PDA

View Full Version : Some Basic Experiments with Economy



Tamur
10-05-2004, 18:36
Over the last couple of days I've used my play time to start and keep detailed track of a Julii campaign's economic status. I kept track of farm, tax, trade, and admin income, as well as population, the management score of the governor, start and completion dates of buildings (especially and including ports, roads, traders, markets, etc).

After running this campaign for fourteen years (through 256 BC), I entered all the data into a spreadsheet, graphed it both across cities and comparatively inside cities.

What I found was both expected and unexpected...

1. Expected: the management score of the governor has a huge impact on all sectors of income - farm, tax, trade, and admin.

For example, I pulled a 4-Management governor out of Segesta to go relieve a seige on Mediolanium, and Segesta's income dropped by 10% within two turns.

2. Unexpected: building traders and trader upgrades in neighbouring provinces ups trade income.

Out of all four trader completions and market upgrades I was able to do, the province building the upgrade or adding the trader did not see a major increase in trade income. Instead, a single neighbouring province saw a spike in their trade income.

For example, in the summer of 267 I finished a trader in Ariminum. Ariminum's income from trade went from 580 to 586 (+6). In the same turn, Arretium's trade income went from 344 to 544 (+200!), and remained at that level or greater for the rest of the game.

3. Unexpected: tax income does not parallel population, total income, or any combination of population and income.

There are other factors at work here. I kept track of governor management skill, and factored in all sorts of combinations of management, income and population, but I've exhausted all the possible combinations without finding a formula that defines tax income. This will need additional info to be "cracked".

Dorkus
10-05-2004, 18:42
Interesting stuff. Thanks for the tests.

Only problem is I don't think you can just click end turn and compare the results from the previous turn.

You need to save, build, end turn.

Reload, don't build, end turn.

Compare those results.

There's natural variance in all the figures. You'll confuse those figures with the effects of the building unless you use the method I describe. (actually, even the method i describe may not work, since the buildings you build seem to affect the "seed" -- e.g. sometimes I build a building and the outcomes of auto battles are different than when I didn't)

RedKnight
10-18-2004, 18:03
Cool work, Tamur. I'm trying to decide how much it's worth it to move my good managers around. You say that they increase income from all sectors. But can you say how much? I can't even find a reference in the manual as to just how much it might be. Did you find anything conclusive? Is one management point worth 5%, 10%, or what? But it sounds like it was quite variable. Can you at least state a range?

Thanks if you can help! ~:cheers:

Tamur
10-18-2004, 19:05
therother has been doing a lot of (far more precise) analysis since I wrote this post. You might do a search for economics-related posts he's started in the last 2 weeks or so, much solid, detailed information available in those threads.

Tamur
10-18-2004, 19:55
edit: double-post removed

Quietus
10-18-2004, 21:19
Hey Tamur,

Nice start there! I'll try to chip in if I can ~:) I'll have to do some test myself later. A couple of quick comments though:

1. Isn't the effect instantaneous? (after you pull out a governor?)
2. How about comparing goods traded from one province to the other? There's a trade detail screen there that breaks down Imports and Exports so I guess you can track the trade income there.
3. I think you have to use un-governored towns and and normal tax just to get the basic data, because governors introduce alot of additional factors there.

Also do you mean to say, you want to get an equation involving the tax income? I'd like to find a trend for that too!

Right now, I've looked at corruption like therother, however, one thing I'm frustrated with is the lack of equation that I can tie with the in-game data. For example: 25% distance always get ~ 16% corruption. 20% distance get ~14% etc but putting a legible equation, I've no such luck.

Anyway, drop some more info once you find some and hopefully, we all can cull resources. ~:)

therother
10-19-2004, 01:45
1. Isn't the effect instantaneous? (after you pull out a governor?) In the first instance, yes, but there may be additional effects of having a good governor in situ for a significant period. I've not found any, admittedly, but that doesn't mean they don't exist.


2. How about comparing goods traded from one province to the other? There's a trade detail screen there that breaks down Imports and Exports so I guess you can track the trade income there.) Problem is that it's difficult to calculate what each good makes by itself, as the trade description screen just gives you totals.


3. I think you have to use un-governored towns and and normal tax just to get the basic data, because governors introduce alot of additional factors there.Yes, this reductionist method seems to be the best at the moment, as there appears to be very few (if any) compounded effects. That is to say that 10% to tax income means 10% to the base income, not 10% on top of the current tax income.


