Re: Power Law of Probability? (The book "Chances Are" by the Kaplans)


W. Watson wrote:
Old Mac User wrote:

I've not read this book. But from yokur citation of "the power law of
repeated successes" it sounds like a loser to me. Worse than a loser...
a source of infection that we really don't need.

In short... it sounds like it was written by a couple of very confused
people.

So who gave it "good reviews". OMU


W. Watson wrote:

I'm reading "Chances Are ... Adventures in Probability" by Kaplan and
Kaplan. It's an interesting book for its background, philosphy and history
of probability. It is not a text book. Their background is in the
humanities. The book is meant for the general reader, and has gotten good
reviews. I wonder about how the general reader fares in some instances.
Equations appear on only about 5 to 10% (if even that high) of the pages.

They mention the power law of repeated successes as (1/6)*(1/6) of the
number of repeated successes. Fine. Later they mention, page 33, "if
something can happen with probabilty a and not happen with probability b in
each of x trials, then we can say, putting the power law into general terms,
that the chance of this not happening at every trial is:
b**x/(a+b)**x "

They then go on to equate this to 1/2 to find the number of trials it will
take to produce a fair outcome. They invoke logs and series expansions, and
finally 0.7 provides a multiplier to odds in any game that will tell you the
number of expected trials. (in dice, 35*0.7 = 24.5 trials) Well, fine.. For a
general audience, this seems quite a bit of math juggling. Has anyone read
this book and wants to comment on this?



Wayne T. Watson (Watson Adventures, Prop., Nevada City, CA)
(121.015 Deg. W, 39.262 Deg. N) GMT-8 hr std. time)
Obz Site: 39° 15' 7" N, 121° 2' 32" W, 2700 feet
--
"I have made this letter [e-mail] a rather long one, only
because I didn't have the lesiure to make it shorter."
-- Blaise Pascal
Web Page: <home.earthlink.net/~mtnviews>


I'll answer both the previous two posts here. See Amazon. In fact, see
<http://www.amazon.com/Chances-Are-Adventures-Probability/dp/0670034878/sr=1-1/qid=1164640918/ref=sr_1_1/102-1905724-8880112?ie=UTF8&s=books>.
12 reviews, and all 4 to 5 stars (the highest possible is 5).If I'm not
mistaken, both or one of, Discover and Sci Am magazines gave it favorable
reviews. Aside from the goofiness I pointed out, the book actually is an
interesting read--so far. I'm only 70 pages into it.

What you've done in posing your questions with snippets of content from
the book was at the old "bait and switch" salesmanship, though probably
unintended.

What you described were the WORST of the explanations anyone could
possibly give on some simple statistical concepts, and you ask for
assessment.
You should get the unanimous review (given what you showed) as mine and
that of OMU, by any statistician who is not completely and totally
blind. ;-)

What you showed NOW are some of the features that made the book
appealing
as light reading for non-statisticians that are amusing factoids or
anedotal
accounts (as in the tablid newspepers) that are appealing to the
general
public. So, there is no contradiction in the different kinds of
reviews over
different kinds of material in the book.


An example of an
interesting historic observation is that Gauss and Poison got the credit for
distributions that DeMoivre discovered. The DeMoivre story is quite
something in itself.

This is the kind of things archeologists (those authors) can do better
in
writing their stories than writing about probability and statistics.
In fact,
what is interesting historical observation is WELL KNOWN to
statisticians
who don't even have ANY interest in the history of statistics
(myself!).

DeMoivre and Laplace both preceeded Gauss, and were generally credited
with the discovery of the normal distribution, which Gauss got the
credit
for the name Gaussian distribution. The fact that almost ALL named
theorems and important statistical results are named by the LAST (and
not FIRST) of the discovers is a phenomenon that was well-known even
before Steve Stigler gave it a name in one of his papers. I knew that
even before Steve got his Ph.D. degree. :-)

So, if the authors had read Steve's historical papers, they could have
gotten a wealth of historical factoids without doing any real research
themselves in historical archeology of statistics.

Another interesting item, to me at least, is the
statement, "LaPlace helped shape the future of science ... he broke forever
Pascal's hope that faith and scientific inquiry could sit comfortably
together." Of course, the Templeton Foundation is trying hard to bridge that
(huge) gap. Richard Dawkins ("The God Delusion") has taken them and others
severely to task for their attempts.

They missed the story of DeMoivre predicted the exact date of his
death:

*De Moivre correctly predicted the day of his own death. Noting that he

was sleeping 15 minutes longer each day, De Moivre surmised that he
would die on the day he would sleep for 24 hours. A simple mathematical

calculation quickly yielded the date, November 27,1754. He did indeed
pass away on that day.

That's the perfect hindsight prediction that only the ultra-gullible
would
bite -- but it makes interesting reading, doesn't it?


One surprising thing about the book is that it has no references or
bibliography. Non-fiction books that do that I usually take one star away
from my personal rating.

The book is at best a college senior term paper on elementary
statistics.
It would get some points for the the tablid stories, and would get very
negative points of statistical and probability SUBSTANCE --which is
basically the COMBINED review of your 12 4-5 star reviews and the
0-negative star reviews OMU and I gave. :-)

Steve Stigler's papers on the history of statistics are much more
substantial, and FREE. I am sure his papers would have contained
ALL the historical tidbits of these authors, and much much more.
Then supplement that with a Statistics for Dummies book and the
combination will make much better read than this mother-son
combo's dabble in statistics.

-- Reef Fish Bob.


Wayne T. Watson (Watson Adventures, Prop., Nevada City, CA)
(121.015 Deg. W, 39.262 Deg. N) GMT-8 hr std. time)
Obz Site: 39° 15' 7" N, 121° 2' 32" W, 2700 feet
--
"I have made this letter [e-mail] a rather long one, only
because I didn't have the lesiure to make it shorter."
-- Blaise Pascal
Web Page: <home.earthlink.net/~mtnviews>

.

Relevant Pages

  • Re: Power Law of Probability? (The book "Chances Are" by the Kaplans)
    ... So who gave it "good reviews". ... It's an interesting book for its background, philosphy and history ... writing their stories than writing about probability and statistics. ... So, if the authors had read Steve's historical papers, they could have ...
    (sci.stat.math)
  • Re: OT: Finesse contest finalists - thanks to all!
    ... Whenever you dismiss someone else's statistics or articles as biases ... propaganda out of hand, that's not fair. ... History has decided to treat him kindly.. ...
    (rec.arts.sf.tv.babylon5.moderated)
  • Re: Consumer Reports digital camera reviews
    ... Their reliability ratings are scientifically unreliable, ... I think many go after CR because their crummy reviews of ... The variable that messes up the statistics is: ... The reason I mentioned ...
    (rec.photo.digital)
  • Re: The Promise of Forth
    ... My problem here is that it appears we have only one history to ... If people at a church picknick get food poisoning and you want to ... Not very good statistics at all. ... certainly the waldorf salad, because all 32 people who didn't eat it ...
    (comp.lang.forth)
  • Re: The Promise of Forth
    ... Randomness versus determinism depends on your point of view. ... So, was the extinction of T. Rex, apparently due to an asteroid impact, determined or random? ... In astronomy we're quite used to using statistics to examine historical data. ... History teaches that logical consistency is neither sufficient nor necessary to establish practical, ...
    (comp.lang.forth)