AU Mod: Re-balancing the power of Armies

[nbarclay]

I think eliminating MGLs would be a bit bigger departure from stock than the stock rules than is appropriate for the AU Mod (to put it mildly).

[jshelr]

Yup, it' would be a big change. But, I suspect none of the fixes mentioned would really change the basic truth that leaders help the human player vs the AI. Personally, I need all the help I can get.

[lockstep]

Originally posted by lockstep
... another (and quite conservative) method to reduce the power of armies is to remove the 'increased army value' flag from the Military Academy. This should reduce the attack/defense bonus by one third (i.e., from 75% to 50% for a 3-unit-army, and from 100% to 67% for a 4-unit-army).

In addition to the other changes proposed for the Military Academy, I suggest to put this change 'under consideration'.

[alexman]

Yes, we should.

It's too bad this change would weaken Armies only after they can already be built by the AI (assuming the victorious flag gets removed from the Military Academy).

[lockstep]

Yes, but armies would still be very attractive after this change. At the moment, they are insanely attractive.

[nbarclay]

(Responding to a post Theseus made in the AU Mod thread and to a couple responses to it by vmxa1)

With the AU Mod, the Military Academy does not require a victorious army to build and it produces a free army every 25 turns. Therefore, AIs do end up with armies.

Whether they use them or not, however, depends on their composition. In AU 503, the only AI armies I ran into were single-unit cavalry armies parked in the Viking capital. When the first unit in an army is a cavalry, nothing else can be added to it because no unit exists that can match a cavalry army's movement rate of four. As a result, the army actually hurts the AI by causing cavalry to sit in a city as defenders waiting for the army to be fully populated instead of being available for offensive use.

But in AU 601, I got to watch a few AI armies in offensive action. The compositions I remember are one guerilla and two cavalry, and one infantry and two tanks. The AIs even made use of the armies' free pillaging capability.

I do have a hypothesis as to what might account for the difference. In AU 503, the Vikings enjoyed an extended period of peace before I took them out in a rapid blitz. As a result, they had a few cavalry left over from earlier in the game that they could use to start armies, and those cavalry were at home in their cities rather than out fighting.

In contrast, in AU 601, there was a period of several centuries in which I kept the AIs involved in almost constant wars with the rest of the world ganging up on a single civ. That situation sent most if not all of the AI cavalry reserves out in the field where they died, and after Replaceable Parts, AIs prefer infantry or guerillas as offensive units. So as new armies were turned out by the Military Academy after Replaceable Parts, their initial unit was a guerilla or infantry, making it possible to add cavalry or tanks to complete the army into something useful. I don't know where the cavalry came from to finish out the armies that included them, but as I recall, two AIs came up with them somewhere, somehow.

Note that under this hypothesis, an AI that loses its cavalry in a war fairly shortly after Replaceable Parts becomes available is likely to have useful armies once tanks become available if not before, while an AI that stays at peace and holds onto a significant cavalry force is more likely to waste its armies initially populating them with cavalry. Thus, the general tone of the game would have a serious impact on whether offensive AI armies will be seen.

Is this hypothesis right? I don't know. At the moment, I'm basing it on only two games, which is not anywhere near enough to develop a scientifically sound conclusion. It would be interesting to hear from others regarding AU Mod games where they have or have not seen AI armies in the field to find out whether their games seem to fit my hypothesis.

Nathan

[Jaybe]

I use similar Army-creation mechanics in my mod (thanx for the inspiration!), and whether armies stick in a city with 1 unit, or maximum-unit armies out in the field can be quite variable. Definitely do not count on them all being used effectively, but yes, sometimes VERY effectively.*
Also, remember that an Army containing one unit gets 1/6 advantage.

*Having replaced artillery with strong "assault" units, I have seen an AI Army containing 4 of the latest assault units take another city.

[Theseus]

Why is this thread not showing in the AU forum?

[Edit: Strike that... this now magically appears. ]

[Theseus]

I just found a 1xMedInf Army in the Ottoman capitol... they never had horses, and had not gotten to Tanks, so the mofo just sat there... unloved, unwanted, and un-used.

[Theseus]

And a second one... tragic.

[player1]

Well, that's why I removed from Military Academy an ability to build armies completly in my mod.

Players will be rewarded with armies for smart warmogering through leaders, but won't have tons of armies after military tradition.

[JimiD]

Just finished a game with the AU mod, standard size, pangea. It seemed most of the AIs had a number of armies in the capital, but only one had more than 1 unit. I dont even remember what the 2 unit army was, I was steam rollering with Tank armies and arts. I doesnt matter what your up against with those..

