X-Message-Number: 9088
Date: Mon, 2 Feb 1998 09:54:51 EST
Subject: probability

Just quickly:

A few people have asked how to obtain my booklet, "Cryonics: the Probability
of Rescue." It is available by mail from the Immortalist Society--just e-mail
me your mailing address. No charge. It will be on the CI web site
(www.cryonics.org) eventually, but that is probably weeks or months away.

A few have also asked me to explain--on cryonet--how it is possible to make an
estimate of probability for this, in light of various difficulties.
Unfortunately, it cannot be made convincing, or even very clear, in fewer
words than found in the booklet itself. Even the relatively full discussion in
the booklet will not be persuasive to those unwilling to pay serious
attention. A couple of points, though:

Charles Platt asks how I can estimate future sociological developments.
Actually, I could (very roughly indeed), but my focus was only on the
SCIENTIFIC odds--the probability of success assuming the continued advance of
science and the continued cryopreservation of the patients.

Paul Wakfer suggests that no probability estimate is possible concerning--for
example--restoration of memory, since we do not understand memory. Again, one
must read the booklet. Actually, there is some experimental evidence
suggesting preservation of at least some kinds of memory after freezing. Aside
from that, we DO NOT necessarily need to understand the details of a process
in order to estimate probabilities involving that process. 

As a trivial example, we make implicit probability estimates dozens of times
daily with little understanding of the underlying dynamics--e.g. whether it is
safe to cross the street now. We "simply" draw on appropriate experience.  A
monkey makes the same kind of implicit probability estimate every time it
leaps for another branch, with no understanding of mathematical physical

I didn't intend to drag on so long, but having come this far I'll throw in an
example I use to underscore the variable and partly subjective nature of most
kinds of probability:

What is the probability that Wayne State University will beat Michigan State
U. in a football game (if they were scheduled etc.)? I can derive at least
three different probabilities, all valid:

Bettor A is an American who reads the AP writers' poll, and finds that the
majority of the sports writers favor MSU. He also knows that the AP favored
team, over a period of many years, has won 70% of the time. For this American,
then, the probability of a victory by MSU is 70%, referred to the following
sequence of experiments: Bet on the team favored by the AP poll, and in the
long run you will be right about 70% of the time.

Bettor B is a Bantu visitor who knows nothing about football and doesn't read
the papers. He might as well toss a coin. For him, the probability of MSU
winning is one half, referred to the following sequence ofexperiments: Choose
a team by some arbitrary method, such as coin tossing or a nicer color of
uniform; in the long run you will be right about half the time.

Bettor C is the coach of MSU, who favors his own team by two touchdowns. In
the past, whenever he has favored his own team by that much, he has been right
85% of the time. For him, therefore, the probability of MSU winning is
approximately 0.85.

Probabilities are objective, in that they refer to relative frequencies of
outcome in a reasonably clearly specified series of experiments. They are also
subjective, in that the choice of the appropriate series of experiments
depends on the information available to the observer. In trying to choose a
series of experiments, the appropriateness of the series is more important
than the size of the data base or the accuracy of the numbers. 

There is much more, but that is enough here.

Robert Ettinger
Cryonics Institute
Immortalist Society

Rate This Message: http://www.cryonet.org/cgi-bin/rate.cgi?msg=9088