I'll start by copying the discussion to date, in chronological order. Each post will be seperated by dashes, and above the post I'll put the poster's name and the thread it was originally in.

---
LGJ, Startup Options

However, the ecology model would haveto be changed to track the use of animals like horses,oxen, etc for the tech model to work because many other places didn't have them until they migrated there. See the Americas developed the wheel and there is much use for the wheel without horses or whatnot, but without that first practical use with animals it seems they don't use it much except for toys. Egypt was the same way, until horses were brought in. Camels don't seem to work with this for some reason....i don't know of any historical places where their was camel driven modes of transportation, but i could be wrong. Anyway much of Africa remained like this because of that.

---
Richard, Startup Options

You wouldn't have to do that in the Earth game, because you would be limiting the tech tree artificially. Basically, the normal routines would be disregarded and the game would choose the tech path automatically.

But you actually make a good case for putting that in the standard game, so I'll discuss it in the ecology thread.

---
Richard, Ecology

LGJ brought up a good point in another thread. He described how the presence of certain types of anumals can affect the development of civilizations. For example, you can't develop cavalry if there are no horses on the continent, and it certainly makes no sense for a city in the arctic to be able to produce elephant mounted troops.
So, we should probably try to define the ranges of some important animals. Hunted animals and domesticated animals kept for food are not very important because those will be handled in an absrtact manner, but draft animals and animals used in war should probably be tracked. These include oxen, donkeys and mules, horses, camels, and elephants.

If a civilization has no access to good draft animals, they will probably not be able to develop intensive agriculture. They will probably turn to hunting and a nomadic lifestyle if there are good hunting grounds, and if they cannot hunt they probably won't advance much, as the agriculture requires so much manual labor that there won't be people available for other things.

We probably can't handle draft animals as an abstraction because technology determines what animals can be used. They had to invent a special collar before horses could be used to plow fields. After they used that collar, they could use horses, which were faster than oxen and plowed fields more effectively. So this means we probably should define the availability of horses, oxen, and mules seperately.

I think that war animals need to be tracked. The availability of horses and camels are definitely important to the way a civilization develops, and if we want to include elephants, their range should be defined as well. I personally don't think we need elephants, because they actually had very minor impact on ancient war. They were not effective and they were not used much.

Do you think we should include dogs and cats? Dogs were important draft animals for the eskimos and they help with farming and other things, but I don't know if this is enough of an impact to justify including them. Cats definitely have an impact on rodent populations and food loss, but again I don't know if this makes them important enought to include. Let me know what you think.

I could define the natural availability of animals by ecological province. Each province would have the following animal related data:
Domesticatable (is that a word?) animals.
Huntable animals.
Oxen
Donkeys
Horses
Camels
? Elephants, Dogs, Cats ?

The original ranges become insignifigamt after a while, however. Once people start domesticating and breeding them, they are spread all over. That aspect of things is best covered in the tech model, however. I'll make a post there to discuss it.

---
Richard, Technology

I think we should discuss technologies relating to the breeding and improvement of domesticated animals and plants. These techs are similar enough to be covered by the standard tech model, but there are a few differences that require a couple special rules. Here is my view of the situation:
I think these techs should be defined as follows:
The quality of the animal, and the civ's knowledge about how to train and use the animal.

First, the tech for a certain animal cannot exist if the civilization has no access to that animal. Native American civs could not develop the Domestic Horse technology.

The first domestic horses were not strong enough to hold a rider on their back. They were only able to pull chariots and carry lighter loads. After people developed better breeds, they were able to hold humans and light cavalry was developed. Contrary to what Civ 2 says, chariots were developed earlier and they are much less effective in warfare than mounted riders. When horsemen appeared, chariots became obsolete. But the horse still wasn't strong enough to carry a heavily armored rider. They had to develop better breeds before cataphracts or knights could be used.

Training is also an important issue. Horses have to be trained well before they will go into combat, and training them to deal with explosions in modern combat is even more difficult.

So in Clash terms, we have the following progression of the Domestic Horse level 3 technology. This chart does not show all advances and the applications have other prerequisites not shown, but it shows the general idea.

Prerequisite: Wild Horses available
Level 10: Pack animal only
Level 25(better training): Chariots available
Level 40(better breeds): Light Cavalry available
Level 55(better breeds): Heavy Cavalry available
Level 70(better training): Riders can use gunpowder weapons.

There are also a few other effects. The level of the technology helps determine the effectiveness, speed, and range of mounted soldiers. The level also does a little to determine the effectiveness of farming and transportation before the internal combustion engine.

There is also a more complicated application. If your cavalry is not trained to deal with gunpowder and you face an eneny with explosive weapons, the effectiveness of your cavalry troops will be greatly reduced. This helps explain why early guns were so effective even though their range and striking power were worse than the bows of the time.

In general, there should be a check for the reaction of your animals in a battle. Inputs are the technology your enemy is using and the tech level of the animal you are using. Depending on the results of the check, some percentage of your animals may bolt. This takes those troops out of the battle and causes confusion on your side. Military model people (if we have any), how would this be implemented in the battle system?

