Talk: Price Catalogue - Starships

Post-Visions

Are we sure the freighter prices are still correct? The freighters I've been on recently (after rescue missions) have all had much lower prices than in the table.

I can confirm the freighter prices don't match Visions 1.77. An A-19 costs 18,400,000 and a B-18 costs 14,100,000Bhocis (talk) 02:35, 7 March 2019 (UTC)

In-Game vs In-Table revisited Dec 3rd

I wonder if we could revisit the issue of which starship parameters exist in-game, as opposed to the ranges allowed in the .pak files. I am still not convinced that all of the starship Class/Slot combinations shown in the Price Catalogue table actually exist.

In particular, I am skeptical that the following combinations actually exist in-game: 36C, 39B, 45C, 48B Haulers; 35C Explorers; 35C, 38B Fighters. Is there a table somewhere that shows which combinations of ship type, class, and slots are allowed? Or has anyone else actually seen any of the seven listed configurations in-game? I know that just because I have not seen a particular ship, that doesn't mean it doesn't exist, but I have looked at a *lot* of ships :) (fwiw, I have never seen a 33C or 32B Explorer either, but based on inferred symmetries in the table, I suspect that those configurations probably do exist.)

Also, there seems to be some discrepancy between the Type Bonuses shown in the Starship Archetypes section (is it ok to discuss that here?) and the values seen in-game. For example, the damage ranges shown for S-class Haulers, Shuttles, and Fighters are listed as 10-20%, 15-20%, and 55-60% respectively. These data appear to come from the ranges in the INVENTORYTABLE.MBIN file. But after looking at about 20 ships of each type in-game, I have only ever seen one damage bonus value for each of the three ship types: 19%, 19%, and 59%. There appears to be something somewhere else in the NMS code that is restricting the values in game. Does anyone else have any ideas of how we could get a better handle on this?

P.S. I rechecked the constants in the price-slots formula vs the unrounded INVENTORYTABLE data. For the six starship types, Ket's wolframAlpha numbers are correct except for Shuttles: A and B should be 0.3751833 and 4.884726 respectively (Anonymous, 21:48, 3 December 2017)

All those ships should only spawn in poor economy systems, 45 slot hauler is a roll between 55.56-66.67 on the spawn. Out of that range 55.56-60 will be C class, and 60-66.67 will be B class in a poor system.

48 slot hauler is 88.89-100 on the roll. out of that range 88.89-90 is B class, and 90-100 is A class in a poor system, pretty rare.

36 slot hauler is 50-62.5. 50-60 is C class, 60-62.5 is B class.

39 slot hauler is 87.5-100. 87.5-90 is B class, 90-100 is A class.

35 slot explorer/fighter is the same rolls as 45 slot hauler

38 slot explorer/fighter is the same rolls as 48 slot hauler

The type bonus multipliers displayed in-game are truncated(cuts of the end). You can see the real number if you inspect the save file on the pc version, it also uses the untruncated number in any math being done. As it never rounds up on the number it is almost, if not actually impossible to get it to display the upper bound. as even a 39.96 bonus(real example) will display as 39.

Also I just checked the shuttle numbers again, you are right, how the hell did I get those shuttle numbers, they don't match at all, going to edit the old table, somehow managed to input 50m instead of 40m on wolframAlpha. Thank you for double checking the numbers --Ket (talk) 06:22, 4 December 2017 (UTC)

Thanks, Ket. Where are you getting the slot range data? I know how to use NMS Modding station and MBIN compiler, but maybe I need to look harder at the EXML files or upgrade my skills a bit further :D.

Also, are there any other factors that determine which NPC ships will show up in a given star system? I think you see bigger ships as you get closer to the centre of the galaxy, and if you have a big ship you are more likely to see other big ships than if you have a small ship, but there also seems to be another random factor that makes some star systems populate with larger ships and some with small ones (besides economy level).

I understand why rounding down will normally always exclude the highest number in the type bonus range, but shouldn't all the other numbers in the range show up occasionally? Why do all S-class haulers have a +19 damage bonus, and never +18 or +17 or any other number in the allowed range? Most of the other in-game type bonuses seem consistent with INVENTORYTABLE, it's just the S-class (and maybe A-class) damage numbers that seem to have restricted ranges. Am I looking in the wrong place? Or is it just me that only sees single values for those parameters? (Anonymous, 09:40, 5 December 2017)

The slot range rolls is as simple as getting the upper and lower bound for each slot by dividing 100 by the number of possible slots. i.e large tier hauler has slots between 40 and 48, aka 9 possible slot counts. 100/9 = 11.11. So the first slot(40) has the lower bound of 0, and upper bound of 11.11, then slot 2(41) has lower bound of 11.11 and upper bound of 22.22 and so on. That same roll is directly used for deciding which class a ship/multi-tool will have. The class ranges are defined in the inventorytable file by the probabilities. The actual roll is not actually a 0-100 range, but it's the equivalent of it so it makes the math easier, think of it like % of possible values. Interestingly enough each planet has 2-4 multi-tools, and that same roll that is used for slots and classes is also used to pick which of those 4 tools it should spawn, so only 1 of them can potentially spawn as S class, the same one that is in the top 25% of possible slots. I suspect that multipliers might be using the same input value, which means that a higher roll will also be in the upper area of the possible multiplier values, but I haven't found any strong enough evidence for it yet as it doesn't seem to use it directly like the others do(possibly split the value between multiple multiplier types?)

