Paradise Planets: A New Guide

[Rexxer]

There are many theories floating around online, guides about how to find Paradise planets. To test these theories, I went through all 256 galaxies searching for paradise planets, planting a base on a paradise planet in each galaxy, on or near a center jump-off. Along the way I collected the star system stats on 509 paradise planets, so in general, I encountered about two paradise planets per galaxy. But Paradise planets are not as rare as that may sound, as this was a speed run as well. The total game time was 252.04 hours, or less than an hour per galaxy. Two planets per galaxy was all I was prepared to find to keep the game time streamlined.

What is a Paradise planet?

Surprisingly, some online guides get this wrong. What causes some confusion is related to the mass of conflicting game development overlays which, while adding new playable function, also cause some naming and functional confusion. Even difficulty level will change a planet's paradise status, and a paradise planet on a normal difficulty may suddenly become non-paradise when the difficulty level is raised. In other words, "Paradise" status is fixed to weather and sentinel status, and if anything changes on the planet that is in conflict with that status, the planet then is no longer considered "Paradise" by the game, (although not necessarilly by Discovery...more on that later).

For a planet to report itself as a Paradise planet when you land, it will always have (1) a "Lush Biome" (star bramble, Pf), (2) have clear weather (ie. no storms), (3) no sentinels. No sentinels means they don't spawn next to you on the landscape but can still exist at secure buildings. Weather also confuses many players, rightfully so, and deserves a bigger explanation:

>>> WEATHER <<<
Weather names, due to various game upgrades over time, have two versions: the original naming scheme and a newer exotic naming scheme.

The original naming scheme was and still is definitive: when you see it, even from space, you know exactly whether the weather is clear or not. Traditional clear weather names (for lush planets) are "Blissful, Beautiful, Mellow, Balmy, Pleasant, Humid, Refreshing Breeze, Mild Rain, Light Showers, and Temperate". If you saw any of those weather names, you had it made: those planets presenting such weather names would be clear weather 100% of the time. Zero storms. And one of those names would be a requirement for any Paradise planet using the classic weather-naming scheme.

The newer exotic naming scheme has bizarre and extreme sounding names, but these names are not fixed to whether it is clear or storming. It could be either one, - the planet's weather status is still spawns with a constant storm intensity from max down to zero internally but that status is hidden, and it's not hinted at by its displayed name. These new names will clue you into other cosmetic features in the atmosphere and other appearance variations, but these names say absolutely nothing regarding weather features that actually affect Paradise status, even though these names are lurid, seemingly specific, and displayed in the weather readout. Stated another way, the exotic weather names are just random names that contain no "tell" or hint regarding actual intensity of storms or complete lack of them in the weather. For example, "Bilious Storms" could in fact be clear beautiful weather. Same for "Poison Cyclones, Azure Storms, or Blood Rain". There are many more, but scary names like that don't rule out clear weather or Paradise planets. As you may imagine, some see a "Poison Cyclone" paradise planet, and automatically think it has storms. It does not.

However, if the weather is listed in red, then it is not a paradise planet. Red type always indicates extreme (read: stormy) weather.

>>> OTHER WEATHER CONFUSION <<<
The game's Discovery tab occasionally reports planet data inaccurately. It has the greatest number of inaccuracies toward the blue end of spectral classes, and few in the red, and essentially none for yellow stars. I have posted previously on this with lots of examples. Basically, if you use a freighter scanner, this scanner uses the Discovery algorithm, and a Paradise planet may be reported when in fact it is not, so if you land, it will be storming but it also will say something like "Rainy" biome, not Paradise. Conversely, Discovery (or the freighter scanner) may indicate a planet is NOT a Paradise planet, when in fact it is. And when you land on the planet, it will accurately report its Paradise status. Making matters worse, the exosuit scanner itself uses Discovery, and therefore may be incorrect. This problem is much more an issue since the Endurance update (July 2022) since the (then) new Freighter scanner just borrows the buggy Discovery algorithm, as does one's exosuit scanner. Prior to the Endurance update, many players never used a freighter since freighters were more trouble than they were worth, and many people traveled the galaxies in their personal ships and flew their space ships up to a planet to ID it, ...and scanning from space has always been 100% accurate.