Also do you mean to say, you want to g et an equation involving the tax income? I'd like to find a trend for that too!Well this thread (https://forums.totalwar.org/vb/showthread.php?t=37831) should help you get started.



Right now, I've looked at corruption like therother, however, one thing I'm frustrated with is the lack of equation that I can tie with the in-game data. For example: 25% distance always get ~ 16% corruption. 20% distance get ~14% etc but putting a legible equation, I've no such luck.Really? Could you send me the data? A word of warning though: make sure you have no Governors with Law traits/characters or Law buildings. Also make sure you take the total income, not the net income the game displays.

The equation I've derived for corruption is not easy to remember, but as a rule of thumb, there is no corruption for 16 squares around the capital, then corruption increases by 8% per 10 squares to ~65% (or ~100 squares) where it is capped.

ToranagaSama
10-19-2004, 15:46
Very interesting post, thanks!


2. Unexpected: building traders and trader upgrades in neighbouring provinces ups trade income.

Out of all four trader completions and market upgrades I was able to do, the province building the upgrade or adding the trader did not see a major increase in trade income. Instead, a single neighbouring province saw a spike in their trade income.

For example, in the summer of 267 I finished a trader in Ariminum. Ariminum's income from trade went from 580 to 586 (+6). In the same turn, Arretium's trade income went from 344 to 544 (+200!), and remained at that level or greater for the rest of the game.

This is *extremely* interesting.

I'm wondering what effect Roads are having in your results.
Did both Ariminum and Arretium have the same level of Roads? Just wondering if there might be some unaccounted for factor resulting in the descrepency.

Something that also *might* be accountable, is whether they both Provinces built Traders and the Trader Upgrades *simultaneously*. Specific to the example above, could it be that Arrretium had a Trader build for a longer period of time than Ariminum, which might account for the discrepency?

Just curious, great info to ponder, thanks again, I'm going to watch this very closely from now on.

[EDIT:] Also, were the Roads build *simultaneously*?

Oleander Ardens
10-19-2004, 16:04
I might join your efforts soon, Tamur

Will keep track of my next campaign, possible with the punic trading power to increase our data...

Let us collect the factors which possibly influence the trade:

Quantity and Quality of the trading goods of the trading partners
Level of the market buildings of the T.P
Level of the port and the T.P
Distance of the trading partners, both on the sea and the land
Quality of the roads and of the T.P
Trade Agreement
Taxes of the trading partners
Population of the T.P
Managment skill of the Governor

So a load of possible factors, so a cheat to which allows you to control also the cities of the TP might be helpful...


In the meantime we could formulate a hypothesis, I'm working on it..

OA

maestro
10-19-2004, 16:33
Just a couple of things from memory (I'm at work)....

Right at the beginning of my latest game (Vh/M Juliiiiiii) I ran a couple of tests to see the effects of the various buildings.... from what I remember.

The first level of trader / market thingy gave a 10% boost to all trade, be it land or sea based.

The first road upgrade (from basic to paved I think) literally DOUBLED land based trade income!

Adding a market meant having more goods appear in the import section of trade and made no impact on sea based trade (don't know if this would impact on sea after on with more trade rights granted)

Adding a port made no difference to income levels but it added sea imports and exports which provided additional income.


edit - I think it's really important that we get this daa between us. To know exactly what impact the various economic buildings have on income will be invaluable for a VH campaign with difficult factions...... ~:handball:

Tamur
10-19-2004, 17:43
Did both Ariminum and Arretium have the same level of Roads?

Blast! :wall: I've been searching my office for the papers I wrote everything down on, but no luck, it's gone. I'll have to re-run this experiment and get everything down again where I can find it.

It sure SEEMS like Ariminum and Arretium both had Paved Roads until something like 264 -- this was before I knew about how to start a campaign with all the cash in the universe, so progression was slow.

Sigh. Guess I'll have to play more...

~D

Quietus
10-20-2004, 03:40
Here you go therother. No governors, no law buildings, no rebels/enemies (just to be safe), normal tax. I put down trade, taxes, farms and mining income as well as other variables, however, i'll just type the total income since it will be messy to look at.

