected in the mind of an author who allowed his choices to be heavily
influenced by the numbers he had been exposed to most recently and most frequently. More specifically, they
suggest that the Bible’s months affected the months he chose for these dates, which in turn affected the respective
days he chose for them. In addition, a very strange anomaly – one that would occur by chance only once in 20
million cases – is present in the dates and is also indicative of human selection. Peer-reviewed studies that have
found effects analogous to these while asking subjects to select numbers are cited. The basic premises used in this
study are listed and supported with extensive real-world data and multiple lines of reasoning. It is shown that even if
these premises did not hold, the paper’s findings would remain valid since there is no conceivable set of alternative
premises that could rescue the book’s dates from statistical implausibility. Anticipated objections and proposed
alternatives to this conclusion are addressed. Links to the tools, spreadsheets, source code, and instructions needed
to inspect my methodology and replicate my findings are provided. These include a small program that can be
installed and used to run the Monte Carlo experiments.

Introduction
The question of Book of Mormon historicity is as old and as fundamental as any in the field of Mormon studies. Is the
book a translation of an ancient, historical record as it purports to be? Or is it rather a fictional narrative – perhaps a
remix of the ideas and literature that Joseph Smith had access to or a creative attempt to settle the doctrinal
controversies of the Second Great Awakening? The implications that this question holds for the validity of the
Mormon worldview could hardly be greater; hence it has been tackled by countless researchers – Mormon and nonMormon – from myriad angles. One clever angle, which has only been lightly explored, is statistical in nature and
involves the substantial difference in randomness between the numbers that arise in various facets of nature and the
numbers that are conceived in the human mind.
You see, we humans are notoriously bad at generating random numbers. When we try to pick numbers randomly, we
tend to favor certain numbers over others and this has been shown in countless experiments. It has also been shown
that when we select numbers without putting considerable effort into doing it randomly, our selections are even
more biased. Other research has found that the numbers we select can be largely influenced by our culture and prior
exposure to other numbers. Now, if the biases in our selections are strong enough, they can be identified with
modern statistical analysis. As noted in a widely cited paper on random number selection by the famous late
psychologist W. A. Wagenaar, “for disproving randomness it is sufficient to show one type of systematic trend in the
series”.
An Example of Statistical Fraud Detection
Now bear with me for a paragraph or two while we do a little thought experiment to see how these points can be put
into practice. Suppose you are a professor at a university and teach a 100-level experimental physics class filled
mostly with naïve freshman. You task your students with a simple experiment and ask them to send their data to you
before the end of the day. The experiment is to hold a Geiger counter next to a radioactive substance and record the
times at which the counter registers a particle of radiation. They are to do this until they have recorded 20 such data
points. The next day, you check your inbox and find, to your surprise, that one of the students who sent you their
data is a student who has rarely been to class and has been inattentive when he has attended. You open the
spreadsheet attached to his email and notice that the time data he entered into it appears to be very evenly spaced –
more evenly, in fact, that any data you can remember ever seeing. This is strange because radioactive decay is a
random phenomenon and your experience tells you that random time intervals are typically much more varied.
You’re feeling suspicious, but then again, maybe you’re just biased by your negative perception of this seemingly
apathetic student. So you run a test to see whether your observations have any empirical significance. You calculate
the length of the time intervals (differences between each pair of consecutive points of time) in his spreadsheet and
then calculate their standard deviation in order to quantify their variation. You then open your statistical analysis
software and run a simulation that randomly selects 20 times of the day (using an appropriate probability
distribution) 10,000 times and calculates the standard deviation of the intervals contained in each of them. What you
find is that, out of the 10,000 trials, only 50 of them produced a standard deviation as low as the one produced by
this student’s data. This implies that the odds of such a low variation occurring in true data are around 1 in 200
(since ⁄
= ⁄ ) – a result 10 times more significant than the universally accepted 1 in 20 threshold for
statistical significance! Now if this was any ordinary student, you might write this off as a rare-but-probablyinnocuous anomaly; however, since you were already suspicious of this student and opened his assignment first as a
result, this finding is almost certainly meaningful. Adding to the significance is the fact that the anomalous pattern
has a simple, plausible explanation – the student fabricated the data, spreading it out evenly in an attempt to make it
look random. In the absence of any simpler, more plausible explanations, you conclude that this one is very likely to
be correct. You email the student to let him know that you have strong evidence that he faked the assignment and
give him a chance to confess to the deed.
Past Analyses of Book of Mormon Dates
Sherlock Holmes was no stranger to this method of fraud detection. In one story from the book Conned Again,
Watson, he used such a method to determine – with odds against chance of 170 to 1 – that a particular list of
birthdays was phony. Duwayne R. Anderson, an author and highly accomplished engineer, summarized this story in
an article he wrote in early 2002 that became an appendix in his book Farewell to Eden: Coming to Terms with
Mormonism and Science (2003). In this article, he showed how a similar technique can be applied to the dates that
appear in the Book of Mormon (BOM). Inspired by this analysis, Lyndon Lamborn, also an engineer and author of the
book Standing for Something More: The Excommunication of Lyndon Lamborn (2009), conducted his own variation of
it and documented it in an appendix of his book, which can be read here (the appendix, not the whole book – scroll
down to the “Mathematical Viewpoint” section).
The gist of these analyses is this: There are a total of eight dates reported in the BOM (see Table 1 below), the events
of which are non-correlated and which we would not expect to be more likely to occur on any particular day of the
month (as we will discuss in more detail later). Normally, eight dates probably wouldn’t be a large enough collection
to find anything significant in, but in this case, the skew in their distribution is so strong that they provide more than
enough to work with. The interesting thing about these dates is that their days of the month are all low numbers.
They are all 12 or lower and 75% of them are 5 or lower. Anderson found that the odds of the sum of these days (41)
being as far as it is from the expected mean (124) by random chance is about 1 in 2,000 (it is actually closer to 1 in
3,400[1]; his analysis relied on the central limit theorem, which only provides good approximations near a
distribution’s peak and is less useful out on its tails, where the phenomenon in question lies). Lamborn found that
the odds of this sum being as low as it is by chance are around 1 in 7,200 (I replicated this simulation with 100
million trials and got about 1 in 6,800[2]). Moreover, simple math reveals that the odds of all eight days being 12 or

less are 1 in 1,526[3] and the odds of them being as tightly grouped as they are (in terms of their mean absolute
deviation) is about 1 in 280.[4]
Table 1. All dated events in the Book of Mormon in order of appearance.

Verse
Alma 10:6
Alma 14:23
Alma 16:1
Alma 49:1
Alma 52:1
Alma 56:1
Alma 56:42
3 Ne 8:5

Year
82 BC
82 BC
81 BC
72 BC
66 BC
62 BC
~64 BC
34 AD

Month Day Event
7
4
An angel converts Amulek
10
12 God delivers Alma & Amulek from prison
2
5
Lamanites begin a war against Nephites
11
10 Lamanites are deterred from attacking a Nephite city
1
1
Lamanites find Amalackiah killed the night before by Nephites
1
2
Captain Moroni receives an epistle from Helaman
7
3
Stripling warriors defeat Lamanites & none are slain
1
4
Jesus begins the post-crucifixion destruction

Also inspired by Anderson’s analysis was one from Ed Bliss, the author of the book What Mormons Don't Know about
Mormonism (2012). His observation was that “the first six events all occurred during the first five days of the first
week”. This is actually not the case, however, as you can see in the table above. What he was looking at was the table
Anderson provides in his article and book, which is sorted by month and then by day rather than by verse (it is
interesting, however, that six of the eight dates are in the first 5 days of the month, and simulations show that the
odds of this occurring by chance are about 1 in 2,300[5]). Now, even though Bliss was mistaken on this detail, he was
on to something. What he really found is that if you sort the events by their place in the calendar year, the first six
dates fall within the first five days of the month. Why does this happen? Is it a fluke? As it turns out, it’s actually a
symptom of a much larger issue, as we will soon see…