Now, after Endurance, people aren't flying up to planets to ID them, just using the freighter scanner. Consequently, more and more players will run into this inaccuracy.

Are There Shortcuts To Finding Paradise Planets?

The short gimmicky answer is: yes, go to a lush galaxy system. It's true Lush galaxies spawn them most frequently, normal and dead galaxies are somewhere in between, and harsh galaxies have the fewest (but they do have them).

That being said, what about within a given galaxy, regardless the galactic spawn rate?

The answer everyone wants to know is whether you can scan for paradise planets by star system information. Spectral class and relative temperature. G5, F6, etc. Traditional guides say yes, but I have pretty decisively found out this is mostly false. There are several reasons why it appears to be true, but I am guessing part of the reason is confirmation bias. Some dudes likely hoped there was a pattern based on some simple reasoning, then jumped in a lush galaxy to prove their point, cherry picking their chosen planets. The problem is, in a lush galaxy, it is difficult NOT to find a paradise planet, so if you go in there looking at any spectral class or relative temperature, you will find lots of systems with paradise planets to match.

That being said, narrowing spectral class to yellow stars is still a good way to expedite searches because:

1. blue, green, and red stars are not as common as yellow stars
   
2. blue, green, and red stars can't be scanned as accurately due to the issues listed previously

The reality is, all classes of stars have plenty of paradise planets, but the method to easily find them is buggier in the non-yellow classes of stars. Consequently, since I was on a speed run, I also biased my data at first, mostly avoiding non-yellow stars due to the buggy guides I was following. And then later, I used the buggy freighter scanner to do the run faster, but of course, that method also misses many Paradise planets in non-yellow stars. So bear that in mind when you see the data:
https://steamcommunity.com/sharedfiles/filedetails/?id=2900810937
The analysis of relative temperature, however, shows the bias toward "3 to 7" is wholly inaccurate, almost the reverse of reality. The correlation to finding paradise planets is essentially zero when looking at the relative temperature:
https://steamcommunity.com/sharedfiles/filedetails/?id=2900811186
For the sake of discussion, let's entertain the idea that there is a little advantage in bracketing your search to "3-7" or some other value range. Let's say there is a 10% higher chance of finding paradise planets in that bracket, - so should you bias your search? Again, the answer would be "absolutely not". But the reason is because statistically paradise planets in most galaxies are kind of rare, and therefore if you miss a paradise planet because you are limiting your search, your search is now longer since the cost of missing a paradise planet is a boat load of time wasted. At some point of concentration, sure, it might be worth bracketing your search, but reality is no where close to that point yet. Think of it this way: if you are betting on whether a coin is heads or tails, a 50% odds sounds pretty good until it isn't. And then you would have to double your bet for every loss in a losing streak...until you run out of money. In the case of paradise planets, time is the thing you will waste massively if you bracket your bet.

[Jaggid Edje]

Really good stuff.
Thank you for sharing it with the community at large.

[jacobellinger]

Originally posted by Rexxer:

There are many theories floating around online, guides about how to find Paradise planets. To test these theories, I went through all 256 galaxies searching for paradise planets, planting a base on a paradise planet in each galaxy, on or near a center jump-off. Along the way I collected the star system stats on 509 paradise planets, so in general, I encountered about two paradise planets per galaxy. But Paradise planets are not as rare as that may sound, as this was a speed run as well. The total game time was 252.04 hours, or less than an hour per galaxy. Two planets per galaxy was all I was prepared to find to keep the game time streamlined.

What is a Paradise planet?

Surprisingly, some online guides get this wrong. What causes some confusion is related to the mass of conflicting game development overlays which, while adding new playable function, also cause some naming and functional confusion.

