I have gone through and collected data on every level for the number of Armies required to complete, the total Mercenaries provided, and the number of Territories. (Note: Base Mercenaries are 20% of the Total Armies required. Additional Mercenaries Advancement will increase this amount).

Total Armies Required are calculated as:
(Total Armies Produced) - (Current Armies) + (Joint Strike) + (Saved by Hospital)

Table is here:
https://docs.google.com/spreadsheets/d/1ug2a2lh7yzCzKuShpFemF1aCEsJw95JJv4OaAKf3d1M/edit#gid=0

Base Mercenaries are not exactly 20% of the total territory cost. Below are the numbers I have.

ID Name Cost Merc #Terr
1 Tutorial 10.4K 2.02K 42
2 Huruey's Castle 23.6M 4.68M 104
3 Floating Rocks 51.4M 10.2M 150
4 War Of The Worlds 108M 21.6M 163
5 Ursa - Luna 188M 37.5M 182
6 Final Earth 181M 36.0M 183
7 Drakemor's World 314M 60.5M 184
8 Peloponnesian War 450M 87.3M 196
9 K-PX 882M 171M 257
10 The Siege of Feldmere 3.42B 678M 395
11 Sengoku - Rise of Oda Nobunaga 8.05B 1.61B 480
12 Copper Creek Castle 9.79B 1.95B 490
13 Geopolitics 13.0B 2.58B 500
14 Breaking Green 21.4B 4.25B 576
15 Far Land 47.9B 9.44B 744
16 Reconquest 1065 45.2B 8.97B 767

I had to back calculate them, and there are also rounding errors that are hard to catch.

For example, a camp with 11.4B may display as 11B, and a camp with 11.5B may display as 12B.

Given the consistency they arrive at 20%, or close to it. It is safe enough to treat them as 20%.

If I'm not mistaken Fizzer did mention somewhere that their numbers are calculated directly based on the number of armies required to complete the level. So I do believe that it is exactly 20% (rounded obviously).

According to my numbers, the ratio is between 19.24% and 19.97%. I agree with Z that we could simply use 20% in most applications.

Oh this is cool. Nice way to see where the big jumps are

A few questions on that to be sure that we speak the same language:

• Don't know how you got to the numbers but is the total number of armies effectively: starting a new level, busting all the fog, and checking the Level Stats for "armies on visible territories"? So, the sum of the costs of all territories? If you did it using FB, where did you get all those FBs from?
  
• With maxed out Additional Mercs, does that mean that one can get 50% of each level purely by mercs (20% + 150% of 20% = 20% + 30% = 50%)?
  
• This combined with JS (assuming JS would work on 100% of the territories) would that mean, that you only had to produce 25% of the total armies and could save/buy the remaining 75% (JS is effectively a 25% discount, more or less)? This would also assume that there were no hospitals at all. Given hospitals the percentage would be even lower.
  
• As a result we would only (!) have to find a source of infinite money (as well as maxed out Additional Mercs and JS) and we would sweep through all the levels?!? ;)
  

* If you have Statistics, you can calculate the total territory cost using the method proposed by Z. I personally keeps track of almost everything when I play a level. As a side effect, I find Statistics almost useless. I do not have Statistics, and I calculate everything myself. A sum over all territories could give us total territory cost.

* Yes.

* Yes, ideally, Merc + JS contribute to 75% of total territory cost. But, I want to mention that the effects of JS and hospital do not stack well. More precisely, effect of hospital is calculated first, and then the effect of JS is calculated.

* This might be exactly what happens to the players who ascends a lot and has large cache modifiers.

Unless I missed something important you can't "calculate" the number of armies on each territory. Unless you write them all down in a spreadsheet and add them all together. So, sure, you can calculate the mercs (by not buying any until you found the last camp and then add them all up) but I think even on the level end screen there isn't a number for "this many armies have you spend in total without all the beneficial effects". Or is this the same as "total armies earned" + "savings by JS" + "savings by hospitals"??? I'm not sure on that one.

I really keep track of almost everything, including the cost of each territory. This may sound as impossible, but it is quite easy to do if we keep track of the details of territories as we conqueror them. By the way, we do not need to wait for conquering all merc camps.

Z's method is much more efficient for the players with Statistics. Please refer to the formula in Z's post. His numbers of total territory cost match with mine surprisingly well.

Okay, you're right, I totally missed this yesterday.
1.) If you always buy all your mercs, then you will know the total number of them given the statistics/level end screen. You don't have to save them for later as the game keeps track of them. If you have some merc camps still populated, you can easily add those remaining mercs to the number in the stats.
2.) I misread the formula in Z's initial post. I thought that given the numbers he presented, I could use this formula to calculate what it takes me to conquer the whole map. While this is also true, the formula was meant as a way to calculate the total number of armies. Therefore, what I assumed could work yesterday was (still my own creation/assumption and) exactly the same formula and therefore the way the numbers were produced. (Except that I omitted the "current armies" part which I was aware of)
3.) You write down the costs of every single territory??? I mean, yeah, sure, I also take notes, for me it's mainly recipes and such. But you manually copy all of the up to 3k territory costs? Are you trying to recreate the map-generation algorithm with that??
PS: Z's numbers should be more accurate, as the numbers you write down are always rounded to some extend. But the game knows the exact numbers and will only round once for the displayed number. Your manual sum can be a bit off sometimes.

Edited 7/19/2021 10:55:10

We can see very accurate territory cost if we select the territory. To get an accurate number of mercenary camp size, we may need to purchase a few times.

I would say that it took me less than 2s per territory if the territory is not important. I am at Europe 1066 AD now, and it only has <1k territories.

The purpose of collecting all the data is to help me (1) make better decisions (2) prepare for the ascension. I would also want to find patterns so that I can collect less data in the future.

Based on the data I had collected, with the help of some codes, I have can determine, with minimal effort, at any moment, (1) the best army camp to upgrade (2) the best mine to upgrade (3) the best recipe to work on (4) the best artifacts to equip. As a side effect, I do not need Statistics, and I can save some AP.

The upshot is that the data allows me to automate many things needed to play Idle efficiently without actively playing it.

Wow, sounds pretty detailed, Functor. I applaud your efforts, as we will all benefit from this.

I'm assuming you're sharing this data with everyone?

Only concern I have is that if the levels are ever regenerated, this all becomes inaccurate and it's a lot of effort rendered irrelevant. That was a problem when WZI first came out; we shared info and map info, etc, but almost every week the levels were regenerated and all the info was immediately wrong.

how about adding a map that is fully fog bustered so that we would know what the map looks like and will know what path to take?

Even if you have collected all 5 possible Fog Busters, you can only defog maps with slightly over 750 territories without having to invest coins. So, if you find someone willing to spend coins for that (which can as well be wasted because it might be outdated if the maps ever change again) or if you are willing to donate the coins for that purpose, this might be possible. All with the asterisk that it's not trivial to screenshot the maps. They are cramped full of information such that there is a lot of overlap (of texts) and a simple image might not be sufficient. Additionally, an image wouldn't tell you the names of the things (mines, recipes, markets, etc). What we needed is an interactive map of some sorts. Where you can zoom and scroll around.
Edit: Ideally, also with cache sizes, initial production of each army camp and stuff.

Edited 7/19/2021 12:29:29

@krinid
The data I collected is all in very ad-hoc format. Can you suggest some info people might be interested in? I can try to format the raw data I had. I also do not really know how to share. For example, I do not have an easy access to google service.

@Shin
It would be nice. I do not have that many fog busters to do that myself though.

@functor
I can share some of the map of Fort Harbor since I revealed all the territories.

I have screenshots of most of the maps ... but they've all changed since they were taken. :(

@functor Are you keeping track of number of mercenaries per camp and cost of mercenaries at each camp? That data could be useful in figuring out average cost per mercenary, which would then help in making decisions on upgrading army camps and hospitals (only pay for a hospital if the increase in armies saved * the number of territories remaining is more than the equivalent cost of mercenaries, only upgrade an army camp if the armies per second * the amount of time you estimate remains in the level is more than the equivalent cost in mercenaries). If you have the base data it shouldn't be too hard for someone to put together a spreadsheet that would take into account mercenary discounts and increase mercenary advancements. Too bad you can't use google services, though, since that's the easiest way to share the work; you could post a spreadsheet (or export to csv and post that) to a file sharing service so that someone else can import it, but it'd be a bummer if people added formulas to make the spreadsheet better and you couldn't benefit from it.

only pay for a hospital if the increase in armies saved * the number of territories remaining is more than the equivalent cost of mercenaries, only upgrade an army camp if the armies per second * the amount of time you estimate remains in the level is more than the equivalent cost in mercenaries

While these calculations in general are a good indicator for whether the upgrade pays off or not, they also have a general flaw that no-one yet has mentioned in any thread as far as I'm aware. This calculation only works if both possible conclusions are viable. If you see that mercs are cheaper that you can buy the mercs and that if the hospitals are cheaper that you can still upgrade the hospital (and the same for army camps). The issue here is that not only money is the limiting factor but also the number of mercs you can still buy. As we've seen now, you have to acquire 80% of the needed armies by other means. If a player now is in the situation that buying mercs is cheaper than upgrading the hospitals or that over the remaining time of this level the army camp upgrade won't pay off fully, BUT the remaining mercs aren't enough to finish the level, THEN I'd argue that upgrading hospitals can save you enough armies to finish the level or that upgrading the army camp earlier would reduce the time the level will take. To make this problem explicit here: There aren't infinite mercs. You can always spend more money because over time you get (theoretically) an infinite amount of money and over time you get an infinite amount of armies (and the time for a level isn't capped and therefore, infinite), but you can only buy as many mercs as there are in one level.