At the moment I have seen no evidence for any probability changes depending on distance from center, just a random spread. And so far have been just as likely to find a large tier ship/tool in a poor system compared to wealthy. Only seen economy directly affect class chances and how fast ships spawn(reduced delay).

Possibly take a look at the multipliers for exotic ships, any roll on them will be S class, since it forces it and ignores the class probabilities set in the inventorytable file.

Also by roll I mean the possible input values, it is likely based on the seed or the coordinates of the thing it spawns at.

You can mod the class probabilities in the inventorytable file to see how it expands the possible slots it can have, which is how i figured out how it works at first. You can set the probabilities of which ship type to spawn at the bottom of the solargeneration global file, nice if you want to test having 100% exotics. The "unknown" structs are in the order of the factions: Gek, Vy'keen, Korvax and then 4 factions that can't own a system atm(sentinels, atlas, travellers and none) --Ket (talk) 12:46, 5 December 2017 (UTC)

Did some more testing with untruncated multiplier numbers. The first multiplier(damage) uses the exact same input as the one used by slots and class, the other 2 does something a bit different. You can convert the damage multiplier back to the same 0-100 range I used for easy of use on the others and they will match exactly, just take the untruncated number minus min value and divide by (Max minus min) and multiply by 100. This means that an S class tool or ship will always be in the top 1-2% of the damage range, and won't use the others. The other 2 multipliers on the other hand seems to use the full range, had values all over the place for all slot counts for those. --Ket (talk) 14:41, 6 December 2017 (UTC)

You still didn't tell me where you're getting your info :)

I did some more ship logging, and have been more diligent about keeping detailed records. OK, you are right: the "missing" ship configurations do show up in poor star systems, and there seems to be no relationship between ship size or distance to centre and ship distribution. But there does seem to be a real difference between different star systems, with entire (and different) blocks of ship configurations being excluded from each system. For example, one system I visited had no small fighters or medium explorers, another had mo medium or large explorers or large haulers, a third had no medium haulers or large explorers or large fighters. The missing blocks do not seem to depend on the star system's economy type or level or dominant lifeform. And one "poor" system I visited had two Exotic ships show up, even though poor systems aren't supposed to spawn any S-class ships. Is there an easy way for me to upload a few spreadsheets to this board to show what I'm seeing? If not, I can send them to dropbox and post the link here. (Anonymous, 16:40, 6 December 2017)

The main source is the class probabilities in the inventorytable file, if you mod the numbers it will expand which slot counts it can spawn with to match. Based on that I logged many ship spawns, many of them are in the price log sheet, and I also got a multi-tool sheet I'm loggin in separately. So far there has been no data to disprove the current hypothesis, and have found more data to prove it in that the damage multiplier uses the same input value in what it picks.

If you didn't already know, it picks 7 shuttle models, 3 fighters, 3 haulers, 3 explorers, +4 race specific type, and 1 exotic model per system. Those picks use the descriptor file of that particular ship type and with the exception of the few parts marked xRARE is equal probability for all. Which inventory tier the ships has is directly linked to specific type of wings, or the body in case of the shuttle. As for exotics in poor systems, I see that I didn't write it clearly, but exotics completely ignore the class probabilities, and is forced to always spawn as S class everywhere. Some people have said they are more common in wealthy systems, but I haven't found enough data to show that yet or any references to it among the unknowns in the globals, but much left to explore. One of the "mysteries" is how it is set to 20 ship variations per system, with the ratio I posted above, but you can still find 21 ship models, maybe related to load location like the multi-tools or just how it handles types with low probabilities like exotics? take a look at the bottom of the solargeneration globals --Ket (talk) 17:51, 6 December 2017 (UTC)

Thanks, Ket. I appreciate you taking the time to walk (drag) me through this. I now understand how the ship size/class distribution works within a specific tier :)

I hadn't known about the "21 ship models per system". I will need to take a look at that more closely to see if I can learn anything about correlating body parts to ship size tiers.

