RTR Kinnn-iggits! (or, how I learned to stop worrying and love the feudal factions)

[Quinn Inuit]

As a mini-modder (I work on the Extended Realism Mod for RTR-PE), I'm always trying to get the AI to make more historically accurate armies. The problem is that the AI's cost-benefit analysis occasionally slights realism when it produces units.

Wien1938 came up with a good way to force the phalanx factions to produce more realistic armies (high-upkeep units that take 1 turn to make that represent recalled phalangites, and cheaper units that take 2 turns to train that represent newbies), and I've been wondering if something similar can be done for the cataphract factions.

There's been a similar debate going on for awhile in the IBFD forum (http://www.twcenter.net/forums/showt...=130882&page=1) about the Savaran, the feudal nobles of the Sassanid Empire. These were very high-quality cavalry units, and they tended to make up 50% or more of Sassanid armies (AFAIK). The question before IBFD developers, then, is how to make the AI produce realistic, mostly-cav Sassanid armies, while at the same time addressing the fact that Savaran are very good units and you don't want to make them uber-cheap.

The solution that Orthanner and Sarmatian came up with was to try to mimic a feudal system with their government. Leave the feudal units (Savaran, in this case) relatively cheap, but reduce the revenue available to the central gov't from areas that can recruit Savaran.

Several methods have been proposed to accomplish this, but all involve linking the feudal unit production building to some sort of income penalty. Is it possible to reduce tax or farming income? If not, adding an unrest penalty to the building would be a sort of back-handed way of forcing a tax income reduction. You could also remove the trade income and law bonus from a royal barracks.

I think this has a lot of potential applications to RTR. RTR has three factions that behave in a similar manner: the Armenians, Parthians, and Sarmatians. Doing this might get the computer to build more cataphracts, which I've noticed it really doesn't (at least with Armenia).

What do you all think?

[We shall fwee...Wode]

That's a sounds a really good idea. I'm no modder, but I would like parthain armies like the one that killed Crassus at Carrhae rather than those cheap infantry found there.

Also two points about the title, I fart in your general direction!

and now I have to go out and buy Dr Strangelove.

[Orthanner]

Actually, I have already suggested to implement this in RTR VII. But I must notice that cataphracts themselves appeared only at the end of the mod's timeframe, in I BC. Earlier, only the Parthians had cataphract-like super-heavy cavalry (yet not fully armoured). I doubt if Scythia will be included even as a counter-faction for the Sarmatians. It might be the only faction having access to cataphracts since the very beginning of the game.

And yes, I like the Wien's idea to tune phalanx factions. IMHO it is also applicable to the Romans.

[Quinn Inuit]

Hmmm, so you're saying that only Scythia, not Sarmatia, Parthia, or Armenia, should have cataphracts for essentially the entire time frame of the mod?

Also, how would you fine-tune the Romans?

[Orthanner]

In both Sarmatia and Armenia cataphracts appeared in I BC. In Parthia - at the very end of II BC. In Scythia - in IV-III BC (see works undertaken by Khazanov, Gorelik, Meliukova, Nefiodkin and others). So, yes. But I've heard that the developers chose the Roxolani as a counter-faction for the Sarmatians (h'mm... "the Sarmatians" stands for the whole tribe formed up by four alliances: the Aorsi, the Sytraks (sp?), the Alans and the Roxolani - ordered here by power descending. So IMHO it is more correct to use the Scythians, not the Roxolani, as a counter-faction for the Sarmatians; the other option is to use, for example, the Aorsi for the faction itself and the Sytraks for the counter-faction).

About the Romans. I've thought just about pre-Marian units. The hastati should be both cheap and quick to recruit, the principes should be a way cheap (yet more expensive) but take long time to train and the triarii should be very expensive but take rather little time to recruit representing the recalled veterans.

[G_the_Great]

This would make things more complex when recruiting and reduce the number of units in the game.
Can you make the recruitment time in decimals? This might be a good way of allowing many units to be recruited while keeping the basic idea of altered recruitment times.
And don't mention the word. We are not allowed to speak of it.

-Edit-
Aargh I said the forbidden word!

[Quinn Inuit]

Well, I've actually been considering making the Sarmatians a non-playable faction in my mod and using them as an amalgamation of all of the steppe hordes that would occasionally rampage through. I figured I could give them a couple of hard-to-reach settlements so they wouldn't be killed, then spawn the occasional army to represent an invasion force (e.g., the Saka invasion of 150 B.C. that pwned Bactria). Do you all think that would make the game more or less interesting?