Methodology and Resources
Before we proceed, I should give you some notes regarding my overall methodology and describe the resources I’ve
made available for download that will allow you to check and replicate everything I’ve done for this project (if you’re
so inclined). Any details not covered here will probably be covered in the remainder of the paper or in the resource
files themselves.
The Monte Carlo Method
The bulk of this research project took the form of Monte Carlo experiments, which are a widely used and universally
accepted method of determining probabilities. In a nutshell, they are randomized simulations that are run
repeatedly a large number of times, often in order to determine the probability that a scenario will produce a
particular outcome. In our case, I ran them in order to determine the likelihood that a collection of eight noncorrelated dates – like those found in the BOM – would have the strange, persistent trends running through them
that the BOM’s dates happen to have. For example, in order to replicate Lamborn’s experiment, I randomly selected 8
days of the month, added them together, and repeated the process 100 million times in order to see how often a sum
as low as that of the days of the BOM’s dates occurs.
To see what I mean when I say that the BOM’s dates are “non-correlated”, look at the table above and notice that
most of these events are separated by multiple years and that there is no apparent, direct connection between any of
them based on what occurred in them (with one partial exception discussed in the “Premises” section). This, along
with many other reasons that will be presented later (including real historical event data), lead us to expect that a
collection of dates like those reported in the BOM will have months and days of the month that are randomly – or at
least nearly randomly – distributed. This is especially intuitive in the case of days of the month, which is by far the
more important premise of this study (since the part of the analysis that employs it (documented in the section
“Trend #1 – Month/Day Proximity”) yields results that are far more significant).
In order to make sure that the Monte Carlo experiments were of the highest possible quality, I used the latest version
of the world’s most widely trusted random number generator for simulations, SIMD-oriented Fast Mersenne Twister
(SFMT), the output of which is practically indistinguishable from true randomness. Moreover, the number of random
numbers this algorithm can generate before repeating its sequence (i.e., its “period”) is greater than the number of
atoms in the known universe raised to the 800th power.
I integrated this algorithm into a Visual C++ program that I developed and used to perform the experiments. Each
simulation included between 1 million and 10 billion trials depending on the rareness of the event being tested and
its importance to the study. The most important simulations were all run with at least 100 million trials. At the start
of each simulation, the SFMT algorithm was given a seed value of 0 in order to show that I did not cherry pick the
value.
These simulations are validated not only by the verifiable soundness of the logic in their programming (see the
source code described in the “Resources” subsection below) but also by the fact that three of them can be precisely
calculated with simple math and, in each case, the simulation results roughly match their respective calculations.[6]
Why the Bible and The Late War?
Throughout this paper, you will see charts that compare various attributes of the BOM’s dates to the same attributes
of the dates reported in the King James Version of the Bible (KJV) and in a book called The Late War Between the
United States and Great Britain. The Late War (TLW) is a history book on the War of 1812 whose first edition was
published in 1816 and was written in the biblical style. With all the thousands of books out there that give dates of
important events, why did I choose these two? I chose them simply because any other choice would have likely been
arbitrary and thus could have been cherry picked. These two books are the only date-reporting books that have been
empirically shown with computer algorithms to be exceptionally similar to the BOM in terms of the rare phrases
they contain (out of over 130,000 books analyzed – see this page for more on this research). These two books also
stand out in that, even prior to this research, they had both been alleged by other researchers to have served as
source materials for the BOM’s author due to the many thematic parallels existing between them and the BOM (for
example, see here and here). Note that unless otherwise specified, all biblical dates that are compared to BOM dates
in this paper include only “documentary dates” (defined below in the “Categorization of Bible Dates & Months”
subsection).
Random Dates in Charts
Another comparison you will see in this paper’s charts is between the BOM’s dates and randomly selected dates. In
each case, the random dates were generated from the first bits (0’s and 1’s) in the random data that was produced
on January 1, 2014 by random.org. This is arguably the least arbitrary of all dates for which data is available on the
site, and I chose it in order to show that I did not cherry pick it. This data file can still be downloaded here. I broke
the first bits of this text file into 4-bit and 5-bit segments and then converted them from binary to decimal to get
months and days, respectively. See the “Random Date Generation” section in the “Other Calcs” tab of Data &
Charts.xlsm (described in the “Resources” subsection below) to see exactly how I’ve done this.
Eight-Date Samples
Several of the charts herein refer to “8-date samples”. In these, the dates contained in the Bible and The Late War are
broken into groups of eight so that they can be compared more meaningfully with the eight dates in the BOM. Before
being broken into these groups, the dates were scrambled using a simple algorithm. Scrambling them (as opposed to
using them in the order in which they appear in their respective texts) was necessary in order to reduce the
correlation between the dates in each sample. This put them on more equal footing with the dates in the BOM, which
are more spread out in time and thus much less correlated (practically non-correlated as explained in the “Premises”
section). The samples are the same in every chart that uses them. For more on this, see the “Clustering” tab of Data &
Charts.xlsm.
Categorization of Bible Dates & Months
Before comparing the Bible’s dates to the BOM’s, it was necessary to determine which of them are actually
comparable. Our analysis of BOM dates is predicated on the fact that the events associated with the dates apparently
could have happened on any day of the month and in any month of the year. Therefore, the only dates that can be
meaningfully compared with them are those that fall into the same category. While some of the Bible’s dates are in
this category, most of them are not. Nearly half of its dates are given in conjunction with the Law of Moses and have

the purpose of either 1) establishing a commandment regarding a particular month or day of the month or 2)
showing that the Israelites were keeping such commandments. Others of its dates are blatantly symbolic, meaning
we have good reason to believe that the numbers in these dates are given for their symbolic value rather than for
their historical accuracy. For example, the dates in the Noah’s Ark story (which, for many reasons, is most certainly
not historical anyway) are full of 1s and 7s, which, as we’ll soon see, are the most significant numbers in the Law of
Moses, indicating that these dates were probably intended by the story’s author to be symbolic.
18 of the Bible’s dates are given in order to put timestamps on the revelations that its prophets received. Four of
these occurred in the first month and five of them occurred on the first day of the month, for a total of nine firsts.
Since 1 is the only number that is important for both months and days in the Law of Moses, this is especially worthy
of notice. The odds of it occurring by chance are about 1 in 6,000[7], making it nearly certain that either 1) at least
some of these dates were symbolic or 2) prophets were more likely to receive revelations on holy days and in holy
months. In either case, there is good reason to not expect these dates to be randomly distributed; thus they should
not be categorized with or compared to the dates in the BOM, which we should expect to be randomly distributed, as
we’ll see later.
Finally, there are a few dates in the Bible that have other reasons for not being comparable with those in the BOM. In
a nutshell, the criteria is simply this:
1. The date appears to be given solely for the purpose of documenting an important event (rather than for a
reason related to the date itself).
2. We have no reason to believe that the event should be more likely to happen during any particular month or
on any particular day of the month than any other.
If a date meets both criteria, we classify it as a “documentary date” (for lack of better terms) and say that it can be
meaningfully compared with the BOM’s dates. All eight of the BOM’s dates meet this criteria since none of the events
they document have anything to do with the Law of Moses or show any signs of symbolic significance, and there is
nothing inherent in their circumstances that should cause us to expect any of them to have occurred on any
particular day or month. They all appear to serve no other purpose than to document an important one-time event.
After carefully going through all 106 of the dates reported in the Bible and all 133 of the months mentioned in it
(some of which are duplicates), I have determined that 20 of its dates should be considered documentary dates and
that 35 of its months should be considered documentary months (not every month has a day attached, so there are
more months than dates). In order to document this classification as thoroughly as possible, I have assigned every
date and month to a category and, in cases where the reasoning for the category assigned might not be immediately
obvious, I have provided explanatory notes.[8]
Resources
The file BOM Date Resources.zip is a small (~6MB) zip file you can download here that contains all files needed to
check and replicate my findings. These files will be referred to throughout this paper, particularly in in-line citations,
most of which point to the “Notes” section at the end. Specifically, the zip file includes the following:

BOMDateSim Installer.exe – A small (0.6 MB) installation package that will install the
program BOM Date Simulator onto your computer so that you can replicate my Monte Carlo
simulations. Just install it onto any computer with Windows that is less than a decade or so
old and find the icon to the right on your desktop or in your Start menu. The program is very
lightweight and fast (some of the simulations will let you run a million trials in the blink of an
eye) and can be uninstalled like any other program. Check the “Repeat” box in order to get the
exact results documented in this paper (this sets the seed value to 0 instead of letting it vary);
otherwise, your results will be slightly different each time. If you have any issues installing it, you might need
to tell your antivirus software to ignore it (since it hasn’t yet had enough users to develop a reputation).

BOMDateSim Source Code.zip – A folder containing all source code files for the BOM Date Simulator along
with a set of in-depth instructions that explains how to use free tools to compile the code into a functioning
program.

Simulator Results.xlsm – A spreadsheet that documents the results of all 17 of the simulations that can be run
in the BOM Date Simulator in addition to five more that can be run within the spreadsheet itself. It also
explains what each simulation does.

Data & Charts.xlsm – A spreadsheet containing all charts and tables included in this paper (plus a few more)
along with the data and calculations behind them. It also contains some of the other calculations mentioned
herein as well as many explanatory notes regarding various aspects of the research.

Historical Event Data.xlsm – A spreadsheet containing several sets of historical event data collected by
research projects at universities. All data sets are arguably relevant to BOM events and include dates of the
events of the US Civil War, starting dates of wars, and dates of other types of militarized conflict, terrorism,
protests, and international alliances. The data sets cover various time periods and geographic regions. The file
also contains charts and calculations pertaining to this data along with links to the data sets and notes on how
I filtered and organized the data.

Raw Historical Event Data.zip – A folder containing all original historical data files as well as the codebooks
prepared by the researchers who compiled them. Also included are spreadsheets showing how I organized
and filtered the data before copying it into Historical Event Data.xlsm.

Note that all three of the spreadsheets contain macros whose code you can inspect by opening the file, pressing
Alt+F11 on your keyboard, pressing Ctrl+R on your keyboard, uncollapsing the “Modules” section of the Project
Explorer on the left side of the window, and double clicking “Module1”.
My goal in providing these resources has been to make this project completely transparent, and I’m confident that
I’ve succeeded in this regard. As a result, the information and arguments contained in this paper can all stand on
their own merits, meaning that one’s perceptions of my own credibility and motives are 100% irrelevant to the
conclusions that should be drawn (take note, apologists). If you think I’ve left anything out or made any mistakes,
please email me and I’ll be happy to fill in the gaps or make corrections.
This completes our lengthy but necessary detour. We can now get back to the good stuff…

Trend #1 – Month/Day Proximity
Let’s return for a moment to the simulations run by Anderson and Lamborn. Both authors note that their results
show that the days of the BOM’s dates are almost certainly not randomly distributed as they should be if the book is
a literal historical record. They conclude from this that the BOM is very likely not historical but rather written by an
author who was biased toward selecting low days of the month for some reason. This conclusion is straightforward
and seems reasonable enough, but even if it’s accurate, the question remains as
Table 2. Month/day distances of the
BOM’s dates.
to why the author had this particular bias and how he could have had it so
Date Month Day Distance strongly. What could have caused him to select such low days so consistently?
1
2
3
4
5
6
7
8

7
10
2
11
1
1
7
1

4
12
5
10
1
2
3
4
Dist Avg:
Dist Std Dev:

3
2
3
1
0
1
4
3
2.125
1.269

Month/Day Distance – Magnitude
Given that there are 30 days in a month, having every one of the BOM’s dates fall
within the first 12 days of the month is indeed strange. It’s almost as though the
numbers had been selected as months instead of days. This raises the question:
“Might these days have any relationship to the dates’ months?” A quick glance at
the table of dates we saw earlier seems to reveal the answer – strangely, each
day is very close to its corresponding month. In fact, on average, the distance
between the day and month of each date is only 2.125.[9] In contrast, the average
distance expected in any collection of random dates – is 10.6[10] and distances
can be as high as 29 (in the case of 01/30).