GcSolarGenerationGlobals tells us that the probabilities are 33.2% for shuttles and 0.3% for exotics regardless of a system's dominant lifeform. Also the probability of a system's preferred design is 33.2% (Haulers for Gek systems, Fighters for Vy'keen, Explorers for Korvax) and 16.2 for each of the non-preferred designs. I have only looked at four systems in detail so far, all Gek, two small and two medium, but *all* the probabilities seem to hold regardless of wealth level or allowed tiers. I saw 1 Exotic in 500 ships in the two Medium economy stars (0.2%) and 3 in 421 ships in the two Poor systems (0.7%). It's not a big sample, but at this end, I haven't seen anything to suggest that there are more Exotics in wealthier systems either. (Anonymous, 07:30, 2017-12-07)

From the systems I visited (very few comparing to you) I always saw 7 shuttle + 7 major type + 3 minor type + 3 minor type = 20 basic ships ... + 1 exotic when I found one... so I would say the ratio is 20 basic ships by system + 1 exotic that ignore the basic ratio. Talinwind (talk) 03:08, 8 December 2017 (UTC)

The 921 ships I looked at (four star systems) ended up distributed as 34.7% shuttle + 32.4% major + 16.6% minor + 15.9% minor + 0.4% exotic. To me that looks pretty close to the expected 33.2% + 33.2% + 16.6 % + 16.6% + 0.3% from the table in GcSolarGenerationGlobals. If you fix exotics at 0.3% (for sure it's not 1/21 or 4.7%), the 7 + 7 + 3 + 3 distribution would predict 34.9% + 34.9% + 15.0% + 15.0%, which doesn't seem to quite jibe with my numbers. To me, it looks like each 33.2% range of the "roll table" gets split into 7 segments, each with one tier determined (?) by the random-maybe selection of wing types (the 16.6% ranges would each be split into 3 segments). That would allow HG to adjust the 100:100:50:50:1 weightings in the GcSolarGenerationGlobals table without changing the 20 ship types per star system model. I don't want to generalize too much from only 4 star systems, but so far it looks to me that the 3 tiers for minor ship types are selected randomly, but that the extra 4 tiers in the major and shuttle types might be biased towards the "small" tier. I had one system where the major ship type data suggested that the ratio of small:medium:large tiers was 6:1:0. The chance of that occurring from purely random rolls is less than 0.2%.

I haven't collected any data on ship spawning rate, but can confirm that the rate at the star station for the last two "Poor" systems was slower than average. The second one was so slow that I moved to a planetary trading post to collect my data, and there the spawning rate was as fast as I've ever seen. So if spawning rate is related to star system economy, it appears that the determining mechanism applies only to space stations, or that space stations and trading posts within a given system are given different rolls. (Anonymous 0200 2017-12-09)

The spawn rate is definitely different for each economy tier, I found the variables for them a bit over a month ago, but as mbincompiler hasn't gotten any updates since they are still listed as unknowns, Its the 3 variables between the shipvariations variable and the spaceship weighting at the bottom of SolarGenerationGlobals. Unknown180, Unknown184 and Unknown188. That value is used in a slightly different formula for stations/freighters vs trading posts, 20/value for stations is time in seconds between ship spawns, it spawns a single ship each time and it can spawn up to 12?(didn't write it down so going by memory). For trading posts its 5/value in seconds between groups of ships, 1-4 per group, there can be a max of 3 groups spawned at once. That same value is also seem to be used in some formula for the weighting on group size, but not sure if that data was just random chance or if it is actually there.

As for the inventory tier, it's directly linked to certain parts, usually wings/subwings, most parts got equal chance between them if they are in the same group, but it's a hierarchy of parts it goes through from the top node, so would have to go down the list in the descriptor file for each ship type, if any part has xRARE at the end of the name then its much rarer than the others, current working theory(based on the assembly code extracted from the exe) is that every part with xRARE has a probability of 1, and every part without it has a probability of 20, which means if there is only 2 choices for that group it has a 1/21 chance. For exotic ships for instance the top node is a choice between royal and squid, squid having xRARE, so 1/21 exotic model picks it would be a squid.

Going through the descriptor files. Shuttle has a default of small(set in canopy) and override by medium tier if it has the double hull, which is 50% of the time.

Exotic technically has a default of small fighter, but its always overridden by medium tier exotic so its not noticeable.

For Haulers: if it has the cockpit with the cargo box below it(Cockpit_E)(1/8), it gets set to medium unless overridden, The Hull sets the default to small(need test, but I suspect this always overrides the cockpit, everything else goes in definition order so would be odd if it didn't), 1/7 "Wings" sets the size to large(WINGSD, that "wing" being the cargo wings, with 2 main variants in the sphere and box version), 1/7 sets the size to medium(WINGSA), and 2/7 sets the size to small(WINGSB and WINGSC), this can be overridden by "subwings". Subwings(C, D, H, I) overrides to small, Subwings(E, G) overrides to medium, and Subwings F overrides to large, A B and None does not override. WINGSA and WINGSB has 1/10 chance for any specific subwing, WINGSC has 1/10 chance 3/4 of the time(subwings are attached to thrusters), but it can't get subwing C if it picks the diagonal type, so 1/9 for that one. And finally for the sphere variant of WINGD(50% of the time) it can use subwings(A,B,I,H) + 2 none variants, so 1/6 for each specific one. Doing the math on those (ignoring the cockpit) seems to be roughly equal to 58.1% chance for small, 20.3% chance for medium and 21.6% chance for large.