The spreading and trading of this tech is another unique aspect. While most techs can be copied and transferred as data, this one cannot. You have to physically deliver the better breeds of horses. Also, it is a lot harder for this tech level to decrease. In fact, it may never do down normally. The genes will stay steady or improve by themselves, as only the healthy and fit horses will breed.

Conversely, the tech will drop more easily in some situations. Ownership of a group of horses is a zero-sum game, so if someone steals or kills all of your best horses the tech will go down. This won't happen if someone steals a normal knowledge based tech. Also, assunming that the best horses are chosen for war as they usually were, extensive battlefield deaths could reduce the tech level.

So we will probably have to define a derived class of Tech objects called Breeding Techs. They will have most of the characteristics of teh base Technology class, but the methods will have to be changed to account for their differences.

Does this sound good?

---
roquijad, Technology

Use, taming and creation of new horse races is IMO such an important factor in warfare history that if a special tech system for that is needed, then we must go for it.
I like all the things you say with the exception of mixing within the same concept the actual technique of breeding/training/taming with the availability of horses themselves. What if we treat horses as a "special" in terms of the economy model?

Specials like copper or oil exist in some mapsquares and people can take advantage of them for economic production. They can also be traded. If we now make some specials, like horses, transportable, that is, a mapsquare can change to "having the 'horse' special" from not having it, then the avilability of horses is IMO modeled better.

Civs, through trade or military appropiation, could obtain horses, which in gameplay terms means changing one of the mapsquares they control to "have horses". The tech on the other side has to do with the technique itself. Together, you can produce military units with the abilities you described.

The number of military units you can produce in a turn is given by the amount of with-horses mapsquares you control, while the characteristics and types of units available depend on the tech.

The tech could also be useful to produce more with-horses mapsquares. The greater the tech level, the greater the chance for reproducing animals and get a new with-horses mapsquare.

---
roquijad, Ecology

I liked your idea for horses in the tech thread. Now I realize it is extensible to other animals. I find it great if availability of oxes, mules, etc have an impact on agriculture as well as warfare. Transportation techs would also be very related to the animals you have (if you have).
My list of important tamable animals:
Horses
Camels
Mules
Elephants
Oxes

I agree Elephants didn't play a big role in history, but I'd include for scenario purposes.

---
Richard, Technology

Rodrigo: This tech does not include the presence of wild or tame horses. Rather, the provinces in your civ must have enough wild horses (as defined in the Ecology model) before you can get the tech. Once you have the tech, horses are assumed to be integrated into the economy so IMO it would not be good to keep modeling them as a natural resource at that point.
You saw the posts out of order, so I understand the confusion. I should have mentioned that this was a continuation of the post in the Ecology thread. We should discuss the map availability of horses in the ecology model, and this thread should be dedicated to the advancement of the tech once you have the horses.

---
Richard, Ecology

Rodrigo mentioned specials in the other thread, so I looked them up on the webpage. They should do a good job of modeling the wild animals available for taming. I can add routines to the ecology model so it generates these animal specials in some of the squares that form good habitat for the animal.
However, I don't think that these specials should still be on the map after they have been tamed and integrated into the economy. At that point, I believe they should be modeled as province infrastructure.
This may seem odd, but as far as I can tell the equations would be the same. FE the number of horses available would be an important part of the farming and transportation infrastructure, as their presence increases the ability to complete economic activity.

But I'm getting off topic. This should be discussed in the economic thread.

I could generate lots of specials like the wild horses. Once the proper tech is researched, the specials would disappear and become a part of the province infrastructure.

However, I still think that Huntable animals should be treated abstractly. There are just too many of them and they are too widespread to be included as unique specials, so they should still be charactistics of the ecological province. This attribute would be used as a food input in the economic models, as labor turns the game into food. The ecology model would then determine the depletion of the herd in a manner similar to the deforestation checks.

Other domesticatable animals are a bit of a quandary, however. Theoretically, they should be treated like horses; they would start as specials and then become infrastructure. But there are just too many of them. I don't think anyone wants a map littered with wild chicken and wild pig specials. Then there is the fact that most common farm animals were domesticated before 4000 BC, so it would be pointless to include their wild counterparts.

But it does need to be modeled somehow. Cultures with domestic food animals are much less likely to hunt, and their diet will be different. I think there should be some modeling of the amount of farm animals in the province. And it can't be strictly economical, as some areas of the world, like the Americas, don't have animals that can be domesticated as farm animals. I need some help with this one.

A very good aspect of the specials plan is that it makes it much easier to add things like elephants, dogs, and cats. All we have to do is create the special and a couple infrastructure rules. They would simply be another instance of an existing class, so we don't have to do anything special to include them.

By the way, I should be working on adding specials to the third version of the Ecology model. While it wouldn't cover geologic specials like iron, it is the best place to model things like grapes. Mark, could you give me a list of the specials that are associated with the ecology?

