Spreadsheets of Manufactory Efficiency by Chapter, Revised and Complete

[MaidenFair]

Okay, I finally got all the manufactories finished and ready to share!

Human Manufactories:
T1:
Marble (or Moonstone)
Steel (or Platinum)
Planks (or Elven Tree Gum)

T2:
Crystal (or Obsidian)
Scrolls (or Arcane Ink)
Silk (or Royal Velvet)

T3:
Elixir (or Silly Soap)
Magic Dust (or Alloy Shrooms)
Gems (or Cosmic Bismuth)

Elven Manufactories:
T1:
Marble (or Moonstone)
Steel (or Platinum)
Planks (or Elven Tree Gum)

T2:
Crystal (or Obsidian)
Scrolls (or Arcane Ink)
Silk (or Royal Velvet)

T3:
Elixir (or Silly Soap)
Magic Dust (or Alloy Shrooms)
Gems (or Cosmic Bismuth)


These spreadsheets analyse the efficiency of each manufactory level in each chapter of the guest races, from Dwarves to Revenge of the Exile. The cultural and residential numbers for the first five chapters are based on the old research tree (which the vast majority of players still have). As before, the first tab of each spreadsheet uses the appropriate chapter-specific numbers to demonstrate the efficiency of the usual path taken by players of updating all buildings to as high of a level as possible in each chapter. Then each tab after that runs the math for all manufactory levels with a focus on a particular chapter. At the end are tabs with the basic residence, culture, and manufactory numbers that I used for calculation.

The manufactory levels that are available in a given chapter are in black text; the levels that would not yet be available for upgrade are still included in a lighter grey text, in case anyone wants to make copies and play around with the formulas or numbers. The highest production values on each tab are highlighted so they can be seen at a glance.

 

[MaidenFair]

If anyone has questions about the data or layout, I'll be happy to explain. I did my best to ensure accuracy but with so much data to copy and collate, it seems almost inevitable that I will have made a mistake in data entry somewhere, so if you see any errors, please let me know so that I can update the sheets appropriately.

Things I left out on purpose: 1) Roads - the ratios of road/production I got doing these calculations seemed to vary greatly based on how many manufactories I dealt with and what chapter they were in. To be accurate, calculating roads also requires working in whole buildings, which is not how the spreadsheet is currently set up. Accordingly, I have, for now, left these out of the equations as a confusing wildcard. I may eventually add a column, but for now they remain uncalculated. Nowhere that I ran the numbers so far have I found that roads altered the balance of whether a given manufactory level was more or less productive than an equally producing number of lower manufactories, so I think the space-saving possibilities of lower manufactory levels as reflected in the spreadsheet is still accurate even when roads have been taken into account.
2) Supplies/Gold Required - The supplies or coins required seem to be almost even through all the levels of any given manufactory (e.g., production of 1 Gem in a Level 1 Manufactory requires 32 Supplies and 320 Gold, while production of 1 Gem in a Level 18 Manufactory requires 32.0087336 Supplies and 320.0087336 Gold, and production of 1 Gem in a Level 28 manufactory requires 31.2 Supplies and 312 Gold), so calculating the cost seemed unnecessary. To produce, say, 850 gems, will cost you almost exactly the same amount in coins and supplies whether you do it with one level 30 manufactory, or with 4 level 17s.

Since my main goal in putting this together was to see if, as I suspected, a larger number of low level manufactories could provide equal or greater production in the same space than a smaller number of large manufactories, I also played around with seeing what effect this might have on the use of MMs. As far as I can tell, in at least some places, the lower level manufactories are so much more efficient than their larger counterparts, that using the same number of MMs in each scenario resulted in an almost equal production amount between the two manufactory sets, with a slight edge even going to the smaller manufactories occasionally. For military players who only use one or two of each factory, this may not come out equal. As a caterer who needs greater production myself, the numbers certainly seem to lie in favour of the smaller manufactories; I'm only going to use 3 or 4 MMs on an average day anyway, so if I get the same production whether I have 4 large manufactories that I sparkle, or 12 smaller ones (of which I sparkle 4), I'd rather take the smaller footprints that give me more ability to rearrange the city, and weren't as expensive to upgrade.

