Making Cash in No Man's Sky

From w.mfi
Jump to: navigation, search


High-Frequency Trading in No Man's Sky

played NMS 8/25-9/20/16. this section created 9/29/16, last edited 11/8/16

There are tons of articles about making credits in No Man's Sky. I consider this YouTube from TerraMantis one of the best because it's more than just another qualitative list of things to do; he actually talks about rates at which you can make U, and gives solid tips along the way. We don't need yet more lists of all the ways to make money. We want to know the one best way to make money. Finding this requires a quantitative compare of all competing methods.

Like him, I found that mining and harvesting exotics are not as good as trading in stations. I went on to find a more robust way of making money than TerraMantis, as outlined below. But first a word on mining.

Mining

Here's a little table on metals you can mine. Mining's not the best way to make money. But I was still surprised that while most money-making sites talk about mining, almost none talk about metal prices. Few even say the best one (and some say a wrong one).

So here it is. The best metal to mine for profit is Emeril. It can be found in huge boulders on some planets:

Metal Units per

piece

Δ

U/p

U/p vs.

Iridium

U/p vs.

Emeril

Emeril 275.00 +186% 100.0%
Gold 220.00 55 +129% 80.0%
Aluminum 165.00 55 +71% 60.0%
Nickel 137.50 27.5 +43% 50.0%
Copper 110.00 27.5 +14% 40.0%
Iridium 96.25 13.75 0% 35.0%
U/p = Units per piece of this metal.
Δ = price difference relative to previous line (in original sort order).

When prices are compared to Emeril, the relative basis becomes obvious. Values are multiples of 13.75 (or 27.5 or 55.0, however you want to say it). Easy breakpoints relative to Emeril. It is 25% more profitable than the next-most expensive metal, Gold.

In some very brief testing of Emeril mining, I could make about 1.3 million U per hour, including trips to go sell when full. (Mining from metallic boulders, not from green crystals similar to Plutonium. Crystals would take much longer to find.) I'm not the swiftest guy; I'm sure some folks can do a bit better than this. Maybe even 2 mil per hour.

But it will never be in the same league as the 5 - 10 mil per hour you can make with trading (below). And if you can mine faster than me, you can probably trade faster than me, too.

High-Frequency Trading

The best way to make money is through station trading - sitting in a station and buying and selling items from one pilot to another. The stationmaster can help, but isn't critical.

This can be done on most any station. But ones with plenty of traffic and better selection will be best.

Stations and Pilots

I've watched three stations closely for up to an hour. Two had 13 recurring pilots, the third had 12. (Maybe there was a 13th who, by chance, didn't show up while I was watching. It's entirely possible because some pilots are extremely infrequent.)

  • The ships and pilots you see happen over and over. It's always the same ~13. Each ship and its pilot always has the same pilot name, ship CGI model, ship model name (Hiyabushi S80 or whatever), and, importantly, always the same items and quantities for trade, including what's starred (double priced). The same holds true for the stationmaster - he always has the same name, and the station terminal items always have the same type, quantity, and stars. (Although you'd have to reload or revisit the station to replenish its terminal.)
    • However, if you buy the same item from pilots several times in a row very quickly, it seems to decrease the quantities available, compared to normal. Sometimes it even seems that there's a rate to how frequently they replenish in the system overall, because elements (40+ per pilot) will be especially hard hit. So perhaps they "drain" more quickly as a result. Items, with only up to 4 per pilot, require a very intense run to become low; this is very rare (once an hour?).
    • The only thing I've noticed that constantly changes for a given pilot is the price and number of slots if you want to buy their ship.
  • Pilots and trade quantities continue to be the same even if you reload or visit a planet and come back to the station. Although I haven't tested it thoroughly, I'm pretty sure these same pilots and ships are also the ones seen at any trade stations or landing pads on planets in a given solar system. (However, apparently prices are never starred for any pilots or terminals anywhere in a system, except at its station.) The same pilots, quantities, etc. persist even if you go to another solar system and come back.

Station Traffic Patterns

A numbering system for landing pads at a space station in No Man's Sky. Click on image and read caption for details.

Everyone's noticed that some pilots repeat. Do ship traffic patterns have useful implications?

