Newsgroups: alt.ufo.reports
From: kymhorsell@gmail.com
Subject: why are more ufos seen in some states but not others

[uploaded 32 times; last 06/09/2024]

EXECUTIVE SUMMARY:
- We run a new s/w over UFO sightings of different types and see which
  state demographics predict local UFO activity.
- Some factors occur in several UFO models. It seems state temperature
  and distance from the sea figure in several key "types".
- Some key factors may relate to the behaviour of UFO witnesses.  But
  almost all seem to relate "mostly" to the reported behaviour and/or
  imputed interests of UFOs.  We can reason in some cases a factor can
  not reasonably be associated with state residents in the way or to
  the extent the stats procedure says it does.
- Finally, we look at a couple of key UFO types and note some
  interesting and some unpleasant associations they may have.  In many
  cases the attributes found by the s/w using only relatively
  unimaginative data science techniques and a wide range of data are
  the same suggested by many UFO witnesses and/or researchers over a
  number of decades. Seemingly, no matter how crazy some of the claims
  may sound, some of them seem to have some backup in the hard data.

The AI's have been chattering away and come up with some new takes on
the patterns of UFO sightings across N America.  They've had some new
tools to play with in the past few months and this report is based on
state by state "lasso regression" that validates the model found.

A lasso regression involves not only trying to find the "best fit" of
a line across a dataset to follow a given target variable (in our
case, here, some kind of UFO count state by state) but it also tries
to make the resulting model "as simple as possible" by also including
a special "regularisation term" to the fitting value to be
optimised. Normally this means as many coefficient as possible in the
model get set to 0 and only the "key variables" are left to explain
the target data as well as they can.

If that wasn't enough of a thing to grasp in one sitting, the
"validation" part gets added on. It turns out just statistical tests
are not always the best way to establish a model is "good".  It may
appear to fit the data very well (typically meaning the so-called
"explanation power" statistic -- the R2 -- is reasonably close to 1)
yet have no predictive power at all. A new set of data -- despite the
claims of statistical tests -- might fail to be well-estimated by the
model because "something changed" between the data you had
when the model was created and the time you got the new data to use to
make new estimates. To better ensure a stat model will work with new
data at some time later we have to at least run a test for that at
least once.  And the way most stats-heads and s/w packages do it is to
hold some of the data aside when building the model to test whether it
actually explains the "holdout data" just as well as the data used to
build the model.

So in the procedure used here the model is created using just a random
1/2 of the data to hand, then it is tested against the other 1/2 of
the data and if both halves are reasonably-well explained by the model
then it is accepted as having not only statistical robustness, but
predictive power.

So these are the models we will now look at. Using all the state
demographics and other data to hand we will try to build robust and
predictive models for UFO activity of various types over each state.
In the UFO dataset I normally use -- the set from the NUFORC -- the
types of sighting are normally broken down by "shape" of the object or
objects seen. The NUFORC have a long-established set of shapes and I
normally follow those fairly closely. Some new shapes tend to get
added over time, following "fashion sense" among witnesses.  E.g. a
"tictac" was added at some point. But I also have chosen to sometimes
add yet other shapes my own fashion sense dictates.  E.g. I look in
the (short) comment section of reports for certain keywords like "faint",
"bright", "cube", or "pyramid", among others.

And so we will look here at how a series of validated lasso
regressions explained sightings of a collection of UFO types across N
America.  Viz:
	pink
	purple
	Egg
	Disk
	Oval
	Circle
	Sphere
	bright
	faint
	gold
	red
	yellow
	black
	Light
	notLight
	Triangle

In general those keywords starting with a capital are NUFORC's chosen
shape names while the lower-case words are ones I've chosen for one
reason or another to locate in the comment section of reports.  They
are typically colours but also can indicate other things like "bright"
and "faint". Most of these can be applied to Lights type UFO's.  But
they also may apply to lights seen *on* some other type of UFO.
Typically a witness reports whether there were lights on the bottom of
some otherwise dark object, or sometimes they see what they think are
"port-holes" or "domes" on object through which some coloured light is
visible as well. And, finally, the notLight category is all of those
UFO's not in the Light category. It make sense to split up what people
typically see as "little lights going across the sky at night" and
everything else they might see typically in the daytime and oftentimes
reasonably up close and personal.

