Kramsib's Value Theory

[rah]

That's why these types of forumlae are mental masturbations.
There will always be turns where an adjustment will have to be made to maximize the city. I.E. in any perfect formula let's say it's kicking out 7 shields. after 7 turns you have 49 shields. that's seventh turn you could have adjusted to 8 to finish that caravan and not wasted the 6 shields the next turn. Despite the less then optimal use of food or trade.

That turn your going to be one beaker short of making monarchy in an odeo year.

That turn you need to finish the wonder before someone else does.

The turn you crank down all science and trade to delay getting a tech that you plan to steal from a neighbor when he develops it so you can leap frog.

The turn you crank your money so you can bribe that barb before it destroys a key city.

The turns you maximize food to the city will be size two prior to an attack.

The turns you minimize food to keep the city from growing and going into revolt.

The turn were you maximize trade for celebrating or delivering caravans.



There are hundreds of situations where your not going to be using that perfect combination. It depends on whats going on in the game around you. So you can all waste your time argueing over theoritical things (especially the vague nature of this one). I'll spend my time playing MP games by the seat of my pants. And I bet my results will be better than yours.

It's always about what is most important to you for that turn. And you don't need a complicated formula to figure out.

[rjmatsleepers]

Originally posted by rah

There are hundreds of situations where your not going to be using that perfect combination. It depends on whats going on in the game around you. So you can all waste your time argueing over theoritical things (especially the vague nature of this one). I'll spend my time playing MP games by the seat of my pants. And I bet my results will be better than yours.

It's always about what is most important to you for that turn. And you don't need a complicated formula to figure out.

If I'm the I'm the person you're betting against, you'll win your bet easily.

But thanks to Solo investigating "theoretical things" we now know about which trade goods a city will supply and how to keep it supplying them. There are a lot of other examples I could choose. You yourself mentioned an Oedo year - another example of investigating "theoretical things". Progress comes from a combination of good practical play and good theoretical analysis.

RJM at Sleeper's

[Kramsib]

Re: Re: Fifth Statement: fc = sc = ac

Sorry I wasn't able to answer till today, I had a lot of work to do.

Originally posted by rjmatsleepers


I'm not sure what you mean by "technically efficient". If you move 1 of the workers that is producing no shields on to the forest and another on to a plain, you would get a 10 shield surplus which is much more efficient for producing units. These 3 additional shields would be at the cost of 3 food and 1 arrow, but would eliminate any shield waste in the final row of a unit build - great for rushing vans.

RJM at Sleeper's

When I speak about "technical efficiency" I mean a kind of "absolute advantage".

Grassland 2 f / 0 s / 0 a = 2 items
Bonus Grassland 2 f / 1 s / 0 a = 3 items
Plain 1 f / 1 s / 0 a = 2 items

A bonus grassland is more efficient than a grassland or a plain as it produces 1 extra shield compared to simple grassland or 1 food compared to plains.

But a square with grassland has the same technical efficiency as a square with plains. Which is better?, it depends on the "value" of 1 f against 1 s. If 1s = 1f, both are equally efficient, if 1s>1f, plains are more efficient and viceversa.

Under technical point of view, squares with the same number of items are equally efficient. Different values for each kind of item make efficience varies. So I'm talking about two concepts, technical efficiency and global efficiency. Ok?

In the example above, if you move 1 citizen from no shield grassland to a forest you are "changing" 2 f + 1a for 2 s in exchange. That is to say, 1 f + 1/2 a for 1 s.

You can change it to plains "changing" 1 f for 1s, which is a cheaper option.

The example was given intentionally to show, in the future, how my theory can be useful to maximize production or even decide were to produce to maximize "value", so be patient.

The picture shows the AI's decission, the distribution of the citizens by clicking the city square on the resource map.

[Kramsib]

Re: Re: Fourth Statement: Trade arrows are money

Originally posted by rjmatsleepers


But until you reach democracy you cannot get 10 gold from 10 arrows, unless the roundings are working in your favour. And with a market (bank, etc) you get more than 10 gold from 10 arrows. My conclusion is that the "value" of a trade arrow depends on the form of government and the infrastructure that you have built, but it is very rarely 1 arrow = 1 gold.

RJM at Sleeper's

I tried to remark that I was talking about cost and not about value yet.

The fact that you cannot get 10 gold from 10 arrows in most governments alters the "value" of trade arrows, but not its cost.

A slide restriction increases the value of trade arrows over the 1a = 1g. But the cost remains the same, just becouse cost depends on the elements used to get production.

[Kramsib]

The Difference Between Cost and Value:

I have to explain these economic concepts before continue explaining my theory for not becoming misunderstood.

Let there be a producer.
Let there be a customer.

The producer, produce output, for example chairs.

To build a chair, producer basicly needs:

wood
tools
time
...

These elements mean a cost for the producer.

Let's assume our customer needs a chair, this implies that a chair has some value for the customer.

Let C be producer's cost.
Let V be customer's value.
Let P be the price.

Producer will never ask for a price which P Customer will never pay P>V. He wants a chair but not at any price, in that way, V is the maximum price the customer will pay.

We can say

P-C >0 is producer's profit
V-P >0 is some kind of "customer's profit" as it buys at a lower prize under the maximum he wants to pay.

So, only when V>P>C, we have negotiation between producer and customer.

Producer wants P=V becouse that is the highest price he could ask and that maximizes his profit (P=V maximizes P-C).

Customer wants P=C becouse that maximizes his "profit" V-P.

Then, producer and customer start to bargain until they come to an agreement wich determines final price.

-----------------------------------------------------------------
My theory tries to study "value", so I need to determine costs first. Then I will estudy value, and finally we can study how can a "virtual" market will work and how prices are set up.