I parked my POV and recorded two streams of visits (32 and 55 minutes). All I did was watch (record), without moving or interacting with any ships. Both recordings were at a station with 12 ships I had previously identified. Some findings:

  • Three of the 12 ships clearly showed up much more than the others, averaging 18.7% of ship visits each (range of 6 values 17.1 - 22.4%) versus the expected 1/12 random average of 8.33%. Conversely, the infrequent 9 were highly variable, with an average of 4.9% (range 0[1] to 10.3% in 2 x 9 = 18 values). The more-frequent quarter showed up 4 times more often than the less-frequent remainder (18.7/4.9). They accounted for slightly over half of visits (3x18.7 = 56.0%).
    • Making two videos showed that the exact same three ships were the most frequent both times. There was no overlap between the high frequency and low frequency group. This effect is clearly for real, it's for specific ships, and everyone can attest to how common some ships are, at a given station.
    • It's possible there was some weighting to how often the rarer pilots showed up, or maybe there wasn't... the number of times they appeared was often quite different in the two recordings, suggesting it's simply hit or miss for them all. Low frequencies need lots of testing for good precision. Meanwhile, noting every single landing for long videos is seriously tedious. You're welcome to do more testing if you want to refine it.
  1. The zero count is from a ship seen twice in the first recording, but not once in the second one.
  • It takes about 60 seconds from the time a ship generates (outside the station entrance) until it lands. This can be easily demonstrated by reloading in a station and waiting until the first ship lands.
    • If you stand close enough to a pad, nothing will land there. And if you move close to an occupied pad, the ship won't leave. In other words, being close to a pad freezes its current ship state.
    • For shits and giggles, let a ship land right on top of you. You'll be trapped in its framework, and it can't leave because you're close. You're screwed! You'll have to reload. :)
  • If ships won't land on pads you are near, do other pads make up for the traffic? In the 32-minute video, my POV was on a balcony above the landing floor, with no pads blocked. For my 55-minute video, I stood right in the center of pads 3, 5, and 7, blocking them all (right below the number 5 in the first inset). With 3 of the 7 pads blocked, this station had 4/7ths of the traffic that the first did (landings per minute for the whole station, averaged across the whole video). The result was clear: You DO simply block pads that you are near, and get less station traffic because of it. So move off to the side of the pads whenever you're waiting.
Distance between landing pads. See image page for more details.
  • How far to the side? To answer this question, you may have noticed that if you stand in the center of four pads, you block the closer two but not farther two. A little math using video capture of walking across the pads and noting video frame timestamp shows that you block pads at about ~1.5 pad-diameter distance. See second inset for more details.
  • For whatever reason, it was also found that pads 1 and 6 only got half the traffic that the rest did. In both recordings. So be near them if you must be near some, while moving back away from pads and waiting for the next ships. Maybe hang out near your ship at pad 0.
  • It is recommended that you save and reload before starting a heavy trading session. So that your ship is at pad 0 and not in the middle of things, blocking your movement and keeping ships from landing there.

Items, Prices, and Stars

Everywhere on this page, prices are relative to you, consistent with how you see Sell or Buy in the bottom of your trading screen, unless otherwise noted.

Item pricing in No Man's Sky is fairly simple, but there are a few wrinkles:

  • There is a base price for each tradable item and element. You can easily see it in your inventory. This price is also equal to the "Galactic Average" against which prices are compared when selling. So the baseline Galactic Average is the same as your base sell price. And like it implies, the Galactic Average is the same across the entire galaxy. For practical purposes, the Galactic Average - which is the same as your base price - is simply the game's base price for the item.
  • When trading, your Buy price is always 20% more than your Sell price for items, and 25% more for elements (Heridium, Emeril, etc.). We will call this the Buy Markup. Apparently, the Buy Markup is a given across the galaxy because if something is, say, +5.0% compared to Galactic Average when you Sell it, the Buy price (which is more) will also be exactly +5.0%. (It would have been +6.0% if the Buy markup was 1.2 x 1.05. No, it's +5.0% of the galaxy-wide base Buy price. Which is 1.2 times the galaxy-wide base Sell price.)
  • The deviation from the Galactic Average that you see when you are trading applies to everyone in that station (pilots and terminal). If a particular item is +1.0% for one guy, it is +1.0% for everyone in the station. No exceptions, except that the asterisk (star) makes it double priced (more on this below). However, item prices will differ slightly from system to system; Dynamic Resonators might be 2.3% higher than Galactic Average for everyone at the station for Solar System A, and 1.7% lower for everyone at System B.
  • The star asterisk next to an item means that its Buy and its Sell price are both exactly double what they would ordinarily be at that station. It occurs on a per-pilot (or terminal) basis; if his Buy price is starred, so is his Sell price. This makes the Buy price 2.40 (items) or 2.50 (elements) times the unstarred Sell price. Except for extreme-markup oddballs.

Exceptions to Buy Markup Rules

There are important exceptions to the Buy Markup rule (+20% for items, +25% for elements).

