What is the most unlikely event that has been documented to have
occurred?  For example when a person wins the lottery the chances of
that may be 1 in 65 million, but people do win the lottery.  I want to
know THE EVENT, in the history of existence, with the least chance of
occuring, that has been proven to happen.  I also want the statistical
chances of that event occuring, both when it happened last, and when
it will happen again (if known and different).  Please do not answer
with an event whose chances of occuring cannont be calculated.  The
event must have happened, must be proven to have happened, and must
have a statistical chance of occurance calculated either before or
after the event happened.

---

Request for Question Clarification by pafalafa-ga on 28 Sep 2005 08:13 PDT

Here's a pretty good candidate:


http://scripts.gophercentral.com/archives/list.asp?a=3&i=BizarreNews&rn=336



One in 23 point something trillion!   Let's see anyone beat them odds...

---

Clarification of Question by jim258kelly-ga on 28 Sep 2005 08:21 PDT

That is an interesting event.  I'd like to know exactly which two
lotteries the person won, it doesn't go into much detail.

---

Request for Question Clarification by guillermo-ga on 28 Sep 2005 08:50 PDT

There is the case of a man -Roy Sullivan- who is said to have been
struck by lightning 7 times, and survived all of them. If true, I
doubt that any recorded case would defeat it as the least probable.
I'll look for verification and probability calculation and get back to
you with the information to see whether that makes an answer or not.

Guillermo

Jim,

I'm glad that event caught your fancy.

Here are some excerpts from the actual newspaper account of the double-winning:


===============
The San Francisco Chronicle
DECEMBER 12, 2002

Double lottery winners beat odds of 1 in 24,000,000,000,000;
Belmont couple spends $124,000 -- $20 a day for 17 years -- then hits
jackpot twice in one day


It had to happen sooner or later for Angelo and Maria Gallina, who
figure they have spent $124,000 over the years on lottery
tickets...What happened was that they won the jackpot -- not once, but
twice, on the same day. An hour after winning $126,000 in the Fantasy
Five game, they won $17 million in SuperLotto Plus.

That's never been done before, lottery officials said Wednesday, maybe
because the odds of its happening are 1 in 24 trillion -- which is a
24 followed by 12 zeros...

Angelo Gallina, a man who does not excite easily, said he celebrated
their twin killing by filling their car with gas and getting a
haircut...

...For the record, the odds of winning the SuperLotto Plus are 1 in
41.4 million and the odds of winning the Fantasy Five are 1 in
576,000. Multiplying those numbers yields 1 in 24 trillion...
===============


Note that the 1 in 24 trillion odds shown are merely for winning the
two lotteries over any time period.  The odds of winning them both
within the hour are no doubt considerably higher than just 1 in 24
trillion.


Despite the astronomical odds involved, this is not the first case of
a double-jackpot winner:


===============
The New York Times
February 14, 1986

ODDS-DEFYING JERSEY WOMAN HITS LOTTERY JACKPOT 2D TIME

Defying odds in the realm of the preposterous - 1 in 17 trillion - a
woman who won $3.9 million in the New Jersey state lottery last
October has hit the jackpot again and yesterday laid claim with her
fiance to an additional $1.4 million prize...
===============


I think this certainly stands are quite possibly the longest long-shot known.

My guess is that the case of the multiple lightning strikes mentioned
by guillermo-ga -- which I believe to be a true story -- are probably
not regarded as fully random events.

Lightning strking, say, a lightning rod multiple times would certainly
not be deemed a rare or unexpected event.  It may well be that some
unfortunate human beings -- for reasons unknown -- act a bit like
human lightning rods, and experience multiple strikes during the
course of their lives.


I trust this is the information you were seeking.

However, please do not rate this answer until you have everything you
need.  If there's anything else I can do for you, just post a Request
for Clarification, and the odds are very good that I'll be able to
help you further.


All the best,

pafalafa-ga


search strategy -- searched Google and several newspaper databases for:

[ odds trillion OR quadrillion ]

---

Request for Answer Clarification by jim258kelly-ga on 28 Sep 2005 10:33 PDT

Alright, can we be sure that we have checked every person that ever
won more than one lottery.

Finally is there any other kind of event, besides the lottery, that
fits my criteria.

---

Clarification of Answer by pafalafa-ga on 28 Sep 2005 11:18 PDT

Jim,

Interesting questions!


>>...can we be sure that we have checked every person that ever
won more than one lottery...<<

100% sure?  No.

But reasonably sure...Yes, I think so (with a few caveats).