Got most of my armies from MGLs though.

Got to admit the omission of MA, although harsh, appeals to me - surgery for cancer.

and I havent even started on artillery....

[Mad Bomber]

With the AU Mod, the Military Academy does not require a victorious army to build and it produces a free army every 25 turns. Therefore, AIs do end up with armies.

Whether they use them or not, however, depends on their composition. In AU 503, the only AI armies I ran into were single-unit cavalry armies parked in the Viking capital. When the first unit in an army is a cavalry, nothing else can be added to it because no unit exists that can match a cavalry army's movement rate of four. As a result, the army actually hurts the AI by causing cavalry to sit in a city as defenders waiting for the army to be fully populated instead of being available for offensive use.

But in AU 601, I got to watch a few AI armies in offensive action. The compositions I remember are one guerilla and two cavalry, and one infantry and two tanks. The AIs even made use of the armies' free pillaging capability.

I do have a hypothesis as to what might account for the difference. In AU 503, the Vikings enjoyed an extended period of peace before I took them out in a rapid blitz. As a result, they had a few cavalry left over from earlier in the game that they could use to start armies, and those cavalry were at home in their cities rather than out fighting.

In contrast, in AU 601, there was a period of several centuries in which I kept the AIs involved in almost constant wars with the rest of the world ganging up on a single civ. That situation sent most if not all of the AI cavalry reserves out in the field where they died, and after Replaceable Parts, AIs prefer infantry or guerillas as offensive units. So as new armies were turned out by the Military Academy after Replaceable Parts, their initial unit was a guerilla or infantry, making it possible to add cavalry or tanks to complete the army into something useful. I don't know where the cavalry came from to finish out the armies that included them, but as I recall, two AIs came up with them somewhere, somehow.

Note that under this hypothesis, an AI that loses its cavalry in a war fairly shortly after Replaceable Parts becomes available is likely to have useful armies once tanks become available if not before, while an AI that stays at peace and holds onto a significant cavalry force is more likely to waste its armies initially populating them with cavalry. Thus, the general tone of the game would have a serious impact on whether offensive AI armies will be seen.

Is this hypothesis right? I don't know. At the moment, I'm basing it on only two games, which is not anywhere near enough to develop a scientifically sound conclusion. It would be interesting to hear from others regarding AU Mod games where they have or have not seen AI armies in the field to find out whether their games seem to fit my hypothesis.

Nathan

I made some comparable modifications in my personal scenerio and I am seeing results similar to yours (Palace produces an army every 100 turns, and the M. Academy produces one every 10 turns). From what I have observed (in limited games) is that the AI has the Army kept in the capitol until the army is completely filled with units tasked with the offensive role, once the army is filled then it sends the army out to attack. the overall results are encouraging; but only to a point. The AI will not be able to build armies as good as the human player does without a new patch which adresses the AI army behavior.

[alexman]

After playing more games under C3C (AU 601 as Rome anyone?), and after player1's discovery that the AI does not populate Armies with a slower unit than the speed of the Army, I am more convinced than ever that Armies are seriously broken in C3C and need a drastic change to be fixed.

So since it looks like there will be no other patch any time soon, what would you think about this drastic change: Reduce Army transport capacity to 1!

This change has been suggested before (I think it was lockstep [edit: ducki], although he suggested a reduction by 1), but back then we had not yet discovered how broken Armies were in C3C.

Pros:

1) Strategic choice of what to do with an MGL. It's no longer a no-brainer to make an Army. Perhaps we will see more Palace moves, or even city improvement rushes.

2) Vast improvement of the AI. The AI no longer has the problem of using Armies once it gets them, since 1 unit is all that can be loaded. Also, unlike what happens now, the AI will have no problem attacking 1-unit armies since they do not get more HP than a regular unit.

Cons

1) Definitely a big change

2) Armies in C3C are fun, as long as you are the one using them.

I think that just the movement bonus to a unit will be worthwhile, and will give a nice dilema of what to do with an MGL. In order to give back some of the benefit lost by militaristic civs, I propose to decrease the number of cities required for an Army to 1. The Military Academy might need a small boost as well. Perhaps more frequent Army spawning.

What do you think?

[JimiD]

I assume the +1 move is hard coded and not editable?

I am at work and cant check.

Stupid question: I am sure it would have been the first thing player1 checked on discovering the reason they dont get filled.