 

[Smooper]

I am not much of a technical Elvenar player but it seems I should upgrade my manufactory to the first upgrade available in the new chapter and leave it. Then 3 chapters later I should upgrade it 4 times to get it to the most efficient level of manufactory again. Does this seem right?

 

[MaidenFair]

@Smooper, it would depend somewhat on what chapter(s) you are in and which manufactory you're looking at, but there is definitely a general trend like you have described. It seems that until you get to the very end chapters, it is almost always better to have the manufactories most of an available set of upgrades, which is most often four levels, behind where they could be (for pure production purposes; I am aware that some elements in the later game are based on working population, required culture, et al., and might be impeded by having manufactories not fully upgraded).

 

[StarLoad]

Can't wait to see the Elf version, good job and thanks

Ed

 

[Smooper]

@MaidenFair After looking at this more closely, I feel more confused. The last column in your spreadsheet is important to me and I will bring an example. I have raised all my steel manufactories to lvl 20 as I am in Woodelves. These manufactories take 18 places and will for the next 3 upgrades. You show that the steel per square goes down every time I upgrade even though I know it will produce more per hour and takes the same amount of room for each upgrade in the woodelves. Can you explain what is happening here?

 

[MaidenFair]

@Smooper, absolutely! It's because this spreadsheet is also taking into account the population required by the manufactory, as well as the culture required by both the manufactory itself and the residence(s) needed to support it. If you happened to be perfectly balanced at 0 population available and 0 culture available, even if you had a beautiful open space of exactly the right size needed for the base of a manufactory (in this case, 3x6), you wouldn't be able to place it, because it requires population and culture. Since you can't build or upgrade a manufactory without population and culture, the space those take up must necessarily also be factored in to the production capabilities of the manufactory. Where it gets tricky is that as the manufactories are upgraded, they require more people to operate them but the amount of steel produced does not increase proportionally to the population required.

So for example, with your steel manufactories, a level 20 requires 2160 people and makes 231 steel (I'm using the base production number here and ignoring boost percentages to make math easy), meaning it takes 9.4 people to make 1 steel, while a level 23 requires 4252 people and makes 290 steel, meaning it takes 14.7 people to make 1 steel. So the problem is that the level 23 requires just about twice the population, while producing a lot less than twice the amount of steel. It's a similar story with the culture requirements. Therefore, the level 23s need more "support" squares in the form of population and culture than the 20s do, hence they are, overall, less efficient.

Does that make more sense?

 

[Mykan]

If anyone has questions about the data or layout, I'll be happy to explain. I did my best to ensure accuracy but with so much data to copy and collate, it seems almost inevitable that I will have made a mistake in data entry somewhere, so if you see any errors, please let me know so that I can update the sheets appropriately.

I have run my own calculations (not a google sheet otherwise I could share) and my numbers that I checked to yours are almost identical, couple decimal places vary. We differ on the cultural building we have chosen whch seems to be the reason for the variances. I tended to favour buildings of size 4 or above so culture buffs could be used where you have some 1 and 3 size culture options that are more efficient on a per square basis, but we still come up with the same factory size in each chapter.

I also ran my numbers for magic residences and that is where things can change (assuming I didn't make any mistakes) depending on the chapter. Magic residences can sometimes mean upgrading further is worth while but sometimes it just makes the loweer factory even better.

If a town is parking or going very slow it also needs to be mindful of where in the chapter they are and what culture options are available to them. For example if you park after the advanced scout tech of a new chapter you will need to consider the best building for the earlier chapter as new culture and residence options aren't open to you. Again magic residences can influence this as they are accessible immediately for most chapters.

I played around with armory training size as well and it follows a similar pattern.

