Transcript Philosophy 100 Lecture 9 Pascal`s wager
Introduction to Philosophy Lecture 6 Pascal’s wager By David Kelsey
Pascal
• • • Blaise Pascal lived from 1623-1662.
He was a famous mathematician and a gambler.
He invented the theory of probability.
Probability and decision theory
• Pascal thinks that we can’t know for sure whether God exists.
•
Decision theory:
used to study how to make decisions under uncertainty, I.e. when you don’t know what will happen.
–
Lakers or Knicks:
–
Rain coat:
•
Rule for action:
when making a decision under a time of uncertainty always perform that action that has the highest expected utility!
Expected Utility
•
The expected utility for any action
: the payoff you can expect to gain on each trial if you continued to perform trials...
– It is the average gain or loss per trial.
–
A trial:
is a an attempt at achieving success. • Example… –
The payoff or value of an outcome:
occurs.
what is to be gained or lost if that outcome • •
To compute the expected value of an action:
– ((The prob. of a success) x (The payoff of success)) + ((the prob. of a loss) x (the payoff of a loss))
Which game would you play?
–
The Big 12:
pay 1$ to roll two dice. –
Lucky 7
: pay 1$ to roll two dice. – –
E.V. of Big 12: E.V. of Lucky 7
:
Payoff matrices
•
Gamble
: Part of the idea of decision theory is that you can think of any decision under uncertainty as a kind of gamble.
•
Payoff Matrix
: used to represent a scenario in which you have to make a decision under uncertainty.
– – –
On the left:
our alternative courses of action.
At the top:
the outcomes.
Next to each outcome:
add the probability that it will occur.
–
Under each outcome
: the payoff for that outcome •
Calling a coin flip:
– If you win it you get a quarter and if you lose it you lose a quarter.
• • • The coin comes up heads: ___ It comes up tails: ___ You call heads ___ ___ You call tails ___ ___
The Expected Utility of the coin flip
• •
So when making a decision under a time of uncertainty
: construct a payoff matrix
Which action:
– Perform the action with the highest expected utility!
–
To compute the expected value of an action:
• ((The prob. of a success) x (The payoff of success)) + ((the prob. of a loss) x (the payoff of a loss)) • •
For our coin tossing example:
–
The EU of calling head:
• … –
The EU of calling tails:
– …
Choose either action…
Another coin tossing game
•
Different payoffs:
what if the payoffs were greater when the coin comes up heads than if the coin comes up tails.
• • • It comes up heads: ___ The coin comes up tails: ___ You call heads ___ ___ You call tails ___ ___ •
The EU if you call heads:
– … •
And the EU if you call tails:
– … •
So Call Heads!
Taking the umbrella to work
•
Do you take an umbrella to work?
You live in Seattle. There is a 50% chance it will rain. – – –
Taking the Umbrella
: a bit of a pain. You will have to carry it around.
• Payoff = -5.
If it does rain & you don’t have the umbrella
: you will get soaked • payoff of -50.
If it doesn’t rain then you don’t have to lug it around:
• payoff of 10. • • • • • It rains (___) It doesn’t rain (___) Take umbrella ___ ___ Don’t take umbrella ___ ___
EU (take umbrella)
= …
EU (don’t take umbrella)
• Take the umbrella to work!
= …
Pascal’s wager
•
Choosing to believe in God
: Pascal thinks that choosing whether to believe in God is like choosing whether to take an umbrella to work in Seattle.
–
It is a decision made under a time of uncertainty:
–
But We can estimate the payoffs
: •
Believing in God is a bit of pain
whether or not he exists: •
An infinite Reward: …
•
Infinite Punishment
: …
Pascal’s payoff matrix • • • Believe Don’t believe God exists (___) God doesn’t exist (___) ____ ____ ____ ____ •
Assigning a probability to God’s existence:
– A bit tricky since we don’t know.
– For Pascal: • since we don’t know if God exists we know the probability of his existence is greater than 0.
– EU (believe) = … – EU (don’t believe) = …
•
Believe in God: …
Pascal’s argument
•
Pascal’s argument:
– 1. You can either believe in God or not believe in God.
– 2. Believing in God has greater EU than disbelieving in God.
– 3. You should perform whatever action has the greatest EU.
– 4. Thus, you should believe in God.
•
Not existence but Belief:
…
Denying premise 1
•
The first move:
–
Deny premise 1:
•
The second move & Pascal’s reply:
–
Believing for selfish reasons
:
Denying premise 2
• Deny premise 2: –
Infinite payoff’s make no sense:
–
Can we even assign a non zero probability to God’s existence?
The Many Gods objection
•
We could Deny premise 2 in another way:
–
Many Gods & the Perverse Master…
The Perverse Master
• • • •
The new payoff matrix:
God exists (__) Perverse Master exists (__) Neither exists (___) Believe _____ _____ ___ Don’t Believe _____ _____ ___ •
Disbelief seems no worse off than belief:
–
EU (believe)
= … –
EU (don’t believe)
= … • What if we thought it
less likely that the perverse Master exists than does God: