There are 400 billion star systems in elite, and I have to wonder... Has anyone come across a system identical* to Sol?
* Identical either to 2019 Sol or 3303 Sol
* Identical either to 2019 Sol or 3303 Sol
Sol system
- Thread starter Jefferson613
- Start date
Well, of course you won't find an exactly identical system (or at least the chances are incredibly tiny) because the Sol system is entirely hand-designed.
Although, considering how many systems some commanders have visited, I'd be surprised if someone hasn't found a system that's at least similar. I personally haven't, but I'm sure someone will reply to this thread having found one
Although, considering how many systems some commanders have visited, I'd be surprised if someone hasn't found a system that's at least similar. I personally haven't, but I'm sure someone will reply to this thread having found one
Well, of course you won't find an exactly identical system (or at least the chances are incredibly tiny) because the Sol system is entirely hand-designed.
Although, considering how many systems some commanders have visited, I'd be surprised if someone hasn't found a system that's at least similar. I personally haven't, but I'm sure someone will reply to this thread having found one
It is one of my few goals.... Just cause it'd be so awesome/ random
Best of luck to you! Be sure to update us if you find anything noteworthy!
Some qualifying questions:
"Identical" is a scientifically imprecise term. How much tolerance or variation is acceptable for it to qualify as an "identical match"? For example, if the third planet is an Earth-like, how close to Earth do the planet's parameters have to be? Do they need to be a three-decimal-points exact match, is within 10% good enough, or is any old planet good enough, so long as it's rated "Earth-like"? Do you need an asteroid belt after Planet 4? What about matching the star type, planet orbital distances, eccentricities and all the other parameters?
How much of Sol system is needed to be copied? The fewer planets you're looking for, the fewer "dice rolls" you need to get to generate planets that are exactly the same as the ones in Sol. If it's just the eight "real planets", then it's much easier than if you include Pluto, Eris, Persephone and friends that are in the ED Sol system.
To give you an idea of the difficulty: I've searched the old EDSM data dump of some 27,500 Earth-like planets; not one of them is an "exact match" for Earth in terms of surface gravity and atmospheric density. Even if you allow a 10% variation (gravity and atmosphere both between 0.9 and 1.1 of Earth-standard), there's only a couple dozen or so matches. That's because the Stellar Forge algorithms tend to make either ELWs of the correct gravity with atmospheres that are much thicker, or the correct atmospheric pressure but are slightly too light; see my graph in this post. And that's just two parameters; by my reckoning, you'd need eight parameters to match (mass, gravity, temperature, atmosphere, volcanism, rotation period, axial tilt, presence of no rings and a single moon) to get a "true match for Earth". So far, the best I've found is 5 out of 8 parameters matching within 10% or a similarly reasonable variation. And that's just one of the planets of Sol system.
TLDR: getting an "exact match for Sol" involves odds that are on-par with rolling two dice and scoring "12" thousands of times in a row. 400 billion is a lot of dice-roll attempts, but so far, I don't think our exploration has found one.
"Identical" is a scientifically imprecise term. How much tolerance or variation is acceptable for it to qualify as an "identical match"? For example, if the third planet is an Earth-like, how close to Earth do the planet's parameters have to be? Do they need to be a three-decimal-points exact match, is within 10% good enough, or is any old planet good enough, so long as it's rated "Earth-like"? Do you need an asteroid belt after Planet 4? What about matching the star type, planet orbital distances, eccentricities and all the other parameters?
How much of Sol system is needed to be copied? The fewer planets you're looking for, the fewer "dice rolls" you need to get to generate planets that are exactly the same as the ones in Sol. If it's just the eight "real planets", then it's much easier than if you include Pluto, Eris, Persephone and friends that are in the ED Sol system.
To give you an idea of the difficulty: I've searched the old EDSM data dump of some 27,500 Earth-like planets; not one of them is an "exact match" for Earth in terms of surface gravity and atmospheric density. Even if you allow a 10% variation (gravity and atmosphere both between 0.9 and 1.1 of Earth-standard), there's only a couple dozen or so matches. That's because the Stellar Forge algorithms tend to make either ELWs of the correct gravity with atmospheres that are much thicker, or the correct atmospheric pressure but are slightly too light; see my graph in this post. And that's just two parameters; by my reckoning, you'd need eight parameters to match (mass, gravity, temperature, atmosphere, volcanism, rotation period, axial tilt, presence of no rings and a single moon) to get a "true match for Earth". So far, the best I've found is 5 out of 8 parameters matching within 10% or a similarly reasonable variation. And that's just one of the planets of Sol system.
