Research costs

[Seravy]

Simply put, what we are missing might be something similar to Sawmill and Marketplace but in research and power. Albeit that would impact the common phase which is not needed. So idk. But wizard's and sage's guilds are clearly too expensive to get enough of those in the uncommon phase and univ/parthenon is not enough to support it (and not every race is getting those anyway).

Any ideas? We'd need to find a way to let players get ~200 more total power(or research) by turn 120.

Unless I'm actually underestimating incomes but considering I looked at 3 saves, that is not very likely now.

(oh, btw the calculations are missing the maintenance costs. So in reality the income needed is higher, even in the common phase. We might really need a "basic power income" building.

[Seravy]

Thinking about it some more...the common phase works slightly better because the 20 income from the fortress can make up for the lack of basic power building. But this doesn't last for the uncommon phase because unlike marketplaces and sawmills, you don't get fortresses in your other cities

...so yeah, I think we should reposition the Shrine into a "basic production" building, maybe do the same to the Parthenon and University/Library to some extent. Alchemist guild can be a fairly decent "base power" building as well, though that would remove the choice involved for getting or not getting the magic weapons so it likely shouldn't do that.

Let's see...

Shrine, cost 120, produces 6 power.
Library, cost 80, produces 6 research. Maybe remove it as a requirement from the Alchemist guild, maybe even wizard's guild.
University, cost 140, produces 6 research
Parthenon, cost 200, produces 6 power

This would raise power/research by 11/city, or 165 for 15 cities. Not 100% adequate to cover the missing amount but comes close enough?

Problem? Shrine/Parthenon produces units. If they are also a basic building, that's not very good.
Ideally, we'd want a new basic power production building, but that would require removing something which I don't think we can. Albeit...do Ship Yards have any real meaning?
...continued in the "races, units, buildings" thread.

[Suriname]

Problem: the AI bonus is compounded by these buildings, that AIs can buy sooner/ in more cities.

Solution: implement a concurrent reduction in AI cheating bonus, aiming at the same amount at wardec turn.

[Nelphine]

Hmm. I would have said 80 skill and 160 skill for uncommon and rare. I also wouldn't expect one combat per turn for both those phases. I would expect more like 2/3 to 3/4 per turn, although I'm not sure which phase has more. Probably rare.


So in general, I expect the phases to go quicker, and each phase to be slightly weaker than you do. I also expect less combat, but my guess is that I'm wrong on that and that actually dictates my picks, so I can control combat because I don't enjoy combat as much.

[Seravy]

We can work on the finer details later. First we need to fix the likely 40-50% missing income for the uncommon phase which starts by gathering actual income data. I wish I had kept all my old save files. There is no way I can play multiple games up to turn 180 in a reasonable time and I sure want to do this update this week.

[Seravy]

Do you think letting the AI play on Normal and checking their income would be a fair approximation of a player's? I feel it would be a bit too low as the AI dumb. But Advanced might have too much advantage for them...

[Nelphine]

It looks like all your math is still based on rp = sp + mp.

As much as I'm not sure about rp = sp = mp, I DO think that's closer to accurate than rp = sp + mp. 

Looking at one number, common total research, if we use rp = sp = mp, we get that the average common should cost 114 mana. (Based on both of us agreeing common tier is 6 commons plus 2 uncommons).

Seeing as we've both said previously we think most of the game is pretty awesome and balanced, and this number is almost exactly reality, I'm inclined to continue looking at using this number.

Applying that to uncommon total income required of 581, we get that the actual income required then becomes 436.

That's a huge chunk of your mythical 'power building' you're interested in, and I'm inclined to think that's actually a lot closer to reality.

[Seravy]

Well 436 is still like 150-200 more than what I earned in the games I have checked.
More importantly, 114 mana for a common creature on average? That way further from reality than 50, even if I include the raised cost of nagas and ghouls (both of which are expermential changes).

But here is the problem. If a common creature costs 113, we get Turns=150 for the common phase. That would be horrible. (Unless you mean 113 RP cost instead of 113 mana. But we can't calculate the RP cost until all tiers are considered because we agreed the tiers mix rarities. So we'd need a number of RP total for all 3 tiers to assign a cost for each tier individually.)