I'll discuss the economic aspects more in the proper thread. What do you all think about the ecology/map generation aspect of this topic?

---
Richard, Technology

I see now how I caused some confusion. When I said that taking horses would lower the tech level, I didn't mean that the level included the quantity of animals. I meant that if the horses lost were of the best breeds, there could be a small drop in the quality of the gene pool. If you had horses of similar quality elsewhere, there would be no drop in the tech level.

---
Richard, Economy

We are discussing horses and other farm animals in the Technology and Ecology threads, and part of the conversation turned to the economic aspects of the animals. Here are some of the ideas I came up with:
There are two ways of modeling animals that I can see: specials and infrastructure.

I think the infrastructure modeling is a better abstraction. This way, the results of the animals' presence can be modeled, and they are "behind the scenes" so they won't clutter up the map. This method of modeling may seem odd, but as far as I can tell the equations for animals and other infrastructure would be the same. There are a few differences, but the "animals" class of infrastructure objects can be derived from the basic "infrastructure component" object class

FE the number of horses available would be an important part of the farming and transportation infrastructure, as their presence increases the ability to complete economic activity. Just like the amount of blacksmiths available, the amount of certain animals affects the general economic ability to do things.

But unlike most infrastructure components, animals can be moved around. They could be transferred from one province to another, and invading armies could take them home and add them to the infrastructure of their civ. Another difference is that animals are bred rather than built. You don't have to expend any inputs other than food to increase the animal infrastructure; you just have to avoid eating too many of them. This leads to another point: animals can be consumed. This is a major difference, but we should be able to take care of it by altering the methods used to calculate depreciation.

While animals integrated into the economy can be tracked like this, wild animals should be specials. But unlike most specials, they are only used once. When your civ discivers the proper tech, the wild animal specials are turned into domesticated infrastructure components.

We should probably have the following infrastructure components:
Horses
Oxen
Donkeys/Mules
Food providing animals
Camels

I also think there are some good reasons for including Dogs and Cats as well. Dogs help farm operation, and cats control granary-draining rodents.

What do you think of this? Is this a good plan, or is there a better way of modeling these animals?

---
roquijad, Technology

Yes, I got all confused.
Here's my new proposal considering also your idea to handle horses as infrastructure. I'll put all the implications for the three models here, just for simplicity.

In the tech model:
I'd suggest the tech itself shouldn't count in any horse-specific info such as genes or horse race. The tech should be only the taming/training/breeding technique. And therefore, the only special thing about this tech would be that animals are needed to develop the tech, while in other senses it's a normal tech.

In the ecology model:
For other animals such as oxes, mules, etc the ecology model indicates if they exist in the province or not and how many of them there're. For horses it's the same, but we can split them, only because of their importance in history, in, say, 3 races. Wild/Ancient/Natural horses, FirstBreed horses, SecondBreed/Nowadays horses (or simply horse1, horse2, horse3). Of course, in the ecology generator only Wild horses appear in the beginning of the game.

In the economic model:
It sure sounds really odd to consider animals as infrastructure, but it makes a lot of sense. However, a couple of differences between "real" infrastructure and "animal" infrastructure make me prefer to simply manage animals as a new feature in the econ model.

People and the govt can invest in real infrastructure, increasing it, while this would be kind of strange for animals. Real infrastructure can't be moved/stolen/traded, while animals can. Real infrastructure doesn't "grow" by itself, while animals do (they multiply).

So why don't we just add animals as a new "thing" in terms of what a province has in the economy model? Having their own properties (such as tradable) ensure a better modeling. I believe for the econ model is really simple to include them as a factor to production. I propose the econ model to have a "animal force" variable. Just one variable related to animals as input to production. Using the info from the ecology model about what animals are present, we compute this "animal force" like, FE:
Number_of_horses3*1
+Number_of_horses2*0.7
+Number_of_horses3*0.4
+Number_of_oxes*0.6
+Number_of_mules*0.7
+...

The econ model then treats animal force as a normal infrastructure variable for production.

How the three models interact?
The ecology generator creates "wild horses" as a special. Once a mapsquare with horses is within an economic/political province, the special disappears from the map and a number of Horse1 appears in the province stock.
The same is done with other animals. The province then has a list of animals in stock like:
Number_of_horses1
Number_of_horses2
Number_of_horses3
Number_of_mules
Number_of_oxes
etc

The ecology model also manages, each turn, their multiplication, increasing number of animals in stock via a base value and a modifier given by the tech system (the breeding/taming tech). Then the econ model uses the number of animals to compute "animal force" as an input to production.

Animals this way can be traded, stolen, killed and moved without problems. It just changes the stock. Everytime a military unit is created that uses horses, the number in stock is reduced and the military unit gets part of its properties based on the horse race and breeding/training/taming tech level.

The higher the tech level is, the higher the probability for a new race of horses to appear, unless the horse3 already exist in the province.