TLDR: getting an "exact match for Sol" involves odds that are on-par with rolling two dice and scoring "12" thousands of times in a row. 400 billion is a lot of dice-roll attempts, but so far, I don't think our exploration has found one.
Regarding finding an "identical twin for Earth", here's a bit more focussed breakdown of that old EDSM data set, to show you how quickly the increasing number of parameters can slash the number of possible matches:
Total number of ELWs in the data set: 27566
Subset of the above that are in single-star systems, and have orbital position number 3: 1901
Subset of the above that also have "silicate magma" volcanism: 256
Subset of the above that also have gravity within 10% of Earth: 183 (there seems to be a hard floor of 0.93 Earth-gravities for Silicate Magma planets; all the planets eliminated in this round were heavier than 1.10G)
Subset of the above that also have atmospheric pressure within 10% of Earth: 0. (as noted in my previous post, getting gravity and atmosphere to both hit Earth-normal simultaneously on the same ELW is nigh-on impossible. So we'll skip this parameter.)
Subset of the previous group of 183 that also have a rotation period (day length) within 10% of Earth: 37.
Subset of the above that also have a surface temperature within 3 degrees of Earth-normal-average (288 K): 4.
These four worlds are in the following systems, with the following star types and links to their system's EDSM pages:
Nyeajoa LZ-F d11-31 : F8 VAB
Nyeajeau AA-Z d17 : F3 VI
Floawns XO-X d2-1295 : A6 VI
Blaa Eaec IN-A d14-146 : G5 VAB
Most of these are orbiting stars that are much hotter than Sol, so have much longer years. Only the last one in the list is orbiting a G-class star, cooler than Sol, with the planet's year at 644 days, it's not that much longer than Earth's (by comparison). Unfortunately, we can't see the entire system map in the EDSM links for any of these systems because their discoverers have not done a full system scan. And none of these four systems are in Marx's ELW database, either, so unless someone else visits them (or their original discoverer sees this thread and posts screenshots of the system map), we won't know exactly how Sol-like their systems are.
Total number of ELWs in the data set: 27566
Subset of the above that are in single-star systems, and have orbital position number 3: 1901
Subset of the above that also have "silicate magma" volcanism: 256
Subset of the above that also have gravity within 10% of Earth: 183 (there seems to be a hard floor of 0.93 Earth-gravities for Silicate Magma planets; all the planets eliminated in this round were heavier than 1.10G)
Subset of the above that also have atmospheric pressure within 10% of Earth: 0. (as noted in my previous post, getting gravity and atmosphere to both hit Earth-normal simultaneously on the same ELW is nigh-on impossible. So we'll skip this parameter.)
Subset of the previous group of 183 that also have a rotation period (day length) within 10% of Earth: 37.
Subset of the above that also have a surface temperature within 3 degrees of Earth-normal-average (288 K): 4.
These four worlds are in the following systems, with the following star types and links to their system's EDSM pages:
Nyeajoa LZ-F d11-31 : F8 VAB
Nyeajeau AA-Z d17 : F3 VI
Floawns XO-X d2-1295 : A6 VI
Blaa Eaec IN-A d14-146 : G5 VAB
Most of these are orbiting stars that are much hotter than Sol, so have much longer years. Only the last one in the list is orbiting a G-class star, cooler than Sol, with the planet's year at 644 days, it's not that much longer than Earth's (by comparison). Unfortunately, we can't see the entire system map in the EDSM links for any of these systems because their discoverers have not done a full system scan. And none of these four systems are in Marx's ELW database, either, so unless someone else visits them (or their original discoverer sees this thread and posts screenshots of the system map), we won't know exactly how Sol-like their systems are.
Last edited:
That last one makes a reasonable stab at being Sol-like albeit still a long way off.