[Seravy]

Anyway, let's try to get that accuracy up since we can't do anything else while waiting on data and deciding on income changes.

Average cost of common creatures currently : 56.25. Note that this is unweighted. However, we were setting goals based on a generic creature, not any specific one. I don't think we should aim for 40 ghouls or nagas, nor skeletons. The 40 is for something in between like bears. So I think this matches my expectations well enough, and I'm willing to sustain the goal of ~40 creatures (or equivalent spells, remember you might spend all of this on wraith forms or flame blades or earth lores if that's your preference) with this number.

Average cost of uncommon creatures : 150. Exactly, to my surprise.
Again unweighted. I'm a bit unsure what to do with those that are halfway between common and uncommon, as they pull the average down here, but would pull it up in commons. I definitely want to keep their research low enough to let players get it early, so these might see use in the common phase - but even if they do, they still keep getting used in the uncommon too, and they do replace commons fairly well - the per unit cost is higher but so is the per unit military value so you only need fewer of them for the same result. So I think it's best to keep these in the uncommon group. Ultimately we are likely talking about a +/- 5 difference in the averages either way, not relevant.

Average cost of rare creatures : 318

Average cost of very rare creature : 447, but Summon Demon is dragging this down a lot. I say we should exclude it, it clearly does not fill the same role as a Colossus.
Average cost of very rare creature, improved : 480.

Not sure if we can even use this for anything, as there is no next tier to base any calculations on here.
Assuming we agree on the goals, the bold are the values we need to set by altering income sources and research costs.

Skill goals
Common 20 to 40, average 30
Uncommon 40 to 100, average 70, assume 21 from amplifying towers
Rare 101 to 200, average 150, assume 49 from amplifying towers (might be too conservative?)

Common calculation
With this and costs known, we can pick a Turns and preferred amount pair for these numbers.
Turns = Creatures / 56.3 * 30. Turns = 1.88 * Creatures.
So we have 75 turns and 40 creatures. As my data request was for turn 72 (6 years), we get 38.3 creatures with that which is close enough. (I rather define things in full years for simplicity)
That puts MP at 2160, SP at 1200. So RP = 3360 for this phase. Income has to be 93.
We ignored combat and maintenance costs for this phase and assumed they pay for themselves through loot or additional cities. Whether that's reasonable or not is actually questionable.
If we also include a small combat and maintenance spending of 500 mana then RP = 3860 and income is 107.
If we include a more severe maintenance spending of 20/turn plus a conservative 400 on combat spells, then RP = 5200 and income is 144.
Honestly, I'm unsure which is more realistic. Combat spending, if done well, should pay for itself, but we are including treasure and cities in the income data we gather and they pay for themselves through that. So we probably need to include these costs? (RP = 5200 looks ugly though. That's like 400 for a common and over 1k for an uncommon depending on the number of spells...but bringing up income to this amount might be the greater problem.)

Uncommon calculation
Assuming 1/2 battles a turn, and a maintenance of 30/turn (10 creatures, yes we used more spells but they got used up in battles or were spells with no maintenance like Water Walking)

Turns = Creatures / 150 * 70 for this phase, so turns = 2.142 * Creatures.
If we want 6 years again, that's 33 creatures. I think that's a fine goal.

SP = 4641
Overland MP = 5040 (70 * turns)
Combat MP = 2520
Upkeep MP = 2160
Total MP = 9720

Thus, RP = 14361 and Income = 399

Rare calculation
Let's stick to 1 combat/turn and 100 maintenance a turn here. It's probably overestimating combat and underestimating maintenance but it's close enough.

Turns = Creatures /318 * 150 = 2.12 * Creatures.
Same as above, 6 years, 33 creatures. Perfect.

SP = 16560
Overland MP = 10800
Combat MP = 10800
Maintenance MP = 7200
Total MP = 28800

RP = 45360, Income = 1260

[Nelphine]

No, 114 for average research cost of a common spell. I was using this to explain why I think it matches current numbers, which we generally agree are pretty good, in order to explain why I think sp = rp = mp is more accurate than rp = mp + sp.