In this way IMO everything has sense. For simplicity instead of having a large list of animals, we can aggregate the least important ones and have only:
Horse1
Horse2
Horse3
Camels
Other1: mules, oxes, etc
Other2: cats, dogs, etc

What about animals for consumption (like pigs)?
In the ecology generator I'd give each mapsquare a "wild life" value according to climate, etc. All mapsquares conforming an economic province would be summed to create a "wild life food value" that would act as a multiplier for food production in the econ model. This multiplier can in time be also affected by the Breeding Tech, so the better the tech level, the greater the multiplier.

What do you think, Richard?

---
LGJ, Technology

Well i don't think we need to model many animals. Like everything else, just those important to history (outside religion...cuz there we can have anything modeled). Horses are definatly one. Dogs are too, though mainly in the beginning. Dogs are also the first animal to be domesticated (or rather wolves were).
Here's the list of what I think is essential to model:
Camels
Dogs (Formerly wolves)
Horses
Elephants
Cars (maybe...Other than reducing mice/rat population and religious importance i don't see any use).
Livestock in general

Remember everything we add makes things more complicated so think carefully what is really needed.

---
Richard, Technology

roquijad:
The problem with taking horse quality out of the tech equation is that it becomes hard to do tech prerequisites that way. You need a certain breed of horse for heavy cavalry, and if that isn't in the tech it becomes difficult to define the time that your civ can make those knights.

It is possible that we could ignore prerequisites and use the horses as required inputs in the production function. So when you order a province to make knights, it checks to see if it has enough horses of the right quality. Those horses are taken out, and the quantity and average quality of the province's horses decreases (they took the best horses).

While that would work, it seems like it would generate more micromanagement than the tech level approach.

roquijad, ecology: I don't understand. Do you agree that the wild animals should disappear from the ecology model once they get domesticated?

roquijad, economy: We seem to agree on the basic features, but not the implementation. It could be better to make a new class rather than a derived class, but I'm not sure. The coder will have to decide this one.

I don't like the splitting of horses into only three species. Variables should change continuously, rather than discreetly. The jump from horse2 to horse3 would be huge and unbalancing, requiring micromanagement. I think quality should increase continuously, in the manner of a technology. Also, horse2 and horse3 does not belong in the ecology model IMO. I think that ecology should deal solely with wild, undomesticated things.

roquijad, interactions: Again, once horses are domesticated, ecology should out of the picture (except for its effects on farm yields, which determine how much you can feed animals). I think the modeling of horse population growth is best done in the population model.

I can simplify that model so it will run less equations for the horses (the one-cohort simplification will do well), and run it at the province level to determine province horse population. That way, aspects of the population model like military recruitment, famine, and disease can be applied to the animals.

While horses are important enough to get this treatment, I'm not sure about the other animals. I could define "farm animals" as a single race and run the population model on that as well, but it might seem funny.

I'm not sure about the "wild life food value" you mention. It seems to be about the same thing as the huntable animals value. Could you explain further?

LGJ: I'm still not sure about elephants. Dogs will probably have to be put into into a bigger category.

Okay, here is an outline of the revised plan for treating animals:

Each set of animals we choose to model has four components that are important for modeling:

1) Initial wild availability. I think they should appear as specials defined by the ecology model and disappear entirely from ecology once domesticated.

2) Quantity of domestic animals. This can be modeled in the population model, or if you all prefer, as an infrastructure component. Rodrigo mad a good case for not doing infrastructure, and I agree.

3) Quality of domestic animals. I think this should be a continuous variable tracked by province in the economy model, similar to the definition of the quality of roads or other structures. Data to determine this would be input from the economy model, and the technology model. This is what is used when determining if horses can be used for a particular purpose, like heavy cavalry. It also alters the effectiveness of economic activity.

4) Knowledge of the animal and its uses. This can be a normal technology, like any other.

So we already could be tracking the data over four models, and discussion about military aspects hasn't really started yet. The military uses the quantity to make units and it can take the quality and knowledge aspects to determine unit stats and combat actions like tendency to bolt.

Here is my list of animals or groups of animals to be modeled as per the previous description:
Horses
Draft Animals (Oxen, Donkeys)
Food Animals (Pigs, Chickens)
Household Animals (Dogs, Cats)
Elephants
Camels

I think we agree that horses need the full modeling, so I won't discuss those. The draft animals should probably be treated the same way, as a group. They were sometimes used in war, and are a vital part of the farming infrastructure.

Food animals can be modeled like this as well. While not used in war, they will be affected by disease, famine, and conquest. Their quanity and quality is important for determining the health of the people, and they should have a corresponding knowledge variable.

I think the mechanisms you have so far are reasonable, pending further details. But I think you are going way overboard on the number of things modeled. Other than horses, and possibly other draft animals, IMO the player won't get much out of the added complication.

We just Can't make a complete world model. Unless the Player gets a Lot of bang-for-the-buck out of something new, we must skip it IMO. Otherwise we will end up arguing about details of cat domestication... This while we have only gotten a handful of comments about the Whole military combat model in the last several months! Which do you think has more bearing on the eventual success of Clash?

Their primary role originally was hunting and later for herding and even security. Only recently has the dog's life become mostly for domestic reasons.

The infrastructure model that I am now completing is capable (I think), if need be, to handle domestic animals via infrastructure classes (well, pretty much everything can be handled via infrastructure classes) but it is not originally designed to get into that much detail. The population model on the other hand can handle non-people populations as well and in a much better way as a matter of fact, but it would be overkill. According to the importance we attribute to the subject, we can fit animals somewhere.

Sincerely though, personally I don't really care much about them. Besides that, your topic title is silly.

I think it is important to make sure that we don't have camels or elephants coming from cities in the arctic. Similarly, the player should not have a complete population of horses or draft animals left after a huge famine. There are some things that cause players to lose respect for a game, and IMO this is one of them.

Now 3 of the modes of transportation i listed are region specific. One needs to ask then why horses can't be used in each of these places.

Elephants: Jungle areas are often too dense and have lots of exposed trunks that can easily trip horses, but elephants can walk over the trees and one can see farther with them in general.

Camels: Although horses are used in the desert, they require much more water and food. Although they are in general much faster, they won't last as well on long journey's like camels will.

Huskies: Well in addition to be domesticated because they are dogs for the original purpose of hunting, horses can't survive in artic regions. Also horses are quite heavy and thus could more easily break thin than the lighter huskies.

Reindeer: We could use this as an alternative to huskies, but either way it should be only one. I think huskies is better since it uses something we already have without adding another type of creature to model.

Why do dogs in general need to be modeled?

Dogs (or early on domesticated wolves) are great hunters. They also, unlike other pack hunters, understand the concept of a master and chain of command. Anyway wolves would help tremedously in getting meat to eat (why else waist time domesticating an animal and feeding it?). Its usually not so big a deal as to whether a population lives/dies, but it could be.
Anyway, just like horses changed what they were to do from just plowing fields to carrying people so did dogs. With the domestication of other animals for sustainable source of food (herding) rather than hunting/gathering dogs took on a new role of helping keep those animals from wandering off. This is almost essential for some animals as they can stampeed at speeds unsustainable by humans for any length of time. Although not nessarily stopping the stampede, the dogs dig keep the livestock from spreading out, acting as a would be preditor(?) allowing time for the humans to catch up and bring them back. Without this, most livestock used today would probably not be domesticated now.
Finnally dogs were/are used as a means of transportation in most artic places where horses (and obviously cars) are too heavy and where other animals would die. Although attempts to domesticate some animals in the artic were done and there are domesticated Reindeer, huskies as far as i know are still for crossing these frozen lands are with huskies.
Finally near the beginning dogs/wolves have been used for security purposes. Although this might not be an important to any models.

Livestock: Ok, we don't need to model every type of animal here? Why? Because in general they provide the same things: food, clothing materials and various bone tools. We do however need some way of modeling them in general. Domesticated animals arrived prior to agrarian societies to reduce the chance of starvation. In many cases, an additional side effect is skins used for clothing and bones that were shaped into useful weapons and tools without having the risk of not finding anything when hunting. Some others like cows provide milk, but these can be linked with total food output. Now i know there are specific types of livestock that can survive better in differnt areas, but these regions are much broader than horses and unlike horses they can with proper facilities be used anyplace with a few exceptions. Another reason is because we are not modelling as far as i know differnaces in food like rice and wheat which is in many ways the differances between the types of livestock.

Draft animals: These include donkeys and oxen. I'm not totally convinced we need to model these for it seems with the exception of specialized tasks that horses (which are also draft animals) can handle any problems that these couldn't handle. The only exception i might see would be adding donkeys/mules for mountainous regions. You'd haveto show me why else they are needed and why oxen are needed at all.
I'm pretty sure someone's wondering now why not just list camels, elephants, horses, reindeer, etc all as draft animals like livestock? The differance is that some of these groups are also used for other non-draft purposes and huskies really aren't draft animals. Elephants are often used in construction of areas. Camels may be the only exception i can think of, but with other regions having specialized means of transportation i see this as discriminating against them otherwise.

Household animals: I can't see any way these would affect the model at all. I feel as though more than anything else listed they would feel 'pasted on' with no real use.

Cats: It is in my opinion unnessary for model reasons to model the domestication of cats. Although important for mousers, especially to merchants and in some places as guards, their signifigance ends for any model to handle. With those two i listed 1> as mousers their impact was signifigant, but not overly so. I think it would be adding more complexity for little gain. 2> as guards, they are few and far between, much less than guards. Those that are used are generally more exotic and less domesticated cats.

One thing about the wild vs. domesticated situation. Almost always there are wild animals. Although the numbers are usually less as time goes on, there are those that escape, those that are not domesticated because they are too hard, religious regions (Hinduism comes to mind), etc.. I think its important that no matter how high the level of domestication, the wild aspect be kept, even if it is at zero for that square because that can change.

