Min-maxing in CivWorld - When You Take a Facebook App Too Seriously

baldfalk

Warlord
Joined
Mar 20, 2011
Messages
242
I'm more or less through my first game; I'm going to list down some insights I've had which I've not seen discussed much elsewhere.

List is subject to change; when I think of something new, or a better way to describe something, I'll change this OP as well to keep it neat.


I) GLOBAL ECONOMY - Capitalism Ho!
(note: As I've only played maybe 90% of one game, this is pretty anecdotal but a lot of it is based on methods that apply to these types of games in general. Feedback/discussion is definitely welcome)
As in all mass-multiplayer environments with a shared auction house like mechanic, how succesful you are depends on how much you can capitalize on other people's stupidity/ignorance/casualness/desperation/whatever you want to call it. Whenever they're close to a new tech or a new food pop target, the market is likely to spike.

That's the obvious part. What is a little deeper is that like all mass multiplayer environments, although there are many players involved, these events are finite. That means, the more crafty capitalists there are in a game, the smaller piece of the pie each one will get. If there are 5 people periodically checking the market, they're all going to nab only a fraction of the amount if only 2 people were doing it.

What's usually important is being extra-watchful early-game. Oftentimes by being faster/more efficient at controlling the market, you essentially 'mark your territory', discouraging other people. Or on the flipside, gauging just how OCD the opposition is, and if it's time to bail on the idea. Since CivW games last a week, I'd wager the first 1-2 days is very often the deciding factor as to who emerges as the dominant capitalistic force late game. (And that's when bubble pop is quite OP as well, I might add)

This is also one of the reasons that the more good players there are in a coordinated group, the more diminishing returns there are in effectiveness in controlling the market. If you're playing with a bunch of friends who are all capitalism-savvy, you can kiss your 500,000 medal goodbye. (Or of course, you could work out stretches of time to cover the market so that you know that someone else is watching when you're at work/sleep/whatever for maximum around-clock coverage)


RESOURCES
The concept is simple. If you have control of the market, you set the prices. If you set the prices, you have control of the market. It's a critical mass of both resources and gold you must reach before it becomes easy. A typical game seems to have 500-500-250-100 for food/production/research/culture, so that's what you should aim for. Keep it as close to those prices as possible, leaning a little to the side of what you believe immediate demand will be like. (e.g. if someone is buying food, try 550-600 instead of 500. If a war is about to pop, let Production drift up to 600-700)

You'll typically be able to sell a bunch at 600-700 and buy at 200-300 pretty regularly, but those prices can get way more ridiculous whenever someone is desperate.

To get a healthy supply/margin of resources you need a hefty population and bank in on trickle income so keep that in mind. There are many other discussions out there on how to position farmers and slingshot through the early population numbers.

A spike upwards in Culture means someone is probably close to popping a great person. Half of the time due to the amount of casuals in the game this means that it will be sold and another commodity will spike upwards from the proceeds. Be on the lookout. Culture's peg can go higher than 100; it's a factor of the price of great persons vs how much culture it takes to pop one. I'd say go with the selling price for the calculations (i.e. 8000 to next GP, 12000 sale price of great person = peg culture at 12000/8000 * 100 = 150)


LUXURIES
If you do emerge as one of the few people who play the market like a hawk, you also can do more with the Luxuries. Forget what you know about supply and demand - stripped to the bone, what controls the prices of luxuries is a simple algorithm that goes up and down, which is THEN modified by the number of units of that luxury that is being held by players. The thing to realize is this - How often or little players buy commodities is seemingly completely independent of how the algorithm chooses when to swap directions. Also, when it starts going up, it's going to go up for a while. When it goes down, it's going to go down for a while. Adhering to the following rules will yield a sick profit:
1) When the algorithm changes from down to up, buy as much quantity until the buying price is 60% of what you've seen the highest luxury price so far (i.e. if you've seen any luxury go up to 400, you'll want to buy up to 240)
2) When the algorithm changes from up to down, you sell whatever you have all the way down to 60% of what you've seen the highest luxury price so far.
3) This is the tricky one. When the price climbs or drops due to other people buying or selling, you discourage by buying just to raise the price enough to discourage them from buying further, or selling enough just to discourage them from selling further. The fewer people play with luxuries, the easier time you'll have with this.

Price changes in 1 and 2 can be differenciated by 3 (i.e. algorithm vs other players) by noting the number of commodities other player's have.

edit: re: 1 and 2, I had someone ask me elsewhere "but Falk, what if I've seen prices up to 500, and Oxen reach up to only 350 then start going down again, should I still sell down to 300? Wouldn't it be more profitable to wait for 500 again?"
The thing to remember is (ideally) there is little-to-no actual supply and demand because you've dominated the market enough that enough people give up trying to turn a decent profit with luxuries. Price is seemingly controlled by an algorithm. If this is true, you'd make a profit by selling 10 units then re-buying them when the price reaches its valley and starts to rise again, rather than holding on to them throughout. The reason you sell only to 60% and not all the way is to help deter the price from going -too- low and thus minimizing 3). Of course, use your judgement.