For a planet to report itself as a Paradise planet when you land, it will always be (1) of the "Lush Biome" (star bramble, Pf), have (2) clear weather, (3) no sentinels. No sentinels means they don't spawn next to you but can still exist at secure buildings. Weather also confuses many players, rightfully so, and deserves a bigger explanation:

>>> WEATHER <<<
Weather names, due to game upgrades, have two versions: the original naming scheme, and the newer exotic naming scheme:

The original naming scheme was and still is definitive: when you see it, even from space, you know exactly whether the weather is clear or not. Traditional clear weather names (for lush planets) are "Blissful, Beautiful, Mellow, Balmy, Pleasant, Humid, Refreshing Breeze, Mild Rain, Light Showers, and Temperate". If you saw any of those, you had it made. The planet presenting that would be clear weather 100% of the time.

The newer exotic naming scheme has rather bizzare extreme sounding names, but are not fixed to whether it is clear or storming. It could be either one, - the weather status is still static internally but its static status is hidden and not reported by its displayed name. These new names have other planetary associations, but none that affect Paradise status or the weather, even though it is displayed in the weather readout. The exotic weather names are just random names that mean nothing regarding weather. For example, "Bilious Storms" could in fact be clear beautiful weather. Same for "Poison Cyclones, Azure Storms, or Blood Rain". There are many more, but they don't rule out clear weather or Paradise planets. As you may imagine, some see a "Poison Cyclone" paradise planet, and automatically think it has storms. It does not.

>>> OTHER CONFUSING OVERLAYS <<<
The game's Dicovery tab occasionally reports planet data inaccurately. It has the greatest number of inaccuracies toward the blue end of spectral classes, and few in the red, and essentially none for yellow stars. I have posted previously on this with lots of examples. Basically, if you use a freighter scanner, or Discovery, a Paradise planet may be reported when in fact it is not, so if you land, it will be storming but it also will say something like "Rainy" biome, not Pardise. Conversely, Discovery (or the freighter scanner) may indicate a planet is NOT a Paradise planet, when in fact it is. And when you land on the planet, it will accurately report its Paradise status. This problem is much more an issue since the last update since the Freighter scanner just uses the buggy Discovery filter. Previously, people just flew their space ships to a planet to ID it, which was 100% accurate.

Are There Shorcuts To Finding Paradise Planets?

The question everyone asks is whether you can scan for paradise planets by star system information. Spectral class and relative temperature. G5, F6, etc. Traditional guides say yes, but I have pretty decisively found out this is mostly false. There are several reasons why it appears to be true, but I am guessing part of the reason is confirmation bias. Some dudes likely hoped there was a pattern based on some simple reasoning, then jumped in a lush galaxy to prove their point, cherry picking their chosen planets. The problem is, in a lush galaxy, it is difficult NOT to find a paradise planet, so if you go in there looking at any spectral class or relative temperature, you will find lots of systems with paradise planets to match.

That being said, narrowing spectral class to yellow stars is still a good way to expedite searches because:

1. blue, green, and red stars are not as common as yellow stars
   
2. blue, green, and red stars can't be scanned as accurately due to the issues listed previously

The reality is, all classes of stars have plenty of paradise planets, but the method to easily find them is buggier in the non-yellow classes of stars. Consequently, since I was on a speed run, I also biased my data at first, mostly avoiding non-yellow stars due to the buggy guides I was following. And then later, I used the buggy freighter scanner to do the run faster, but of course, that method also misses many Paradise planets in non-yellow stars. So bear that in mind when you see the data:
https://steamcommunity.com/sharedfiles/filedetails/?id=2900810937
The analysis of relative temperature, however, shows the bias toward "3 to 7" is wholly inaccurate, almost the reverse of reality. The correlation to finding paradise planets is essentially zero when looking at the relative temperature:
https://steamcommunity.com/sharedfiles/filedetails/?id=2900811186
For the sake of discussion, let's entertain the idea that there is a little advantage in bracketing your search to "3-7" or some other value range. Let's say there is a 10% higher chance of finding paradise planets in that bracket, - so should you bias your search? Again, the answer would be "absolutely not". But the reason is because statistically paradise planets in most galaxies are kind of rare, and therefore if you miss a paradise planet because you are limiting your search, your search is now longer since the cost of missing a paradise planet is a boat load of time wasted. At some point of concentration, sure, it might be worth bracketing your search, but reality is no where close to that point yet. Think of it this way: if you are betting on whether a coin is heads or tails, a 50% odds sounds pretty good until it isn't. And then you would have to double your bet for every loss in a losing streak...until you run out of money. In the case of paradise planets, time is the thing you will waste massively if you bracket your bet.