There are of course many other shapes and colours. But these are the
boundaries we're using today. :)

I currently have 915 state-by-state datasets. These cover a very
wide range of things including social factors, economy, environment,
weather, medical, law, crime and many other things.  All of these are
thrown into the mix and all the above UFO types are cross-checked
against all 915 factors and the best ones "lassoed" and validated.

We then can check to see whether some factors are common across all
the selected UFO types. And there are a few.  Ignoring the factors
that only apply to one type we find the common factors are:

Variable	Number of UFO types with this var as a relevant factor
		in NUFORC sightings since ~1950
coffeerank		5
maxtemp			5
murder2019pc		5
coastdist		4
diversity		3
infgasgatherint		3
medage			3
birthrate		2
carspc			2
casespc20210408		2
cattledeathspredpct2015	2
coalconsint		2
coviddailycasespc20211121 2
dope			2
evpct			2
lung			2
maskapr2021		2
minpre			2
missingdoenetworkpc	2
pctgenrenew		2
pctnonwhite		2
planeaccidents-personal	2
pre6			2
suiciderate		2
totdeathspc14feb2021	2
under18			2
usoutagesvandalismpc	2
wind			2

Hopefully most of these things are self-explanatory. Some factors are
related to conditions in the US at a certain time. Some are factors
each month (e.g. "pre6" is avg June precip). And some are averages or
totals over an otherwise-unstated period of time (e.g.  "wind" is the
avg monthly NOAA-designated "windstorm"s since record began in the
1950s upto the present).

But first off we see some things appear to apply to 5 of the 16
categories we've selected to look at. These "biggest factors" are how
much coffee do state residents drink, what is the state's max
temperature, and how many murders per million residents were recorded
for 2019.  At this point let's leave open the sign of any of these
factors.  The relevant UFO's may be more numerous associated with
either larger or smaller values of these demographics -- we're not
saying yet which or what.

But we can at least have a laugh. It seems one of the chief reasons
people make reports of seeing things in the sky or water -- they drink
or do not drink a lot of coffee. So we might reason coffee drinkers
might see more things because they are more alert, they work shifts and
are up and about at night, or it may suggest the kind of people that
do (or don't) drink coffee are just also the kind of people that have
to report weird stuff to someone or they will bust.

The other interesting items on down the list cover weather
phenomena and some folklorish-type things also appear -- e.g. cattle
deaths, COVID, plane accidents, missing people, power outages, young
people, old people, the white/nonwhite thing.

The unusual factors include fossil fuels and renewables. Dope smoking.
(How can this be at all relevant to whether someone things they saw
something are got a telepathic communication from Neptune? :)

So first up -- it seems some of the "usual suspects" have shown up out
of a casting call of nearly 1000. The more we look at the data the
more it seems many of the claims of witnesses have something to them.
Maybe not in the way they claim or as many others think, but at least
"some" thing.

So let's look in detail at least at a few of the UFO categories and
see what tickles their fancy, so to speak.

First up. UFO's in general. What states see overall sightings and why?

The relevant model looks like:

 REWEIGHTED LEAST SQUARES BASED ON THE LMS
 *****************************************

 VARIABLE     COEFFICIENT    STAND. ERROR     T - VALUE     P - VALUE
   ----------------------------------------------------------------------
       x1     -22.46857       3.66484      -6.13084       0.00000
maxtemp
       x2     36.35508        6.56023       5.54174       0.00000
murder2019pc
       x3     -2.72445        0.57501      -4.73811       0.00004
coffeerank
       x4     -11.86986       2.08848      -5.68348       0.00000
coastdist
       x5     -20.65764       4.15430      -4.97260       0.00002
povertyavg
       x6     14.28373        4.56078       3.13186       0.00363
medage
       x7     755.07172       277.74136     2.71861       0.01037
cattledeathspredpct2015
       x8     -0.86123        1.30898      -0.65794       0.51514
lung
       x9     3.96524         7.41957       0.53443       0.59663
cfr20211219
 CONSTANT     644.73535       211.37297     3.05023       0.00449

 WEIGHTED SUM OF SQUARES =     81807.57812
 DEGREES OF FREEDOM      =        33
 SCALE ESTIMATE          =        49.78973
 COEFFICIENT OF DETERMINATION (R SQUARED) =        0.90595
 THE F-VALUE =       35.318 (WITH   9 AND   33 DF)   P - VALUE = 0.00000
 THERE ARE    43 POINTS WITH NON-ZERO WEIGHT.
 AVERAGE WEIGHT          =         0.84314

We see immed that all factors are statistically significant except 2.
But we are using a validated model here. While "statistically" lung
disease (cases per capita) and "case fatality rate" from COVID as at
19Dec2021 don't go up and down across states in close parallel with
the way UFO sightings go up and down across states -- but when added
to this model they increase its predictive accuracy.  Remember, we
have built this model on 1/2 the states (selected at random) and
tested it on the other 1/2, finding the average error on the
predictions were comparable to those of the 1/2 used when the model was
built. So all the variables listed contain "useful information" whether or
not a statistical test can decide whether it should be there or not.