M/D Distances Compared
30
25
20
BOM

15

KJV

10

TLW

5

Rnd

0
1

2

3

4

5

6

7

8

Date
Figure 1. Month/day distances of the BOM’s dates compared to those of
representative eight-date samples from the KJV, TLW, and a large random population.

Is it plausible that such a low average could have arisen by chance? In order to find out, I ran a Monte Carlo
simulation that picked eight days at random a billion times and counted the number of instances in which the
average distance between these days and their corresponding months was 2.125 or less. Out of a billion trials, the
phenomenon occurred only 15,310 times, implying that it has about a 1 in 65,000 chance of happening.[11] This is
comparable to your odds of winning a raffle in which every resident of a suburban city entered their name.
M/D Distance Averages

M/D Distance Averages – 8-Date Samples

12.0

14.0

Rnd

12.0

10.0

KJV

TLW

10.0
8.0

8.0
6.0

6.0

4.0

4.0

2.0
2.0

BOM

0.0
B

0.0
BOM

KJV

TLW

R

K1 K2 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11
Sample

Rnd

Figure 2. Average month/day distance of the BOM’s dates compared to that of the KJV, TLW, and a large random population 1) in their
entirety (left) and 2) after being randomly broken into eight-date samples (right).

Month/Day Distance – Variation
While looking at the first chart above (“M/D Distances Compared”), you surely noticed that the BOM’s line (red) is
substantially lower than the others on average. Now take another glance at it and see if you can find anything else
that sets it apart. You might notice that while the other lines are varying wildly, the BOM’s is quite flat in comparison.
In other words, the BOM’s month/day distances are unusually consistent. The question is, “Are they consistent to an
extent that’s statistically significant?” We can find out by first quantifying this consistency using the most widely
used measure of spread, the standard deviation, which comes out to just 1.269[9] – a fraction of the expected value
for eight random dates, which is 6.8.[12] Then we can run another simulation to determine the likelihood that such a
low value would arise by chance.
M/D Distance Std Devs

M/D Distance Std Devs – 8-Date Samples

8.0

10.0

7.0

Rnd

8.0

KJV

TLW

6.0
5.0

6.0

4.0

4.0

3.0

2.0

2.0

BOM

0.0

1.0

B

0.0
BOM

KJV

TLW

Rnd

R

K1 K2 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11
Sample

Figure 3. Standard deviation of month/day distances of the BOM’s dates compared to that of the KJV, TLW, and a large random population
1) in their entirety (left) and 2) after being randomly broken into eight-date samples (right).

In this simulation, the computer again selected eight days at random a billion times, but this time, it looked for cases
in which the standard deviation of the month/day distances was about 1.269 or less. Out of a billion trials, the
phenomenon happened 16,992 times, implying that its odds of occurring by chance are about 1 in 59,000[13] –
nearly equivalent to the results of the first simulation.
Combined Significance & Implications
We have now identified two persistent trends running through our data. By running yet a third simulation that looks
for cases in which both trends occur, we can quantify the overall strangeness of the situation we’re looking at. Out of
10 billion trials, we get 22,760 hits, implying odds of about 440,000 to 1.[14] While it’s surprising to find two
incredibly rare trends in a data set, what makes these particular patterns even more interesting is that they come
together to provide a single insight – that the months and days of the BOM’s dates are consistently close to one
another to an extent that seems to defy historical plausibility (hence these trends will be hereafter be referred to as a
single trend).
Now let’s consider what the implications of this might be. If the dated events reported in the BOM are truly
historical, how can such a spectacular oddity be explained? Here we have events of various types instigated by
various people and groups (and divine beings) from various cultures during multiple seasons spanning multiple
decades of time. How could it make sense for such a diverse
collection of events (or any collection of events for that matter)
to occur on days of the month that are directly linked to the
month in which they occur? It stands to reason that the
likelihood of someone coming up with a plausible explanation
for such a state of affairs is miniscule.
Since there appears to be no good reason for the days and
months of a collection of historical dates to follow each other so
closely and persistently, can we say with some confidence that
the dates are therefore not historical? Given the lack of a
sensible alternative interpretation, I would say, not only that we
can, but that we are compelled to do so. But if the dates aren’t
historical, where did they come from? Arguably, the only
reasonable possibility is that the dates were selected by an
author who, probably subconsciously, had a tendency to allow
his choice of day to affect his choice of month or vice versa. Most
likely, the choice of month affected the choice of day rather than
the other way around since the days are all 12 or less (recall that
having all eight days be 12 or under by chance should only
happen once in 1,526 cases[3]).

Analogous Peer-Reviewed Findings
Is there any precedent for such a psychological effect? It’s clear to me that there actually is. There is a welldocumented and uncontroversial cognitive bias called anchoring that causes us to choose numbers that are heavily
influenced by numbers we’ve recently been exposed to. As one article puts it, “when we are faced with making a
decision involving numbers, we tend to become stuck on the first number we see.” In other words, when a person is
shown a number and then asked a question with a numerical answer, their answer is often much closer to the
number they were shown than it would have been otherwise. According to one theory, this happens because they
subconsciously use the number as a starting point – or an “anchor” – and then adjust away from it to obtain their
answer rather than freely choosing their answer with a clean slate. Anchors can be either self-generated, i.e.,
conceived in one’s own mind, or provided by an external source.
Research has shown this effect to be extremely robust. Even if a number is randomly selected, irrelevant, or absurd
in the context in which it’s used, subjects’ numerical judgments are still affected by it. Even if people are made aware
of the anchoring phenomenon and are explicitly advised to avoid it, they often still fail to do so. Even when they are
offered financial incentives to avoid it, they can’t help themselves. The effect even works on a subliminal level, i.e.,
when a person is exposed to a number so briefly that they don’t consciously realize it.
Some might question the relevance of this bias on the grounds that the research on it all seems to involve numerical
judgments rather than random number selection (from what I have been able to find). However, we do not know
that the BOM’s author was trying to select numbers randomly or that no judgment went into them. If the book was in
fact conceived by a 19th century author, then the inclusion of dates (as well as the fact that the book was written in
16th-17th century English) was likely an attempt to make the book sound more like the Bible since they appear in the
same peculiar, antiquated language. For example:
Bible: “And it came to pass in the seven and twentieth year, in the first month, in the first day of the
month…” (Ezekiel 29:17)
BOM: “And it came to pass in the thirty and fourth year, in the first month, on the fourth day of the
month…” (3 Nephi 8:5)
This makes it possible that the months and days the author selected were part of this attempt. If the author felt like
certain numbers or dates were more bible-like than others – which he very well might have as we’ll see in the next
section – his choices would have more closely resembled judgments than random or haphazard selections. Perhaps
with each date the author selected, he selected a month and then adjusted away from it to obtain a day (recall that
anchors can be self-generated).
Another Way to Slice It
Just for fun, let’s consider one more way to quantify the strangeness of the BOM’s month/day distances before
moving on. You may have already noticed in the table above that all eight dates have a distance of four or less. A
simulation with 100 million trials places the likelihood of this occurring by chance at about 1 in 133,000.[15] A
precise mathematical calculation (which cannot be made to obtain the earlier figures) reveals that the odds are
exactly 1 in 133,333.[16] This method of quantification is not as statistically rigorous as combining the average and
standard deviation since it uses a binary metric (a criterion that each date either meets or doesn’t meet) that was
devised ad hoc rather than continuously variable metrics that are commonly used, but it’s still interesting to note.
% of M/D Distances 4 or Less

# of M/D Distances 4 or Less – 8-Date Samples

100%

10

80%

8

BOM

6

60%

TLW

4
40%

Rnd

KJV

2
20%

0
B

0%
BOM

KJV

TLW

R

K1

K2

T1

T2

Rnd

T3 T4 T5
Sample

T6

T7

T8

T9 T10 T11

Figure 4. Percentage of the BOM’s month/day distances that are four or less compared to that of the KJV, TLW, and a large random
population 1) in their entirety (left) and 2) after being randomly broken into eight-date samples (right).

Trend #2 – Month Correlation
Now let’s turn our attention to the months of the BOM’s dates. As with the days, the randomness of the months looks
questionable at first glance. Half of them are 1 or 2 (three are 1; one is 2), two of them are 7 and the other two are 10
and 11. So there appears to be a degree of clustering in the numbers, leading us wonder whether the author might
have had a bias towards these clusters for some reason. Since the dates’ days were influenced by their months, could
their months have been influenced by something else?
The First Place to Look
Many works of literature that were available in the early 19th century have been alleged to be source materials for
the Book of Mormon and many believe that a combination of source materials were used by its author(s). The most
commonly alleged of these works is undoubtedly the King James Version of the Bible, which
Table 3. Month counts in
the BOM’s dates & the KJV. reports dates of important events using language very similar to how the Book of Mormon
reports them. It has recently been shown that the BOM shares more rare phrases with the
Month BOM KJV
KJV than with any other book. We also know that Joseph Smith studied the Bible extensively
1
3
32
from his youth and was intimately familiar with it. For these reasons, it is the most logical
2
1
14
place to look for signs of influence. Could Joseph have spent so much time with it that the
3
0
6
months mentioned in it might have influenced the months he chose for dates in the Book of
4
0
6
Mormon?
5
6
7
8
9
10
11
12

0
0
2
0
0
1
1
0

10
4
27
5
7
11
3
8

If we search the KJV for the phrases “first month”, “second month”, “third month”, etc, we
can find out how many times each month is mentioned in the Bible (you can do this here).
We can then compare these numbers to the number of times each month appears among the
dates in the Book of Mormon and see if there is any correlation. When we do this, what we
find is staggering – a correlation of .90 between the two[17], which is tremendous by any
standard (correlation is on a scale from -1 to 1). Indeed, the same clusters that we see in the
Book of Mormon’s months are present in the Bible’s. Moreover, the most common month in
Correlation: 0.90 the Bible is the most common in the BOM, the second most common in the Bible is the
second most common in the BOM, and the third and fourth common months in the Bible are
tied for the third most common in the BOM (see the highlighted rows in Table 3 to the left; interestingly, all six of the
symbolic dates given in the Noah’s Ark story contain months that are among these four).
What are the odds of such a high correlation occurring by chance? To find out, I ran a simulation that selected eight
months at random 100 million times to see how often such a correlation occurred with the Bible’s months. It
occurred only 29,392 times, which implies odds of about 1 in 3,400.[18] These are comparable to your odds of
winning a raffle in which every student of a community college entered their name.[19]

Correlations w/ KJV – 8-Date Samples
1.00
0.75

BOM

TLW

Correlation

0.50
0.25
0.00
-0.25
-0.50

Rnd

-0.75
-1.00

Sample

Figure 5. Correlation with the KJV’s months of the BOM’s dates’ months compared to
that of eight-date samples of the TLW and of a large random population. The purple
lines depict the average positive and negative correlations of random date samples.