Now as far as the techs for them go, i think it is nessary to have a breeding/training/taming (maybe not taming) basic tech. Everything else about the animals should be handled elsewhere. Quality and quantity is important as well as something to model genetic structure in a very simplistic way for disease and adaptation purposes (this is ness for anything including crops...(greater yield esp. through genetic engineering usually means less adaptation and also results in similar genes which allow disease to spead faster. Same is true for humans). This can be handled by a simple 1-10 scale of how similar in structure they are.

Also quality & quanity for wild and domesticated needs to be different. Having tons of wild horses and few domesticated but well bred horses would otherwise make the horses you have domesticated seem average at best. I also must agree with rodrigo on the specific levels for animals although 3 is too few. This models the variation in quality better than a continuous variable which tends to average things out.

For horses here's my idea:
horse1: wild horses during early civilization
horse2: modern wild horses
horse3: early domesticated horses, unable to do anything
horse4: horses that can plow and be chariotted (early still)
horse5: horses that can be ridden
horse6: horses that can be ridden by armored fighters and perhaps wearing armor (warhorses, pre-gunpowder ages)

See this is where it comes confusing....should horses capable of staying calm under gun/cannonfire be after horse6 or horse5 since during this age those heavily built war horses weren't needed and indeed speed was much more essential.

I don't think we need to worry about wild animals after there are domesticated ones. The wild animals are only important because they define what domestic animals you can have. Once their role as prerequisite is done, we don't need them.

Wild horses are not included in the quality rating. Once you have that rating, the wild animals don't exist anymore. It only averages up trained, domestic ones, which will always be about the same quality anyway. The discrete horse1 through horse6 system is a lot of trouble and fuss, and it invites micromanagement. IMO a simple "Quality" rating does the same job a lot better.

Each animal or group of animals we choose to model has the following data:

Tech level, by civ, determines how well you can train them. This is what determines the ability to deal with gunpowder, etc.

Quantity, by province. I assumed we were going to track this anyway.

Quality, by province. Determines the physical shape of the animal and what they are capable of. But if you can't train them this won't help much.

For sure horse modeling is the only one that needs detail. All other animals can be divided in simply two categories: Food Animales (animals you eat) and Production Animals (animals that help in production, such as oxes or dogs). I'm willing to forget about camels and elephants as warfare tools, just for simplicity. Neither had an important role in warfare anyway.

So, having Food Animals, Production Animals and Horses is IMO a reasonable middle point. We need the quantity in each province for each and handle quality only for horses.

The amount of Food Animals would act as a modifier for food production, while the amount of Production Animals + Horses would serve as an input factor for economic production. A (unique for all animals) Breeding/Taming Tech would also act as modifier to the amount of Food and Production animals.

Another animal tech handles training of horses for wars or the development of new horse races. But it isn't necessary to do this for other animals.

The ecology generator puts in the world FA, PA and horses according to climate, vegetation, etc as "specials". These specials disappear when a civ has a province controling the mapsquare-with-special and from that point ahead the econ model handles quantity of animals. They are said to be "domesticated".

You're right about the problems with discrete horse races, Richard. I just wanted to keep concepts "pure". I didn't like the idea of a tech representing at the same time a technique and a genes quality or anything like that. If you can somehow separate them, it'd be great, if not, doesn't matter.

That sounds like a pretty good analysis and compromise. I agree with most of it, especially the addition to gameplay. Horse stealing was often the prime motivation for raids in ancient times, and animals were very important to nomadic civilizations.

However, I still think we need the quality for all three types. the quality of draft animals will have an impact on how much work you can do, and the quality of food animals will have an impact on the amount and quality of the food they produce. I don't know if we are modeling the quality of food, but I think we should. It would have a big effect on disease and happiness.

I don't know if we need an additional "war training" technology. We will already have the basic training/breeding technology and a military technology that determines your civ's level of cavalry tactics. I think the war training is assumed in both of those.

I think we might be able to include camels and elephants. We could define a category called "mounts" that includes horses, camels, and elephants. They all can be used as draft animals, but they can also be mobilized as for war, exploration, or trade. The military model treats them as seperate units, but the province economy model treats them as an aggregate.

The unit produced by the province is the one most suited to that province's climate. So a desert province produces mostly camels from the war animals pool, a jungle province produces elephants, and most other places produce horses. Provinces that cannot support any of these, like arctic areas, cannot make any cavalry or mounted units.

So to summarize:

Each set of animals we choose to model has four components that are modeled:

1) Initial wild availability. These appear as specials defined by the ecology model and disappear entirely from ecology once domesticated.

2) Quantity of domestic animals. This will be modeled in the population model and tracked seperately by province.

3) Quality of domestic animals. This is a continuous variable tracked by province in the economy model, similar to the definition of the quality of roads or other structures.

4) Knowledge animals' breeding, training, and uses. This is a normal technology, like any other.

Note that concepts are no longer combined. Each component is modeled seperately in the appropriate model.

And we are modeling the following:
Mounts (Horses, Camels, Elephants)
Draft Animals (Oxen, Donkeys, Dogs)
Food Animals (Pigs, Chickens, etc.)

