Sign In

SHARE QUESTION Bao

u
yh
esfo
'tn
yo pW iu
srihat RELATED QUESTIONS

Twitter
caliorsthib
hestco
dviertoh o telo maras What is an intuitive explanation of Bayes'
Facebook
nb
ed
yiw
olh
ad
u
stveaoro fn aeixpatl Rule?

Google+ ?erR cC
ih
drair What is the significance of Bayes'
Theorem?
Re-Ask Follow 6
RELATED TOPICS Why is Bayes' Theorem important?

How has Bayes' Theorem been used

Richard Carrier 5 Answers throughout history?

Tim O'Neill, Atheist, Medievalist, Sceptic and amateur Historian Bayes' Theorem: In laymens terms: how
do you compare the validity of
hypotheses in a Bayesian world?
To begin with, it's illustrative to note who uses Bayes Theorem to analyse
history and who does not. In the first category we have William Lane Craig, How is Bayes' theorem used in robotics?
the conservative Christian apologist, who uses Bayes Theorem to "prove" that
Jesus actually did rise from the dead. And we also have Richard Carrier, the What is the best book about Bayes'
Theorem?
anti-Christian activist, who uses Bayes Theorem to "prove" that Jesus didn't
exist at all. Right away, a curious observer would find themselves wondering Reviews of: Bayes' Theorem
how, if this Theorem is the wonderful instrument of historical objectivity both
Craig and Carrier claim it to be, two people can apply it and come to two What is Bayes' Theorem?
completely contradictory historical conclusions. After all, if Jesus didn't exist,
How do I solve the Monty Hall Problem
he didn't do anything at all, let alone something as remarkable as rise from the using Bayes Theorem?
dead. So both Carrier and Craig can't both be right. Yet they both use Bayes
Theorem to "prove" historical things. Something does not make sense here.

Then if we turn to who doesn't use Bayes Theorem to analyse history we find
this category includes ... pretty much every single historian on the planet.
Again, this should strike the objective observer as distinctly odd. After all, if
Bayes Theorem can genuinely be applied to determine the truth or otherwise of
a historical event or proposition, it's exceedingly strange that thousands of
historians all over the world are not applying this remarkable tool all the time.
Richard Carrier maintains that this is because every historian on earth, except
him, is too ignorant and mathematically illiterate to understand the wonders of
this remarkable tool and only he has been clever enough to realise that it can
be applied to history. Given that Thomas Bayes ' theorem was first published
in 1763, our objective observer would be forgiven for finding it remarkable that
no-one noticed that it could be used in this way until Richard Carrier, an
unemployed blogger, came along.

Bayes Theorem is used in probabilty theory and statistics as a tool for

determining statistical probability. When used to compare two competing
statements or probabilities, it is expressed like this:

Where:
P(A),the prior probability, is the initial degree of belief in A.
P(-A), is the corresponding probability of the initial degree of belief against
A: 1-P(A)=P(-A)
P(B|A), the conditional probability or likelihood, is the degree of belief in A,
given evidence or background B.
P(B|-A), the conditional probability or liklihood, is the degree of belief
against A, given evidence or background B.
P(A|B), the posterior probability, is the probability for A after taking into
account B for and against A.

For an idea of how this can be applied, let's look at a real world example
(originally found in this post ):

converted by W eb2PDFConvert.com
Marie is getting married tomorrow, at an outdoor ceremony in the desert. In
recent years, it has rained only 5 days each year. Unfortunately, the
weatherman has predicted rain for tomorrow. When it actually rains, the
weatherman correctly forecasts rain 90% of the time. When it doesn't rain, he
incorrectly forecasts rain 10% of the time. What is the probability that it will
rain on the day of Marie's wedding?

Solution: The sample space is defined by two mutually-exclusive events - it

rains or it does not rain. Additionally, a third event occurs when the
weatherman predicts rain. Notation for these events appears
below.

Event A1. It rains on Marie's wedding.

Event A2. It does not rain on Marie's wedding.
Event B. The weatherman predicts rain.
In terms of probabilities, we know the following:
P( A1 ) = 5/365 =0.0136985 [It rains 5 days out of the year.]
P( A2 ) = 360/365 = 0.9863014 [It does not rain 360 days out of the year.]
P( B | A1 ) = 0.9 [When it rains, the weatherman predicts rain 90% of the
time.]
P( B | A2 ) = 0.1 [When it does not rain, the weatherman predicts rain
10% of the time.]
We want to know P( A1 | B ), the probability it will rain on the day of Marie's
wedding, given a forecast for rain by the weatherman. The answer can be
determined from Bayes' theorem, as shown below.
P( A1 | B ) = P( A1 ) P( B | A1 ) P( A1 ) P( B | A1 ) + P( A2 ) P( B | A2 ) P(
A1 | B ) = (0.014)(0.9) / [ (0.014)(0.9) + (0.986)(0.1) ] P( A1 | B ) = 0.111

So, despite the weatherman's prediction, chances are Marie won't be rained on.

