Announcement

Collapse
No announcement yet.

Design: Logistics (Abstract)

Collapse
X
 
  • Filter
  • Time
  • Show
Clear All
new posts

  • #16
    Na being at work..........still haven't got connection at home

    I am actually afraid to get a connection at home.........then, me thinks, I will get no sleep at all anymore..............

    Comment


    • #17
      Seeing how much time you spend here during working hours, I can only agree to that
      Administrator of WePlayCiv -- Civ5 Info Centre | Forum | Gallery

      Comment


      • #18
        J Bytheway,

        My apologies, but at the time I was unaware of the population modelling thread. It does provide an interesting problem...


        D.
        "Not the cry, but the flight of the wild duck,
        leads the flock to fly and follow"

        - Chinese Proverb

        Comment


        • #19
          Originally posted by Gen.Dragolen
          J Bytheway,

          My apologies, but at the time I was unaware of the population modelling thread. It does provide an interesting problem...
          Don't worry about it - it can be tough to keep track of what's going on, even in a relatively quiet forum like this one.

          Comment


          • #20
            1) Looking at the additions to movement cost. I think a nice touch would be when you use the "go to" function that tells how many moves it takes that a box would come up as well telling you the gold/PW/food cost as well. This way when you plan army moves it would calculate all the costs for the moves. It would be a great war planning tool as you determine how expensive an attack/invasion would be.

            2) I'm thinking that ancient units like warriors, should mainly have a move cost in gold since foraging was common in older units. More specialized units would have a more expensive logistics system (legions vs barbarians). i guess that can all be worked out if it possible to implement this.

            3) I'm debating how ships logistics should be done, since they donthave a logistics tail like land units. You may have to pay up front for all movements from harbor/cityport to next port. This will make the galley sinking thing different since movements would have to be hopping from port to port with a supply load.

            4) linking it to above ships should have a range as well almost like aircraft where they have to travel to ports, have a ship floating far off shore for long periods isnt a good abstraction of naval war. If you plan patrols from port to port, i think its a better abstraction (nuclear power would change this).

            more later....
            Formerly known as "E" on Apolyton

            See me at Civfanatics.com

            Comment


            • #21
              Originally posted by E
              1) Looking at the additions to movement cost. I think a nice touch would be when you use the "go to" function that tells how many moves it takes that a box would come up as well telling you the gold/PW/food cost as well. This way when you plan army moves it would calculate all the costs for the moves. It would be a great war planning tool as you determine how expensive an attack/invasion would be.
              I am not sure about it.......This shall come a bit as a surprise. If you feel you have enough resources, make war, but don't be surprise if you haven't To have an overview in the nation manager is OK, but only AFTER moving, me thinks.

              2) I'm thinking that ancient units like warriors, should mainly have a move cost in gold since foraging was common in older units. More specialized units would have a more expensive logistics system (legions vs barbarians). i guess that can all be worked out if it possible to implement this.
              Ancient units no PW/Food....., if you see it as such a big problem, we can make them 'cheaper'. But having exceptions, this we shall have only for some 'special' units like explorer (which makes real sense) and maybe spy and abolonist (sp?). Even a slaver for me shall pay maintance cost.

              3) I'm debating how ships logistics should be done, since they donthave a logistics tail like land units. You may have to pay up front for all movements from harbor/cityport to next port. This will make the galley sinking thing different since movements would have to be hopping from port to port with a supply load.
              New unit? But I see problems with the AI. And again in an abstract model, what we want to use, the logistic tail could be seen by invisible transports.

              4) linking it to above ships should have a range as well almost like aircraft where they have to travel to ports, have a ship floating far off shore for long periods isnt a good abstraction of naval war. If you plan patrols from port to port, i think its a better abstraction (nuclear power would change this).
              only nuclear? WWII submarines were already for extended periods on missions. Main reason why ships had to get into port was repair (including normal maintenance repair).
              But I agree: being able to have a bloody sail/trireme out of harbour for centuries doesn't make sense.....The ship shall have already rotten away....

              Maybe we could implement a 'fuel'-counter equivilent, which is being refreshed every time you enter a city. With following rules:

              Normal city: 34% of max. being refilled
              City with harbour: 100% being refilled.

              This would make harbour more useful. But again: Can the AI handle it?

              more later....
              Let's see

              Comment


              • #22
                Just had another brain fart about the units: we really need to shorten the production times.