[Kramsib]

Originally posted by rah
That's why these types of forumlae are mental masturbations.
There will always be turns where an adjustment will have to be made to maximize the city. I.E. in any perfect formula let's say it's kicking out 7 shields. after 7 turns you have 49 shields. that's seventh turn you could have adjusted to 8 to finish that caravan and not wasted the 6 shields the next turn. Despite the less then optimal use of food or trade.

That turn your going to be one beaker short of making monarchy in an odeo year.

That turn you need to finish the wonder before someone else does.

The turn you crank down all science and trade to delay getting a tech that you plan to steal from a neighbor when he develops it so you can leap frog.

The turn you crank your money so you can bribe that barb before it destroys a key city.

The turns you maximize food to the city will be size two prior to an attack.

The turns you minimize food to keep the city from growing and going into revolt.

The turn were you maximize trade for celebrating or delivering caravans.



There are hundreds of situations where your not going to be using that perfect combination. It depends on whats going on in the game around you. So you can all waste your time argueing over theoritical things (especially the vague nature of this one). I'll spend my time playing MP games by the seat of my pants. And I bet my results will be better than yours.

It's always about what is most important to you for that turn. And you don't need a complicated formula to figure out.

Perhaps I cannot state a formula for players to play perfectly, but I can state a formula to play better.

But, one of the aims for this analysis is to find a way to program a better AI. If the AI were able to use a mathematical model to maximize the effects of its decissions, it could rid of cheats and become another amazing challenge for humans to play.

[Kramsib]

Originally posted by rjmatsleepers


If I'm the I'm the person you're betting against, you'll win your bet easily.

But thanks to Solo investigating "theoretical things" we now know about which trade goods a city will supply and how to keep it supplying them. There are a lot of other examples I could choose. You yourself mentioned an Oedo year - another example of investigating "theoretical things". Progress comes from a combination of good practical play and good theoretical analysis.

RJM at Sleeper's

Totally agree with you

[Straybow]

Yes, this kind of analysis is what would make a competent AI.

We can dream, ya know.

Any prospect of another Civ2 Progresivo?

[rah]

The two examples you guys quote are Oedo years and Solo's trade thread. There are actually many others. Xin's science and his square conversion tricks. The extensive attack manuals and barb guides. And most of these have contributed to strat, BUT, for solo's thread, and others there are only a couple of features that are used in fast paced MP games. Oedo years is easy to incorporate since it adds no time. Most of Xin's tricks aren't used because they're so time consuming and you'd never find a game if you played that slow every game. Calculating attack success % during a game are usually just approximated to the point of good/bad attack type thinking not taking the time to calc the exact percentage on every attack. Rough guidelines and not mind numbing calculations are the things that end up being used in MP games. For SP games you need not worry about anything since the AI is so easy to beat and almost any strat will work.

I agree that it could help the AI since anything would have to work better than the "MORE FOOD IS GOOD" type programming. But since we never play with an AI, it's still all mental masturbation.

[Peaster]

Rah - I take your point, but I think you go too far with it. There are other ways to play Civ2 besides MP, most of which allow some time for thought -

1) GOTM's: The goal is not just to defeat the AI, but to do it quickly, or with maximal score, or etc.
2) PBEM's and PBEM duels.
3) Playing Scenarios
4) Those of us who are just curious about the best Civ2 strategies, and play SP games as experiments.

I think most (non-MP) players don't do calculations very often, but you are probably at a disadvantage if you NEVER do them.

[Kramsib]

When I talk about AI, I am not thinking about AI as an enemy but also as an assesor, imagine how fluid your MP games will be if you can trust on your automatized citys because they are performing an optimal micromanagement for you.

[rah]

Don't get me wrong, I do use the calculations. Thanks to Solo I know when a demand can change so I plan better. I calc rough attack percentages. I know when a caravan delivery is on the optimal path. I squeeze every beaker to get monarchy on an oedo year as early as possible.
But as I said, they're usually quick approximations. That's all you have time for. What we're doing here is way to vague and would not have many applications. As I stated earlier, there are too many times when you would want to go against the formula to maximize what you're attempting to accomplish in a turn.

As for playing SP, if you check back in the archives you'll see that I participated in earlier SP challanges. I was the third person (i think) to win at OCC. I participated in some of the early landing and highest score contests. I have also played many of the scenarios. But guess what, they still got boring and I no longer participate. I don't begrudge the people that haven't done those things completely and are still having fun doing it. I no longer do. Trying to win the scoring record is probably the most boring thing anyone could ever do. Doing it more than a couple times is enough to make you never want to play the game again. I only tried 3 time because my brother was also doing it so it became a macho thing. I finally gave up when he topped my most recent record because I couldn't face doing it again even though I knew I could top it. That was years ago.

And you're right, I am slightly MP biased since it's the only thing that provides a challange. The AI will never be as challanging as a human opponents, even bad ones.

No matter how good they programmed the AI, I would NEVER let it manage my cities in an MP game without oversight. At least early in the game where a difference of one beaker, shield or food could make a BIG difference. Later in the game it might be nice, but most of your opponents are dead by then.

[Straybow]

So the answer then is while we're reprogramming the AI we reprogram things like making advance costs to be probabilistic, somewhat like MoO2.

I don't know if food and growth could be handled that way... Oh, what they heck, we're dreaming here. We'll reprogram the growth algorithm, and the production box to boot. Don't ask me the nit-picky details.

Then your precious beaker/food/shield squeezing won't matter so much.

[rah]

Yes, you are correct.
But I really do like reading the different theories. Any new trick that your opponent is not familiar with yet is an advantage to be exploited. I don't know how many times I've figured we've sucked new knowledge dry on this game only to be surprised. But I just don't see enough potential in this one. Of course, after saying that, prove me wrong. I've been married for almost 20 years, so I'm used to it.