Fighters are fairly simple in comparison, 10 different wings, 5/10 are small, 4/10 are medium, and 1/10 are large, so 50%, 40% and 10%.

For Explorers: default is small tier, and out of 12 wings, wings(A) overrides to small, Wings(D,G,K) overrides to medium, and Wing B overrides to large. but the wings can be placed as either asymmetric or symmetric, and in either with a joint facing up or not. With the right wing overriding the left wing if going by definition order. Symmetric wing sidejoint got all 12, upjoint got wing(A,C,F,G,I), Asymmetric wing sidejoint got all except wing L, asymmetric upjoint got wing(A,C,F,G,I). Doing the math, that should be roughly 63.2% chance for small, 30.3% chance for medium, and 6.5% chance for large. --Ket (talk) 12:35, 9 December 2017 (UTC)

Oops there might be a couple of mistakes with the hauler and explorer math, screwed up some of the fractions when going up the tiers, so that chances are skewed towards the side with more parts. I'll redo it properly once I got a bit of time --Ket (talk) 21:50, 10 December 2017 (UTC)

@Annonymous: I think you are miss-lead on some points. First, could you please log-in before posting :p Second, you have to be very careful with statistics. I mean you can throw a coin 1000 times you will never find a perfect 50/50 result, there is always a bias in experiments comparing to math. I personally visited 8 star systems, and I confirmed the distribution of 7+7+3+3 in 6 systems, plus 1 exotic in 3 systems (you can see my logs in my userpage). It is clear that it is a question of time you spend, but you will find 7+7+3+3 basic ships + 3 freighters (and eventually 1 exotic) in every systems. Third: When we talk about 7+7+3+3(+1) we talk about ship model, I means 20(+1) ships with different names. I guess if you look at your 920 ships you will confirm that. As as understand, in each systems, 20(+1) ships are generated. Then the stats of each ship is reroll at each landing. Talinwind (talk) 02:17, 11 December 2017 (UTC)

@Ket: Amazing info about the tier rates... I am looking forward the revised maths :). Please be sure that all these info are summarized on starship. Talinwind (talk) 02:17, 11 December 2017 (UTC)

And fixed. Hauler should be 70.83% chance of small, 12.94% chance of medium, and 16.23% chance of large. Explorer should be 64.5% for small, 29.8% for medium and 5.7% chance for large. I can post the sheet I did the math on if required --Ket (talk) 13:47, 11 December 2017 (UTC)

Synch Nov 12th

• synchronized all price from @Ket sheets to the table (merge Poor + Average + Wealthy system)
• change background to "transparent"
• add a note about the wealth
• get rid of the horizontal tables at the bottom

I get a backup of the previous version, so if you disagree with one of this change, let me know and I revert
It could be nice to add a note about the wealth involve in the slot count on starship. (better that the link to economy I currently add). Talinwind (talk) 01:41, 12 November 2017 (UTC)

Many of Ket's prices are "save edited/modded and may not appear ingame (correct prices)". Also, price for 36-slot A Explorer on Ket sheet looks wonky? New numbers for 46 A Hauler and 33 B Fighter are ok though. - Anonymous 21:15 2017-11-12

Obs that was supposed to be a 25m, not 35m, fixed. The prices are correct as I didn't mod those, I only save edit the the ship slot count to check each possible slot quickly, the available slots on the other tabs should be correct based on the % values defined in the inventorytable file, though I haven't logged all the appearances ingame yet. --Ket (talk) 21:59, 12 November 2017 (UTC)

Format Nov 9-10

Do we know for sure which slots can be found in which class?

For Hauler and Shuttle it seams we have all the prices. For some cell, it is obvious like: Explorer C22, C24 > we miss the C23 but it should exist. I guess Explorer and Fighter are similar in slots distribution, so I mimic the pattern according to the price we have. (do someone disagree with that ?) For Exotic, we have multiple feedback that the range is now (v1.38) adjust from 15 to 20. The only problem that left is for the Capital freighter. We don't have all info. I fill the slots according to what I saw In game. We need to adjust the availability as we find more ships.Talinwind (talk) 07:04, 9 November 2017 (UTC)

The available slot % I posted on the class page and on the bottom of my price log should be correct, its based on the numbers in the inventory definitions and I haven't found anything that contradicts it, as far as I know its used by ships, freighters and multi-tools. The game does something similar to a 0-100 roll when deciding the slot count, and depending on that roll it also decides the class.