                Why in heaven's name should it take a century to produce an artillery unit when you have a modern factory in a city wth a decent production level ? The game balance should be in how many units you can actually field with the logistics costs involved which should keep the players and AI from having a raging horde of units that took only a few years to make...

                Was playing for a little while last night and it struck me as odd that I was able to produce machine gunners by about 1200AD (went postal on two of my continental neighbours early in the game and had about 25 cities by 1000 BC) yet the first ones weren't in service until almost 1400 AD.

                This has always annoyed me because it represents a gross over-simplification of reality. Real life has shown that you need hordes of gold to raise an army in a relatively short time, they cost a lot to keep in the field and are usually disbanded at the end of the war.

                This may seem complicated, but if you can only afford to keep a small garrison in each town, and only one or two armies in the field, it certainly makes the job of invading and conquering someone a lot trickier. Especially if the enemy's army shows up on your door step and you have most of your forces commited at the front, or on the other side of the kingdom.

                You should be able to crash together some conscripts or levies and put the in the field, with the understanding that the casualties will be severe and you may learn to really hate the people who designed the AI's logic and stategies. Not to mention that it should really bankrupt the Exchequer.

                The decision tree shouldn't be that hard to workout either:
                - Assign a garrison strength as a default
                - Units cost so much per turn to maintain in the field per turn
                - Target cities are so many turns away
                - Send out the most number of units we can afford in as large a group as possible
                - If battle is successful, continue if money holds out
                - If battle unsuccessful, retreat to strongest defensive position to await reinforcments
                - If no reinforcements, then return to friendly territory and begin cease-fire negotiations

                Shouldn't be too bad to code given that they have some decision trees to work with already.


                D.
                "Not the cry, but the flight of the wild duck,
                leads the flock to fly and follow"

                - Chinese Proverb

                Comment


                • #23
                  [QUOTE] Originally posted by Gen.Dragolen
                  Why in heaven's name should it take a century to produce an artillery unit when you have a modern factory in a city wth a decent production level ? The game balance should be in how many units you can actually field with the logistics costs involved which should keep the players and AI from having a raging horde of units that took only a few years to make...
                  [quote]

                  Don't really disagree, BUT as Locutus and me discussed in here armies were having fairly huge numbers. I do agree that the AI needs some 'polish' to really improve their handling of the armies, but completly restrict it Yes I think we shall make support more costly, but not too costly, me thinks it would favour again the human.

                  Was playing for a little while last night and it struck me as odd that I was able to produce machine gunners by about 1200AD (went postal on two of my continental neighbours early in the game and had about 25 cities by 1000 BC) yet the first ones weren't in service until almost 1400 AD.
                  I'll try to find some links, but in the moment, just quickly out of memory.

                  Whenever somebody invented/discovered something it took 'ages' to implement. Taking your example of a machine gunner: The first working 'machine-gun' has been invented in America (AFAIR) (Gatling). That was somewhere around 1862. But full usage it only experienced far later in WW1 were you really had platoons equiped with machine guns (no don't remember the names anymore, expect one was called maxime). So thisdelay makes sense. Even other stuff took ages to implement. Think of the steam engine or just the automobile. I could go on for ever So this kind of behaviour I consider normal.

                  The rest is a bit hard to reply upon, but I try:


                  This has always annoyed me because it represents a gross over-simplification of reality. Real life has shown that you need hordes of gold to raise an army in a relatively short time, they cost a lot to keep in the field and are usually disbanded at the end of the war.
                  Let's gather some examples again: I'll try to post them in the above mentioned thread.
                  So long you are defending your country, the leader hardly ever needed hords of gold. For invasion yes. (to say it short)

                  When you implement a new idea and people are willing to follow you, the costs are fairly low. (like in selfdefence or crusades) It also depends on the kind of government.

                  This may seem complicated, but if you can only afford to keep a small garrison in each town, and only one or two armies in the field, it certainly makes the job of invading and conquering someone a lot trickier. Especially if the enemy's army shows up on your door step and you have most of your forces commited at the front, or on the other side of the kingdom.

                  You should be able to crash together some conscripts or levies and put the in the field, with the understanding that the casualties will be severe and you may learn to really hate the people who designed the AI's logic and stategies. Not to mention that it should really bankrupt the Exchequer.
                  Yes, now there you come to the point where armies get costly:

                  Paying somebody to do it, or in other words mercenaries. But conscripts are cheap, even if you have to train them.

                  This principle has dragged down (as well) the roman empire. The extreme cost to pay for their mercenaries.