What’s the Problem?
At this point, some readers might be thinking, “Ok, so what? The Bible and the Book of Mormon are both scriptural,
so we should expect some similarities between them.” While it might seem reasonable at first glance, this simplistic
view is hard to maintain after taking a closer look at the situation. What we find is that having the distribution of the
BOM’s dates’ months mimic that of the Bible’s months actually differs sharply from what we should expect to see.
This is the case for a few reasons, the first being that the Israelites and the Nephites used different calendars. The
Israelites used the Hebrew calendar while the Nephites measured time relative to significant events such as the time
Lehi left Jerusalem, the commencement of the reign of the judges, and the sign of Christ’s birth. Furthermore, Jesus
was crucified on the 14th of the month on the Hebrew calendar
(John 19:14-15, Leviticus 23:5) but on the 4th of the month on the
Nephite calendar (3 Nephi 8:5). A second reason is simply that
the two civilizations had different cultures (all BOM dates are of
events that occurred more than 500 years after the Nephites’
ancestors separated themselves from the Israelites) and lived in
different parts of the world with different seasons, climates, and
geographies – all things that could affect the timing of important
events. Yet another reason is that there is a third (radically
different) culture in the mix – the Lamanites – who were easily as
responsible for the timing of the BOM’s dated events as the
Nephites were (see Table 1). Moreover, three of the events
cannot be blamed on either culture since they were divinely
caused.
There is an even more important reason why we should not
expect the BOM’s months to mimic the Bible’s if they are
historical: the vast majority of the correlation is contributed by
date-specific commandments in the Law of Moses and symbolism
that are prevalent in the Bible but entirely absent from the BOM.
The main reason for the clusters at the first and seventh months
in the Bible is because these are important months in the Law of
Moses and are thus mentioned often. Therefore, the only thing that should cause us to expect similar clusters in the
BOM’s months would be if the BOM’s dated events depended on these same decrees. What we actually find, however,
is that none of the dates given in the book have anything to do with the Law of Moses (and are clearly not meant to
be symbolic). In each case, it is evident that the date’s purpose is nothing more than to help document a significant
one-time event.
If we separate the Bible’s months that share this purpose (of simply reporting a noteworthy event), i.e.,
“documentary months”, from those that do not (giving or keeping the Law of Moses, symbolism, duplicates, etc.) and
compare the former to the BOM’s months, we find a correlation of only .31[17], which a simulation of a million trials
shows to be statistically insignificant (its odds of occurring by chance are 1 in 6[20] – well below the generally
accepted 1 in 20 threshold). This means that the months of the dated historical events in the BOM are not
significantly correlated with the months of comparable events in the Bible but are extremely correlated with the
entire collection of biblical months, the vast majority of which are not comparable (recall our discussion on biblical
date categorization in the “Methodology and Resources” section). This is the opposite of what we would expect to
find if there was a reasonable faith-promoting explanation for the correlation.
Month Distribution Correlations
BOM vs KJV (All Months)

TLW vs KJV (All Months)

3

15

2

TLW Count

BOM Count

12

Corr = 0.90
1

Corr = -0.11

9
6
3

0

0
0

5

10

15
20
KJV Count

25

30

35

0

BOM vs KJV (Doc Months)

5

10

15
20
KJV Count

25

30

35

30

35

Random vs KJV (All Months)

3

15

2

Rand Count

BOM Count

12

Corr = 0.31

1

Corr = 0.25
9
6
3

0

0
0

1

2

3
KJV Count

4

5

6

0

5

10

15
20
KJV Count

25

Figure 6. Month correlations between the KJV’s full collection of dates and 1) the BOM’s dates, 2) those of TLW, and 3) those of a large
random population as well as between 4) the KJV’s “documentary” dates and the BOM’s dates (lower left). Each dot represents one of the 12
months and some dots are on top of each other. Only (1) is statistically significant.

Another way to visualize the contrast between the BOM’s connection with all biblical months and its connection with
documentary biblical months can be seen below. Notice that the former is impressively strong while the latter is
imperceptible and consistent with what would be expected from a random distribution.

Month Distribution Comparison
BOM Month Counts

KJV Month Counts (All Months)

4

40

3

30

2

20

1

10

0

0
1

2

3

4

5

6
7
Month

8

9

10

11

12

1

2

3

4

5

6
7
Month

8

9

10 11 12

KJV Month Counts (Doc Months)
5
4
3
2
1
0
1

2

3

4

5

6
7
Month

8

9

10

11

12

Figure 7. Distribution of the BOM’s dates’ months compared 1) to
that of the KJV’s full collection of months (upper right) and 2) to
that of only the KJV’s “documentary” months (lower left). The
BOM’s distribution is clearly similar to the former but not
noticeably similar at all to the latter.

Implications
To phrase it another way, if the BOM’s dates were truly historical and if we could really explain the correlation of its
months with the Bible’s by pointing to common culture, climate, commandments, or inspiration, we would expect
the biblical events that are actually comparable with the BOM’s to be providing the bulk of the correlation. But since
this is far from what we find, we are once again forced to consider the alternative – that the book’s dates are not
historical. Either that or the correlation arose from chance, but of course, the odds of that are 1 in 3,400.
The correlation we’ve found can be simply explained as follows: when the BOM’s author was selecting dates to
include in the book, his choices were skewed by the months that appear most commonly in the Bible. This influence
may have been subconscious or the author may have intentionally used the months he was most accustomed to
reading in the Bible in order to make the BOM sound more Bible-like. If this was his aim, his mistake was failing to
realize that forcing superficial similarities could create problems if someone ever looked more closely. In his
defense, however, this would have been difficult to foresee for someone who lived before the rise of computers and
modern statistics and who had "a very limited understanding of the elementary rules of arithmetic", as Orson Pratt
put it. Either way, this finding seems to confirm what many scholars have long believed and have collected other
forms of evidence to support – that the BOM was strongly influenced by the KJV Bible through a relatively modern
author like Joseph Smith (possibly with the help of one or more others).
Analogous Peer-Reviewed Findings
Again, the implied psychological effect is not without precedent in the academic literature. For example, when asking
subjects to randomly select letters, researchers have found that the distributions of letters they select correlate from
.70 to .83 with the distribution of letters in the English language. Since people tend to choose letters with a frequency
that depends on the frequency with which they’re exposed to them, it seems likely that they would do the same with
numbers if they were exposed to some more often than others.
Research has borne this out at least partially by showing that the numbers people tend to choose depend on their
culture. For example, Americans and Europeans tend to pick 7, while Nigerians prefer 9, Turks prefer 3, and people
from Taiwan favor 8. The more significant a number is in a person’s culture, the more often they see and hear it, and
this could very well be what causes them to select it more often. Researchers have suggested that “the repetition of
the symbol 7 in the Bible… could well explain its predominance” in America and Europe. If ordinary modern folks
can have their number choices influenced by the Bible, why should we doubt whether a 19th century student of the
Bible who participated in Christian revivals in a community where Christianity was in the air could have his choices
similarly influenced?

Noteworthy Anomalies
Up to this point, our examination has focused on persistent trends that run through the entire collection of dates
contained in the BOM. However, there are also a few interesting anomalies in these dates that are worth noticing
even though they each involve only a subset of the dates. Let’s look at them briefly.
Calendar Clustering
The BOM’s dates are much more clustered on the calendar than what would be expected by chance. They include the
1st, 2nd, and 4th days of the 1st month as well as the 3rd and 4th days of the 7th month (all in different years). It’s as
though one date was placed at the beginning of the calendar, one date was placed at the middle of it, and then half of
the remaining six dates were placed right next to them. The simplest way to quantify this degree of clustering is by
noting that the two clusters of dates we just identified have a combined distance of four days between their dates
((2-1) + (4-2) + (4-3) = 4) and an average of 1.33 days between them ( ⁄  1.33). We can generalize this metric to
any collection of dates by sorting them in chronological order (ignoring their year components), calculating the
distance (in days) between adjacent sorted dates, and averaging the three shortest distances. For lack of better
terms, let’s call this metric 3DC (“three-distance clustering”).
Clustering – 8-Date Samples
25.0

KJV

3DC

20.0

Rnd

15.0

TLW

10.0
5.0

BOM

0.0
B

R

K1 K2 T1 T2 T3 T4 T5 T6 T7 T8 T9 T10 T11
Sample

Figure 8. “Three-distance clustering” (3DC) of the BOM’s dates compared to that of
representative eight-date samples from the KJV, TLW, and a large random population.

In order to determine the odds of a collection of eight dates having a 3DC as low as 1.33, I ran a simulation that
randomly selected eight dates and calculated their 3DC to see whether it was 1.33 or less. Out of 10 million trials,
only about 1 in 300 of them produced this level of clustering.[21] Now the significance of this finding is diminished by
the fact that, to a large extent, the phenomenon arises as a natural consequence of the month/day proximity and the
month correlation identified in the two sections above. However, it still could have been partly caused by the author

having a subconscious preference for the dates at the very beginning of the year and those in the very middle of it.
This would possibly be in line with research that has shown that people do tend to prefer certain numbers over
others when asked to make random selections.
New Year’s Week
Without looking at the table below, scroll up and take one more look at the BOM’s eight dates in Table 1 to see if
anything strange in them jumps out at you – something that you don’t even have to do any math to notice…
Did you see it? If not, I’ll give you a hint: it’s in the second half of
the dates. Now go ahead and give it another shot if you’d like…
The title of this subsection might have given it away or you
might have already caught a glimpse of the answer in the table
below, but at least now you have a little game you can play with
others. The oddity that appears in the last half of the BOM’s
dates is the fact that if you were to change only one of the eight
numbers in them, you would have the first four days of the year
in chronological order!
Verse
Alma 52:1
Alma 56:1
Alma 56:42
3 Ne 8:5

Month
1
1
71
1

Day
1
2
3
4

1/1
1/2
1/3
1/4

Figure 9. The BOM’s last four dates are just one digit away from being the
first four days of the year.