Lotteries are pretty public events, and even the winners of single
jackpots receive a good deal of media attention.  Double jackpot
winners are certainly a rare-enough, and newsworthy enough, event that
one would expect them to be well-covered in the news media.  Since I
did a fairly thorough search, I think I've captured the key most
unlikely lottery events with the longest odds of occuring.

The caveats are, lotteries have been around for centuries, so I can't
really vouch for historical circumstances.  The news databases I
checked generally extend back to the 1970's or so.

Also, the databases are largely limited to English language news, so
again, I can't vouch for what may have happened in, say, Italian
lotteries.

Still, given the wild popularity of lotteries in the US in the past
few decades, I think most of the strange events that have occured in
lottery history have probably taken place here in the US in the past
25 years or so.



>>...is there any other kind of event, besides the lottery, that fits
my criteria...<<

Good question.  Your criteria were very specific:

"Please do not answer with an event whose chances of occuring cannot
be calculated.  The event must have happened, must be proven to have
happened, and must have a statistical chance of occurance calculated
either before or after the event happened."

Very few events in the world can be calculated with reliable and precise odds.

Don't get me wrong...people can certainly >>create<< estimates,
saying, for instance, that the odds of life evolving are a zillion to
one, or whatever.

However, no two statisticians are likely to agree on the stated odds,
since there are so many non-quantifiable variables that come into
play, that attempting to pin a number on most events becomes just a
fool's game.

There are some exceptions, though, and the best ones involve gambling.
 Two independent statisticians should pretty much come up with the
same numerical odds for winning a lottery game, or being dealt a hand
of blackjack from a fresh deck of cards.

Gambling games are set up deliberately to be able to 'find' the odds
involved, either through calculations (of random number occurences, as
with lotteries) or through compilation of many individual bets (as
with the 'odds' assigned to a horse in a race).

Therefore, I do think it very likely that almost the only class of
events that are going to meet your criteria would be gaming events. 
And the 'purest' form of gaming is the lottery, where nothing other
than pure chance comes into play (as opposed to the horse races, where
a knowledge of the horses can help one in placing a bet).

Therefore, I'm tempted to say that -- other than lotteries -- only
other gaming type events would possibly meet your criteria.  Perhaps
someone has hit the jackpot multiple times in a slot machine -- this
would be a very similar, very long-shot occurence to winning the
lottery jackpot twice.

However, my search on trillion-to-one odds turned up the lottery
events, but not other gaming events.


Having said that, let me also add that there are many ways to
interpret your original question.  Others (or you!) may reasonably
disagree with the points I've made here.

If that's the case....just let me know through a further Clarification
request.  I'll be happy to continue discussing this with you, or
conducting further research, if that is your desire.

Let me know,


pafalafa-ga

I wish we had information that extended further into the past so we
could be sure that this was the least likely event.  But I'll except
it for now.

How about the evolution of intelligent life on earth:

"...The odds are against the right combinations of circumstances
occurring to evolve intelligent life on earth. The odds are about
400,000 trillion trillion trillion trillion to one. Evolution is
fantastically improbable. I believe that it did occur, but that it
could never occur again on any planet or any other solar system..."

http://www.socc.org/archive/Apolegetics/IsThereAGod.html

I don't believe that the scientific community can come to an agreement
of how probable Evolution is.  If so I would like to see the
documentation behind that.

However, the origin of the Universe was the starting topic for this
question.  One of the main points in Intelligent Design is that
Evolution is too unlikely to have happened.  And so failing knowing
how probable Evolution is I decided to ask what the least likely event
is.

If Google can show that the least likely event is Evolution, and give
a statistical probability of it happening, well then I think that
proves that Google is the next comming of Jesus, and the Internet is
god.

I believe the 1 in 24 trillion calculation might be valid only if the
winner played each lottery game once and only once.

The odds (chance?) becomes less if the winner played multiple times,
though I would still expect such a double winning to be a rare
occurrence.

Try this extreme situation: a person wins big on one lottery with 1 in
a million odds, then turns around and buys 100,000 tickets on a second
lottery where the odds are 1 in 50,000. The chance for winning both
lotteries is not the product of 1 in a million times 1 in 50,000.

Some people are very exact in their meaning of "odds" and "chances"
and this might have an effect on the calculation. I could be wrong in
ignoring that distinction here.

A simpler example might be if I had two dice, one red and one green.
If the odds of rolling a six with the red die are 1 in 6 and the odds
of rolling a five with the green die are 1 in 6, the odds of rolling a
six with the red die and also rolling a five with the green die on the
same day are 1 in 36 only if I roll each die once and only on that day
and do not try again on multiple days. If I roll both dice multiple
times on a day and do that every day for 17 years, the odds of finally
hitting both the six and the five on the same day are not 1 in 36; I
would expect that the combination of a six and a five would have
occurred multiple times over those 17 years, that is, the probability
of getting both the 6 and the 5 on a single day is actually quite
likely.