 

[Smooper]

@MaidenFair So steel produced per square doesn't mean per square of the steel manufactory but of all the buildings that will be needed to support that manufactory. I think I got it. Since, as I said in my first post, I am not much of a technical player, this probably is not that important to me. If I have population, I will build it; if I don't I won't. I have a player in my fellowship who builds the residences and manufactories and then gets rid of the residences for the rest of the round and runs at a negative pop. Is this a good strategy and why is this even possible?

 

[Gkyr]

So for example, with your steel manufactories, a level 20 requires 2160 people and makes 231 steel (I'm using the base production number here and ignoring boost percentages to make math easy), meaning it takes...

Your valuable contribution is masterful and appreciated. If any new players are reading this and planning to apply these analyses from the start of their Elvenar gaming, it must be said that they should apply these principles to the establishment of their boosted manufactories and not to unboosted ones.

 

[Zoof]

I have a player in my fellowship who builds the residences and manufactories and then gets rid of the residences for the rest of the round and runs at a negative pop. Is this a good strategy and why is this even possible?

It's an excellent strategy in order to reuse space that would otherwise be occupied by space-consuming residences, so long as you can swing back and solve the negative population problem when the time comes to upgrade anything that needs it. Which depending on your chapter, might not happen for a very long time.

It's possible because the game only checks population or culture when you build a building that requires it, and at no other time. Buildings that don't require either resource (including AWs, event buildings, guest settlements, and maybe others) can be placed whether or not everyone is homeless or the city's dirtier than Detroit.

If you don't mind going a few rounds of Elf Game Tetris every so often (or even better, enjoy it), it's fairly effective at boosting that city's capabilities in things like Spire and Tournament. It just takes a bit of work to square things away and still be able to use what you built. And to undo things if you absolutely have to do upgrades to progress the game.

 

[StarLoad]

Depending on your chapter you might also run into Coin issues that will require instants to refill ... But in lower chapters that might be a big issue, in later chapters it might be doable with teleports.


Ed

 

[BrinDarby]

So for example, with your steel manufactories, a level 20 requires 2160 people and makes 231 steel (I'm using the base production number here and ignoring boost percentages to make math easy), meaning it takes 9.4 people to make 1 steel, while a level 23 requires 4252 people and makes 290 steel, meaning it takes 14.7 people to make 1 steel. So the problem is that the level 23 requires just about twice the population, while producing a lot less than twice the amount of steel. It's a similar story with the culture requirements. Therefore, the level 23s need more "support" squares in the form of population and culture than the 20s do, hence they are, overall, less efficient.

My Q is .... did you explore the efficiency difference between say, ohhhh ...
(2) lvl 20 vs. (1) lvl 23 , because thats the same amt of ppl ??
Space is doubbled just for the manu and is culture more/less ,
since population is a constant between the two.

my Q might be, what is the metric for defining efficiency.

 

[Ihrlaen]

2) Supplies/Gold Required - The supplies or coins required seem to be almost even through all the levels of any given manufactory (e.g., production of 1 Gem in a Level 1 Manufactory requires 32 Supplies and 320 Gold, while production of 1 Gem in a Level 18 Manufactory requires 32.0087336 Supplies and 320.0087336 Gold, and production of 1 Gem in a Level 28 manufactory requires 31.2 Supplies and 312 Gold), so calculating the cost seemed unnecessary. To produce, say, 850 gems, will cost you almost exactly the same amount in coins and supplies whether you do it with one level 30 manufactory, or with 4 level 17s.

Quick question: Are Divine Seeds costs scaling linearly as well?

 

[MaidenFair]

I tended to favour buildings of size 4 or above so culture buffs could be used where you have some 1 and 3 size culture options that are more efficient on a per square basis, but we still come up with the same factory size in each chapter.

I actually ran the numbers initially with all the smallest culture buildings in each chapter, on the assumption that no one is going to try to plop a 5x6 pure culture building into their city, no matter how good it is on a per-square basis. XD I hadn't thought about the availability of neighbourly help. It's good to hear that the culture numbers chosen don't affect the outcome very much either way.