As you can see in the image to the right, a simulation of 10
billion trials implies that the odds of this happening by chance
somewhere in a collection of eight dates is roughly 1 in 25 million.[22] However, with a phenomenon as rare as this
one, we would need even more trials to get an accurate figure (notice the large variation in the “Results”, which
range from 31 to 54). Luckily, there are so few ways that this pattern can occur that its exact probability can be
calculated with simple math, which puts the figure at almost exactly 1 in 20 million.[23]
Given the extreme rareness and strangeness of this occurrence, it seems more than plausible that it was caused, at
least in part, by an effect that various researchers have documented – the human tendency to select sequences of
numbers in their natural order. It has been theorized that the frequent encounters most of us have with ordered
sequences in our daily lives contributes to this bias.
Contagious Contrivance
If it seems strange that the prophet Mormon, who served as the primary author of the BOM would report only those
dates that meet a particular set of arbitrary criteria, how likely do you think it would be for multiple other BOM
prophets to do the same thing with the same criteria? As it turns out, Helaman (Alma 56:42) and Amulek (Alma
10:6) also give dates whose months coincide with the most common months in the Bible and whose days are very
close to these months – the third and fourth days of the seventh month. Moreover, these dates are right next to each
other on the calendar, which can’t be blamed on any causal link between the events since they occurred about 18
years apart from each other and are completely different in nature.
This proximity on the calendar is significant because the odds of two people reporting adjacent dates by chance are
less than 1% (1 in 120)[24], but as we just noted above, studies have found that a particular person will usually show
preferences for some numbers over others even when they’re trying to pick them randomly. So based on these facts
alone, there is reason to suspect that these dates were both selected by a single person rather than being genuinely
historical and independent of one another. Add to this the unlikelihood that all three of the BOM characters who
report dates would happen to give dates that conform to the peculiar set of criteria we’ve discussed, and we have yet
more reason to believe that all of the book’s dates were conceived in the mind of a single human author who was
putting words into the mouths of these figures.

Summary of Results
To err on the side of conservatism, let’s ignore the strange anomalies above for a moment and focus only on the
systematic trends we discussed earlier that involve the entire set of dates contained in the BOM. What are the odds
that, out of all the events that occurred in the book, the ones that would be significant enough to have their dates
reported would all have dates 1) whose months strongly correlate with those in the Bible and 2) whose days are
very close to those months? Since these two phenomena are independent of one another (i.e., the presence of one
does not affect the probability of the other), we can multiply their odds against chance together to obtain the
probability of both of them occurring:
x

 1 in 1.5 billion

=

In other words, if we had 1.5 billion different history books that each documented the dates of eight non-correlated
events, we would only expect one of these books to contain the patterns we’ve identified in the BOM’s dates with
equal or greater strength. Thus if one of these books was published every single day, we would expect to wait over 4
million years before seeing one with such trends.
Now some might be led to conclude from this that the odds of the BOM being historical are 1 in 1.5 billion, but this
wouldn’t be exactly right. There’s a subtle but significant difference between the probability of an event occurring
and the probability that the event has actually occurred, and the way to get from one to the other is with a
universally accepted formula called Bayes’ theorem, which looks like this:
=
In our case, the terms in this formula can be interpreted as follows:




is the probability that the BOM’s dates are historical (H) given the trends (T) we’ve identified in them
is the probability of such trends occurring among the dates if they are historical
is the prior probability that the dates are historical (the probability before the trends were found)
is the probability of such trends occurring among the dates if they are not historical
is the prior probability that the dates are not historical

Some of these terms are easy to assign values to.
is what we just finished calculating, so it’s ⁄
or
0.0000000007.
depends on where one stood in regards to the historicity of the BOM before beginning to read
this paper, so it could range anywhere from 0 to 1. For the sake of fairness and objectivity, and since it’s common
practice with controversial matters, it makes the most sense for us to approach it from the perspective of someone
who’s been agnostic on the question, so we’ll set it to 0.5. This would cause
to be 1 – 0.5 = 0.5 as well.
The equation might be a little intimidating, so let’s simplify it a bit. The first thing we can do is recognize that the
term
is an extremely tiny number compared to the term it’s being added to in the denominator (if this
isn’t clear now, it should become so shortly). Because of this, it has a negligible effect on the value of
and can
thus be ignored, which gives us
=
Next, since we’ve decided that
=
and since one of these is in the numerator while the other is in the
denominator, they cancel each other out. This leaves us with

=
which is of course much nicer to look at than what we started with. Since we already know what
is, the only
thing left to do is determine the value of
. Again, this is the probability that the trends would occur if we
were to assume that the dates are not historical. In other words, if the portions of the BOM that contain dates were in
fact concocted by a storyteller who pulled the dates out of his head, what are the odds that he would have chosen
dates with the trends we’ve found?
This is a very difficult question to answer because we don’t have access to this potential storyteller or even a reliable
way to know who he is and, even if we did, we wouldn’t know which relevant cognitive biases he was subject to
during the phase of his life in which he wrote these portions of the book or what his objectives were, if any, while
selecting the dates. Was he especially prone to anchoring or any of the other biases we’ve considered? Was he
superstitious about dates whose months and days are close together? Was he trying to make the months resemble
those that seemed to be most common in the Bible? We just don’t have any good way to answer these questions or to
even guess at probability distributions for their potential answers. I therefore propose that we should again take an
agnostic approach and assign a probability of 0.5 to each trend that we’re considering for a combined probability of
0.5 x 0.5 = 0.25. Now some might argue that we should instead use 0.5 as the combined probability, but again, for the
sake of erring on the side of conservatism, I will use 0.25 or ¼ for the value of
. This yields

=

=

=

= 1 in 375 million

This suggests that given the extremely anomalous trends present in the BOM’s dates (and given the agnostic starting
points we’ve assumed), the odds that the dates are historical is about 1 in 375 million. This is less than one’s odds of
winning the Powerball lottery, which, as of October 2013, are 1 in 259 million (before that, they were 1 in 176
million), and it is several times lower than the odds that one of your children will someday be elected President of
the United States (assuming you’re an average American).

Premises
In determining the probabilities herein, we have relied on the premises below. Note that the supporting points given
for each premise are intended to show that we should expect the premise to be at least close to true. If it turns out
that the premise is only close to true rather than 100% true, this would only somewhat decrease the significance of
our findings rather than invalidate them (as evidenced by the experiments documented in the next section “What if
Events Were Far from Randomly Distributed?”). Also note that #3 does not apply to Part II of our analysis and #4
does not apply to Part I:
1. The Nephites used a 12-month calendar, each month containing 30 days – a reasonable assumption since:

They considered the 11th month to be “in the latter end of the… year” (Alma 48:2, 21, 49:1).
They observed the Law of Moses (Alma 25:15-16), which would have required such a calendar since
many of its commandments were specific to particular months and days of the month (as we pointed
out earlier), and this was the type of calendar used by the Israelites. To be exact, the months on the
Hebrew calendar alternate between 29 and 30 days in length. While it’s true that an additional month
is added to leap years, these years are in the minority and no 13th month is mentioned in the Bible.

2. These events are non-correlated – a reasonable assumption since:



There is no obvious, direct connection between any of them (with one partial exception – there is a
visible, although indirect, connection between Amulek’s conversion and his prison break with Alma
over three months later; however, there is little reason to believe that the month and day on which this
escape occurred became more likely than any other when Amulek was converted).
We would not expect any of them to make a particular month or day more likely than any other for any
other of the events.
They differ from each other starkly. Only one pair of them can be argued to be the same type of event –
when the Lamanites begin a war against the Nephites and when they are seen marching toward a
Nephite city to start a war.
All but one of them occurred in a year unique from the others

3. These events would be expected to happen on days of the month that are randomly distributed – a reasonable
assumption since:


There is no reason to expect that events such as these would be more likely to happen during any
particular part of the month than another, especially not month after month throughout the year.
Even if some of the events were somehow more likely to happen during particular parts of the month,
the events are so different from each other that we would then expect these parts of the month to be
randomly distributed.
I have collected several large, high-quality data sets on various types of historical events, including
starting dates of wars, acts of terrorism, protests, the forming of international alliances, and the events
of the US Civil War and, in each category, the days of the month on which these events occur appear to
be randomly distributed (with one minor exception that we’ll discuss later). Moreover, with each data
set, the average and standard deviation of its month/day distances are in line with those of random
dates. See the section “What Should Historical Dates Look Like?” (WSHDLL)
There are two other books commonly studied in conjunction with the BOM that report dates using
similar language – the King James Bible and The Late War. The days of the dates in both books exhibit
the same signs of randomness as the data sets mentioned above (note that the biblical dates in
question are “documentary” dates as explained earlier). See WSHDLL.

4. These events would be expected to happen in months that are randomly distributed – a reasonable
assumption since:


Nothing in the BOM states or implies that any events are more likely to happen during some months or
seasons than during others. In fact, three or all four of the seasons are represented among the given
dates of events. Similarly, five of the events took place in the context of war, yet these are scattered
among three of the seasons.
The BOM says nothing of seasonal weather fluctuations, cold, or snow. LDS apologists have pointed to
this in order to support their theory that the BOM took place in Mesoamerica rather than in the Great
Lakes region of North America. John L. Sorensen has written "Where is the snow in the Book of
Mormon? Where is the cold in the Book of Mormon? Not a single word that indicates anything other
than warmth and even tropical heat."
Of these events, the only type of event that arguably occurs more than once did not occur during the
same month on both occasions but rather on months that are three apart from each other (the
eleventh and second), which is the average distance between all possible pairs of months.[25]
In each category of historical events that I have collected data on (described in #3 above), the months
during which the events occur appear to be randomly distributed. Moreover, with each data set, the
average and standard deviation of its month/day distances as well as its correlation with biblical
months are in line with those of random dates. See WSHDLL.
The months of the King James Bible and The Late War exhibit the same signs of randomness as the
data sets (note that the biblical months in question are “documentary” months as explained earlier).
Again, see WSHDLL.