Another thing to note. In the lasso regression program the order of
variables listed is usually the order of importance the program
assigns to them from all relevant statistical information (too
voluminous to print here!). So the first variable -- and most
important to predict UFO sightings -- appears to be the state average
temperature (these come from a database of all state weather stations
the NOAA maintained between at least 1950 and 2000, with many
maintained by other agencies back to 1900 and in some cases back to
~1850 as well by certain individuals rich enough to have one of them
new-fangled thermometers in their barn).

The coeff is -ve meaning higher max temps mean fewer sightings.  So
this suggests UFO's like it cold. Typically its hard to decide whether
a factor is "talking about" the witness or the UFO behaviour, but in
this case we can imagine cold weather is not likely to encourage
people to go outdoors more, or look up in the sky more.  If anything,
cold weather should be associated with less UFO's if it were just up
to what we understand of human behaviour.  So it seems this factor
says something about the UFO's inherent behaviour.  And given previous
posts that suggest much UFO behaviour seems goal oriented or rational
then we are talking about "desired" here.

The 2nd most important factor seems to be the murder rate.  The coeff
is +ve. The more murders seen in a state in 2019 the more the UFO
sightings coming out of that state (all ufo counts used in this report
are per million capita) since those records began.  This might be a
factor that applies to witnesses. The more likely you are to be killed
-- no matter how small the objective prob is -- the more people watch
out, maybe even looking at the sky.  But it could also apply to UFO's.
Perhaps UFO's don't like the heat. But perhaps they like
gunfire. Perhaps even the murder itself is some kind of attractant.
"The kind of people we need to watch more closely", maybe.

At number 3 is coffee rank. The lower the rank the more that state
loves coffee! So the coeff is again -ve meaning the more residents
drink coffee the more UFO's they tend to report.

At No 4 is distance from the coast. Again -ve. The closer to the coast
the more sightings. I.e. UFO's seem to be associated YET AGAIN with
the sea.

At No 5 is poverty. Again, this might be something that applies to the
behaviour of witnesses and also UFOs. Since the coeff is -ve it may be
states with higher levels of poverty have fewer people with the time
to look around them or watch the sky. They are out working 2 or 3 jobs
or are unconc in a gutter.  Or it might be the UFO's. States with low
poverty rates are the ones that attract them. Maybe they perceive a
threat from those well-off and less of a threat from people that are
poor.  And as the Pentagon says daily, you have to monitor people that
maybe threaten your way of life. Maybe even just your ideals.

The next factor is population age. Coeff is +ve. The states with older
populations report more UFO sightings. This may be related to the
habits of older people. Maybe they spend a lot of time looking up.
(I sure know I do). But it also may be associated with UFO wants.  They
are "attracted" to older populations for some reason.

And the last factor that *is* stat significant is percent of cattle
deaths in a state that have been attributed to predators of some
kind. Hugely +ve. For each 1% of cattle deaths "authorities" have
tagged as caused by a coyote or some-such, there have been an
additional 700-800 sightings per million capita in that state since
NUFORC data began. Again. This MIGHT relate to witnesses.  If cattle
ranching is a big thing then there might be more people employed in
the industry or living out in rural areas.  Maybe such people have a
lot of time to watch the sky and see more UFOs.  But it also starts
alarm bells ringing. Given the long long history of stories about
cattle being killed mysteriously overnight with no witnesses and
having unusual marks or damage to their bodies, sometimes appearing to
have been dropped from a height, and with no footprints either human
or animal anywhere nearby.  Recent documentary footage also suggests
when UFO's go overhead -- whether visible to human eyes or only to IR
cameras -- cattle sometimes react like they are scared out of their
wits.  For some reason.