Recruitment time in decimals? I don't understand.

I see your point about the cataphracts, Orthanner. I think I could replicate that by tying their recruitment to the Marian Reforms (it's not perfect, but it's a reasonable facsimile of it). But what would I do for pre-cataphract units? I don't know where to acquire any, and I don't have the time or knowledge to create new ones from scratch.

As for the pre-Marian Romans, I thought traditional Roman armies were comprised of roughly equal numbers of hastati and principes, and that none of the types were always recalled veterans. By that I mean that an army was delineated based on the equipment a unit could afford and its experience in battle, as opposed to being newly-trained vs. recalled.

That said, I would like to encourage players to play more like the ancient Romans and maintain large armies only in wartime. Therefore, what about reducing Roman unit recruitment costs but cranking up their upkeep? That will encourage keeping around only the troops you need, but will it damage the AI?

P.S. I'm glad you all have like my thread title.

[pannonian]

That said, I would like to encourage players to play more like the ancient Romans and maintain large armies only in wartime. Therefore, what about reducing Roman unit recruitment costs but cranking up their upkeep? That will encourage keeping around only the troops you need, but will it damage the AI?

Do a bunch of EDUs specific for each faction, and get the player to run a batch file to swap in the right one.

[Quinn Inuit]

That would work, but it would be a pain in the neck for both players and me. I'd hate to do it just to tweak a few unit costs. If there's any way I can avoid it, I'd like to.

[Orthanner]

I see your point about the cataphracts, Orthanner. I think I could replicate that by tying their recruitment to the Marian Reforms (it's not perfect, but it's a reasonable facsimile of it). But what would I do for pre-cataphract units? I don't know where to acquire any, and I don't have the time or knowledge to create new ones from scratch.

You can use, for example, cataphract archer model (there is a texture for the armenian ones remade by RTR team). Also the Sarmatae Katafraktoi model can be used.

As for the pre-Marian Romans. You're right.

[G_the_Great]

I was thinking that it might be possible to recruit multiple units in one turn, eg 2 or 4, by using, eg 0.5 or 0.25, for the recruitment value. (it is probably wise to stay away from 0)
If you were to impliment larger recruitment times for certain units while others have large upkeeps instead then taking 2 or 3 turns to produce basic units would decimate the number of units all factions field. This, I believe, would harm gameplay.
If decimals were used for recruitment times for units which have high upkeep then you could recruit multiple units in one turn and this would mean that you would not need to use large recruitment times (say 2 or 3 turns) for units with low upkeeps.

Another potential problem with this idea is that I have never heard of the AI disbanding units, just like it never (or rarely) retrains them. This means that it could become more vulnerable to the human player's ability to use strategy, something the AI has never been good at.....

[Quinn Inuit]

Orthanner: that makes sense. I'll do some more research and, if it seems appropriate, implement that in a future release.

G: I'm not convinced smaller armies would harm gameplay. In the ancient world most standing armies were very small anyway, with expansions being on a seasonal basis only.

[Wolfstorm]

Do the AI recruits units that take more than 2 turns to train?
For example, during my bactrian campaign, the seleucids brought more golden shields than silver shields. After I reduced the attack of the golden shields by one point, its troops varied more, now with more bronze shields in it. But the silver shields were still uncommon. Cavalry is also uncommon, and I fought stack after stack of phalanx armies. Using cavalry tactics, I was able to kill twice of what the AI was able to kill, routing the enemy frequently.
If units that takes 2 turns are already uncommon, I guess those that takes 3 or 4 turns will be even more so. The player can, for example, train a very strong sacred band that takes 10 turns, but I doubt the AI will do the same.

[Quinn Inuit]

The AI will train multi-turn units, but you have to give it a kick in the pants to do so. Upping the recruitment time on basic phalanx units helped the AI make more varied infantry armies, but I still haven't figured out how to make it recruit more cav yet.

[MasterOfThessus]

Perhaps the AI can't see the difference between cav and inf? Cav generaly has worse stats due to their use of the chargebonus...

[svramj]

I Think The AI identifies units using the class and catagory of the units i.e

infantry, cavalry, etc

light, heavy, medium, spearman, etc

Units with light and heavy infantry are the most recruited followed by spearmen and cavalry, changing the category of the unit may solve the problem.

[Quinn Inuit]

It would, but then the AI would use them like spearmen or swordsmen, which wouldn't really work.

I think reducing their costs might work. I'll test that out and see what happens.