john <vega...@accesscomm.ca> wrote:
Tell me how someone hoaxes publishes the
same complex diagram that will not be dug up
for ten years and in both appearances
it is connected with extraterrestrial humans?
john
The fact that you need to be told, shows how clueless you are about
science.
The first post of yours I read worded it "almost the same complex drawing"
(or something like that). Now that people have doubted your belief, you
have switched to "the _same_ complex diagram" in a stupid attempt to defend
your own lack of understanding.
The answer is simply John. If you look at 1,000,000 drawings made by
people over thousands of years, you will see duplicates. Oh my god, I
drew a circle, and look, the guy 40,000 years ago drew a circle. Gee, that
must be proof of UFOs. Do you have any clue how totally stupid such an
argument is?
Ok, so everyone draws circles so you expect to be lots of duplication. So
lets draw a circle with a line in it. Not as many duplicates in those 1,
000,000 drawings we have collected over the years. The more complex the
drawing gets the less likely you will be able to find a duplicate. So we
have to think about how complex the drawing has to get before we expect not
to find any duplicates in every drawing anyone has ever found on the earth.
Now, from experience with such things, we know our drawing doesn't have to
be all that complex, before it becomes extremely hard to find a matching
drawing. Once you add a few more circles, lines, and a random strokes, the
drawing becomes so unique that you have never seen anything like it. And
from the fact that we have never seen anything like it, we tend to extend
that belief to the idea that "there is nothing like it". But we haven't
ourselves, taken the time to look at every one of the billions of drawings
available to be looked at. So in fact, we have no real clue how common our
"unique" drawing really is. Our instincts based on our highly limited
first-hand personal experience is VERY deceiving. In fact, if we had a
better picture search system, we would find that the drawing has to get far
more complex than we might expect before it really becomes unique in the
set of all pictures ever drawn by man over the past thousands of years.
So when we are are shown a match, in the set, we need to try and understand
the true odds of that match happening by chance and whether the odds are so
out of line, that we can truly justify the argument that it didn't happen
by chance. But how to you scientifically measure the odds of such a match?
Well, if you look at it and just use your personal instincts to produce a
measure of "seems far too complex to have happened by chance" our own
personal experience will bias our view, and make us believe the odds of it
happening by chance are far too small. But our own personal experience, is
far too limited in these data sets that include billions of examples. WE
have never personally searched such huge data sets and as such, our own
personal experience with small data sets will deceive us. What looks
"impossible" often isn't even usually when dealing with these large data
sets (the set of all drawings every made by any man).
But the effect I talked about above, the effect of trying to judge how
complex a drawing we have to draw, before the odds of finding a duplicate,
is only one of 3 statistical problems at work in your "science".
The next one comes from that fact that you are searching for a match to ANY
TWO pictures every drawn by man, vs picking one, and then trying to find a
match for that one, and no other. The statistics in these two cases is
VERY different.
Let me give an example. Lets say we generate a set of a million random 12
digit numbers. Then you make up a random 12 digit number. What are the
odds that the number you made up, is in the set of numbers we generated?
The odds are not very good. It's about 1 in a million. You can make up
lots of 12 digit numbers, and most of the ones you make up will not be in
the set.
But what if instead, you search the set, to see if there are any duplicates
in the set? Without formal training in this, you might think the odds of
someone being able to show you a duplicate, is the same as the odds of
someone making up a single random number, and then finding it in the set.
But it's not. It's not even close. The odds of there being at least one
duplicate in the exmaple I gave above, is highly likely. Almost certain.
(sorry I'm not going to bother to try and calculate the real odds right
now).
The reason is that you aren't just looking for one number in the set being
a duplicate with every other number, but instead, you are looking a million
numbers (every number in the set) being a duplicate with some other number
in the set. You are in effect, doing the "pick a number and see if it
matches something in the set", a million times, and then asking, did any of
those million numbers match one of the million numbers?