  • The Atlas Stone Buy price is 41 times its Sell price. (The base price is 68,750 U, and the mean Galactic Average Buy price is 28,256,250.) Atlas Stones are a special item in the game.
  • The entire Warp component line. Apparently, the designers want it to be difficult to warp at first:
Item Base

Sell Price

Normal Buy

Price (x1.20)

Actual G.A.

Buy Price *

Actual Buy

/ Sell Ratio

Ingredients
Warp Cell 46,750 ** 56,100.0 N/A N/A 1 Antimatter, 100 Thamium9
Antimatter 5,232.5 6,279.0 53,000 10.13x 1 Electron Vapour, 50 Heridium, 20 Zinc
Electron Vapour 4,950.0 5,940.0 43,325 8.75x 1 Suspension Fluid, 100 Plutonium
Suspension Fluid 1,100.0 1,320.0 20,625 18.75x 50 Carbon
* Actual Galactic Average Buy Prices are estimates only good to four significant digits; the fourth and fifth digits shown are guesstimates.
** Warp Cells can't be sold; this is the base value you see in your inventory.

Just so it's clear: Only these items are known to have extreme markups. The markup occurs for all pilots and stations everywhere, throughout the whole game, as far as I know.

These items can't be traded for profit, because the buy price is greater than the sell price (MUCH greater).

Trading For Profit

Throughout this analysis, "pilot" equals "ship", since each unique pilot always has the same unique ship (see above). Also, this strategy excludes extreme-markup items like Suspension Fluid and Atlas Stones.

The price scheme leads to a few conclusions for making profit:

Because items have a 20% Buy Markup, and some pilots have starred items, you can get 67% profit if you buy unstarred items and sell them starred. For example:

  • an item with a base Sell price of 10 will have
  • an unstarred Buy price of 12 (Buy at this) and
  • a starred Sell price of 20 (Sell at this).
20 / 12  =  5 / 3  ≈  a 1.667x return on your money.
  • This holds true for all items except extreme markup items.
  • Elements are worse with their 25% markup (2000/1250 = 1.60x), but they won't do for rapid-fire trading anyway. More on this in Optimizing, below.


The key to this strategy is: first, find pilots who are starred for high-priced items, then find others you can Buy it from (unstarred). At the same station.

Then just keep bouncing between ships as they land, rinsing and repeating for as many hours as you want to make 10 to 20 million Units per hour.

For starters, target items worth more than 33k U starred Sell price / 20k unstarred Buy price. If you're so busy or fast that this level rarely fills up your free inventory slots, then lower the cut-off. If you get full too much, raise it higher or change which ships you hit. More on this, below.


There are lots of deals at every station, but finding paired items takes time (starred Sell and unstarred Buy). You can't just buy everything expensive, or you'll have to eat the 20% Buy Markup if nobody's starred for it.

Still, as soon as you see something that's starred, you can start buying it from subsequent pilots already (unstarred, of course).

That said, Dynamic Resonators (DRs) appear to be special. All three stations I tested had two people starred for it, and everyone sold it. So it is a fair bet to always buy unstarred DRs already. It is definitely possible for the station master to have starred DRs too, which gives you a great backup... if you have been wheeling and dealing and your inventory is full of DRs because no starred guys have shown up in a while, you can always run to the terminal. Usually you wouldn't, though, because it takes so long to run over to the station terminal. Profit per hour = Profit / time.

Paired items vary greatly from station to station. Not only including what's available, but also the quantity. Each pilot always has the same quantity, which varies from 1 to 4 items. So Pilot A always has 3 of them. Pilot B always has 1. Etc.

This approach takes 10-30 minutes to set up (figure out what to buy). Then you can make up to 20 million units per hour for as long as you want.

  • Start taking notes on which ships have starred items above a certain price. I use a cut-off around 20,000 (unstarred Buy price), which equates to a 33.33k starred Sell price. You can adjust, depending. (Not extreme markup items, of course.) Check the station master, too... while I don't mention him much, consider him a part of all this writing. Because he is farther, though, you would only go to him during down times, or if your slots are full of something you can sell him, starred. Otherwise, he's too far to be a part of high-frequency trading.
    • When I say "write down every pilot", I mean any sort of description that will let you quickly identify them without having to open their menu to see what they have. Remember, it will always be the same 12-13 ships with the same pilots and same inventory. You want to be able to identify them by sight (without opening dialogue). Then you can skip the less-profitable ones in the heat of trading, all depending. I write down ship descriptions like "big boxy yellow" or "X-wing sleek red fighter". For the names, I abbreviate titles (which are sometimes a mouthful) and mainly go by the last name. Only the last word is an actual name per se, and always unique in my experience.
  • In theory, you could write down every item above your set point, right from the start of evaluating a station. This way, you already know who you can Buy from, if you find a starred trader later. If you do this, you waste time making a bunch of entries that won't pay off. Whereas in the other approach, you have to go back through to find who sells things you found starred. Either way, it takes time to get up to speed. I prefer the latter, but it's your choice. If you get heavily into this, you might make a template of high-priced items and use a checklist approach. Someone could even write a script to show who's most profitable (more on this below).