Also note the following, which show that the timing of each event was not merely dependent on any one culture. We
would expect these complicating factors to add yet more complexity and thus randomness to the timing of the
events:

Five of the eight events involved both the Nephites and Lamanites, who had very different (practically
opposite) cultures.
The three remaining events were at least mostly divinely caused:
o Two of them were mostly done by God and/or an angel with the participation of Alma and/or Amulek.

o The last event was Jesus’ doing immediately after his crucifixion (and the date given for it differs from
its date on the Israelite calendar, as noted earlier).

What if Events Were Far from Randomly Distributed?
Let’s assume for a moment that we live in a world where the third and fourth premises described in the previous
section have very little validity, that is, a world in which we would expect the dates of the significant events that
occurred among BOM peoples to be far from randomly distributed. In fact, let’s go beyond that and assume, not only
that these dates were nonrandom, but that they were nonrandom in ways that favor the distribution of the eight
dates reported in the BOM. This would be a strange assumption indeed, but let’s go ahead and make it for now to
show that we’re willing to consider the notion for the sake of argument. We’ll make it in stages with the assumption
in each stage being stronger and less plausible than it was in the last. We’ll also make it in two sets of stages: one set
that would increase the odds of the dates’ days being consistently close to their corresponding months and one set
that would increase the odds of the dates’ months correlating strongly with those in the Bible.
Assumption Set #1
First, let’s assume that important events among BOM peoples were, for some reason, twice as likely to occur during
the first half of any given month (recall that all of the BOM’s dates fall into this category). Simulations show that, in
such a scenario – a scenario that would be inexplicable and highly implausible, to say the least – the probability that
the book’s dates would have month/day pairs that are as consistently close together as they are is about 1 in
45,300.[26]
Next, we’ll assume that such events only happened during the first half of the month. In other words, no important
event ever took place among the Nephites or Lamanites during the second half of any month in any year. Even in this
preposterous scenario (which, let’s face it, isn’t remotely plausible), the likelihood of the month/day phenomenon is
only 1 in 1,750.[27]
If that isn’t ridiculous enough, let’s imagine that every significant event took place on one of the days of the month
reported in the BOM (there are seven of them – the 1st, 2nd, 3rd, 4th, 5th, 10th, and 12th) and in the same proportions
(so that events would occur on the 4th twice as often as they occur on the other days). Even under such impossible
circumstances, the odds come out to only 1 in 280.[28]
To top it off, let’s really go crazy and propose that the creature allegedly responsible for the dates in the BOM
(whether it was God or the prophet/historian Mormon) was strangely fixated on the particular numbers that appear
in the dates and was determined to select dates with those exact days and months. In this case, the only variable
would be how the days and months were matched up with each other. Even in this absurd situation, the likelihood of
the trend we see is just 1 in 124[29] – still well into the realm of statistical significance.
Assumption Set #2
Now onto assumptions that would make the month correlation phenomenon more likely if they were true. First,
we’ll suppose that significant occurrences were twice as likely to happen during the 1st and 7th months (the most
commonly reported months in both the BOM and the Bible) as they were during any other month for BOM peoples.
In this bizarre state of affairs, the probability that the months of the BOM’s dates would correlate with those of the
Bible as strongly as they do is 1 in 300.[30]
Lastly, let’s imagine that the kinds of events reported in the BOM occurred in the New World in the exact same
months and in the exact same proportions with which comparable events (the “documentary” events discussed
earlier) are reported in the Bible. Even under this outlandish proposition, the odds of finding the month correlation
phenomenon in the BOM’s dates are only 1 in 208.[31]
Combined Assumptions
While it’s almost certainly too generous to grant that even one of the above assumptions could be valid, considering
the notion that two of them could be valid is even less realistic by orders of magnitude. Nevertheless, we’ll do it here
for the sake of argument. If somehow an assumption from each of the two sets were valid, the odds of both trends
being present in the dates would be somewhere between 1 in 363,000 and 1 in 13,500,000 (since ~ ⁄
x ⁄
 ⁄
and ~ ⁄
x~ ⁄
 ⁄
; note that I have ignored the third and fourth assumptions in Set #1
because, while I believe that all of the assumptions are virtually impossible, these two are obviously absurd).
Implications
We can see here that even if our premises regarding the randomness of significant dates, for some strange reason,
didn’t apply to BOM civilizations, there would still be no conceivable set of alternative premises that could rescue the
book’s dates from statistical implausibility. Even under the most extreme and generous assumptions, it is incredibly
unlikely that the dates are historical.
Before we continue, let me make sure we’re clear that this section is intended only to demonstrate the robustness of
our conclusions, not to provide any kind of realistic lower bound for the significance of our findings. Therefore, to
say that the trends in the BOM’s dates are significant at odds against chance of anywhere between 363,000 to 1 and
1.5 billion to 1 would be a grave mistake, for this would not account for the fact that the circumstances required to
get anywhere close to this lower bound are practically impossible. Even a lower bound of 13,500,000 to 1 would be
unrealistic since it would require two implausible scenarios to coexist. Let’s also be clear that the propositions we’ve
considered here are extreme, hypothetical assumptions that we have no rational basis for making, especially in light
of the justifications given in the previous section for expecting BOM dates to be randomly distributed.

What Should Historical Dates Look Like?
Recall that in order to determine the statistical significance of a particular trend (i.e., the probability that it would
appear by chance) in a set of dates of important events, we run a Monte Carlo simulation that selects dates randomly
and sees how often the trend arises. Of course, the more random the dates of important events are in real life, the
more realistic the simulation will be and therefore the more accurate its results will be. So naturally, the question is
“How random do the dates of important real-life events tend to be?”
In the “Premises” section of this paper, I argued that we should expect the eight dates reported in the BOM to be
either randomly distributed or close to it. I offered several lines of reasoning in support of this view (a view that
many will find to be intuitive and not needing of support), one of which was that the populations of historical event
data I have collected all appear to have dates with months and days that are at least close to randomly distributed. In
this section, we’ll take a look at some of this data and compare it to data that was created by a trusted random
number generator.
As discussed in the “Methodology and Resources” section (“Resources” subsection) near the beginning of this paper,
I have compiled this data – which was collected by research projects at universities – into a spreadsheet. Each of the
data sets is arguably relevant to the types of events that are reported throughout the BOM. They include the dates of
significant events of the US Civil War, starting dates of wars, and dates of other types of militarized conflict,
terrorism, protests, and international alliances, and they cover a variety of time periods and geographic regions.
Scroll up to this section for more details on the files Historical Event Data.xlsm and Raw Historical Event Data.zip
and for a link to where you can download them.
Before we get to this data though, let’s look at the dates of events reported in the Bible and in the book The Late War
and compare them to randomly generated dates (see the subsections of the “Methodology and Resources” section
titled “Why the Bible and The Late War?”, “Random Dates in Charts”, and “Categorization of Bible Dates & Months”
for explanatory details). Note that the value “n”, given in the title of each chart specifies the number of dates in that
population:

The Bible
KJV Months (n = 20)

Random Months (n = 20)

5

5

4

4

3

3

2

2

1

1

0

0
1

2

3

4

5

6
7
Month

8

9

10

11

12

1

2

3

KJV Days (n = 20)

4

5

6
7
Month

8

9

10

11

12

Random Days (n = 20)

4

4

3

3

2

2

1

1

0

0
1

3

5

7

9 11 13 15 17 19 21 23 25 27 29
Day

1

3

5

7

9 11 13 15 17 19 21 23 25 27 29
Day

Figure 10. Comparisons between the distributions of 1) the KJV’s “documentary” dates’ months and random months (top) and between
those of 2) the KJV’s “documentary” dates’ days and random days (bottom).

The Late War
TLW Months (n = 92)

Random Months (n = 92)

16

16

12

12

8

8

4

4

0

0
1

2

3

4

5

6
7
Month

8

9

10 11 12

1

2

3

TLW Days (n = 92)

4

5

6
7
Month

8

9

10 11 12

Random Days (n = 92)

8

8

6

6

4

4

2

2

0

0
1

3

5

7

9 11 13 15 17 19 21 23 25 27 29
Day

1

3

5

7

9 11 13 15 17 19 21 23 25 27 29
Day

Figure 11. Comparisons between the distributions of 1) TLW’s months and random months (top) and between those of 2) TLW’s days and
random days (bottom).

As you can see, both the months and the days of the historical events in these two books are roughly as evenly
distributed as their corresponding random populations (which were not cherry picked, as explained in the
“Methodology and Resources” section). In addition, for each population, I have calculated the average and standard
deviation of the month/day distances as well as the correlation with biblical months, and in each case, the figures are
in line with those of random dates.[32]
Now let’s look at some more recent historical event data:
Armed Conflict from 1946 to 2008
Armed Conflict – Months (n = 166)

Random Months (n = 166)

25

25

20

20

15

15

10

10

5

5

0

0
1

2

3

4

5

6
7
Months

8

9

10 11 12

1

2

3

4

5

6
7
Month

8

9

10 11 12

Armed Conflict – Days (n = 166)

Random Days (n = 166)

14

14

12

12

10

10

8

8

6

6

4

4

2

2

0

0
1

3

5

7

9 11 13 15 17 19 21 23 25 27 29
Days

1

3

5

7

9 11 13 15 17 19 21 23 25 27 29
Day

Figure 12. Comparisons between the distributions of 1) the months of historical armed conflict events and random months (top) and
between those of 2) the days of historical armed conflict events and random days (bottom).