Slot range for each class depending on economy in % of inventory range, 0% being minimum slot, 100% being max slot for that inventory tier + ship type

	Poor	Average	Wealthy
C	0-60%	0-49%	0-30%
B	60-90%	49-84%	30-70%
A	90-100%	84-99%	70-98%
S	n/a	99-100%	98-100%

--Ket (talk) 08:03, 9 November 2017 (UTC)

For example, a medium tier hauler: 0-12.5% = 32slot, 12.5-25% = 33, 25-37.5%= 34, 37.5-50% = 35, 50-62.5% = 36, 62.5-75% = 37, 75-87.5% = 38, 87.5-100% = 39 --Ket (talk) 09:55, 9 November 2017 (UTC)

Made another tab on my own sheet to show which slots should be available for each economy level, the % numbers are from the inventorytable file, and i know they act as advertised from testing and modding the values, should I update the background colors on this page?: https://docs.google.com/spreadsheets/d/1oH05PLVbaW3Uz3Rwst9FhkMZvIwGU9NHsBvVruJww6o/edit#gid=874975169 --Ket (talk) 16:55, 9 November 2017 (UTC)

That is a very useful info ! If you could find a way to incorporate that in the table that would be awesome. I would say different shade of background colors (but I am not sure DDfairchild will approve ^^". I will check more closely the wealth of systems now ! Talinwind (talk) 01:10, 10 November 2017 (UTC)

Using shading for some of the spaces improves readability which is borderline ok. Using colors to mean something is a different story - let us avoid that path. Ddfairchild (talk) 04:25, 10 November 2017 (UTC)

I know I know... What about put a background as it is in all slots available. @Ket are you ok to do so or should I edit the page as I am maybe more familiar with the code? And adding a text note about the slots that vary according to the wealth of the system.

It might also be nice to add the price range on starship page.

Looking closer, that is nice to know that rate change, but for the patient one, it do not matter that much... The only breaking news that is important to be written on the price catalogue page is No S class will appears in a Poor star system...Talinwind (talk) 04:43, 10 November 2017 (UTC)

I added a few new prices today, including 2 haulers and 1 shuttle. I don't think we've seen all the available ships yet, but I think we're starting to get close.

The table showing the "roll ranges" for each class and economy type is very useful to me. I think having a 1-in-50 chance of getting an S-class is much better than 1-in-100, so for sure I want to be in a wealthy economy if I'm looking for a new ship. Any idea what determines the chance that a ship will fall within the small, medium, or large size tiers? It used to be that bigger ships appeared more frequently as you got closer to the centre of the galaxy, didn't it?- Anonymous, 18:29 11 November 2017

I don't think there is any specific weighting for which part(that decides inventory tier) appear in which system, other than the regular random chance, though going to the matching race for the ship you want gives you a greater variety in models, and haulers got 3 different parts that all grant large inventory tier vs only 1 each for explorer and fighter, most of the other things that are decided by economy or race is fairly straightforward in how its defined in the files, though there are many unknown variables that might hide more, or just unintended bias with the seeds. Also since this page is called price catalogue, should it include multi-tool prices? I made another sheet for those using the same slot availability numbers as the ships(they use the same system for inventory), save edited my inventory to check all the prices after removing the 30% price loss with a mod, so the prices should be accurate: https://docs.google.com/spreadsheets/d/1NQJpQ0CSqzKaINDKhnsoNeOpJBHWJZIj0WNvaZlSFL4/ --Ket (talk) 19:13, 11 November 2017 (UTC)

@Barret (Anonymous is you right ?): As the sheets by wealth level of @Ket match with the In-Game price so far, I compiled the tree sheets in the current Price catalogue. hope everyone is ok with that.

@Ket: I think we should separate ships price catalogue and tool price catalogue. the pattern are the same, but the price different right? Talinwind (talk) 01:41, 12 November 2017 (UTC)

Format Nov 7-8

As for table formats, I think the best way to decide this is to create a sample table or two in each format on the page. Yeah, some of it would be throw-away work, but unless people can actually SEE a real version of how it would look in the wiki, we can talk all day about it. It's been said more than once that "I don't know art, but I know what I like." One other suggestion - I have a great deal of faith in Artifexity as far as what looks good and what doesn't. (He's the editor who did all the icon designs for the front page, plus keeps the maps updated for the CivSpace page.) I think getting his input on this issue would be well worth the time if someone wants to approach him on the subject. Ddfairchild (talk) 01:57, 6 November 2017 (UTC)

Thank you for your trust. I think you are discussing at a high level. The table already looks fantastic as it is. [...] I can not comment on the table formats, as I have no examples. Artifexity (talk) 06:45, 7 November 2017 (UTC)

Here a test of the tables separated by archetypes and horizontal... Now we can judge by sight. Waiting your comment ;) Talinwind (talk) 10:01, 7 November 2017 (UTC)

The archetypes in a single table are clearer and more compact. Nevertheless, difficult decision. I would prefer the overall view, so do not split. Artifexity (talk) 10:30, 7 November 2017 (UTC)

