the game of jet and ufo #3

[kymho...@gmail.com]

We've been looking at simple game theory models that might explain the
observed ongoing interactions between UFOs and conventional aircraft.
In the NUFORC database (up to end 2023, anyway) 3976 reports (just shy
of 4% of the total) mention an unusual object that is either being
observed, chased or is just harassing a conventional aircraft,
typically a military jet or chopper. (The "interaction" reports
mention "military" aircraft about 10% of the time; jet (aircraft)
about 20% and chopper about 20%).

The question is -- what's going on here? The beauty part about our
question, if we use some relevant theory, is we might be able to tease
apart the intentions or motives of each "side" in the perceived
contest as well as their methods. (As well as how "hot" the contest is :).

Up to this point we tried to apply simple game theory -- the math
theory that deals with situations where different "agents" have
mutually contradictory aims or goals. The usual application of the
theory is to determine what strategy in a particular game/scenario is
likely to produce the best results for either side and e.g. whether
some kind of stalemate is inevitable and everyone should just give up
and go home now.

But we'll aim higher with the help of a little AI programming.

What we propose to do here is assume some simple things about the
contest between jets and ufos -- informed by what may be our erroneous
interpretation of what we have seen in the skies for the past 70+
years -- and figure out what 2 "sides" with approx equal levels of
intelligence would learn to do to most mostly achieve their goals.
We'll allow our program to enumerate a large number of possible goals
and determine which set most closely matches the observations we have.
Ergo deducing (maybe!) what the h*ck is going on.

The "AI" part of the equation comes as a general learning algorithm
that will operate each side of the contest. Each "agent" or "side" in
the game will be given a suitable level of intelligence that it will
use to try to out-compete the other side. The simulation will run for
a certain number of rounds -- we might imagine these rounds as being
in "months" or "weeks" but we wont try to pin that down too closely :)
-- where they turn up to the office, pull some levers available to
them, and the universe renders its verdict and awards points to each
side. For the first 1/2 of the simulation each side is in "learning
mode". For the 2nd 1/2 of the game we'll gather some data about what
each is doing (assuming it has learned enough about the contest the
moves after then are not total nube) and finally look through the
observations and pick which levers and actions have resulted in a
situation that looks like the one we see in the IRL datasets. From
that we will see "what is going on". Or at least we'll have some kind
of mental model of what is going on even if it is a little off in some
respects. :)

As outlined in a prev post we'll assume there are 2 sides we label
here "jets" and "ufos". Each side can control the weather to some
extent. Each side can do nothing, or call in clouds in one of 3 layers
in the atm ("low", "medium" and "high"). Each side in a given round
can decide to fly at one of those levels or decide not to turn up at
all. Like many game-theory models we'll have all decisions made AHEAD
of the round, the actual round only determines which levers each side
decided to pull based on past experience. But seeing what the opponent
did in a given round will inform future actions.

At the end of the round each side is awarded points by the universe.
Points are determined by a simple function based on the various
features of the round. E.g. was there clear line-of-sight from the
ground to the ufos (if they were flying). Was there clear
line-of-sight from the ufos to any orbiting sat. Did the jets guess
right and there were ufo's to chase at their chosen level. Etc.

The learning algorithm will try to maximise the points returned to
each side as well as it can. With conflicting goals the situation can
converge to a stable (Nash) equilibrium where the chosen actions of
each side perform as well as possible for them and any change by them
will result in less cheesecake. Or it can be unstable meaning each
side chases mostly its own tail for eternity. There is also a
consideration whether the scenario is "zero sum" or not. In a zero sum
game the gain for one side comes as an equal loss for the other. It
models "maximum competition" or "finite resource" scenarios. We will
try both and see which seems closer to observed data.

In the 2nd 1/2 of each simulation we'll gather some stats on what was
going on. Which layers each side generally called for clouds. Where
each side decided to fly. How many times the ground had clear sight of
either jet or ufo or -- maybe more importantly -- both. And a few
other details.