I hope the devs fix the classification issues you mentioned with scanners and nameing.

[Guyver8]

For your 'shortcut' section.
yes...the easiest (read cheesey) way to find a paradise planet is pin starbulb in your catalogue this will either detect a planet in your system or direct you via the galaxy map to a system which has starbulb

NB. this is the simpliest method to find planets of particular types...if it has a known mineral/flora/substance that you can target

[Xautos]

i don't even need the data, with a little sense you can figure out that there is always paradise planets in any galaxy no matter how big or small the chance is. a chance is a chance.

if you'll encounter them by cheesing it like above or by random chance without cheesing it, you'll eventually run across such planets out in those billions of stars.

[Krash Megiddo]

Good information. I spent a month searching for Lush anomaly planets in Lush galaxies. Its easy for me to fall into the confirmation bias trap. I searched all star types and was mostly disappointed with yellow and red. I found really interesting planets in green and blue star systems. My conclusion: only green and blue star systems were worth searching

[Rexxer]

Also, just as a point of record, of whenever I encountered more than one paradise planet in a star system, I included a separate data point for each Paradise planet, as if it was in a separate star system. Of the 509 planets discovered, 20 were from systems with 2 paradise planets, and 3 were from one system. Here is the F8 star system which had 75% paradise planets...and note, it is not in the traditional search "bracket" of 3-7:
https://steamcommunity.com/sharedfiles/filedetails/?id=2900967861

[Rexxer]

Originally posted by Xautos:

i don't even need the data, with a little sense you can figure out that there is always paradise planets in any galaxy no matter how big or small the chance is. a chance is a chance.

if you'll encounter them by cheesing it like above or by random chance without cheesing it, you'll eventually run across such planets out in those billions of stars.

In many of the "harsh" galaxies, I occasionally spent less time looking for a paradise planet than in "normal" galaxies. It could have been luck, but this is also why I never bracket my search parameters when looking for paradise planets. Any star system can have them.

[Krash Megiddo]

@Rexxer, what did you think of #19 Isdoraijung? I spent many hours in many lush galaxies but I seemed to have the most luck in #19 for some reason. Its as if 40%-50% of the systems there have paradise worlds. Is that possible?

[Icewolf]

Thank you Rexxer, very interesting stuff.
I always thought G5 has the best chances, but apparently temperature doesn't matter and F class is even slightly better. That's good to know.

I use a very simple "shortcut" that you didn't mention: I only visit 5 or 6 planet systems. That saves a lot of time, because vistiting a 6 planet system is like visiting two 3 planet systems (or three 2 planet systems).

[Krash Megiddo]

Originally posted by Icewolf:

Thank you Rexxer, very interesting stuff.
I always thought G5 has the best chances, but apparently temperature doesn't matter and F class is even slightly better. That's good to know.

I use a very simple "shortcut" that you didn't mention: I only visit 5 or 6 star systems. That saves a lot of time, because vistiting a 6 star system is like visiting two 3 star systems (or three 2 star systems).

I don't know about star count, but you reminded me of a tactic that works well for me and that's only visiting systems with 4 or more planets = higher chance of paradise roll.

[Icewolf]

Sorry, i meant planets of course. I will edit my post.