For me, I find the separate horizontal tables at the bottom easier to find what I want, as long as I'm looking at a specific type of ship. One thing that might make the "big" table easier to read would be a very minimal blank column between each of the archetypes to separate the sections. There is one additional solution that nobody has mentioned. Once this table is done, it's unlikely to change, correct? If that's the case, build the table in excel, format it to match the general wiki appearance format, then include a screen shot of the table instead of having actual wiki code to mess with. Ddfairchild (talk) 00:11, 8 November 2017 (UTC)

Including a thin, coloured vertical bar (rather than a blank column) to separate archetypes might be a cleaner way to improve readability. A horizontal bar every 3 or 4 rows might help too, especially for those scrolling across small screens. (If someone can tell me how to do this in wiki language, it would be much appreciated - the Gamepedia tutorial page instructions don't seem to work!).

I made some modif to improve the "big table" readability: add empty column between archetype and remind slots on the right. I think it helps, but don't hesitate to revert if you disagree.Talinwind (talk) 06:36, 8 November 2017 (UTC)

The empty columns do help, imo. If other people like this, maybe consider an empty row every fifth line or something? Or changing the empty column / row background to a different colour?

What about splitting off just the freighters to a different table? They really are different from starships, and the info should be linked to the freighters page rather than to the starships page. - Anonymous, 22:59 8 November 2017

The empty columns and slots on the right are a big improvement. I do wish there was some way we could get rid of the "big gaps" in the table for slots (see 1st pic) - there's a lot of empty space there <g> But it is what it is. I think the table now is much more usable then the initial version was. So nice work, folks. On the other hand, it's rather convenient how the space at the top not used by Haulers is just about equal to the space not used at the bottom by the other types. Don't know how it would look if we just slid the haulers up? (second pic) Ddfairchild (talk) 23:44, 8 November 2017 (UTC)

• 

  empty space

• 

  possible solution

I get your point but I think presenting different slots on the same line is not a good idea, that will confuse the reader. What can be done, is to deplete the border of the empty cells, at least is the "blank area" top of hauler and bottom of others. That will result with a black hole in these portions of the table and probably ease the reading. a bit like the black cells in the original table (image I posted previously) Talinwind (talk) 01:31, 9 November 2017 (UTC)

Available slots Nov 6

@Ket I agree we should "mark" the available slots at it is in the image i post or your sheets. I think add a background color at the cell is the way to go (Visual edit is much easier for table on wiki) Talinwind (talk) 01:27, 6 November 2017 (UTC)

we try to avoid colored background in our tables as much as is absolutely possible. The periodic table pages are just about the only place where colors are "ok"... Ddfairchild (talk) 01:32, 6 November 2017 (UTC)

even a light transparent gray as it is for the headers ? just one way to emphases the cells Talinwind (talk) 01:36, 6 November 2017 (UTC)

I would probably pretend not to notice that shade of gray if it was used sparingly.

I like the background color (#384640) at the cell. Transparent it would be even better. If it were possible, I would change the font for the numbers. The overall picture appears a bit vague due to the different size of each number. Artifexity (talk) 06:45, 7 November 2017 (UTC)

At first I tried with "transparent" but it was too mild (almost not visible), that is why I picked the #384640. Changing font is a good idea. I don't know well about font used on this wiki, do you have a suggestion for a monospace font that could improve the things ? Talinwind (talk) 09:04, 7 November 2017 (UTC)

#384640 is fine enough. If you could change the font just for the numbers, then Arial and many others would fit. The numbers would be uniform. But I think that is difficult to implement. Artifexity (talk) 09:27, 7 November 2017 (UTC)

BTW, some of the cell backgrounds in the table have been changed to pale green and some have been left black. Is there a particular style that the Gamepedia NMS group wants us to adhere to? I will try to make my future edits consistent with the "house style", but first I need to know what that is! :) - Anonymous, 03:42 8 November 2017

The background color come from me, the point is to emphasis on which slots can be find in which class. Sadly I did not find this option in the visual edit mode. I recommend to fill the cells in visual edit mode (much easier). The background color have to be set in source mode but it can be update only once in a while (that can lead very easily to mistake, careful with that)Talinwind (talk) 06:36, 8 November 2017 (UTC)

Format Nov 5-6

Have a suggestion that may help with future maintenance. Instead of having slots on the left and all the types and classes on the top, I think rotating the table and then also having separate tables for each archetype would make things a bit easier.
1) with separate tables for each archetype and four rows in each table (one for each class), it helps separate the information instead of being in one huge table. Most people who would use this page are likely going to be interested in prices for a certain type of ship and so could jump straight to that particular table.
2) let's say in the future that HelloGames adds new classes (class D maybe?) for each type of ship. It will be much quicker to add a new row for the new class, then to figure out where to add multiple new columns in a single big table. (rows are much easier to add than columns, as I'm sure you've discovered <g>)
3) if a new archetype is added, it's also going to be a lot quicker to copy and add a new small(er) table than it is to add multiple columns to a big one
4) if the maximum number of slots is expanded, then yes, you would have to add new columns but they would at least all be at the end of the table definitions instead of stuck in the middle somewhere <---- Ddfairchild