The first thing our objective observer should notice here is that we have hard
data to plug into the equation. We know how often it does rain in this region,
how often it doesn't rain and how often the weather forecast is right or wrong.
So we can get a meaningful answer out of the equation because we can plug
meaningful data into it in the first place.

So there are two problems here when it comes to trying to apply Bayes
Theorem to history: (i) Carrier and Craig need to treat questions of what
happened in the past as the same species of uncertainty as what may happen
in the future and (ii) historical questions are uncertain precisely because we
don't have defined and certain data to feed into the equation.

Bayes Theorem only works in cases where we can apply known information.
So, in the example above, we know how often it rains in a year and we know
when the weather forecast is and isn't correct. So by inputing this meaningful
data, we can get a meaningful result out the other end of the equation.

This is not the case with history.

Bayes Theorem's application depends entirely on how precisely the parameters

and values of our theoretical reconstruction of a real world approximate
reality. With a historical question, Carrier is forced to think up probabilities for
each parameter he put into the equation. This is a purely subjective process -
he determines how likely or unlikely a parameter in the question is and then
decides what value to give that parameter. So the result he gets at the end is
purely a function of these subjective choices.

In other words: garbage in/garbage out.

So it's not surprising that Carrier comes up with a result on the question of
whether Jesus existed that conforms to his belief that Jesus didn't - he came up
with the values that were inevitably going to come up with that result. If

converted by W eb2PDFConvert.com
someone who believed Jesus did exist did the same thing, the values they
inputted would be different and they would come up with the opposite result.
This is why historians don't bother using Bayes Theorem.

So what exactly is Carrier doing by applying this Theorem in a way that it can't
be applied? Apart from being incompetent, he seems to be doing little more
than putting a veneer of statistics over a subjective evaluation and pretending
he's getting greater precision.

Not surprisingly, despite his usual grandiose claims that his use of Bayes
Theorem is some kind of revolution in historiography, his book Proving
History: Bayes's Theorem and the Quest for the Historical Jesus (2012) has
pretty much sunk without trace and been generally ignored by historical Jesus
scholars and historians alike. His failure to convince anyone except a gaggle of
historically clueless online atheist fanboys of his vast genius means that Carrier
is most likely to remain what he is: an unemployed blogger and general
nobody.

(For those who are interested, his mathematics and statistics is about as bad as
his historiography. See the Irreducable Complexity blog's rather scathing
review of Carrier's use of Bayes: A Mathematical Review of "Proving History"
by Richard Carrier The blogger is actually inclined to agree with Carrier on
the existence of Jesus, which makes his criticisms of his misuse of the
mathematics even more trenchant.

And for an amusing critique, try this blog post that applies Bayes Theorem and
Carrier-style reasoning to "prove" that, in fact, Richard Carrier doesn't exist:
Does Richard Carrier Exist? )
Updated 48w ago
Upvote 29 Downvote Comments 4+ 1,341

Salvator Vella, Epicurian, skeptic, humanist and free thinker

As mathematical law, Bayes Theorem is absolutely true and, manipulating

multiple uncertain hypothesis together using reasoning that violates Bayes
Theorem will result without any doubt into wrong conclusions. Because its
truth is absolute, it is valid even for matters that are not fully scientific, like
historiography.

Dr. Aviezer Tucker, Associate at Davis Center, Harvard University, specialized

in philosophy of historiography and history wrote a book, A Companion to the
Philosophy of History and Historiography . In it, he says: "Bayesian analysis
can explain most of what historians do and how they reach a consensus on
determined historiography." Also "But even where reasoning about the past is
entirely qualitative, it often involves judgments of comparative
probability; and the norms governing such reasoning can often be modeled
well using Bayesian resources."

In other words, the use of the Bayesian reasoning is seldom explicit

(something R. Carrier made explicit) but a sound historical reasoning respects
the Bayesian Theorem.

This is like the law of gravity. People did not wait for the gravity law being
discovered by Newton to use its law implicitly to design catapults and make
ballistic predictions. Bayes Theorem is not different. Good historians make
argumentation and reasoning that are compatible with Bayes Theorem in
order to infer sound conclusions from archaeological discoveries for example.
In other words, what Dr. Aviezer Tucker is saying that, even if historians do
not use Bayes's Theorem explicitly, their reasoning is "ballistically" correct and
in line with this theorem.

There is however at least one domain of history where Bayes Theorem is often
violated, this is what concerns the study of the historicity of Jesus, because the
matter is obviously loaded. Furthermore most scholars that have a view on the
historicity of Jesus are most often theologians or doctors in Biblical studies, no

converted by W eb2PDFConvert.com
historians and they are not trained accordingly and their reasoning is often
loaded by a confirmation bias.