Terrorism in Western Europe from 1950 to 2004
Terror in W. Europe – Months (n = 3472)

Random Months (n = 3472)

350

350

300

300

250

250

200

200

150

150

100

100

50

50

0

0
1

2

3

4

5

6 7
Month

8

9

10 11 12

1

2

3

Terror in W. Europe – Days (n = 3472)

5

6 7
Month

8

9

10 11 12

Random Days (n = 3472)

160

160

140

140

120

120

100

100

80

80

60

60

40

40

20

20

0

4

0
1

3

5

7

9 11 13 15 17 19 21 23 25 27 29
Day

1

3

5

7

9 11 13 15 17 19 21 23 25 27 29
Day

Figure 13. Comparisons between the distributions of 1) the months of historical terror events in Western Europe and random months (top)
and between those of 2) the days of historical terror events in Western Europe and random days (bottom).

We can make the same observations with these dates – their distributions closely resemble those of random
populations and their figures regarding month/day distances and month correlation are very close to those of
random dates.[33] This is also true of the other historical event data sets, but for the sake of conciseness, I will refer
you to the spreadsheet if you’d like to see those. As mentioned earlier, a couple of the data sets show one minor
exception to this randomness: wars in recent history have often started on the 1st and 15th days of the month, which
could very well be explained by the fact that these are the days on which soldiers are paid. Perhaps it is easier to put
new soldiers to work at the beginning of a pay period.
It is apparent that significant historical events tend to be dispersed throughout each year and each month in ways
that are random or nearly random. From this, and in light of the points made in the “Premises” section, we can amply
induce that the dates given for important events in the BOM should be consistent with random dates if they are truly
historical. At the very least, they should not deviate from randomness in ways that are tremendously statistically
significant and indicative of cognitive biases on the part of the book’s author.

Objections & Alternative Explanations
Haven’t we used too small a “sample size”?
It has been suggested that this study can’t be significant because it relies upon a relatively small “sample size” of
eight dates. Now, if we were trying to use a sample to form conclusions regarding a much larger population of dates,
this could be a valid criticism, but this is not what we’re doing. Rather, we are simply determining the probability of
such a peculiar collection of dates occurring by chance by simulating it billions of times. We can determine the
probability of all kinds of collections of events in this way regardless of their size, even collections that only contain
one event. For example, we can determine the odds of getting tails when we flip a coin by doing the same thing –
simulating a coin flip a large number of times. Because of this and because we are using all of the dates contained in
the BOM rather than a sample of them, the term “sample size” is not even applicable to this study. For similar
reasons, the commonly misused term “selection bias” is not applicable either.
Can’t God put whatever he wants in his scriptures?
The question here isn’t about what God can do but rather about what he is likely to have done. If God is to blame for
the suspicious trends we’ve found in the BOM’s dates, he must have seen some benefit in arranging them this way.
But what benefit could possibly come about from such a contrivance? What could be learned from it? Who could be
helped by it? What good could God bring about by planting evidence within his own holy book that discredits it –
that makes it look like his servant fabricated the book under the influence of cognitive biases and other literature?
This would create needless confusion and “God is not the author of confusion” (1 Corinthians 14:33).
Could it have been Mormon’s fault?
Some readers might wonder whether the biases I’m attributing to Joseph Smith could have actually been held by
Mormon, the historian and prophet who allegedly abridged the volumes of Nephite records onto the small set of gold
plates that became the BOM. In this view, either Mormon’s selection of the events he reported or his selection of
which events to provide dates for was influenced by these biases. However, I find this explanation to be highly
implausible for the following reasons:

As we saw earlier, not all of the dates can be attributed to Mormon since two of them are given by others –
Amulek and Helaman – as part of stories they tell about important events of their recent past. And as we saw,
these aren’t just any two dates; they’re consecutive dates on the calendar that exhibit the very problems
we’re trying to blame on Mormon. It is therefore likely that they were conceived in a single mind – the same
mind that chose the dates that Mormon allegedly reported. This would make them the product of an author
rather than a historian.

It is practically certain that there is no known precedent for a historian or a history book cherry picking the
historical dates it reports based on the numbers in those dates, and we have good reason to expect that we
will never find one. There is, however, plenty of precedent for people who pull numbers out of their head
being influenced by biases and exposure to other numbers (as evidenced by the numerous studies linked in
this paper as well as countless others), especially when they’re not trying to be random in their selections.

To blame Mormon for the biases we have identified is to allege that a historian and prophet who God charged
with compiling “the most correct of any book on earth” and then told what to write in that book (3 Nephi
26:12, Words of Mormon 1:7):
o Had a strange preference for 1) dates whose days are very close to their months and 2) dates with
months that appear most frequently in the Bible.
o Allowed the dates he reported for events in this book to be strongly influenced by these preferences.
o Had no purpose in his reporting of dates in general. Seemingly, the only real value that readers of the
BOM could receive from his reporting of the dates of select events is in the form of an indicator of
which events deserved to be most carefully documented. If Mormon decided which events to report
dates for based on which numbers he liked rather than on which events were most important to
carefully document then there was really no purpose in providing the dates at all. This would not be in
line with what the BOM tells us about the plates it was written on – that they had limited space
(Mormon 8:5, 9:32-33, Jarom 1:2, 14, Omni 1:30), were hard to engrave (Jacob 4:1-3, Ether 12:24-25),
and could only contain those choice words that were of the utmost importance to mankind (1 Nephi
6:6, Words of Mormon 1:5).

What about the “loose translation theory”?
In response to the discovery of dozens of anachronisms in the BOM, the fact that the translation errors of the 1769
version of the King James Bible were copied into it by its author, and myriad other problems with the book, some
LDS apologists have advocated for the “loose translation theory”. This theory suggests that when Joseph Smith was
“translating” the gold plates from reformed Egyptian into English, rather than being shown the correct translation
word for word, he merely had ideas revealed to him through spiritual impressions from God and was then allowed to
put those ideas into his own words.
Some will probably propose that this was what allowed Joseph Smith’s cognitive biases to influence the dates that
appear in the BOM. The main problem with trying to explain away the problems with the book’s dates in this way,
however, is that the theory simply isn’t applicable to this kind of issue for the same reason that it isn’t applicable to
the hundreds of names of people, places, etc. given in the book or to words like “curelom”, “cumom”, or “deseret”,
which are given in the book and are claimed or implied to be words in the native language of the book’s peoples.
Highly specific, arbitrary details like dates, numbers, and never-before-seen words/names could not have been
conveyed through vague ideas; they must have been clearly communicated in a way that left little to no room for
error. If the BOM is a genuine translation, its translator must have vividly seen or heard such details either physically
or mentally thus making them immune to his own biases.
In addition, notice that the BOM’s author(s) were very diligent about allowing the reader to roughly keep track of
when events took place. There are 415 uses of the word “year” (or “years”) in the book, most of which are given in
order to pin its events to a timeline. In nearly all of these cases, only a year or approximate part of a year is given for
the event. For example:
And Helaman died, in the thirty and fifth year… (Alma 62:52)
… in the commencement of the eighty and first year they did go forth again against this band of
robbers, and did destroy many… (Helaman 11:30)
… Moroni did arrive with his army at the land of Bountiful, in the latter end of the twenty and
seventh year… (Alma 52:18)
This tells us that whoever authored or translated the book was more than willing to use very rough chronology,
which indicates that in the few cases when he used exact chronology, he did so deliberately, intending the date he
gave to be relied upon. So if Joseph Smith had only been given vague impressions regarding when the eight
precisely-dated events occurred, why would he have needlessly made wild guesses at specific dates that were
probably wrong, especially when he dated the rest of the events in very nonspecific ways? Such sporadic bouts of
feigned precision would be pointless, untruthful, and inexplicable.
Another problem with invoking the loose translation theory is that it is at odds with recent official statements from
the church on how the BOM was translated. In an essay that was approved by the First Presidency and then posted
on the church’s website on December 30, 2013, the church describes a tight translation process on no fewer than
three occasions:
According to witnesses of the translation, when Joseph looked into the instruments, the words of
scripture appeared in English.
[T]he scribes and others who observed the translation left numerous accounts that give insight into
the process… According to these accounts, Joseph placed either the interpreters or the seer stone in a
hat, pressed his face into the hat to block out extraneous light, and read aloud the English words that
appeared on the instrument.
The principal scribe, Oliver Cowdery, testified under oath in 1831 that Joseph Smith “found with the
plates, from which he translated his book, two transparent stones, resembling glass, set in silver bows.
That by looking through these, he was able to read in English, the reformed Egyptian characters,
which were engraven on the plates.”
There are also many other statements in this essay that support a tight translation while nothing in it suggests a
loose translation. So there you have it: the loose translation theory is at odds with the official position of the church
and its current prophets as well as with the testimonies of those who were involved in the translation process.
Two of the Three Witnesses of the BOM – Martin Harris and David Whitmer – as well as Church Historian and
Seventy B.H. Roberts and a close friend of Joseph Smith, Joseph Knight Sr., all reported a translation process that
ensured that every symbol on the plates was translated correctly. For example, David Whitmer, in whose home most
of the translation took place, wrote the following:
I will now give you a description of the manner in which the Book of Mormon was translated. Joseph
Smith would put the seer stone into a hat, and put his face in the hat, drawing it closely around his face
to exclude the light; and in the darkness the spiritual light would shine. A piece of something
resembling parchment would appear, and on that appeared the writing. One character at a time
would appear, and under it was the interpretation in English. Brother Joseph would read off the
English to Oliver Cowdery, who was his principal scribe, and when it was written down and repeated
to Brother Joseph to see if it was correct, then it would disappear, and another character with the
interpretation would appear. Thus the Book of Mormon was translated by the gift and power of God,
and not by any power of man… I, as well as all of my father's family, Smith's wife, Oliver Cowdery and
Martin Harris, were present during the translation… (An Address to All Believers in Christ, Part First,
Chapter 1, p. 12)
It stands to reason that at least some of these men could not have described such details without having been told
them by the translator himself. Therefore, even if the loose translation theory somehow turned out to be right, it
would mean that Joseph Smith was dishonest in his description of the translation process, which would substantially
harm his credibility and compel us to wonder what else he might have lied about, why he felt a need to lie, and why
God would have chosen a deceiver to do such important work for him.
Royal Skousen, an accomplished professor of linguistics and English at BYU who has served as the president of the
Utah Association of Scholars for 15 years and counting, is the man behind the Critical Text Project of the BOM, and,
according to Yale University Press, is the world’s “leading expert on the textual history of the [book]”. He has spent
decades studying the original manuscripts and has written this on his discoveries regarding the book’s translation:
…I began to see considerable evidence for the traditional interpretation that witnesses of the
translation process claimed: the text was given word for word; Book of Mormon names were
frequently spelled out the first time they occurred in the text; and during dictation there was no
rewriting of the text except to correct errors in taking down the dictation. Joseph Smith was literally
reading off an already composed English-language text. The evidence in the manuscripts and in