I would agree with you that split the table by archetype will make it easier to maintain... but I think most of the "ship-people" are used to the sheets bellow and that is why the prices is on this format... what other contributors think ? (Ketila, Thamalandis, Barrett ?)
Talinwind (talk) 07:22, 5 November 2017 (UTC)

DDF is right, we should split the tables. But we can keep its format (class above and slot down) as it feels more natural. The main point for this is rather that we don't need to cross it out. I did the crossing, but curse derped and so it wasn't saved. In theory we have to cross out all slot boxes classes don't reach as the Haulers 48 slots is insane, especially for exotics. We should split them by archetype to reduce clutter and avoid X X X X for tons of slots for non-haulers. Keep the table format as is. Thamalandis (talk) 15:59, 5 November 2017 (UTC)

Agree with slots vertical and type/class horizontal, easier to scroll down than sideways, though depends a bit if its 1 big table or split by type(though wiki tables are a lot different to maintain than a sheet, which changes things). I think there should be some way of showing which slot counts are for each inventory tier per type, like the colored sections on my sheet, but not sure how to best do it on a wiki, shuttle is also an issue as small and medium overlap somewhat. There is also the fact that some slot-class combinations are only valid for some economy levels. For example a 48 slot B class hauler is only obtainable in poor systems, as B maxes out at 47 in average and 46 in wealthy --Ket (talk) 17:42, 5 November 2017 (UTC)

I hadn't seen the googledocs page when I prepared the orig post; to me having slots listed on the left is *much* more user-friendly (ie assuming the user is me, and I use the table to see "what can I buy with X units", or "how many units do I need to get Y".) As far as maintenance goes, it's pretty easy to add new values manually to the table as is. If the table has to be completely reformatted in future, it's not that hard to write a script to make the conversion. The advantage of having everything in one table is that it's easy to compare prices between (say) freighters and fighters. I'd be happy with people adding any prices validated in-game, but would rather not see prices that are solely mathematically generated, like 60 million units for a 48-slot C shuttle. BTW, ship prices within a given archetype and class follow a power law (not linear, not as bad as exponential). I can post the formulas if anyone is interested (empirically derived, not extracted from code, but should be very close). - Anonymous, OP for this table, 17:09 5 November 2017

Would love the formula, was kinda the intention when I started the price log, which is why there are several tabs related to testing editing the various values the formula is based on, switched focus to the repair cost formula halfway as that would be more useful and can't just make a sheet for those numbers, as they are different for each ship. Sadly ended up giving up on the repair cost formula and have been focusing my time on other things so never finished getting the rest of the log finished. As I now know how the inventory slot count that spawns are decided its very easy to just add those and then check the price of that particular slot count using save edit. --Ket (talk) 17:26, 5 November 2017 (UTC)

The formula is Price = A * Slots ^ B. Assuming the info in the googledocs page is correct about different classes having fixed percentage costs above C class, then the values I get for A and B are:

Haulers: B = 4.21

S Class: A = 10.5454

A Class: A = 9.4348

B Class: A = 7.5724

C Class: A = 5.8622

Shuttles: B = 4.88

S Class: A = 0.5572

A Class: A = 0.4579

B Class: A = 0.4153

C Class: A = 0.3712

Explorers: B = 4.37

S Class: A = 4.8196

A Class: A = 3.9507

B Class: A = 3.5487

C Class: A = 3.1985

Fighters: B = 4.26

S Class: A = 10.4695

A Class: A = 8.9572

B Class: A = 6.2611

C Class: A = 5.2475

Freighters: B = 1.57

S Class: A = 1846023

A Class: A = 1732900

B Class: A = 1357129

C Class: A = 1177547

Except for freighters, these give the price in units, not thousands of units as in the table. Remember that these are empirically derived, not extracted from the NMS code. But still, most of the numbers should be accurate to within a fraction of a percent, with the freighter parameters being much less accurate due to the limited data. - Anonymous, 20:09 5 November 2017

I put the 2 unrounded datapoints into wolframAlpha using your function and got these numbers, fits pretty much perfectly as long as you account for the rounding on all numbers except the 2 that are used to define the price for each ship/multi-tool(underlined on my sheet). Multiply the result by the class multiplier for each type to get the non-C prices

type	A	B
Hauler	5.91419	4.20713
Shuttle	0.375183	4.88473
Explorer	3.226151721	4.370056296
Fighter	5.34893	4.26026
Exotic	1317.99	3.04318
Freighter	1122329.633	1.575532122
Pistol	315.385	3.08969
Rifle	5721.25	1.96562
Alien/Experimental	47091.9	1.46791

--Ket (talk) 16:24, 11 November 2017 (UTC)

Updated with correct shuttle numbers: --Ket (talk) 06:26, 4 December 2017 (UTC)

If the choice is to keep the slots on the left, then splitting the table by type is no longer a good idea. With literally dozens of rows (i.e. slots) for most archetype, it would make the page very long and you would have to do an awful lot of scrolling to get to the bottom of the page. To do the comparison, between the "Fighters vs Freighters" mentioned above, you would end up scrolling way down, then way up, then way down, then way up etc. If you put your slots along the top, your scrolling is significantly reduced.