A couple of thoughts:

quote: I don't know if we need an additional "war training" technology

1. Quantity: The population model handles that, but the tech influences it, right? Better breeding -> faster quantity grows.

2. Quality: As far as I know the econ model doesn't handle quality. If so, then there's something unsolved here yet...

3.

quote: The unit produced by the province is the one most suited to that province's climate. So a desert province produces mostly camels from the war animals pool, a jungle province produces elephants, and most other places produce horses. Provinces that cannot support any of these, like arctic areas, cannot make any cavalry or mounted units.

I'm not sure what you're saying here. It's right, but I don't know what the point is. If you were trying to define what the role the ecology model has after generating the world, then my stand is it should check for coherency for human-induced animal movements. An ecology province during play should only determine the chance for an animal to exist in it if is brought from somewhere else via trade, FE. The military units you can build should only depend on the quantity of animals you have in stock without doing any check with the ecology model. If you couldn't bring elephants to Alaska (eco model's check), then you have 0 elephants in your province. As a consequence, the mil model cannot build an elephant mounted unit in that province. That's how I think is correct. The mil model only checking animal stock in the province.

4. The categories Food and Draft are abstractions, right? There'll be never in the game something as particular as an oxe, right? If so, what we're really modeling is:
Horses
Camels
Elephants
Draft (including all, except the above)
Food

I just separated your "mount" because Draft and Food were used as an abstraction, while Mount was not, if I'm understanding correctly... If I am, then I'd insist in not having a tech for each. 5 techs only for animals sound like too much to me. Can't we just use the same for all of them?

2) I thought that the quality of things like roads and bridges would be modeled. It seems like that should be in there.

3) "Mount" is an abstraction as well. The province stores the date for "Mounts" as an abstraction. When you make a unit, the "mount" pool is reduced and the proper unit is formed. It is assumed that the animals you have in stock will match the province's ecology. Of course you have to have gotten the proper animal somehwere, but once you have it, the model automatically assumes you moved it to the provinces it is best suited for. So "mounts" in a desert province is assumed to mean camels, even though the model treats it the same as the mounts in other provinces.

The ecology model determines what you will have in stock and the military model uses that data to make units.

I thought this abstraction would let us put in elephants and camels without any extra modeling effort. If it doesn't make sense I can abandon the idea and switch back to horses only.

Yes, draft animals can sometimes be used in war. Before they had good horses, they fought with "battlewagons" drawn by oxen. These were heavily armed and armored but slow.

4) Mount is an abstraction, just like the others. We are modeling:
Mounts
Draft
Food

Now I am not so sure. Horses, camels, and elephants would act exactly the same. Thus mounted units consisting of elephants would for instance be capable of entering any square that horses could enter.

It would add flavour to the game to have the three types of mounts, but at the cost of some rather unrealistic results.

Then there is the possibility that rodrigo mentioned, where there must be a class for each type of mount. This would allow restriction of movement of elephants into unsuitable terrain or climate zones. Basically we could have a mount interface that is implemented in three different ways, unique properties and methods overriding the default ones from the interface. The question is: Is the more complex modelling worth it in order to add a little flavour? In my opinion camels and elephants have not made a great impact on history - at least nothing compared to the horse.

quote: Originally posted by Richard Bruns on 09-14-2000 07:55 AM ...Horses alone will be modeled.

And draft animals and food animals, right?

I'm not really able to get too into the depths of this discussion right now, but can I throw out a quick thought?

'Animals' are a resource. In other words, 'animal' objects will implement 'resource' interfaces (as will metals, woods, edible plants, etc). Then there's no reason we couldn't have any number of animal 'subclasses' such as 'horse', 'elephant', etc. Each would have distinct behavior. It would even be a small matter to have different 'breeds' of each animal -- an 'African Elephant' class and an 'Indian Elephant' class, a 'warhorse' class and a 'appaloosa' (sp?) class.

F_Smith: I don't think we should go too far. Anything i think we do in this area should be for ness. reasons only. I think this area more importantly than others because ecology is very complex, more complex than anything else almost.

I'm just pointing out that from a coding standpoint, the approach I detailed is the most efficient way to go.

Animals will be 'objects'. Each type can be as different as we want, with no added difficulty.

Ya'll can build your model with or without that in mind.

Actually, from a 'coding' standpoint, we should have objects for individual crop types.

A 'crop' will be a 'plant' object. No reason not to subclass it into 'rice' and 'wheat', etc.

1) The concept "Draft" isn't good. I orginially proposed "Production Animals" and Richard changed it to "Draft". I didn't know this word, so I assumed it was just another way to mean "productive", but now I understand Draft is more like "Pulling a Plow". My initial idea was way wider than that. Productive Animales are all animals helping to production. This involves draft animals such as oxes, "helping" animals such as dogs in herding and animals that provide raw materials such as sheeps providig wool. In this way, the econ model can increase production in an abstract way using all kinds of animals and with all their usages as a source. This way dogs and other animals LGJ mentions would be included in this abstraction.

