All the strategies above assume a UFO captain starts from some subset   
   of planets and has a free choice to go to any other in the solar   
   system.  For completeness I also include the Sun (ouch ouch ouch!)  as   
   both a source and destination.  His destination choice is weighted by   
   how close each other planet is at the time.  Closer planets are   
   weighted proportionally more likely to be chosen for the next trip.   
   Depending on the distance the trip takes a proportionate time. If the   
   destination planet is Earth then we assume on approach to Earth the   
   UFO is seen, to be included int the sightings on that day. If the UFO   
   is *leaving* Earth for some other planet we assume it is sighted   
   leaving on the day the trip starts.   
      
   The strategies above then only differ by which planets are assumed to   
   be the possible starting points for trips. E.g "only6" means "only   
   Saturn". "dis1" is "everywhere except Mercury".  "only76" is only   
   Saturn and Uranus. &ct.   
      
   So we see the stand-out best match of our "Let's Play UFO Captain" is   
   "Only Saturn".  The set of correlations at the top of the post seem to   
   unambiguously point at Saturn -- of course more likely one of its more   
   interesting moons -- as the major origin of almost all UFO objects   
   sighted on Earth.   
      
   This is consistent with what we found using much simpler methods some   
   weeks back when we just looked at the correlations between sightings   
   and the distance of each planet. Saturn appeared at the top of that   
   (simple) list.  So we have "simulated UFO flights" and found the same   
   answer so now have slightly more confidence that finding is correct.   
      
   What seems to have been ruled out is something that had seemed at one   
   time more likely -- UFO's originating on most planets in the solar   
   system. It seems the correlations we obtain for other planets --   
   either their distance from Earth or their apparent position in the sky   
   (as seen from Earth) -- are "by products" of the movement of all the   
   planets being coordinated by the laws of gravitation. :)   
      
   Of course this latest work is predicated on the simple assumptions   
   we've made are "correct", or approximately correct.  If the true   
   situation is much more complicated then all bets are off and maybe   
   UFO's might originate from other planets than Saturn.   
      
   Finally, we can do the same tricks for different types of UFO's.  Some   
   shapes or colors cause the AI used here to give up and not decide from   
   the observations which flying strategy is more likely.  There is not   
   enough data (the high quality UFO sighting data I use from NUFORC   
   starts in early 2006) or the data we have is too noisy with mistakes   
   and hoaxes to choose between even the simple alternatives we have used above.   
      
   But a couple of key UFO types run through the simulate-and-match s/w fine.   
      
   E.g. the "light in the sky" UFO's are likely a grab-bag of all kinds   
   of things including mundane satellites, planets, and LED kites.  But   
   also a few odd objects.  But let's see whether the AI can drill down   
   into the data and find a planet where they might be coming from.   
      
   The list of correlations between planetary params and "lights" type   
   UFO sightings is:   
      
   Planet	Param		R2   
   saturn	rg		0.52515325   
   uranus	rg		0.48244359   
   mercury	Dec		0.45859890   
   neptune	rg		0.39373586   
   venus	Dec		0.37491284   
   jupiter	Dec		0.34234073   
   jupiter	RA		0.30687365   
   mars	rg		0.22915010   
   pluto	rg		0.18996626   
      
   Wow. Right at the top is says the distance to Saturn is the major   
   correlation. Surely this *must* mean the AI will pick "Saturn only" again...   
      
   Using the simulate-and-match s/w the top strategies that match the   
   above list of correlations is:   
      
   Strategy		R2 (likelihood of strategy producing the   
   			correlations we see above)   
   only80-t		0.68358740   
   only8-t			0.60636297   
   only83-t		0.52888418   
   only84-t		0.51587928   
   only6-t			0.33140172   
   only87-t		0.30025828   
   only86-t		0.28551806   
      
   Interesting! Now it says the pattern for "Lights" suggests they are   
   coming from Neptune and the Sun! In 2nd place it says "Neptune only"   
   is a likely explanation for the observations. In 3rd place it says   
   "Neptune and Earth" and 4th is Neptune and Mars. It's only down in 5th   
   place it "takes the bait" of the large corr between distance to Saturn   
   and daily UFO sightings and says it might be all coming from   
   Saturn. But the relative chance of "only saturn" versus "only neptune"   
   is around a 2-to-1 difference in favor of Neptune.   
      
   Finally, we can take the complement of the "lights" and look at "Non   
   Light" UFO's. At least these should reasonably be expected NOT to be   
   confused with a long list of star-like suspects. Of course it's also   
   expected daytime or dark UFO's can be confused with aircraft and   
   drones and other mundane things. But we might expect a lower noise factor.   
      
   The list of planetary correlations for Not-a-Light UFO sightings is:   
      
   Planet	Param		R2 (planet param predicting daily sightings of   
   			Non Lights)   
   mercury	Dec		0.56434431   
   neptune	rg		0.53736315   
   uranus	rg		0.50838866   
   venus	Dec		0.48216660   
   saturn	rg		0.45602463   
   venus	RA		0.39064275   
   jupiter	RA		0.26042923   
   mars	rg		0.25687112   
   pluto	rg		0.24839742   
      
   This time Neptune comes 2nd in the list. Will the AI pick Neptune as   
   the source for Non Lights. Or is it "more complicated than that"?   
      
   The list of strategies that match the above correlations is found to be:   
      
   Strategy		R2   
   only83-t		0.46722888   
   only86-t		0.35696505   
   only6-t			0.31287376   
   dis1-t			0.25339020   
      
   The most likely simple strategy that produces the observed pattern of   
   correlations between planetary params and UFO sightings is "only83" --   
   that Non Light UFO's are coming from Earth and Neptune.   
      
   But now we have a problem. The AI picks Saturn for "all UFOs" but has   
   picked Neptune for Lights and Not Lights -- i.e.  each 1/2 of "all   
   UFOs". It has not spotted (because it wasn't told about the   
   relationship between the UFO types) that these predictions are   
   somewhat logically inconsistent.   
      
   One resolution for us (apart from going back and making the s/w   
   understand that its solutions for "all UFOs" must jive with "Light"   
   and "not Light" being 2 non-overlapping subsets of the same thing) is   
   immediately apparent. The No 2 pick for "Not Lights" is "only86" --   
   Neptune & Saturn.   
      
   So the overall solution might be "most UFO's appear to come from   
   Saturn", but "Light" UFO's mostly come from Neptune (and maybe the Sun?).   
   "Non Light" UFO's (mostly) come from Saturn and the rest from Neptune.   
      
   At least that's seemingly the simplest explanation for all the   
   correlations we have, above.   
      
   --   
   Upcoming events:   
      
   "Online Haters" Are Psychopaths, Recent Study Finds.   
   The Debrief, 6 Apr 2021   
   Recently published research finds that "online haters" or people who   
   post hateful comments on online forums and social media demonstrate   
   high levels of personality disorder traits associated with psychopathy.   
      
   Starlink satellites are beaming super-fast internet to Australians below,   
   but not everyone's happy   
   ABC Science, 13 Apr 2021 19:56Z   
   Starlink promises to give people living in the bush internet that's just as   
   fast as in the city, but it's yet to prove itself.   
      
      
   [continued in next message]   
      
   --- SoupGate-Win32 v1.05   
    * Origin: you cannot sedate... all the things you hate (1:229/2)