I also tend to disagree with the idea that adding new values is easy. Single values, yes. Adding a row is a simple copy-and-paste. But when you need to add new columns to the middle of a table it's a much more tedious task. New columns require careful counting to make sure you are adding the new parm in the right place, and that you get it added to every row on the table. And if the examples I used above (new class, new archetype) happen, it won't be adding just one new column, but four or five; all of which need to be exactly placed.

Also trying to wrap my head around how people tend to use this table of information. I know for me, when I'm looking at a table, I look at row first, find the row I want, then scroll over to the proper column. If that's the case for most people (and I don't know the answer to that), then they are looking for the number of slots they want, then checking the column to see what a ship with that many slots costs? For me, I'd probably be saying "I want a high-class hauler, let's see how much a high-slot number is going to cost me" rather than "I want a high-slot number, let's see how much a hauler will cost with that number of slots."

But whatever the group decides is the way it will be. You folks will be the ones doing the maintenance in either case <g>, and I was basically looking forward to ways to make that easier. Ddfairchild (talk) 17:41, 5 November 2017 (UTC)

No the split has to happen! If need be by flipping the table - as you want DDF - but just for one reason; a full table has 1/3 of CROSSES for the non existing ships. No type reaches 48 aside of the Hauler. So putting the 20 slot Exotic in the same gives you 28 cross sections and at least around 14 for all others. Why would you do that? No you would split the tables per archetype and save that useless space. Do it by flipping the table, I don't care. If that is the concession necessary. Actually if I think about it, the sort by class... the problem is it goes really far right instead of down for the Hauler. Which for mobile users might be "meh". But as a PC user I don't really care. Now adding hundreds of empty [X] entires... that is nasty to the eye. Thamalandis (talk) 18:03, 5 November 2017 (UTC)

To a database programmer, blank cells in the table are certainly inefficient and ugly. But to the average user logging into this page, I think lots of blank cells are aesthetically quite clean. Adding a new column is onerous if you go through the code line by line, adding a new line every (say) fifth line. But there are faster ways to automate the process. This was my very first attempt at building a table in wiki, and it only took about an hour. - Anonymous, 20:09 5 November 2017

One thing I want to make clear - while there ARE some formatting and text designs that I will "highly recommend" because of wiki standards, the formatting of this table is going to be whatever works best for you guys. I don't necessarily want one design over the other per se; the main reason I jumped in with my opinion is because I'm looking down the road at what may be your maintenance issues. I have just found through more than one painful edit that you can never depend on HelloGames leaving things they way they are <g>. Ddfairchild (talk) 23:53, 5 November 2017 (UTC)

Well I like the "one table for all" but it get very hard to read. Splitting by archetype will "gain" in readability by reduce the empty cells, and the size of table for each type. We could 1> tryto edit the style of the table to make it more readable. 2> Split by archetype... still not sure what is best.

As for the orientation... I will prefer to stay as it is (1 row = 1 slot). Splitting by archetype will reduce the number of rows for each table (up to 24 I guess). and as said having 24 columns of ~7 digit (xxx,xxx) will be not readable at all. Web site are much more easy to scroll vertically.

Log

I got a large amount of the prices logged at https://docs.google.com/spreadsheets/d/1oH05PLVbaW3Uz3Rwst9FhkMZvIwGU9NHsBvVruJww6o It is possible to figure out which slot counts spawn in each economy level for each class using the numbers posted on the class page. For example with a 0-100 roll, a 50 roll would be right in the middle between the min and max for that particular inventory tier and would be c class in poor systems and b class in average and wealthy. Class and slot count are based on the same "roll" but doesnt depend on each other. --Ket (talk) 22:05, 4 November 2017 (UTC)

Do someone mind if we copy the prices from Ketila log in the current page (at least the prices validated in game) ? Talinwind (talk) 07:22, 5 November 2017 (UTC)

Updated the price log, checked the price of every single slot that should be available according to the numbers in the files, also added tabs that is filtered by economy level and 2 that shows the available slots for each slot range, could be used for multi-tools as they use the exact same system, just different prices --Ket (talk) 18:32, 9 November 2017 (UTC)

I love your tabs by economy level ;). Are you sure that the slots vary that much between economy ? I would be great to put all that in the wiki page, but I don't have good solution for integrate all info yet... Talinwind (talk) 10:29, 10 November 2017 (UTC)