2) I think F_Smith has a point in what regards implementation, at least for mount animals. In the coding side, we should have Productive Animals, Food Animals and any number of mount animals. For the default game we should stay only with horses, but the flexibility to add more will be there for scenarios or people who want to see more mount animals.

What i want to know is why shouldn't we have a seperate listing for dogs and we should have one for horses and other mounts. Can you tell me how you can model abstractly what the domestication of wolves did for humanity? I wouldn't think so because the 'production animals' really are more for ones that work the land and dogs are much more versitile since they can be used as both draft and production (if it includes hunting and herding)...only mounts could do that and a wolf really isn't a mount.

Yes, I know that all of the animal groups are just objects, and that we could theoretically make thousands of them without difficulty. Basically, this discussion is to find a way to model all the animals using existing objects, so we can take advantege of OO flexibility. (I have learned something by studying the Java book.)

But the problem is the interface and processor time. Each animal group object has a lot of methods to run every turn and will demand attention from the player. So we can't have too many in the default game. Sure, the functionality will be there for scenario designers to do whatever thay want.

(Slightly OT) I really think it would be bad to model individual crop plants. It would force excessive player management and take a lot of computation time. All crops are modeled as a single object. People are assumed to pick different crops automatically, but that is not modeled. The output of the "crops" object simply varies by terrain and climate.

LGJ:

You keep misunderstanding what I meant by "household animals." I must be saying something wrong, so let me try to explain the concept again. The term does not mean "pets." Instead, they are animals that do important things that are not covered by the other groups.

You have listed many good reasons for dogs to be included. Every one of those reasons is a reason to include "Household Animals." Every one of those reasons is a reason I came up with "Household Animals." Dogs ARE Household Animals. Just like "Mounts" was intended to include horses and some other animals, "Household Animals" was intended to include dogs and some other animals. The concept was supposed to include wolves, cats, dogs, lions, etc.

This was meant as a way to explicitly model what dogs can do for humanity. I never said we should have a seperate listing for horses and other mounts. I wanted to combine them all into larger gropus.

You ranted against household animals, and then proceeded to rant about how the exact same concept shoud be included. This is surely the symptom of a communication problem.

I definitely messed up by coming up with those hybrid concepts. While the computer (and my computer-like brain) would deal well with them, normal (non-deranged) humans found the concepts bizarre. So nobody knew what I was trying to say. Sorry for confusing you all. They seemed like such good ideas at the time. . .

Roquijad:

1) Actually, I meant "Draft" to be a synonym for "Production." Those animals were always meant to do many things. Don't assume that my English will be good just because it is my native language

2) I agree. I was trying to fit the exotic mounts into the normal game, but the idea flopped. We should just leave them out and show scenario makers how to implement them. if they want them.

3) I never meant the battle units to be the same. I specifically said that the military model would treat them differently. We will make the resulting units act very differently.

Okay yea...seems like you had both me and Rodrigo confused there....

Anyway we still need something else other than horses for mounts unless u want use dogs (huskies) for the artic regions as i think allowing horses to live in nothern Siberia and Alaska, etc. is just stretching truth a little too far. It might be possible when there are seasons, but i don't like the idea.

Other than that i think the groups are fine. Although the we shouldn't have 'wild dogs' but just 'wolves' instead.

As for wild creatures we i think they should be modeled even after full domestication in that square and it shouldn't be gradual levels like you have mainly for population reasons of the breeds. FE you could wipe out the entire best breeds in a war and simply lowering the average doesn't model that right because it would take a long time to rebreed to the same level without brining in new animals.

Richard -- when you subclass an object like

'animal'
'horse' extends 'animal'
'beaver' extends 'animal'

there aren't any additional public methods, and there is no additional drain on computational resources, unless the object in question has behavior so complex it needs heavy number crunching.

In this case, all animals would have 'behavior' type objects. In a horse, that object would have different methods from a beaver (or whatever).

But in both cases, the same public methods get called. That's kind of the point of an object hierarchy. You can loop thru 'animal' objects without worrying about what kind of animal it is, and call public void migrate() (or whatever). The 'animal' subclass will know what to do.

I don't think we should model lots of discrete breeds. That would be a mess. The quality scale would work fine; if someone killed all the good horses and dropped the quality by 30%, then it would take a long time to recover. If you traded for good horses, the quality level would quickly rise to that level. I think the quality scale can do whatever the breeds can, because the quality is basically a way to model and average lots of diffrent breeds.

The "Genetic Adaptability" variable could be a good addition to the crops and animals attributes. Sorry I didn't see it when you posted it earlier.

'Clever with the code'?

No, that's standard OO programming. The game *will* be coded that way, in every 'component'.

For example, we have a bunch of 'TurnHandler' classes ('GovtTurnHandler', 'EGTurnHandler', 'CivTurnHandler') that implement a 'TurnHandler' interface with the method 'oneYear()'. So doing 'turn' logic is extremely clean and easy.

Can we code all animals as abstractions and allow particular animals to be instances (in OO terms) of animal classes? That way we'd have:
Mount
Production
Food