The other 2 factors are -- as mentioned -- not stat significant but
nevertheless are proven predictive of UFO sightings. Lung disease
seems to be a weak repellant (-ve coeff). And COVID fatality rates
seem to be a weak attractant (+ve coeff). From other work we have some
indication that, overall, UFO's of many different types seem to "scare
off" COVID fatalities, but there seem to be a select few types that
are stat sig associated with more cases and more fatalities.  This is
the "all UFO's" group here, so the overall summary is "weak association
but predictive; maybe a minority of the population is involved in some
non-trivial way".

And, finally for today, let's look at another interesting category.
"Light" UFOs. These are of particular interest to me because I seem
to see so many of them. Over the years of the pandemic I got used to
seeing them every night. It became so usual that "something"
would go over and act strange at night that when visitors called
around late in the day I would invite them to sit out in one of
several yards I have on the property, have a cup of coffee and a chat,
and watch for the light show around 7-8-9pm. Was never
disappointed. Sometimes visitors saw the lights. Saw
little planes chasing the lights, sometimes in and out between the
clouds. But sometimes made not much out of it apart from "what is the
big deal -- doesnt seem unusual to me; Chinese satellites y'know".

Light UFOs:

 VARIABLE     COEFFICIENT    STAND. ERROR     T - VALUE     P - VALUE
   ----------------------------------------------------------------------
       x1     -0.00709       0.00377      -1.87855      0.06761
deathspc20210408
       x2     -1.05109       0.32227      -3.26148      0.00227
pctblack
       x3     -8.10105       1.24383      -6.51300      0.00000
torn7
       x4     55.61506       10.06568     5.52522       0.00000
carspc
       x5     -0.55233       0.16195      -3.41046      0.00149
coffeerank
       x6     7.81158        1.77127      4.41016       0.00008
stlifex
       x7     3.90940        0.96536      4.04970       0.00023
medage
       x8     -4.53755       1.30144      -3.48656      0.00120
missing2019int
       x9     1.45956        0.40913      3.56746       0.00095
missingdoenetworkpc
 CONSTANT     -689.87598     149.56705    -4.61249      0.00004
 WEIGHTED SUM OF SQUARES =      8028.60742
 DEGREES OF FREEDOM      =        40
 SCALE ESTIMATE          =        14.16740
 COEFFICIENT OF DETERMINATION (R SQUARED) =        0.88672
 THE F-VALUE =       34.791 (WITH   9 AND   40 DF)   P - VALUE = 0.00000
 THERE ARE    50 POINTS WITH NON-ZERO WEIGHT.
 AVERAGE WEIGHT          =         0.98039

The first-up variable is total covid deaths up to a given date per
capita.  There were several dates like this to choose from. But the
program found 2021Q1 was the only one it needed. Coeff -ve. So the
fewer deaths upto this point the more likely to report a UFO since
1950. So -- we assume -- people that take precautions during a
pandemic; wash hands, wear masks, stay away from public areas and
don't go visiting friends and relatives.  These are the people that
see a lot of UFOs. Or they are the people the UFO's want to
see. Probably a mix of the 2.

2nd up is percent of state African American. Coeff -ve. Either black
Americans are too busy working to look up or white folks are more
interesting to UFOs. There is other data that suggests some kinds of
UFO's have an "affinity" for Jews and Africans and outside of the USA
they do cluster around the relevant areas. So maybe this attractant is
only available in the US. Whatever it is.

The 3rd factor we have seen (and used) before. Tornadoes. In one of
the peak months -- July. Coeff -ve. More tornadoes in July mean fewer
UFO sightings. You might expect people to report more UFO's if
tornadoes are throwing unusual things up into the sky to look at. But
it could run the other way. This might suggest tornadoes more influence
UFOs than state residents. UFO's are repelled by tornadoes, we might
suspect.  We have seen elsewhere that UFO sightings "pile up" on
one side of a line of storms in the US, while they "thin out" on the
other side of the line. Meaning perhaps no matter how well they appear
to hold position during high winds they do not "like" to travel in
very stormy weather.

Again the next 2 factors relate to pop age. Older people apparently
report more UFOs. But also states with higher life expectancy report
more UFOs. Maybe less to do with how people with high life expectancy
behave and more what UFO's may be looking for.