                  Coming to modern times like WW1 and 2: The armies were mainly concisting out of conscripts with a 'small' overhead of trained people. But keeping the cost low. Actually is was cheaper to replace people than equipment. That this has a huge disadvantage is another problem. (serious one).

                  The decision tree shouldn't be that hard to workout either:
                  - Assign a garrison strength as a default
                  - Units cost so much per turn to maintain in the field per turn
                  - Target cities are so many turns away
                  - Send out the most number of units we can afford in as large a group as possible
                  - If battle is successful, continue if money holds out
                  - If battle unsuccessful, retreat to strongest defensive position to await reinforcments
                  - If no reinforcements, then return to friendly territory and begin cease-fire negotiations

                  Shouldn't be too bad to code given that they have some decision trees to work with already.
                  This decision tree is already assuming too much. The little part called strategy/tactics. Which city to attack? Which units to send? and so on........

                  Humans could say: No I don't take the big city in front of me, instead I take the city on the mountains with more production and decreasing the AI's production or similar. Or just taking a city in his network to disturb his 'traffic'. Also why shall anyone bother about taking size 1 cities? (including the AI). Those you would normally take while going after the big game.

                  It is not that it is bad, but too many parts missing before.......

                  Comment


                  • #24


                    Actually while responding here, I was thinking of something completly different:

                    Instead of having cities producing units, why shall we not change this principle completly:

                    Every city (can) allocates a certain percentage to unit production and there will be just ONE empire queue for which units to produce. Those units will be 'produced' were you have barracks or the capitol. Starting with the capitol to field units. This would be also a bit more reallistic.

                    Comment


                    • #25
                      Gilgamensch,

                      I like the last idea about having an empire queue working in conjunction with improvements like barracks. Troops from barracks, ships from shipyards, tanks and aircraft from factories.

                      But I beg to differ about the expense of keeping conscripts or levied troops in the field. It isn't the cost of paying the people, but the cost of feeding them, and keeping them in arms and armour.

                      A man marching for 8 hours a day will need at least 1kg of food and 2 litres of potable water. A horse will need about 5+ kg (any horse owners out there ?) of fodder and 4+ lites of water When you start adding it up for an army, and it's a big cost. To put it in perspective, a Roman Legion of 4000 men would then need 4 tonnes of food and 8000 litres of water every day. Even if they were able to carry food in their packs, they could at most carry enough for 3-4 days before it was too heavy to manage with the 30-40 kg of kit they also had to carry. This is why troops to this day will often forage along the way.
                      Sun Tzu commented that 1 cart load from the enemy was worth 20 carts brought from your own land.

                      Consider the cost of equipping troops: a good sword smith might turn out 20 swords a month, assuming he had access to enough iron and charcoal, it makes his work worth a lot of gold. Most of the traditional weapons used by the common troops were spears, polearms and wood cutting axes, because weapons like swords were so expensive that only nobility could afford them. The same applies for the armour.

                      Mercenaries were often used as garrison troops and they were expensive, but so was the cost of maintaining armour and horses.
                      But they were cheaper than paying to have a large number of your own people under arms since they were no longer working the farms and businesses that made the kingdom wealthy.

                      Fast forward to WWI, and the cost of keeping the army in the field bankrupted most of the countries who fought in it, with the exception of Canada and the US. The single biggest cost for the British forces was fodder for their horses, and they had more horses than men on the Western Front. It follows that the Great Depression was partly due to the inflation resulting from the debts incurred from WW I.

                      Now as for the bleeding edge of weapons technology, yes there were many early instances long before they came into general use. The Chinese had things like gunpowder about 1000 years before it came into common use in Europe. But they didn't make massive use of it on the battlefield.

                      So it may be that we need to have a few more units to represent the different technological advances. Jumping from Pikemen to the British Infantry from the Boer War to Machine Gunners in WW I is a bit of a leap. If we made the units faster to produce, then we could actually get them into play before someone comes out with modern troops like Hover Infantry.

                      And yes, the decision tree I wrote up is very incomplete. In code form, it would likely be a couple of hundred if { } then { } statements just to decide which city to attack first. And then it would be tied into alot of other decision trees. I have not looked at the source code to see how they did it because I'm not a programmer by trade. Much of the detail in the code they used to build the AI would be lost on me. I have to rely on what passes for logic in my brain to figure how it could work.

                      I know that the challenge is to make the game balanced and playable, but I think it can be done fairly well with a good level of historical detail. Much of this is simply a matter of deciding on a model and making it work.