I also ran my numbers for magic residences and that is where things can change (assuming I didn't make any mistakes) depending on the chapter. Magic residences can sometimes mean upgrading further is worth while but sometimes it just makes the loweer factory even better.

My city is almost entirely blue, thanks to the Spire (one lonely little Dwarves residence left that will be teleported as soon as I feel like sparing the blueprints for upgrades on the others), so I should run those numbers too. It's interesting that the outcome is variable.

I played around with armory training size as well and it follows a similar pattern.

Good to know! As a caterer, I have no armories currently but if I ever need them, I'll be sure to check the math before making assumptions.

 

[MaidenFair]

Since, as I said in my first post, I am not much of a technical player, this probably is not that important to me. If I have population, I will build it; if I don't I won't.

Sounds perfect. Whatever playstyle you enjoy is the right one.

I have a player in my fellowship who builds the residences and manufactories and then gets rid of the residences for the rest of the round and runs at a negative pop. Is this a good strategy and why is this even possible?

@Zoof and @ed1960 hit on the main points, I think. I haven't tried this playstyle myself, because I wouldn't want to get attached to the space and then discover I needed to delete some three expansions worth of buildings to fit in residences again just to change chapter.

The main cons I can see to this strategy would be: 1) scarcity of coins, as Ed mentioned, 2) the impossibility of making spur-of-the-moment decisions, like, "Oh, I'll just throw up a few low-level workshops to get through these quests faster", 3) frontloading costs of upgrades to a chapter; you can't spread out upgrades over a few weeks or months, they have to be all lumped together as fast as possible, which could strain resource levels, and 4) requires reworking your city on a massive scale relatively often (unless you're moving through chapters very slowly, which is often the case in later chapters), thereby also necessitating the possession of lots and lots of teleport spells.

I suppose the main pro is more space to play with...I have no idea if you can pull a regular building like a manufactory out of inventory with negative pop, since summoning things from inventory doesn't use a builder. If so, then you could have a city full of manufactories or armories without worrying about the population or culture to support them, which could be very efficient. I'm sure someone who has tried this would have a long list of other pros that I am not aware of.

 

[Zoof]

I suppose the main pro is more space to play with...I have no idea if you can pull a regular building like a manufactory out of inventory with negative pop, since summoning things from inventory doesn't use a builder. If so, then you could have a city full of manufactories or armories without worrying about the population or culture to support them, which could be very efficient. I'm sure someone who has tried this would have a long list of other pros that I am not aware of.

Popping a stored building (that would normally be obtained from the Buildings menu) from inventory actually does check if pop/culture is both required and met in the same manner as though you were actually building it. Such buildings also actually require a builder for the 10 seconds it takes to pop out in inventory. If all your builders are busy, you will not be allowed to place that fancy new Magic Residence or Workshop, or whatever other normal building, even if you had infinite culture and population. The drop grid will appear but it will stay red no matter how much empty space there is.

The problem isn't noticed because the vast majority of summons are truly magical; they don't require a builder. It becomes noticed when you do a fair bit of teleporting and summoning more mundane things, or are just unlucky the first time you get that fancy Magic Whatchamacallit from Spire and can't place it because everyone's off busy elsewhere.

 

[MaidenFair]

My Q is .... did you explore the efficiency difference between say, ohhhh ...
(2) lvl 20 vs. (1) lvl 23 , because thats the same amt of ppl ??
Space is doubbled just for the manu and is culture more/less ,
since population is a constant between the two.

my Q might be, what is the metric for defining efficiency.

I did not put that math into the spreadsheet because there are wayyy too many permutations. I have done that math in many scenarios, however, including multiple times in my own city. In the above-mentioned scenario, the level 20s would take up a combined 92.73 squares with a 3-hr production of 462, while the level 23 only uses 61.86 squares (67% of the area) but also only produces 290 steel (63% of the production). So it would depend whether you are shorter on space or on resources.

It's probably quite rare that a player in Woodelves only has one T1 manufactory, however, so if the numbers are run with a larger set of manufactories, the earlier levels gain more efficiency. Also to take into account in this scenario, neither of those is the most efficient level for Woodelves, and both become available in the same set of upgrades, so there is less difference between them than between upgrades that became available in different chapters.

