Two finds-one is rare-and one may be rare....

After finishing DW2 I travel through the Sagittarius-Carina Arm region towards home. For most of the journey in the area didn't see much to write home about.
Excited to see Stellar phenomena make an appearance the moment I entered the region-around 1 in 3 systems had at least one NSP.....but after finding the umpteenth Peduncle Tree and Pod the joy of that faded pretty quick.:confused:

Then I found an ELW orbiting four stars in the system. For all my previous ELW finds, a three star orbit was my best so know this is a rarity :)

131255

Within 6 jumps I came across a system with a binary HMC and WW. Nothing special about them but the HMC has a moon orbiting it

131256

I'm sure this is the first time I've seen a binary pair with a moon orbiting one of them......but I may be having a senior moment :unsure: and need to know how rare this is!
 
Orvidius will tell you this for sure. He knows everything that can be deduced from EDSM database. ;)

However, I'm pretty sure I've never seen such a configuration. Nice find!
 
Yup, an ABCD ELW is rare. Congrats on your find!
As for the binary with a moon... I've seen them rarely, but what looks interesting here is that with 0.47 EM, the planet looks fairly light. Would be interesting to know the characteristics of its partner and its moon.
 
That will be slightly harder to datamine. Not impossible, but not a quick lookup either. ;) Anything involving binary worlds gets tricky. Though in this case, looking at them specifically, rather than filtering them out, might be easier. However detecting larger numbers of barycentric objects (trinaries, etc) might not be possible from the orbital data that's included. I'll put some thought into it.
 
Congrats on the find Commander.

Until now I was not aware of the rarity of such a configuration. I checked all of my ELW find's (About 400) and I have found only one with ABCD. It's actually a 1A (Earth-like moon).

That is what I like on ED. Exploration can be boring as hell, and then suddenly it hits you with an extraordinary system.


Fly/land safe.

CMDR Steyla
 
I don't think binaries with moons are that rare - Chrysus system in Colonia has a nice one where one of the binaries is a landable giant ice world with rings and two moons (naturally, it's where they put the surface base) - though as marx says, ones with the planet that light must be unusual.

That ELW must be exceptionally rare, though.
 
You can find 21 ELWs around ABCD in the ELW list. Also there is 1 BCDE, 1 CDEF and 2 DEFG ELWs. That gives you a chance of 0.16 % to find such a combination. But it means 1 of 1000 ELWs, not 1 in 1000 systems! If you say you have an average of 1 ELW / 100 systems, you have to search through 100 K systems to come across one of those. But it can always get worse: there are 2 ELWs around ABCDE...
 
Yup, an ABCD ELW is rare. Congrats on your find!
As for the binary with a moon... I've seen them rarely, but what looks interesting here is that with 0.47 EM, the planet looks fairly light. Would be interesting to know the characteristics of its partner and its moon.
Pics of Moon and WW as requested-

131301

131302
 
Great find :) Haven't found sth. like this myself.

You wrote:
I'm sure this is the first time I've seen a binary pair with a moon orbiting one of them......but I may be having a senior moment :unsure: and need to know how rare this is!
That was a nice task, because it was very specific: A High metal content world, paird with a water world and one of them has a moon.

Well, I took the latest EDSM data (the bodies.json file, updated 2019-05-03) and looked for (almost) exactly that.

I found ca. 137,500 of HMC worlds in a binary system with a water world.
In 542 of these systems, one of the constituents has a moon.

The "almost" is because:
1.: my algorithm counted likely also trinary / quaternary / etc. systems if there was a HMC and a water world. So this may increase the number of total binary systems. On the other hand, probably not many of such systems exist, so the results will likely not change significantly if these are taken out.
2.: for the second number my algorithm has the condition, that at least one of the planets must have at least one moon. I have seen some systems where one of the planets has several moons. Both planets having moons is possible, too. However, all of these are included in the 542 stated above.

So the task was not exactly fullfilled, but can you live with these numbers?

And then there is of course the possibility that I did a mistake, so it may be a good idea if CMDR Orvidius checks these numbers ;)
 
Last edited:
I haven't had a chance to work on it yet, so that's really cool to see some numbers here already. :)

I wouldn't be surprised if we get different results since there are a lot of ways to approach the data, and comparing orbits is always tricky since the EDSM data doesn't include the barycenter orbits, even though the game treats it as though the barycenters orbit the parent body, and the paired worlds orbit the barycenter. So we have huge holes in the data for this sort of thing.
 