Prov. (TotIncome)(Corruption)(%Distance)(%Corruption)

Lug (2034)(203)(15)(10)
Mas (3259)(289)(15)(8.9)
Luv (1306)(104)(15)(8.0)
Car (2857)(257)(15)(9)
Nar (3994)(559)(20)(14)
Aq (1828)(255)(20)(14)
Tri (745)(119)(25)(15.8)
Mog (1334)(226)(25)(17)
Ales (2348)(375)(25)(16)
Cart (4767)( *)(25)(16)
Osc (3893)(1012)(35)(26)
Tha (3705)(963)(35)(26)
Iaz (1729)(449)(40)(26)
Por (2117)(592)(40)(28)
Cir (3713)(1076)(40)(29)
Get (2322)(928)(55)(40)
Num (2970)(946)(55)(38)

I didn't include other provinces with rebels, as well as province I can't compare with "%distance" wise. Total income, of course, is income before any deduction (devastation, corruption, army, generals/agents).

*I missed writing number down. I'll check the game again since this is from paper.

therother
10-20-2004, 03:58
Here you go therother.Thanks Quietus. Those numbers look okay, with one proviso - what is this per cent distance? My equation uses squares. For instance, if my formula is right, then Lug should be ~28 squares away from your capital. You can get coords using the show_cursorstat RomeShell command. It's then a case of simple trig to get the distance.

EDIT:By %distance, do you mean the PO penalty for being away from the capital?

therother
10-20-2004, 04:10
Actually your numbers are quite cool.

For instance, I can work out the number of squares your city is away from the capital:

Distance = (%Corr+12.26)/0.7935 = 28.05

And from there, regenerate your public order % (assuming that is what it is):

PO% = 1.0537*Distance - 13.917 = 15.866 = 15%

therother
10-20-2004, 04:15
Just tried it for Por (2117)(592)(40)(28)

Distance = 50.73724008
PO% = 39.54482987 = 40%

Makes me feel a lot happier to see all this work out from other numbers generated from a completely separate game. I've always been a little worried about all the dubious things I've been doing to mine...

therother
10-20-2004, 06:27
All right, so you want a equation for % corruption in terms of % Distance to Capital:

Corruption % ~= ((0.7935*%DtC)-1.2248)/1.0537

Quietus
10-20-2004, 11:06
Yeah, I didn't know where else to get a distance value to study corruption (hehe). I'm pretty sure I'll use your finely generated equations! However, what I really wanted is to be able to read all those Icons and make formulas out of those. Why? Because, knowing how CA uses the Icons can potentially spill into reading the other icons, such as economic ones.

One other example is unrest. I always get 15% unrest, but where does CA get this value? :dizzy2: Then, I've read what JeROME said of the culture penalties, however, he was fuzzy about unrest. So, I will continue to do tests on unrest.

Anyway, I'm not able to do economic tests yet (aside from how rebels affected trades affected sea exports and land trade but not sea imports, in the old trade disruption thread). Will start testing on it soon.

Keep on testing everyone! ~:)

:charge:

therother
10-20-2004, 11:10
One other example is unrest. I always get 15% unrest, but where does CA get this value? :dizzy2: Then, I've read what JeROME said of the culture penalties, however, he was fuzzy about unrest. So, I will continue to do tests on unrest.The only other thing I can think of is checking your Governor's traits/retinue. Some of them cause unrest.

RedKnight
10-20-2004, 20:51
tr, there's been a bunch of numbers flying back and forth - if you could clarify -

for your "corruption percent" equation,

the percent corruption is a percent of total income for the city, right? (everything on the plus line of income including e.g. administratiive, and nothing on the negative side?)

what was your correlation coefficient for the regression line?

in summary, it's very clear that it's only related to distance, yes? (barring governor's corruption retinue or traits)

thanks!! ~:cheers:

therother
10-20-2004, 21:38
for your "corruption percent" equation,

the percent corruption is a percent of total income for the city, right? (everything on the plus line of income including e.g. administratiive, and nothing on the negative side?)Hmm, I've never had a Governor in the city to see if added anything. Should be quite straightforward to take a raw Governor a with either just management ability, or one that you could give a management endowing retinue character to, and have no other effects on income/corruption.

But yes, all the other incomes added together, with no deductions - gross income.


what was your correlation coefficient for the regression line?All the coefficients are on their respective graphs.


in summary, it's very clear that it's only related to distance, yes? (barring governor's corruption retinue or traits)AFAICT, that's is indeed the deal.