Despite all this, I'm not saying winning two lotteries is likely; it's
quite unlikely; it's just that I don't think lottery officials
statement on combined odds is correct.

Assuming that the $20 a day was all spent on $1 tickets for the
$126,000 prize for 17 years would mean the winners bought 20 * 365 *
17 = 124,100 tickets. Of course the prize was $126,000, so they would
come out ahead. If they place only half the bets on the Fantasy Five,
they certainly beat the odds, having won $126,000 after spending
$62,000 on tickets over the years, but I don't think their winning the
lottery was still 1 in 576,000.

brix24, you are wrong.  You are basically say that just by playing the
lottery a person's chances increase every time they play.  You are
saying that if I buy one ticket, every time the lottery is drawn, the
more drawings I play in the better my chances get.  This is absolutely
false.  If it were true, the lottery officials would have to restrict
people from playing, because they would all eventually win.

Everytime the lottery numbers are drawn, the chances are the same, for
each number played.  That's a fact.

If you truly believe what you've stated, I suggest you play the
lottery every day from now on, then one day when you're an old person
you are certain to win.

I think Brix was saying that if someone bought a whole batch of
tickets to a lottery he you have a better chance of winning that
lottery.

Back to the original question and Jim's wish for older data:
Lotteries are bigger now with many more tickets sold, so the chance of
winning on one ticket is less.  Ergo, older data would be dealing with
events in which the chance of winning was greater.

The odds of someone winning two lotteries at the same time are not one
in 23 trillion. It is not in the least remarkable for someone to win
the lottery, it happens every time the game is played. The improbable
event here is that the same person won the lottery twice. If the odds
for each lottery are 1 in 5 million (square root of 25 trillion), and
the winner played the lottery only those two times, than the odds are
1 in 5 million (not that unlikely given the thousands of lotteries
that are played every day). Now if you are the guy, you could say,
wow, that was improbable, but you are not the guy, it is some random
guy at the end of a web link, one of the millions and millions of
people who play the lottery.

kottekoe, you are wrong.

In many lotteries it is not guaranteed that there is a winner
everytime.  For example the Powerball Lottery (www.powerball.com) has
about 147 million possible winning lottery numbers everytime it is
played.  There is only a winner if someone bought the winning number. 
Many times there is no winner.  In which case the jackpot is rolled
over till the next drawing.

As for the rest of your statement, it is just mathematically wrong. 
Do you honestly think "Stanford University statistics professor Tom
Cover" (http://www.freerepublic.com/focus/news/805438/posts)
miscalculated the chances of Angelo winning the two lotteries?

When I first saw that article I thought it was comical that they
contacted someone at Stanford University to multiply two numbers
together.  But apparently, for the general public to accept a basic
math calculation, they needed to get an expert.

jim258kelly-ga, I'm sorry for the confusion. I am not saying "that
just by playing the lottery a person's chances increase every time
they play." Nor am I saying "play the lottery every day from now on,
then one day when you're an old person you are certain to win."

I agree that "Everytime the lottery numbers are drawn, the chances are
the same, for
each number played."

Just because someone didn't get a six in five throws of a die doesn't
mean that that person will get a six the next time he throws. (This is
where we both agree.) However, the a priori odds of observing at least
one six in throwing a die six times are 67% (calculated as "1" minus
the odds of not throwing a six, "(5/6) to the 6th power)." There are
times when multiple plays do affect the odds of winning. (We may or
may not agree here.)

The only way to be certain of winning a lottery is to successfully
carry out the plan that an Australian syndicate used in a Virginia
lottery in 1992, that is, to try to buy one of each possible number
combination for a single lottery. Even then, the prize has to be more
than the cost of buying all the tickets and paying people to buy so
many tickets. And you still  have to take into account the probability
that someone else will also have the winning combination and that the
prize will have to be split. I believe the Australian syndicate ran
into logistic problems in buying tickets couldn't physically buy all
the tickets they planned to. Because of that, they could have won
nothing for the $5 million they spent.

It remains true that a lottery is where you are likely to observe a
rare event and know that it is truly rare. You know both the odds for
a single play, and you have millions and millions of tickets played so
there is a better probability of observing a rare event.