Optimizing High-Frequency Trading

One you know all your buyers and sellers, it's time to make money.

If you keep your info in a spreadsheet, it's easy to calculate how much you can make off of pilots you buy items from. Just take the profit per item times the number of each item that they have. (You wrote down how many each pilot has, right?) Then sort or organize them according to how much each is worth. There's often a big difference between the pilot with the most to offer and the least.

  • Pilots that you sell to are in a different league. If they're only good for selling (and they don't have anything to buy), only sell if you have enough of their trade item to make it worth the effort. Otherwise, just skip them. (Memorize your ships!)
If they are infrequent, remember that it might be awhile before they show up again. This is why it's really nice to have more than one pilot starred for something.

Each trade has a steep time overhead. Here's informal data from a few video-capture timing tests:

~19 seconds: Time spent running to, opening dialogue, waiting for Buy/Sell options, and closing after buying. In other words, all the "overhead" time not actually spent in the act of Buying per se. 
  1 second:  Time to buy exactly 1 of an item. (Just the act of Buying, itself.)
  4 seconds: Time to buy 4 of an item. The extra time is to get the quantity spinner run up.
  6 seconds: Time spent to buy all 48 of an element. Once the spinner gets going, it moves fast.
 Also add an additional second or two for every additional item bought from same person. (Time to find and cursor to it.)
Total time spent per ship you Buy items from: 20 - 28 seconds

It takes 10 seconds from the time you initiate dialogue ("click on" a ship and start the little white spinning circle) to the time the Buy and Sell options are clickable!

Now consider what you might have traded, and profit you might make, for those 20-28 seconds:

Type of Trading Profit for Each Item Number Total Profit Seconds Profit/Second MU/Hour Notes   (Prices: Sell / Buy / Starred Sell / Profit per item)
Heavy 22,000 8 176,000 28 6,286 22.63 Heavy duty trading: Buy 4 Dynamic Resonators and 4 Gravitino Balls from one pilot in 28 seconds. Both items are 27,500 / 33,000 / 55,000 / 22,000.
Very Light 16,500 1 16,500 20 825 2.97 An example of very light trading: 1 Geknip, 20 seconds. 20,625 / 24,750 / 41,250 / 16,500.
Elements 61.875 44 2,723 25 109 0.39 Element trading: 44 Chrysonite in 25 seconds. 82.50 / 103.125 / 165.00 / 61.875.

The best probable deal makes over 20 million U per hour, and is many times better than blindly selecting pilots with very little to buy. These are extreme examples, and it doesn't include the time you would need to sell (so 20 MU/hour is probably a practical limit). But still, the point is clear: It makes a real difference who you choose when multiple pilots are waiting (and the ones you don't choose will probably leave). From your point of view, it's the same amount of time and practically identical amounts of hand waving and mouse pointing. But who it's done with makes a big difference.

Thrown into this wrinkly mess is the fact that there are steep differences in how often some ships show up. This is especially important if you can buy a lot of something, but the starred sell-to is very rare. You might trade fine for an hour, then a rare guy won't show up for two straight hours while you fume and fret over your full inventory and wasted time.

Your Worksheet for Fast Money

I did all the work for you and put together a Google spreadsheet so you can jump into high-frequency trading a.s.a.p.

Bonus tab: Complete list of all tradeable items in No Man's Sky! (though only a fraction are good for high-frequency trading)


All right then, intrepid profiteer.

In the final analysis, the amount you make per hour is a function of many things:

  • How frequently ships land at your station. Does it have long lags, or is it so busy you miss important ships?
  • The exact mix in terms of who you can buy from (at what quantity) and who can you sell to, including the stationmaster.
  • How frequently particular pilots show up. This is where luck comes into play.
  • How good your notes are. Can you glance at them and know, in an instant, which of the three waiting ships are the best to hit when the station is crazy busy?
  • How quick-witted you are. Can you keep the general state (fullness) of your inventory, and the right pilots to hit next, at the fore-front of your mind for an hour or more?

High-frequency trading is a game to itself, where quick thinking can be highly rewarded.

And sometimes bad luck can screw you.