UNITS/G-PEOPLE
For Military units and Great People, other than the army stack you're building and the city type you're focusing on, they're really just another commodity to trade, except with a 60% sellback rate instead of 100%. Thankfully, prices for military units have an even wilder algorithmic fluctuation, it seems. I've made a tidy additional profit selling at (>hammer cost * 5) and buying at (<hammer cost * 3). Having a tech lead is sweet because when it does happen, you'll be the first to see that unit type on the market. Keep an eye out for that and buy a bunch of to reasonable prices. Screw the other guys in your nation - unless you know them personally you're probably smarter than them and the units are better off in your hands anyway.

Prices for Great People and units will also spike depending on battles and if nations are close to completing wonders. Great People usually come down whenever casuals who log on to do the puzzle game get great persons and sell them to spike food (which is $$$ for you)

I guess I'd still keep 1-2 of each in reserve for Wonders. (I constantly had a stack of Generals and 2 Builders to emergency-create Great Wall which I did a few times in this game)

Great Artists are generally cheaper, unless two nations are fighting over something like Shakespeare's Theater for that one culture win or something. It also seems to generally go up once Magna Carta comes into play which seems to be the most rebuilt wonder ever. Great Generals start cheap then start going up in price towards the end of the game, so building a small stockpile of them can be serious money.




II) PUZZLE SWAP
Exploit this. You want to do as many double swaps with the '5' button enabled. What you do is set up a double swap at '1', switch over then pop it. This nets you 125 culture per 6 swaps, plus 25 to each person in the rest of the country. The more people in your country you can convince to take the time to learn how to do this, the better. I believe with enough people there would be a critical mass that can be reached where you get an infinite number of puzzle swaps. (I'd need to figure out how much culture gives 1 swap then do the maths I guess) Obviously this technique takes a lot more concentration and can be a lot harder when there are other people working on the puzzle. On the flipside you burn through your swaps much faster so you can let them stack higher and take a shorter time tearing through them.

A healthy stash of Great People in the country can set you up for a succession of cultural wins, or even otherwise they just help out your nation's resources anyway. Even if there are idiots who just use the additional culture to play more puzzle games the terribad way and sit on random techs and contribute absolutely nothing to the nation, and then sell all the Great People they get, you still can capitalize on that (see: Global Economy section above). so it's not a complete loss. As an aside the Prince and Princess in my game, undeservingly ranked #2 and #3 at 435+ fame at the moment literally do nothing but exactly what I described above. I've wanted to throttle them for quite a while now. (They're in my country; I can't kick them even though I've wanted to)

I usually also try to keep my culture under the cap and only pop over when the GP I want is highest. (I didn't realize earlier that you -always- get the most expensive GP, hurr) Keep in mind that puzzle swap moves reset to 0 when you pop a GP, sometimes it means overselling culture to make use of your remaining moves. (Doesn't matter if someone else gets it cheap - you gaining a little bit > you gaining nothing)




III) CARAVAN
Maximizing the returns of this is simple. You want to do this at 1000 every time, and you want as good a board as possible to complete in as few moves. Once you have enough $ for the reward to hit 1000, you complete it once, then look at the next board. If it's terrible, you reset it while the reward is still 50. Resetting simply means just grabbing any tile and spamming it till the reward hits 0, which re-randomizes the board and locks you out for a minute. This essentially means you can get a new board per minute until you get one that looks reasonably good and can be completed in ~8-10 moves with few road type repeats. When you get a good board, you leave that board up till 1000.

You're probably also likely Finance Minister so this will also maximize the amount of extra gold that you give out to the rest of your nation. It's small, but it adds up.
 
Um.. not sure if you know this.. but the puzzle swap is better done without the 5x.

I am not 100% sure of the number of completes it takes (without an error) to increase but it goes up

If your the only person doing the puzzle you can get up to x9 and since your not using up your 5x the amount of moves to accomplish this it nets you a MUCH larger amount of culture in the end.. but to do this you MUST be able to do the puzzle with 0 errors.
 
Not when you're spamming +25's to the rest of the nation.

In a 10-person nation, each 6 moves nets 125 + 9(25) = 350 culture. That's 58 culture per swap, or a flat-out >11x 'modifier' throughout. Of course, doing the puzzle the combo way will inadvertantly give you a few double swaps here and there anyway, but like you said it relies on you being the only person doing the puzzle (since it's exponential-ish, even 1/5th the puzzle gone is a massive loss)
 
Min-maxing would mean me moving out if you showed any form of persistence. :) (Assuming I wasn't playing a group game)
 
ya you have to be the onlly person on the puzzle for the other strat to work (i still give +5 to people doing it that way) its just more efficient way of doing it. Mind you takes longer to do because now your having to deal with 5x less the moves

(ill try an do the math once i save enough moves to tell you the exact number vs the 5x and see which method is better)

I guarantee mine is thou.. since you can get above x5 per and pretty quickly.