I may have some bugs, and haven't had time to verify any of these, so take this with a grain of salt. My script found:

3213109 total binaries
1159694 total binaries with moons

80374 HMC/WW binaries
6475 HMC/WW binaries with moons

I don't have the spreadsheet in the index since it's unverified, but here's a link for anyone who wants to look through it. The spreadsheet only has those HMC/WW binaries with moons.

 
May I ask how do you identify binary systems? Curious since I struggle with that myself.
I'm taking two approaches. I'm looking for names with adjacent numbers first (such as "Foobar AA-A h1 A 1" and "Foobar AA-A h1 A 2", then comparing either:

1. The "parents" field, if it has data. Binaries should both start with a "Null:X" where "X" is the same number, referring to the same barycenter. Accurate when the data exists, but not everything has this data.

2. If no "parents" data, then I'm looking to see if they have the same orbital period. This will have false positives from trojan planets. And maybe miss some due to floating point rounding issues. It's hard to say how precise this one will be, as a result.

EDIT: And this of course misses trinaries and more complex systems, because they all have different barycenters and periods. It would only catch the innermost binary from one of those sets.
 
Last edited:
Thnak you for the answer.

How many trojans are there? Can you give me a link to an example?

It's hard to test since it doesn't show at EDSM or in the data, but for the very few confirmed binaries I had (due to screenshots here on the forum and own data) it seemed that binary candidates have the same value for:
  • orbital inclination
  • orbital period
  • orbital eccentricity
Is this wrong?
I'm asking, because in my algorithm all three need to be the same before a binary is accepted.

If the above holds, these values are usually exactly the same. The only error I have observed seemed like floating point errors but these were rare and tiny; so the "delta" to be allowed is rather small.
However, I did find pairs of planets that could be binaries which had a really small difference in orbital period but for the two other parameters were totally off. So just checking for orbital period seemed to be not good enough.

I've never checked for ternary / quaternary etc. systems since I could not find such confirmed configurations when I wrote my program.

I'm asking because I found significantly more HMC/WW binaries even though my attempt seems to be more strict.

I also wonder how you discovered all these moons. I suspect that my algorithm is not working properly.
What i did was I first identified the HMC/WW binary systems and then I checked if there are bodies in the whole system that have a name that would correspond to a moon to one of these two. I tested it for some of the results and it seemd to be OK, but of course can i not check what the algorithm didn't discover.

That is fun … I truly enjoy talking about that, even though I know that it is just a game :)
 
Thnak you for the answer.

How many trojans are there? Can you give me a link to an example?

It's hard to test since it doesn't show at EDSM or in the data, but for the very few confirmed binaries I had (due to screenshots here on the forum and own data) it seemed that binary candidates have the same value for:
  • orbital inclination
  • orbital period
  • orbital eccentricity
Is this wrong?
I'm asking, because in my algorithm all three need to be the same before a binary is accepted.

If the above holds, these values are usually exactly the same. The only error I have observed seemed like floating point errors but these were rare and tiny; so the "delta" to be allowed is rather small.
However, I did find pairs of planets that could be binaries which had a really small difference in orbital period but for the two other parameters were totally off. So just checking for orbital period seemed to be not good enough.
You're totally right. All three of those should match. I didn't check them because I hadn't taken the time to verify that they would actually match in the data, rather than having 180-degree reversed inclinations, for instance. It's probably safe to test period and eccentricity together in any case.

I think the trojans are pretty rare. I haven't kept track of any, but there was a thread about them:



I've never checked for ternary / quaternary etc. systems since I could not find such confirmed configurations when I wrote my program.

I'm asking because I found significantly more HMC/WW binaries even though my attempt seems to be more strict.

I also wonder how you discovered all these moons. I suspect that my algorithm is not working properly.
What i did was I first identified the HMC/WW binary systems and then I checked if there are bodies in the whole system that have a name that would correspond to a moon to one of these two. I tested it for some of the results and it seemd to be OK, but of course can i not check what the algorithm didn't discover.