If I run two manufactories of the most efficient level, 16, against the 23, they come out ahead: in the pure math, 2 lvl 16s would take 56.46 squares (or in other words, 30 squares for the manufactories themselves, 2032 people, 506 culture) and produce 298 steel. 1 lvl 23, as above, uses 61.86 squares (only 18 squares for the manufactory footprint, but 3317 people and 967 culture; more space) and produces 290 steel (less production).
In an actual city, where we can't deal with fractions of squares or buildings, that means the 16s need two (fully upgraded Woodelves) residences at 10 squares apiece, while the 23 needs 3 residences (an extra ten squares). The 16s need three Mana Fountains (9 squares total) to cover the culture for both residences and the manufactory, while the 23 needs 4 (12 squares). This makes for a total of 30(15x2 manufactories)+20(10x2 residences)+9(3x3 fountains)=59 squares for the 16s, and 18(x1 manufactory)+30(10x3 residences)+12(3x4 fountains)=60 squares for the 23. 1 more square and 8 less base production per 3-hr run, comparing the minimum number of factories. If you start adding those up, assuming the average player has 3 or 4 manufactories, realizing the production number will have a boost applied, and so on...it becomes a definite space saver to use more lower-level manufactories (in this case specifically, and in many other cases).

So, the base metric of the spreadsheet was resource production per square. Of course the application of the math will vary for each individual, depending on what they need most/are in short supply of. If someone always has plenty of space but never enough people, they could focus on people per production, for example. I wanted to keep my city small, have plenty of population available in case of something I hadn't foreseen, not have a visual space full of idential residences and workshops like most developed cities seemed to have, and yet remain productive enough to have good tourney scores and top the Spire. All of those boxes were ticked for me by using more of the earlier manufactories instead of continuing to upgrade.

When I realised that the manufactories were starting to require more and more population but not increasing production proportionately, I started investigating with the question of "how can I get more production in the same amount of space?" If you start with a different question, "how can I get more production out of the same population?" or "how few manufactories can I get by with?" then the answer to "what is efficient for me right now" may vary.

 

[MaidenFair]

Quick question: Are Divine Seeds costs scaling linearly as well?

I don't think so. I didn't check all the manufactories but I tried one of each tier. It appears that the Tier 1 rises more sharply in cost of Divine Seeds per good manufactured before plateauing for a few levels and then rising again, with one slight drop from the penultimate to the ultimate level. Tiers 2 and 3 rose less sharply initially but did steadily become more expensive until again a slight drop at the last level.

Representative numbers for Marble - level 24, 1290 seeds divided by 350 Moonstone equals 3.69 Seeds to produce 1 Moonstone
level 29, 2540/635=4 Seeds/Moonstone
level 30, 2760/690=4 Seeds/Moonstone
level 33, 9600/1990=4.82 Seeds/ Moonstone

As you can see, it's 30% more expensive in terms of Seeds at level 33 than at level 24. Silk is 32% more expensive and Gems 31%, so I would guess all the manufactories follow the same pattern.

 

[Mykan]

Good to know! As a caterer, I have no armories currently but if I ever need them, I'll be sure to check the math before making assumptions.

Once you hit chapter 8 or 222 provinces you will need them. There are a orc nests that can be crafted (maybe build some in chapter 7 and plan to upgrade with RR) as an alternative. I have not run the numbers for orc production, only training size. The interesting thing with orc production is an orc nest provide culture as well so even if an armory were better it may still come down to how close and the value of the culture if it saves using other spots for culture.

It's probably quite rare that a player in Woodelves only has one T1 manufactory

Some players run with a lot of event/evolving buildings that provide goods so it is not too unrealistic. However, personally they can run their own math on the myriad of other options to see if they are worth it from an efficiency point of view.

Have you ever done this for supplies? I have all the tables I need but just haven't had the motivation to do it yet. Apparently I did run a table for it, just never studied it obviously.