There's a well known parlor odds is always trick that takes advantage of
how far off our intuition is on these types of problems.
If you get a group of people together, and ask them all what their birthday
is, how many people do you think you need in the group, before the odds of
finding a duplicate becomes 50/50? (aka highly likely). Our intuition
makes us think we might need half of 365 or around 180 people before the
odds hit 50/50 that there will be a duplicate. In fact, once you get to 23
people, the odds of there being a duplicate becomes greater than 50/50.
This is known as the birthday paradox. Google it.
It's one of many demonstrations of how far off our intuition can be when
dealing with statistics of collisions in large data sets.
Ok, that's only two points I've made so far. Lets move on to the third.
How close do two drawings have to be, before you will start to say "look
they are the same!"? In the birthday paradox, or by pick a random number
example, there's no doubt as to whether there was a match or not.
But when comparing two drawings, (likely to include lots of noise), the
human mind will _try_ to find similarities. That's what it likes to do -
recognize common patterns. They don't have to be anywhere near exact
before we will start to similarities. Hell, we can see a circle and a
square as being "the same" if need be. You draw four small circles in a
square pattern (the circles fall at what would be the corner of a larger
square), and then see a second drawing of four squares, of approximately
the same relative size and spacing, and we think of the two drawings as
being "very similar", even though one was 4 circles, and one was 4 squares.
So, what happens when this natural tendency of the brain to make things
that are different, seem to be "the same" gets applied to a birthday
paradox problem? We have billions of drawings made by man over 100's of
thousands of years, and someone finds two drawings in the set that "seem"
very similar?
This ability to see drawings that are actually very different, as "the
same" creates a huge amplification effect on the statistics of what we
intuitively were already getting way wrong to begin with. It makes the
ability for two items from the set to "match" millions of times more
likely.
So, back to science. Are these two drawings (which you have not shared a
URL with us), that is such good proof for you, truly good scientific
evidence? To answer that, we have to actually calculate the true odds of
the match, instead of trusting some misguided, and known to be way off the
mark - personality intuition on this. If we can't accurately calculate the
odds, then it's not science, it's just crap which is of a type of
statistics we know for dead sure is highly7 deceptive to humans - a type of
problem that we intuitively get wrong every time.
To calculate the odds, we need to know the size of the set we are dealing
with. How many pictures were searched to find a duplicate? Your example
of course doesn't include this number. It can't. How many idiot UFO
"researches" went looking for duplicate pictures, found none, that we never
heard from? But without it, we can't calculate the odds. And without the
odds, we have no way of knowing whether our intuition of it being "highly
unlikely" is in the ball park, or like the birthday paradox, of no use to
science. Evidence that seems good to our intuition, often can be shown to
be worthless, with a little examination. This is such a well known problem
(the fact that human intuition in these matters is almost always wrong), is
why the tools of science has been created in the first place. They are
tools, to show us the truth, even when our intuition wants to make us
believe something else. What science has made it clear, is that
institution is NEVER to be trusted, when looking for the truth.
Real science, NEVER includes a measure of human intuition in it's
arguments.
Your example, is pure human intuition. What is the odds of those pictures
happening by chance? We can't calculate it because the data was not
collected in way that allows us to know the odds. But what we do know
about your "evidence" is that it's a type of evidence well known for
fooling human intuition, and it's just the type of evidence, anyone
actually trained in the scientific method, knows to stay far away from.
It's the WORST type of evidence you can have in science. It's the evidence
which is MOST LIKELY to fool you into believing crap. No matter how
unlikely to THINK the odds of those drawings happening by pure random
chance, you are probably way off the mark.
But now, let me take one step further in the errors you are making here.
In all the above, I was only talking about the odds of two pictures being
"the same" in the set of all pictures ever drawn by man. Your error goes
even further than that.
What we are dealing with, is the set of all possible proofs that
intelligent life from the rest of the universe visited us ...
read more »