The Stanford University professor quoted at the web site may not have
known that the person winning the lottery bought more than one ticket
for each lottery on that day. It appears that the winner was buying
about 10 tickets for each lottery each day. The number of tickets the
winner bought that day affect the probability of his winning that day;
the odds of his winning both lotteries on the same day are made
somewhat better by those multiple plays - but a far cry from making
the double (or even a single) win likely.

Despite my believing that the odds are not calculated correctly for
that specific win, I agree that the odds of winning both lotteries by
that person is the rare event you are looking for.

I certainly do not want to encourage anyone to make multiple plays. If
I recall correctly, some couple once spent $10,000 on a single lottery
in hopes of making their odds better. They lost everything - which is
the expected result.

I also believe it is a bad decision to spend $124,000 on lotteries
over the years and expect to come out ahead. The most likely result is
that you would be out $124,000.

I guess I didn't make myself very clear. I am not claiming that the
Stanford professor miscalculated the odds that a random pair of
entries in these lotteries was one in 23 trillion. Certainly, for the
winner this is a remarkable event. My point is that the event itself
is not especially unlikely, given the large number of lottery games
that result in a winner. The original questioner was looking for
highly improbable events, not for mildly unlikely things. The fact is
that someone had to win the jackpot (yes I understand that they keep
playing until someone wins). So the probablility that someone won the
larger lottery is 100%. It turns out that this guy also entered
another lottery. According to the above article, the probability that
he won the other lottery is about 1 in 500,000. Thus, the probability
that these two lotteries had the same winner is 1 in 500,000 times
100% or 1 in 500,000, still a highly unlikely event, right? No, not
really, when you consider that there are probably many such lotteries
that conclude every day. I would guess it is quite likely that the
winners of the big ones had also entered smaller lotteries concluding
on the same day. If one of the larger lotteries concludes every day
for three years, that is 1,000 lotteries, so under these assumptions,
such an event would have a probability of 1 in 500 every three years.
Still not likely, but there is nothing especially remarkable about it,
even though it would be extremely unlikely for it to happen to you.

This is similar to the situation where two kids in the same grade
school class have the same birthday. The odds of any two kids having
the same birthday is only one in 365, but in a class of 30 kids, it is
quite likely that two of them will have the same birthday.

kottekoe, it is very clear that you don't understand anything about statistics.

Wow!  If I have a 1 in 500,000 chance of winning $17 or even the $5.3
million I'll be rich very soon.

That's ludicrous, that means you can invest $500,000 to win $17
million every time.  That's a 3400% return on my investment or a 1000%
in the case of the $5.3 million!

Cool off guys, let's not cast aspersions about each other's statistics
abilities. Read my message carefully. I did not say that an
individual's odds of winning two lotteries on the same day are one in
500,000. The odds quoted in the article above for winning the smaller
jackpot (which paid $126,000) are one in 576,000 (which I rounded to
half a million). My point is that the existence of such an event is
not especially remarkable, though the odds of it happening to any
given lottery player are indeed low.

Let's consider a simpler example. Jim258Kelly is searching for an
unlikely event. In his original question, he cites a lottery with
chances of one in 65 million. But there is nothing the least
remarkable about the fact that someone won the thing. This is not an
unlikely event at all. It is an unlikely event for any given lottery
player, though. If you had entered and won, you would be amazed.
However, you are not amazed, just envious, when you hear on the news
that someone else won. According to the rules of the game, they keep
playing until someone wins. The odds of someone winning are 100%.

The example of one person winning twice in one day is much better. At
least we are no longer talking about an event that is guaranteed to
occur.

This is not a complex point, but it is a common fallacy when people
talk about coincidences.

I now understand your point kottekoe.  But are you saying that an
"event" which only happens to one person does not count?

I guess I'm saying that your question is a very difficult one to
answer and one should be careful about assigning probabilities to
events. It is clear to me that a winner of a lottery is not a good
example of an unlikely event. Somebody winning two in one day is much
better, but not a one in 23 trillion proposition.

Here is another interesting one, the woman who was hit by a meteorite
in Alabama in 1954:

http://www.xenophilia.com/zb0005.htm

But I read somewhere else that according to one calculation, someone
in the US should be hit by a meteorite every 9000 years. If this is
accurate, the odds of someone in the past century being hit are one in
90. It has been 50 years since Annie Hodges was hit, so this shouldn't
win a prize for being a highly unlikely event, even though the odds of
anybody you or I know being hit by a meteorite are vanishingly small.

I like your question because it is so difficult to answer.

i think everyone has neglected that the least likely event to ever
happen to anyone and that has happened is to be born as someone else.
for example you are never going to be born as Danny DeVito, and this
is never going to happen again. however Danny Devito was born,
therefore thgis event has happened, it just will not happen again.