Also people who have read Thinking, Fast and Slow by the Nobel Prize Daniel
Kahneman and Predictably Irrational by Dan Ariely perfectly know that
human brains are very bad in manipulating probabilities, even among
scholars. Bayes Theorem helps correct most common errors when
probabilities are involved.

It is therefore untrue to say that Bayes Theorem is only useful in exact

sciences. For example, in medicine, it is recognized as being extremely useful
and, when Bayes Theorem is not used, diagnostic errors are more common.

This is the reason why the historian R. Carrier decided to make Bayes Theorem
and Bayesian reasoning explicit and to apply it explicitly to debunk all the
fallacies in reasoning they saw applied by theologians and doctors in Biblical
studies who call themselves historians. In other words, besides this peculiar
domain of historiography, making this theorem explicit is not so needed as it is
the case for Jesus.

Now it is easy to ridicule Bayes Theorem, like some posters did in this thread,
by using absurd probabilities as input to make fun of the results. If you take
the meteorological example in Tim O'Neill posting, the weatherman predicted
rain 10% of the time there is no rain ... in a country where there is no rain in
360 days on 365! In other words, this weatherman predicted rain more than
41 days in a year in a country with rain in only 5 days in a year. Such a bad
weatherman would be fired in a real world and, therefore, it is totally not
surprising that, based on those data, Bayes Theorem gives correctly the result it
gives.

For people willing to better understand the correct use of Bayes Theorem in
diverse matters, I recommend the following book Bayes or Bust?: A Critical
Examination of Bayesian Confirmation Theory by John Earman or the
following site: Yudkowsky - Bayes' Theorem
Written 46w ago
Upvote 2 Downvote Comments 2+ 352

Richard Acton

to the point about subjectively assigning probabilities this is a problem of the

method but there are ways to do this that are more valid than others. Possibly
the most exacting is solomonoff induction see: An Intuitive Explanation of
Solomonoff Induction
this however is generally computationally intractable so there are attempts at
various heuristics to solomonoff induction.
other more rigorous approaches to objective determination of priors include
Empirical Bayes (Empirical Bayes method ) and maximum entropy
(Principle of maximum entropy ).
The extent to which these approaches can be applied in the field of history
seems yet to have been explored in academic history circles, it would be the
direction I would expect Carrier to persue next if he wishes to make a good
case against the choice of priors made by people like Craig.
Written 4w ago
Upvote 1 Downvote Comment 1 39

Nathan Ketsdever, Education entrepreneur, social entrepreneur &

researcher. Former Academic & d...

1. Bayes theorem is an attempt to use naturalism to prove naturalism. It

begins with naturalistic assumptions about how the world works. This is
fundamentally a circular enterprise from start to finish.
2. Events like miracles are on the order of the Big Bang and Creation in
terms of probability--they are on the order of unique, once in a millenia

converted by W eb2PDFConvert.com
 type actions. And in one, sense all individual actions are once in a millenia
type actions--given time and place.
3. Probability doesn't tell us what happened in history--only really agency
can. Even if probability could tell us that--it would suggest we have access
to lots of other information we couldn't have.
4. Its incredibly hard to predict individual choices--whereas its much easier
to predict aggregate or average choices.
5. In the case of history, particularly of this time period, we may not have all
the relevant data points.
6. That X event has 20% probability and another event has 80% probability--
tells us NOTHING if they both occurred. Even using the theory to look
back at history runs up against some serious problems.
7. At the end of the day, we're still left with subjective assessments of
probability.
Written 38w ago
Upvote Downvote 212

Quora User, literature lover

It seems like the best practical approach yet devised to sift theories for
likelihood, appropriateness and relevance. From what I've read of Dr. Carrier's
works, I believe the application of the theorem does the job.
Written 48w ago
Upvote 1 Downvote Comment 377

1 Answer Collapsed

Write an answer

Related Questions
How is Bayes' theorem used in robotics?
What is the best book about Bayes' Theorem?
Reviews of: Bayes' Theorem
What is Bayes' Theorem?
How do I solve the Monty Hall Problem using Bayes Theorem?
How might the Bayes' Theorem be used to locate a missing submarine?
Why does Bayes' theorem work?
I just can't understand Bayes' theorem, even after reading it multiple times. What
is an explanation for how it works?
What are the uses and importance of the Bayes' theorem in statistics?
Do events have to be mutually exclusive in order to use Bayes' Theorem?
Is it possible to calculate the probability of receiving an illegal shipment using
Bayes' Theorem ?
What is the probability that I am drunk given that I went to a music concert using
Bayes theorem?
What are some criticisms of Bayes' Theorem?
How can I apply bayes theorem to my everyday life?
How do I intuitively know when to use bayes theorem in solving a question? Are
there any things I should watch out for in the question?

converted by W eb2PDFConvert.com
Top Stories

Sitemap # A B C D E F G H I J K L M N O P Q R S T U V W X Y Z
About - Careers - Privacy - Terms

converted by W eb2PDFConvert.com