*note.. hehe.. up to 67 prophets.. seriously this game they are just giving them to me. Got to love min/maxing the market

(SECOND NOTE)
I have noticed that the great peoples algorithm has it set up in such a way that the number can keep going up. I have seen a great person at around 40-50k. Not really an ebb and flow as much as a Arrow.

My limited experince with it is this. Lets assume that this graph below me is a great person Graph .. + equals the up swing and - equals the downswing

..... 1000 ..... Right now the great person is stable but when you buy one.
....+4000 ..... Now the price is going to start climbing slowly. (buy another one)
..++8000 ..... now price will rise dramaticly (sell one)
....+4000 ..... price will still climb but slower (sell two more)
...... 500 -..... Now price will drop slowly

etc. I notice that when a great person gets really high the next culture great person is based off the top priced great person on marketplace.. and most people will sell this great person starting the chain reaction. This is by no means a 100% method.. just somthing i have noticed in my playthrus
 
(ill try an do the math once i save enough moves to tell you the exact number vs the 5x and see which method is better)

You don't need to save the moves to do the maths. Prove the maths, yes.

I don't see how going from 1x to 9x (which is an average of x5 for ~25 a swap assuming a linear combo gradient, which is generous in my experience) with a few double swaps is going to beat 125 per 6 swaps not even counting the +25 spam to nation.

etc. I notice that when a great person gets really high the next culture great person is based off the top priced great person on marketplace.. and most people will sell this great person starting the chain reaction. This is by no means a 100% method.. just somthing i have noticed in my playthrus

Anecdotal. It's completely random from my own limited experience. I'll be glad to be proven wrong though. ;p
 
The market is definitely the most powerful thing in the game right now.
 
Ok.. dont forget thou.. your using 30 moves to achieve 125 points vs me using 30 moves to do an entire puzzle (or about) plus when I do get doubles near the end i get 9x my doubles vs x5 yours.

but ill prove the math to you by tonight.. :p

or tommrow.. going to do it both ways and see which one grants more culture both to youself and to the civ (since it will show up as +25 or +5 on the message board)
 
You realize that it's 6 moves for 125, and not 30?

edit: To be honest, I was looking forward to your testing. Now I'm not even sure anymore, on the premise that you don't even seem to know what the 6-move double swap is about.
 
Finishing a puzzle using the turbo swap (5x), without errors and other players, gives ~5000 culture.
 
Ok.. let me try and break this down

if you use the x5 button

6 Puzzle moves grants you a total of 25 points Per correct guess (125 if its a double swap) at a cost of 5 moves Per use

If you use the normal x1 setting

1 puzzle move grants you 5 points per correct guess (25 if its a double swap) at a cost of 1 move per use


BUT. if you stretch this out a bit.


40 Puzzle moves at x5 should (in theory) grant you around 3000 points at a cost of 200 puzzle moves (+about 300 to the civ via the bonus)

Now if i can prove that 200 puzzle moves make MORE than 3000 points (including the bonus to the civ) without using the x5 button.. then my point is proven
 
ok.. Let me try and break this down

if you use the x5 button

6 puzzle moves grants you a total of 25 points per correct guess (125 if its a double swap) at a cost of 5 moves per use

if you use the normal x1 setting

1 puzzle move grants you 5 points per correct guess (25 if its a double swap) at a cost of 1 move per use

no turd sherlock

edit: I've never hated the autocensor feature more than right now
 
Complete combo, 44 moves, 4 of them double swaps for ~+1280 and +20 to rest of nation
The puzzle is 48 squares, each of which requires a swap and a double swap clears 2 squares, so the above number of moves is correct at a glance.

I'm not going to bother doing 6-move double swaps because I can tell you the maths for that (it's not particularly rocket science)
doing the 6-move double swap 7 times = 42 moves, for 7x125 = 975, and +175 to rest of nation.

For 2 moves less, if you have 4 or more people in the nation 6-move double swap comes out at a net gain.
975 + 175(3) = 1500
1280 + 20(3) = 1340

If you have ~15 people in the nation
975 + 175(15) = 3600
1280 + 20(15) = 1580

Also I had to restart 3 times just to get a perfect clear because someone hopped in. (And I screwed up once. Yeah I'm bad, etc) so your utopia of solo swapping is pretty much an utopia.

Now shut up and sit down.
 
^^
In theory you're maybe right.
But no one is a 100% loyal team member or as good and there's always a chance someone wants to puzzle at the same moment too.
 
Yeah, which is why relying on having the entire puzzle is a bad idea (not to mention it generates less anyway as my previous post mathematically proved)

6-move doubles can be done any time pretty reliably.
 
You may be wrong about the puzzle game, but you may actually not be off the mark regarding Great People gravitating towards the highest value on the market.



Like I mentioned, plural of anecdote isn't data, so I'm not instantly convinced but a 1/5*1/5*1/5 chance -is- pretty odd. Enough to pique my interest to pay closer attention, at least.


edit: Um.
 
In the case of the game I'm playing now GP are hanging out way LOW in value.... right now 4 out of 5 of them are below 10k in the late Independence Era.... :O
 
Top Bottom