                      D.
                      "Not the cry, but the flight of the wild duck,
                      leads the flock to fly and follow"

                      - Chinese Proverb

                      Comment


                      • #26
                        Gen.Dragolen

                        I'll reply tomorrow, but already to be said now:

                        I fairly agree with you

                        Comment


                        • #27
                          I like the idea of cheaper to produce units, but more costly (in pw/food/good) in maintaining. IRL any country that had a large army, either went banckrupt or make a (profitable) use of that army.

                          For ancient and medieval age an easy to implement (I hope) idea about food for troops : if in cities/fortifications radius, the units are supported by the empire pool. If between empire's border, but not in cities' radius, they got only half from empire's border, the rest from the tile where they are (foraging). When outside the empire's borders, the troops are only on foraging. All tile will have a value of food that can be foraged by troops, and if that isn't enough for the unit, that unit will lose some HPs (starvation, sicks, desertion etc). That value should be tied in a way to the total HPs of troops. A small unit or half of a bigger unit will be able to sustain itself. In real life seldom we find large army marching in close ranks. Larger armies were in long colums or separate parallel colums that reunite when a battle was expected.

                          Ships can have a value similar with fuel, but named "food/fresh water provision". A city with harbour will full refresh it in a turn, a city without a harbour, in 2-3 turns, and ending the turn near coast will give a few points (the amount consumed in a turn or - better - less)

                          We can also have a "supply depot", very cheap that will have the role of extending the area where units are supplied fully from empire pool. There should be a complete chain to a city to function. That and the fact that will not provide any defence bonus (unlike fortifications) will make the supply line vulnerable.

                          Modern units could cost more production (ammo, fuel, spare parts etc) and for that a supply line will be absolutly necessary.

                          I don't know how coherent I was, but I hope that you understood me.
                          "Respect the gods, but have as little to do with them as possible." - Confucius
                          "Give nothing to gods and expect nothing from them." - my motto

                          Comment


                          • #28
                            Mihai,

                            You have an excellent bunch of ideas there. And you beat me to the punch with your post: as I was skimming through the previous posts, the many of the same ideas occured to me.

                            I really like the idea that each terrain type should have a 'forage' value to rate how much food and water it could supply of units in the field. It would be simple to program and would make for a significant strategic concern: no more single warrior units wandering around the world.

                            It would also make for a more interesting game if your movement had to include keeping your unit health up so they can fight when they get to the battle field.

                            I would suggest that there is a better way to determine just how much support a unit can get when in garrison. They would have to live off the land the same as if they were in the field, but every improvement to the land or the city or the government increases the capacity to support units. Happy peasants/workers are more productive and therefore food and goods more plentiful.

                            And it would mesh nicely with the tennants of the different governments: they can be rated for corruption which would increase the costs as you move further away from the Capital. Perhaps a more localized tax rate to make up the short fall would be in order since and empire wide one glosses over the problem. That or some more improvements that would lessen the local corruption other than just 'Court House' ?

                            And yes, the need for a pair of ports to have a supply line would be a good idea too. It would make the need to capture a city with a port essential for operations just as it has always been in the past. At least until the civ developes 'Amphibious Warfare'.

                            I think what we need to start doing is deciding on what would be some useful and meaningful tests for the people doing the programming to use. Been introduced to the concepts of extreme programming and that is usually the first step after you establish your requirements.

                            Any suggestions ?


                            D.
                            "Not the cry, but the flight of the wild duck,
                            leads the flock to fly and follow"

                            - Chinese Proverb

                            Comment


                            • #29
                              When I thought at the tile forage value, I was thinking that Hannibal lost some troops when he crossed the Alps (most of his elephants) and that many armies disapeared in the deserts. Also that would force players (and AI - that could be a problem) to think more about how to move their armies - how to arrange the colums of the armies.

                              I think that single warriors will be able to wander around the world, as they represent a small army that can live on the field (exept montains, deserts and polar), but if they enter on your land you can attack them with superior forces as you have less problem with supply.

                              When is a shortage of food civils suffer first, not military. Maybe the food formula should be like that : gathered - corruption - units upkeep = what is left for civilian population. If the result is more that necerary, then the population grow occurs. Starvation should come with a high happiness penalty.

                              I am very poor at programing. Last program I made was in highschool in Turbo Pascal.
                              "Respect the gods, but have as little to do with them as possible." - Confucius
                              "Give nothing to gods and expect nothing from them." - my motto

                              Comment

                              Working...
                              X