Balancing Unit Costs

[Dol Guldur]

Balancing unit costs is a nightmare. I have been working and reworking a very long formula for months and it never seems to work out right. Today I added a bracket too many and my poor old Excel 97 spreadsheet crashed...

I'm going to go back and study any trends in vanilla and see if I can work out a formula.

How have you balanced costs in your mod? What tips can you give me?

[Bardo]

There is an interesting topic about this issue here: Total War Center Forums/Rome: Total War Hosted Modifications/DarthMod (DM)/Balancing unit costs.

Sorry if it is from other forum, I haven't found any similar topic in this forum.

[Dol Guldur]

Thx. Bardo - they do hide things away on TWC ;)

Some great discussion, and a few principles gleaned though no real conclusion other than that no one's come up with a formulaic method :)

[econ21]

It would be great if people could share their formula for costs.

I remember WesW saying he constructed one for his MTW MedMod when it was just in its infancy. But I never saw it and get the impression RTW modders often rely more on judgment than mechanical formula.

CA does not seem to cost things according to their effectiveness - the SP upkeep is based on assumptions about troop types, rather than stats. And even the purchase costs, which you would think should be based on stats to balance MP, don't always seem to be justifiable on those grounds.

[nikolai1962]

Well, my way is probably crap for custom battle or multiplayer but what I did was...

a crappy region with a town of 1800 population and no real trade gives about 600 income.

So i made the cheapest garrison unit cost 600 with an upkeep of 300 as my baseline cost. I then ranked faction units according to the barracks level they were built at and added 200 cost and 100 upkeep for each rank with an extra 200 for cavalry.

(Sophisticated system huh )

The point of my system isn't to make the cost an accurate reflection of the unit's capabilities but to make the AI produce armies that make sense for the faction. Plus i don't care if a faction has a better unit for the same cost if historically they had an advantage in that area. e.g the numidian light cav is better than, but costs the same as, everyone else's javelin cavalry.

Needs tweaking obviously but as a base system it is working quite well with the AI producing armies that "fit".

Also your base cost needs to reflect how much you like having lots of full stacks in the game. I prefer everyone having fewer, smaller armies and so my costing is quite extreme to make all the factions struggle to pay for their garrisons.

[Makanyane]

I don't have an answer to any of this but am not sure about having such a high proportion of upkeep to build cost, e.g.

So i made the cheapest garrison unit cost 600 with an upkeep of 300 as my baseline cost....

having played mods that had similar ratio, it soon became apparent as human player that it wasn't worth maintaining units that weren't immediately required for battle / essential garrison, so it was more economical to disband units not needed in back areas and rebuild later on front-line instead of moving them forward.

I can see how system would work between AI factions to produce desired army. But am worried that the human player may be more adept at avoiding the higher maintenance costs (as above) and hence add to the AI's disadvantage ?!

[nikolai1962]

Yeah. I didn't mention i've increased unrest everywhere too so you *need* those garrison units to stop revolts. I'm basically aiming at trying to make it difficult to expand your empire even if the AI wasn't there as opposed to having huge numbers of AI armies, as I get a bit bored with that. Not finished yet though so i may find you're right.

Regardless of the upkeep issue though I think setting your costs purely to manipulate the AI into producing good armies is worth thinking about. Have a very simple formula to start e.g 200 per level of barracks, and then test and tweak it until the AI factions are producing the kind of armies you want them to have.