That is fun … I truly enjoy talking about that, even though I know that it is just a game :)
Yeah, I'm not sure why the hugely different results either, but it's possible I have some glaring bugs that I didn't think of. For the moons, I also started by finding all of the binary pairs, and then checking for the presence of moons. For this, I just take the name of both planets in the binary, and then look to see of there's something with an appended lower-case letter. That is, if the planets are named "Foobar 3" and "Foobar 4", then I do a regexp search in the database for names like "Foobar 3 [a-z]" and "Foobar 4 [a-z]" with the same EDSM star system ID number. (EDIT: This can also be done with the "parents" data if it's there, looking at their ID numbers, but the names are usually fine for proc-gen systems)

But yeah, this is a lot of fun to do sometimes. Playing with the data is a whole additional way to play the game. :D
 
Last edited:
180 degrees reversed inclinations???? o_O
i haven't thought about that! :( … but then I would think one would rotate clockwise and the other counterclockwise. Would that be stable?

Regarding the number of HMC/WW binaries with moons … would you mind putting the names of the binaries with moons into one of your famous CSV-files?
Seeing the moons my algorithm didn't find may help me to figure out what's wrong with my program :) … I just don't know when I'll do that.
And … mhmmm … if it isn't too much, I would also like to see your list of HMC/WW binaries … Then I can maybe figure out systems that I categorize as binary but which actually aren't.

And I have the impression we have derailed this thread a bit ;)
 
Yep, the spreadsheet was linked above. :)

https://edastro.com/mapcharts/files/binary-worlds-with-moons.csv

The bodies would have to orbit each other the same direction, so if the inclination was reversed, other numbers would have to be shifted around to make that work. Kinda like planetary rotations where the rotation is retrograde, but the axial tilt is more than 90 degrees (or less than -90), so technically it still rotates in a prograde direction. Or the cases where the axial tilt is more than 360 degrees, or whatever. I don't think the orbits do that sort of nonsense in the data, but I try not to assume anything. :D

Heh, as far as derailing goes, at least we're still talking about HMC/WW binaries with moons. ;)
 
Thank you very much for providing your data.

I was going line by line through my code and I indeed found a mistake. It didn't count different HMC/WW binary systems (with moons) if they were in a system which already had another HMC/WW binary with moon. So my number increased from the above stated to 667 such systems.

But I could not close the gap to your more than 6,000.
I've checked my code twice more, ran more tests … nothing. Everything seemed to be correct.
So I took all your systems that were not in my collection and looked some of them up.

For example in the Eok Pruae EI-H d11-6156-system planets CDE 2 and CDE 3, and planets CDE 4 and CDE 5 are listed in your list to be HMC / WW binaries with at least one moon.
Interestingly enough, the "parents"-value is in all four cases: [{"Null":4},{"Null":0}]

However, the differences in their orbital period, orbital inclination and orbital eccentricity are all far larger than I would expect for floating point errors.
I don't test for your parents-condition, which evaluates as True in these cases, when I determine if two binary-system-candidates actually are in a binary system. Hence I wonder again if my assumption that these values need to be equal is true. I would abandon that at once if necessary. Quite some numbers in my galactic statics thread hinge on that and I want it to be right.

Also, even taking all of them into account, I'm still some hundred short of your number … but then, this could be because your database is more up to date than mine and I can live with it. I never expected to have an exact match, but I would like it to be in the same range at least ;)

I've manually checked three more which have the same issues.
Right now I'm running a program which checks all of the binary systems which I haven't found, but that may take a while. So if this comes back with everything OK for most of them I'm back to square one :(

P.S. If you want my data / code, I will make it available. For the time being the former are binary files and the latter is messy.
 
Unfortunately it's super hard to verify these if you haven't visited the system yourself. I'm really hoping that the "parents" data isn't completely broken. This is the first time I've tried relying on it. So I may just have to see if any of them are somewhere near me and fly out to them. If it proves unreliable, then I'll have to strip that back out of the code. Realistically, you can't have a binary pair whose period around the barycenter is different. It's possible I may need to check for that anyway, even if the parents field says they should be a binary. I suppose it's entirely possible that I've misinterpreted how it's structured. The "nulls" definitely indicate that it's not a direct child object of a scanned planet or star, but the null numbering might not be reliable, whereas the planet/star IDs should be. So now I'm wondering if unscanned bodies are creating false positives here too. Ugh.
 
Getting binary systems right is hard for me, too … Hence it's nice sharing thougts :)
One of the systems was (somewhat) on my way to my next target: Thueche MC-P c8-1. In the way of science I made the detour. Here BC 5 and BC 6 are listed to be binaries, but they are not:
131961


I'm too tired for today and did not check their parent-values. If I dont forget I'll do that tomorrow.
 
Top Bottom