Over the week or so I tinkered with this I came up with 8 different
types of simulation. 4 of the programs were "dumb" and did basic
learning; the other 4 were copies but with a "smart" learning model in
control. 4 of the sims assumed 0-sum scoring; the other 4 used a
different score chart for each side. Some events awarded points to
each side; some awarded a punishment to each side. And sometimes the
punishment to one side was much different from the reward awarded to
the other. And, more interestingly, some of the simulations allowed
the players to use more than the simple "cloud" and "fly" levers. In
one pair of sims each side could choose or not to turn on some kind of
additional stealth tech. If the clouds weren't enough to hide one side
from the other as they might like, they could switch something else
on. In one pair of sims the ante was upped even further. Might one
side actually be trying to ATTRACT THE ATTENTION of someone? As well as
a stealth mode, there might be a "dazzle" mode to the tech. And,
finally, in 4 of the sims we include a more realistic model of the real
world. While the *intention* of one side might be to call in low-level
cloud before an encounter, the real world doesn't always obey the
instructions it's given. In the 4 relevant sims every action called
for by either side could fail 1/2 the time. We might return later to
also include perceptual errors as well -- "what happened" might not be
open and shut. The jet side may or may not be able to determine
accurately "did the ufos turn up"? Noise can occur on the input/perception
side to each agent just as it can occur on the output/action side. None of
the 8 models here allowed for that.

So it's a lot of material to cover and I for one am champing at the
bit to know what happened in all that crunching.

Here is ep 1. The simplest sim with some noise on the agent output
side, no stealth mode, no dazzle mode, 0 sum rewards/punishments (i.e.
"maximum competition:).


Noise flags Percent of times
ground had clear sight of ufos+jets
N2 N3 0.0907029
N2 0.0907029
N1 N2 N3 2.35828
N1 N3 2.9932
N1 3.67347
N1 N2 12.585
N2 N3 N4 14.898
N1 N4 21.9274
N2 N4 25.7143
N4 31.7687
N1 N2 N4 32.9252
N1 N3 N4 34.898
N1 N2 N3 N4 38.1406
N3 N4 40.5442
(other combinations of noise flags returned 0 cases of jet+ufo observed
from the ground).

The various noise flags are as follows:
N1 whether a jet/ufo travelling *in* a layer with cloud
might be observed "below" is 50/50.
N2 whether a jet/ufo travelling *in* a cloud layer might
be observed from "above" is 50/50.
N3 whether a cloud layer called up by the ufos
comes in at the right time is 50/50
N4 whether a cloud layer called up by the jets
comes in at the right time is 50/50

The rewards are opposite in each round for the 2 sides. I.e. the
program calculates the reward for e.g. the jet side and the reward for
the ufo side is just the -ve of that.

The ufo side is rewarded:
1 point if it can see the ground
1 point is subtracted if the sats can see it
1 point is added if it can see the jets

The amazing thing was the first run of the sim (with no noise active)
found there was NO case where both jets and ufos were visible at the
same time from the ground. They learned from their series of
encounters to anticipate that and avoided it. For whatever reason
dictated by the reward function. :) So we had to wait for the noisy sims.

The next amazing thing was almost the first noisy sim -- where
"looking down" is randomly true when flying in the same level as cloud
-- scored 3.7% of the time jet+ufo both visible from the
ground. EXACTLY the value we found in the NUFORC database.

All other combinations of noise produced vastly different results from
expected -- from 0 chases visible up to a situation where 40% of rounds
produced a visible jet+ufo interaction.

We are forced to conclude (so far) that this simple sim captures
something about what is going on with jets and ufos. The "maximum
competition" aspect may be worrying. But the possibility both sides
actually can control some aspect of weather is supported with the jet
side apparently not wanting to produce too many ground-side witnesses.
If anyone witnesses a chase that stands as a "+1" for the ufo side --
i.e. it's "what they want" and (under 0-sum assumption) what the jet side
doesn't.