I didn't scan them but those first two planets are going to be HMCs (so no Mercury equivalent) and both are rather massive (1.5 and 1.7 EMs). 'Mars' is also larger than the ELW (1.2 EMs). The gas giants are 81 and 28 EMs so no Jupiter equivalent although I guess the first one isn't a bad approximation to Saturn.
On the general issue of finding Sol-likes, it's going to be tough to find a Mars equivalent as it's actually quite rare with the Stellar forge to find a terraformable planet beyond an ELW. It does happen occasionally but it's not something I've seen often.
I didn't scan them but those first two planets are going to be HMCs (so no Mercury equivalent) and both are rather massive (1.5 and 1.7 EMs). 'Mars' is also larger than the ELW (1.2 EMs). The gas giants are 81 and 28 EMs so no Jupiter equivalent although I guess the first one isn't a bad approximation to Saturn.
On the general issue of finding Sol-likes, it's going to be tough to find a Mars equivalent as it's actually quite rare with the Stellar forge to find a terraformable planet beyond an ELW. It does happen occasionally but it's not something I've seen often.
Thanks for that, Allitnil. And yes, it's a better approximation of Sol than I was expecting. If we gave the ELW a moon, it would be much like how the ancients viewed Sol system (with the planets only going out to Saturn).
On top of all the improbabilities mentioned earlier, the Stellar Forge doesn't really seem to like creating "Sol-like" systems, with a few HMC planets, then an asteroid belt, then some gas giants. The old FE2/FFE Forge was better at doing this, which is why the hand-carved real-star systems that are legacies of that old game tend to be more "Sol-like" than the ED procedurally-generated systems; I think observations of real-universe exoplanets has made them tweak the algorithms away from Sol-like systems. Far more probable, in my observations of the Forge, are:
- A string of similar-sized HMCs and ice worlds, with no gas giants;
- A string of gas giants, with no HMCs;
- One or two super-hot metal-rich/HMCs surrounded by asteroids, and then a bunch of gas giants much further out.
Last edited:
You forgot about one essential requirement, a moon. There must be a 3a moon before it can be even remotely considered like Earth, and that moon must be around 1-2 seconds way.
Most Region XX-X XN systems seem unrealistic. The exoplanet database is so imcomplete and biased that anyone basing their algorithms even remotely on it will end up with an unrealistic system.
By the way did you consider water worlds? After all, they're just earth-likes with high sea levels.
But I wonder if someone has found or will find the Kerbin System
The Thargoids did, which is why they're coming our way, to put us between them and Kerbin, before Jeb finds a flower ship and tries to reverse engineer it.
I didn't forget; it's just that the game journals don't flag the absence or presence of a moon, so there's no way to tell if a moon is present from the EDSM data dump and we can't see the system maps to visually check.
And while I would regard it as essential for qualifying as "Earth's twin" ( I did mention "presence of no rings and a single moon" in my list of parameters in my first post), I wouldn't regard the size and distance as being important; so long as the moon wasn't stupidly large, or insignificantly small, when viewed from the surface of the planet, I would consider it a "match".
It is, however, a further barrier to creating a proc-genned Earth-twin. The Stellar Forge doesn't really like giving Earth-sized planets a moon at all, let alone a moon as large as Earth's Moon. An ELW with a co-orbiting planet is far more common, and so long as that co-orbiting planet is far enough away that it looks "about Moon-sized", I'd consider that as having met the "single moon" parameter. After all, Earth and Moon only barely miss out on qualifying as a "double planet"; our barycentre is only 1700 km below the Earth's surface. Make the Moon slightly bigger, or the Earth slightly smaller, and you've got a double-planet scenario by ED's definition.
True, but it's all we've got. The original FE2 Forge was written back in the early 1990s before any exoplanets had been discovered. With no exoplanet-systems to use as a model, all they had to use as a template was Sol system - so the star systems which that Forge generated did tend to all look much more like Sol system. That being said, in terms of real-world exoplanets, the number of alien star systems in which all the detected planets have similar masses is surprising; either there's a flaw with our mass-measurement theory, or there's a general universal bias towards making all the planets in a star system have similar mass. The Stellar Forge programmers have chosen that latter option.