And, finally, missing people. Seemingly contradicting each other the
program finds the density of missing people per km2 (and 2019 was the
only year I could find where FBI missing persons were broken down by
state -- other reports have them all aggregated nationally) is -ve
correlated with UFO reports but missing persons reported in the DOE
Network are hugely +ve correlated.  The DOE Network is a collection of
agencies around the world that are trying to identify dead bodies.
Usually only parts are found.  While witnesses may behave
differently in states where many people go missing and only end up in a
morgue years later and cant be ID'ed; it sounds like it's more a case
of either Lights are attracted to states were more people go missing
and are never seen alive again, or maybe the relationship is in some way
causal. Not only are missing people predictive of Light UFO sightings,
but the links are very stat significant as well (the P-VAL's indicate
99.9% probably not just chance).

In a later post we'll come back to look at some factors that are
common to multiple UFO types.

--
The US government portrays itself as the world's preeminent
superpower, so to acknowledge that there are things in their
airspace, whatever they are, that are faster and more manoeuvrable
and run rings around fast jets doesn't play very well.
So there's the embarrassment factor, and maybe a little bit of
fear that either an adversary has made a quantum leap in
development, which has left the US in a poor second place, or, as
some believe, this really is extra terrestrial, in which case we're
not at the top of the food chain anymore.
-- Nick Pope, 02 May 2023

Don't worry, we'll never run out of oil
Interesting Engineering, 9 Nov 2022
That being said, at current consumption, we have by some accounts an
estimated 47 years of oil left to be extracted. That equates to somewhere in
the region of 1.65 trillion barrels of proven oil reserves. Other sources up
this estimate a bit, but most agree we have around 50 years left, give or take.

Physics Thinktank Proposes Method for Detecting Extraterrestrial Spacecraft
Using Gravitational Waves
The Debrief, 16 Dec 2022
An international team of scientists has written a paper showing how to
detect extraterrestrial spacecraft using gravitational waves.
[The reason LIGO hasn't been looking for "warp signatures"?
Nobody thought of it].

The most extreme life-forms in the universe
New Scientist, 26 June 2008
There's hardly a niche on Earth that hasn't been colonised. Life can be
found in scalding, acidic hot pools, in the driest deserts, and in ...
[Interestingly, if life is *not* found in the warm salty sub-surface
oceans of some of our system's moons it gives more weight to the idea
that life could not have formed spontaneously on Earth but came from
"outside" e.g. via meteorites aka Panspermia].

But what is true and I'm actually being serious here, is there are, there's
footage and records of objects in the skies that we don't know exactly what
they are, We can't explain how they moved, their trajectory. They did not
have an easily explainable pattern.
-- Pres Barack Obama, "The Late Show", 2021

"[F]or the few cases in all domains--space, air, and sea--that do
demonstrate potentially anomalous characteristics, AARO exists to help the
DOD, IC, and interagency resolve those anomalous cases. In doing so, AARO is
approaching these cases with the highest level of objectivity and analytic
rigor. This includes physically testing and employing modeling and
simulation to validate our analyses and underlying theories, and
peer-reviewing those results within the U.S. government, industry partners,
and appropriately cleared academic institutions before reaching any
conclusions."
-- Dr Sean Kirkpatrick, Senate Hearings on UFOs, 19 Apr 2023.

How big a deal is NASA's new UFO study?
[image] An unidentified flying object, as seen by a U.S. Navy jet.
Space.com, 18 July 2022
In early June, NASA announced that it's commissioning an independent study
on unidentified aerial phenomena (UAP), as UFOs have recently been rebranded.
The intent is to move the scientific understanding of UAP forward, said
Thomas Zurbuchen, NASA's associate administrator for science.
"NASA believes that the tools of scientific discovery are powerful and apply
here also," Zurbuchen said in a statement (opens in new tab). "We have
access to a broad range of observations of Earth from space - and that
is the lifeblood of scientific inquiry. We have the tools and team who can
help us improve our understanding of the unknown. That's the very definition
of what science is. That's what we do."
The UAP study team will be led by astrophysicist David Spergel, previously
the chair of the astrophysics department at Princeton University.
"Given the paucity of observations, our first task is simply to gather
the most robust set of data that we can," Spergel said in the NASA
statement.  "We will be identifying what data -- from civilians,
government, non-profits, companies -- exists, what else we should try
to collect and how to best analyze it."
[The investigation of why NASA has missed something obvious to many
for the past 65+ years will come later. But it argues that classical
approaches of Organised Science might draw another blank on UFO's].