In later posts we will look at the other multitude of sim results.
But fair warning -- very few of them produce a value as close as this
first simple sim does to the observed "percent of times a chase is
seen". But OTOH we will also look at some cases where it comes
"close" and also gives a hint at weather ufos are actually trying to
attract the attention of ground observers with "dazzle mode"
anti-stealth tech. Why are they so often so bright if they are trying
not to attract attention? Answer -- those ones are not.

--
"Nothing in life is to be feared, it is only to be understood.
Now is the time to understand more, so that we may fear less."
- Marie Curie

There is something there -- measurable light, multiple instruments -- and
yet it seems to move in directions inconsistent with what we know of physics
or science more broadly. And that, to me, poses questions of tremendous
interest, as well as potential national security significance.
-- Rep. Adam Schiff, D-Calif., 2022 House Intelligence Committee hearing
on UAPs.

[Secret UFO recovery program blown open:]
I hope this revelation serves as an ontological shock sociologically
and provides a generally uniting issue for nations of the world to
re-assess their priorities.
-- David Grusch, 05 Jun 2023
[Talking to Les Kean et al for The Debrief, Grusch called for an end to
nearly a century of global UFO secrecy and warned that humanity needed to
prepare itself for "an unexpected, non-human intelligence contact scenario"].

We are not afraid to entrust the American people with unpleasant facts,
foreign ideas, alien philosophies, and competitive values. For a nation that
is afraid to let its people judge the truth and falsehood in an open market
is a nation that is afraid of its people.
-- JFK

'UFO Revolution' docuseries shows UAP flying over military base, 'blows up
decades of conspiracies': expert
YAHOO!News, 13 Jan 2024 09:36Z
A UFO docuseries "blows up decades of conspiracies," said Jeremy Corbell, an
investigative reporter featured in the series, ...

Floating 'Jellyfish' UFO haunted US military base in Iraq for years, says
former US Marine intelligence analyst shown infrared video by colleagues
Daily Mail, 13 Jan 2024 05:05Z
Marine broke his silence on a 2017 UFO that became 'the ghost story' of his
base. The ex-intel analyst stated that 'the ...

Video shows jellyfish-looking UFO in the sky over Iraq with troops told to
"hunt it down"
Irish Star on MSN, 09 Jan 2024 18:10Z
This viral video, supposedly recorded by the US military in 2018, shows a
jellyfish-looking UFO flying through the sky in ...

Inside secret UFO briefing in DC: Congress finds 'many' claims about US
govt harboring aliens and spaceships 'have merit' after grilling top
US ...
Daily Mail, 13 Jan 2024 08:04Z
The US Intelligence Community's Inspector General, America's top spy watchdog
, gave Congress a top secret UFO briefing Fri ...

Classified UFO briefing: House members emerge with mixed feelings
The Hill, 13 Jan 2024 08:03Z
House lawmakers left a Fri classified briefing on UFOs, referred to as
unidentified anomalous phenomena (UAP) by the ...

Classified UFO briefing provided `validity to what's been said': Burchett
News Nation on MSN, 13 Jan 2024 04:00Z
Tenn., said Fri's classified UFO briefing "provided some validity to
what's been said" in priority testimony about UAPs.

[NOP!]
U.F.O.s Remain a Mystery to Lawmakers After Classified Briefing
NYT, 13 Jan 2024 01:26Z

Coulthart: Classified UAP briefing yields `significant' results
NewsNation, 13 Jan 2024 01:02Z
Thomas A. Monheim, the inspector general of the intelligence community,
briefed the House Oversight Committee after a surge ...

Why is America's Congress being briefed on UFOs?
The Economist, 11 Jan 2024 16:25Z
AMERICA'S SPIES brief the country's lawmakers on sober topics. In 2023 they
gave classified briefings on artificial intelligence, Israel, Russian and
Chinese misinformation, Sudan and Ukraine--in other ...

Secret UFO meeting prompts speculation over what govt is hiding
Metro, 05 Jan 2024 22:18Z
Retired Navy pilot Ryan Graves said that UFOs are an `open secret' among
fighter pilots, and shared a third-party account of ...