the language of the text itself supports the hypothesis that the Book of Mormon was a precisely
determined text. I do not consider this conclusion apologetic, but instead as one demanded by the
evidence.
The opposing viewpoint, that Joseph Smith got ideas and he translated them into his own
English, cannot be supported by the manuscript and textual evidence.
Perhaps the simplest point we can make on this topic is this: if the BOM is a transcription of vague spiritual
impressions rather than a direct translation of the symbols on the gold plates, then why did the plates exist at all?
The extraordinary amount of time, sacrifice, and painstaking effort that reportedly went into engraving and
preserving those records over thousands of years would have served little purpose if neither the plates nor the
writing they contained was used to produce the BOM. The whole thing would arguably amount to little more than a
cruel game that serves only make the story of the restoration of God’s church much more problematic and harder to
believe than it needs to be for no apparent reason.
Some will counter the points I’ve made here by alleging that the BOM could have been translated through both tight
and loose translation processes at different points in time. But let’s think about this for a moment: if God was both
able and willing to give an exact translation of portions of the book, what legitimate reason could there be for him to
not do the same with the entire book? What could cause him to deliberately compromise the accuracy of the
scripture he was giving to his children, which Joseph Smith said “was the most correct” and important book ever
produced? This would gratuitously create confusion for thoughtful students of the book and, as we noted earlier,
“God is not the author of confusion.” But even if we forget all this and assume that the idea can be fit into a sensible,
coherent theory, it still conflicts with the church’s essay, which, from what I can tell, says nothing that would allow
for such a possibility. Moreover, for the aforementioned reasons, arbitrary details like precise dates would need to
be communicated clearly and would thus need to be among the portions that were translated tightly.
Some LDS apologists have resorted to tactics that can only be characterized as either deceptive or extremely sloppy
to make the loose translation theory appear to be evidence-based. For example, Michael R. Ash has authored an
article in the Deseret News as well as a book that both call upon an 1872 statement from the early apostle Orson
Pratt to support the idea:
Joseph “received the ideas from God, but clothed those ideas with such words as came to his mind.”
Between the context in which the quote is given and the fact that it includes no ellipses at the end, the reader is led to
believe that the quote is sufficiently complete and that it supports a loose translation of the BOM. However, it turns
out to be a careful cropping of Pratt’s words. Here’s the full statement as reported in the December 9, 1872 minutes
of the School of the Prophets:
Joseph the Prophet in writing the Doctrine and Covenants, received the ideas from God, but clothed
those ideas with such words as came to his mindbut in translating the book of Mormon by the use
of the Urim and Thummim [a divine object], God not only revealed the ideas but the words also.”
So not only does the small portion that Ash quotes refer to the D&C rather than to the BOM, it actually goes on to
support a tight translation process for the BOM! So rather than support his position, the statement actually
contradicts it. While I’d like to give Mr. Ash the benefit of the doubt, it’s clear that this misdeed is an example of
either gross negligence (i.e., not taking the time to locate the full quote and read it in context) or willful deception. In
either case, we have yet another point in favor of a tight translation.
Please excuse me for spilling so much ink on this subject. I have felt it necessary because this theory – which I see as
mostly a contrived apologetic construct – keeps coming up and being used as the silver shield to deflect all sorts of
evidence against the BOM. It was an interesting idea and the protection it seemed to provide was nice while it lasted,
but as you can see, we now have more than enough cause to let the notion go and give it a proper burial.

Conclusion
While examining the full collection of dates that appear in the Book of Mormon, we have seen that multiple trends
run through them that are extraordinarily statistically significant – even under the most generous of assumptions –
while also being easily explained by an appeal to psychological effects that have been demonstrated in peerreviewed studies. It appears likely that the mystery highlighted by earlier studies – “Why are the days of the book’s
dates so consistently low?” – has been solved: the selection of these days was heavily influenced by the selection of
their corresponding months, which was heavily influenced by the distribution of months in the Bible.
This implies that these dates are not historical, which strongly suggests that the events they correspond to are not
historical either. We therefore have cause to seriously doubt whether any of the events, places, or people in the book
are historical. Note that we haven’t even considered the various other serious problems with the BOM raised by
archaeology, anthropology, biology, geology, geography, genetics, linguistics, history, and literary analysis that
decades of strenuous apologetic efforts have failed to explain in any way that can withstand scrutiny or that can even
be fit into a coherent faith-promoting theory.
For some, these findings might be the straw that breaks the camel’s back. Others will see them as merely a shovel of
dirt thrown onto the grave of a camel whose back fatally broke long ago. At the very least, I believe that those who
have carefully read this paper, considered it with openness and intellectual integrity, and have kept up with the best
available evidence and reasoning regarding BOM historicity will acknowledge that this is a heavy straw being
heaped onto the back of a weary camel whose burden is growing heavier by the year with little to no sign of relief
forthcoming.
Whether you see the Book of Mormon as an inspired but non-historical volume of scripture or as a contrived remix
of the ideas and literature that Joseph Smith had access to, the conclusion that it is not a historical record has
arguably long been somewhere on the spectrum between being consistent with the evidence and being demanded
by it. The findings and resulting implications that we’ve explored here certainly don’t help the situation. As the
saying goes, “The numbers don’t lie”, and the numbers have finally spoken.

Notes
1. Note in the “Other Sim Results” tab of Simulator Results.xlsm.
2. “Days Sum” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
3. The odds of one date having a day on the 12th of the month or earlier is ⁄ . The odds of eight dates all
having such days is ⁄
 ⁄
= 1 in 1,526.
4. “Days MAD” section in the “BOMDateSim Results” tab of Simulator Results.xlsm. The reason I used mean
average deviation (MAD) as a measure of spread rather than standard deviation for this calculation is because
I believe it captures a larger portion of what makes the spacing of the dates' days peculiar. It captures not
only the fact that all eight days are grouped into the first 12 days of the month, but also, to an extent, the fact
that six of them are grouped into the first 5 days of the month. This latter fact is not accounted for as well by
standard deviation.
5. “6 Days <= 5” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
6. Calculations referred to in notes #3, #16, and #23.
7. “Revelation Dates” section in the “Other Sim Results” tab of Simulator Results.xlsm.
8. Notes at the bottom of the “Proximity” and “Correlation” tabs of “Data & Charts.xlsm”.
9. “BOM Dates” table in the “Proximity” tab of “Data & Charts.xlsm”.
10. “Distance Avgs” section in the “Other Sim Results” tab of Simulator Results.xlsm and the “Distances of All
Possible Dates” section of the “Other Calcs” tab of Data & Charts.xlsm.
11. “Dist Avg” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
12. “Distance Avgs” section in the “Other Sim Results” tab of Simulator Results.xlsm.
13. “Dist Std Dev” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
14. “Dist Avg & Std Dev” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
15. “All Dists <= 4” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
16. “Distances of All Possible Dates” section of the “Other Calcs” tab of Data & Charts.xlsm.
17. “Month Counts & Correlations” table in the “Correlation” tab of “Data & Charts.xlsm”.
18. “Month Correlation” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
19. For the sake of completeness and full disclosure, note that there are two events reported in the BOM for
which the month but not the day is given (Alma 56:27, 3 Nephi 4:7, 11). This significant difference in format

(being only a month rather than a full date) suggests that the author may have selected these months through
a different thought process, while in a different state of mind, or for a different purpose than the months
contained within full dates. It’s even possible that a different author wrote them altogether. For this reason,
these two months are in a class apart from the others and should not necessarily be studied in conjunction
with them. Hence I have not included them in this analysis. For those curious, the correlation with biblical
months that results when they are included is 0.82, and the odds of such a figure arising by chance are about
1 in 480.
20. “Doc Month Correlation” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
21. “Clustering” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
22. “1st 4 Dates of the Yr” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
23. “Odds of the New Year’s Week Anomaly” section of the “Other Calcs” tab of Data & Charts.xlsm.
24. For any particular day on the calendar, there are three days that lie within one day of it – the day itself, the
day before it, and the day after it. And since we are assuming 360-day years (12 months with 30 days each –
see the “Premises” section), the odds of a day being one of the three are ⁄
= ⁄
= 1 in 120.
25. “Month Pair Distances” section of the “Other Calcs” tab of Data & Charts.xlsm.
26. “Skewed Population D1” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
27. “Skewed Population D2” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
28. “Skewed Population D3” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
29. “Avg & Std Dev” figure in the “M-D Combos” tab of “Data & Charts.xlsm”.
30. “Skewed Population C1” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
31. “Skewed Population C2” section in the “BOMDateSim Results” tab of Simulator Results.xlsm.
32. “KJV Documentary Dates”, “The Late War Dates”, and “Rand Sample Dist Avgs” tables in the “Proximity” tab of
“Data & Charts.xlsm”.
33. “Date Figures” table in each tab of “Historical Event Data”.xlsm and the “Distance Avgs” and “Correlation
Avgs” sections in the “Other Sim Results” tab of Simulator Results.xlsm.
— TruthIsReason
Please send questions, comments, and corrections to TruthIsReason7@gmail.com