The planet-wide-ocean current appearance of Water worlds is largely because of a disagreement between FD's art department and its science department. "Water Worlds" aren't supposed to be planets covered with ocean; they're supposed to be worlds with liquid water and life on their surface, but a non-human-breathable atmosphere - which is what we see in the statistics of every Water world. Just like an "Ammonia world" isn't covered in an ocean of ammonia, but has liquid ammonia (and life) on the surface in the form of seas and lakes. Water worlds in ED originally looked much the same as ELWs, only with less green on the continents; they were changed to 99.99% ocean planets sometime around version 1.4, I believe. Now that Explorers have an at-a-glance way of telling WW and ELW apart, via the FSS, I'd prefer ED now went back to the earlier appearance of Water worlds.
And I agree that a Water World in the Planet 3 spot should make for a good Sol-analogue too. After all, if current theories are true, then for most of Earth's history, Earth itself would not have qualified as an "Earth-like" by ED's definition: it would have been too hot, or too cold, or too much CO2 in the atmosphere, or too much oxygen. Earth would have been either a terraformable HMC or a terraformable Water world. However, no-one's done an EDSM data dump for Water worlds and few people keep records of them, so it's probably less likely anyone would be able to cite one that happened to be in a Sol-like system.
Last edited:
I could probably make a spreadsheet of water worlds, but there are 10x as many of them as ELWs, and the ELW spreadsheet is already pretty big. I hesitate to make spreadsheets that have over a million rows.
Maybe one that just includes WWs with G-class stars would be useful? Not sure how much that would shrink the list though.
Maybe one that just includes WWs with G-class stars would be useful? Not sure how much that would shrink the list though.
Last edited:
There is now, as of 3.3: the journal logs an event when a system is fully scanned. I don't know if EDSM stores this (yet?), but it would be possible from now on to tell from the journals if a body has any moons or not, but only if a full system scan is done.
In practice though, that still leaves potentially a hundred thousand ELWs with an uncertain "moon status". Not to mention the "star status" too - as I mentioned a long time ago, the number of stars present in an ELW's system might be of some importance too. Think of the day-night cycle.
On the bright side, stars are now automatically scanned and stored, so that part's solved.
You can download the dumps from here (although for the bodies, I'd recommend passing gzip compression to your downloader!) and filter out water worlds only.
Good questions!
I'm not as fussy to demand exact mass / rotation period etc etc.
Just 8 (10) planets, the first one hot HMC? Second one however they claimed Venus is( technically Venus can be terraformed and easier than mars), ELW/WW with or without a moon( because stellar forge hates moons
) .... The Mars planet should probably be rocky ice, although terraformable HMC would work too.Since Neptune is an ice giant, does that translate to elite as an icy body or a gas giant?
Unfortunately I don't think ED takes into account the ice giants. It probably treats them as gas giants, but the ice giants sit outside of the Sudarsky classification system. They have a lot of "supercritical fluid", and aren't really the same thing as a gas giant, but they're not terrestrial worlds either. The StellarForge seems to have a bit of a gap here.
They have Neptune as a class 1 gas giant, so I guess ice giants get lumped there lol
ED doesn't have a separate classification for "ice giant"; Up until the 1990s, Neptune and Uranus were traditionally classified as "small gas giants", which is how they are classified in-game.
However, it seems to me that there should be very little practical difference between a "gas giant" and a "large ice world with a super-thick atmosphere", yet the Stellar Forge considers them separate planet classes an generates them via different mechanisms. Some extreme "ice worlds" in ED have atmospheric pressures in the millions of atms; such a world "should" look just like a gas giant (by comparison, the "atmospheric pressure" at Jupiter's core is about 100 million atmospheres). At the top of Neptune's mantle - the closest thing it has to a "surface" - the pressure is about 100,000 atmospheres, whereas Neptune's core is at about 7 million atms.
So, in our hunt for a Sol-match, either a small gas giant or a large-ish ice world with super-thick atmosphere should suffice. However, if the physical appearance of the planets on the system map is important, then pseudo-Neptune at least should be a blue gas giant. It should perhaps be pointed out that you won't get an exact gas giant match for Uranus; the artichoke-coloured planet in ED is hand-edited to match the greenish colour the planet has IRL, but the Stellar Forge has never been observed to create a gas giant with this colour combination. You might have better luck getting a colour match with an ice world, though the Stellar Forge rarely swaps back and forth between ice worlds and gas giants like that.