[chipmonk]

I think there are two other variables that need to be mentioned. 1. The type of save your playing will help to determine the number of paradise planets you will potentially come across. When I do a cross save galaxy jump (Thanks Krash) I find that a planets designation in a system on my normal save is not the same on my survival save. A paradise planet on my normal save may become a stormy planet on my survival save. 2. One thing that determines if a paradise planet is really a paradise for me is the lack of predators which can only be determined by exploring the planet in question. When using the freighters scanner I always also look for the type of minerals a planet has because if they're activated then it's likely got storms as well, irregardless of what the scanner might say.

[Djsstorm]

I have noticed that "Humid, Light Showers, Mild Rain" to name a few, on any planet that is labeled as "Paradise planet" is false advertising. These planets will get Superheated Rainstorms. They are rare but they do happen. I hate when I see a paradise planet, but the weather has storms when you land, even if they are rare, that is not a paradise.

[Rexxer]

Originally posted by Krash Megiddo:

@Rexxer, what did you think of #19 Isdoraijung? I spent many hours in many lush galaxies but I seemed to have the most luck in #19 for some reason. Its as if 40%-50% of the systems there have paradise worlds. Is that possible?

I have another set of data I will put out eventually which deals with that. I also recorded how many systems I had to visit per discovery. But the short answer is yes, locally. The distribution of paradise planets is irregular, as I would always search a polar coordinate sphere around a starting point (usually an edge system) and record each system visited. Some search "spheres" had a ton of paradise planets, others were sparse, and this is a separate issue from just whether the galaxy is normal, empty or harsh. Lush galaxies though almost always have many many paradise planets.

In the beginning, the NMS cosmos simply had lush worlds which occurred at whatever frequency the Designer allowed by his initial procedure. Then came planet diversification, and out of all the preexisting lush worlds, swampy and paradise planets were seeded. If you had a base on a lush world, it may have become a swamp or paradise overnight after the upgrade. Other planets were converted to exotics, and so on.

Since the SYSTEM seed comes first, I cannot yet tell if the first tap for Paradise is made at the system seed level, or whether it is applied on a planet by planet basis further down the line. Or combination of both, a paradise planet procedure which simply looks at another frequency algorithm summed over another. But ultimately it is a randomization. At a point chosen by the randomization procedure, it likely drills down to select the next lush planet in the "chosen" star system and then converts it to a paradise planet by stripping out weather and sentinels. If it misses the coin toss several times in a row, as randomized trials sometimes do, then you will have a relative lack of paradise planets.

Worse, it appears there are other functionality overlays which can subsequently wipe out the newly minted paradise planet. This may behind the Discovery reporting errors also. For example, if the activated metal overlay is thrown onto a paradise planet, it neutralizes the Paradise conversion: it adds back extreme weather and randomizes sentinels again. Same for the "Infested" overlay which was added in a recent update when big worms were put back in the game. Now, if there exists a paradise planet, it has the potential to be decommissioned as "paradise" and become activated and/or infested.... which means it is then no longer a paradise planet.

[Rexxer]

Originally posted by Icewolf:

Thank you Rexxer, very interesting stuff.
I always thought G5 has the best chances, but apparently temperature doesn't matter and F class is even slightly better. That's good to know.

I use a very simple "shortcut" that you didn't mention: I only visit 5 or 6 planet systems. That saves a lot of time, because vistiting a 6 planet system is like visiting two 3 planet systems (or three 2 planet systems).

Actually that is not a good strategy. I am only conjecturing, but it's based on observation: many of the only paradise planets I found in a search area occurred in two planet systems. The randomization algorithm doesn't go just by the planets separately....it seems to consider star system frequency. If it parses by star system, so should you. Whenever you bracket your search, you inevitably make your search longer.

Remember the planets in a system are not individual things until they are DERIVED from a SYSTEM seed. Most operations you are assuming happen at the planetary level likely first happen on the star system level.