Are you, in fact, a pregnant lady who lives in the apartment next door to Superdeath's parents? - Commodore

Create an account  

 
AI difficulty multiplier tables

As most players are probably aware, there is a table of multipliers for the advantages the AI receives.
This includes :
Population Growth
Outpost Growth
Gold income
Power income
Research from cities
Production in cities
Upkeep costs
Food produced

I noticed the Outpost growth multiplier seems to not work as intended. It's a % chance to receive a fixed additional growth each turn, meaning 100% is the highest meaningful value and 0 is the lowest. Yet the game uses 100-400% for this.
I've decided to eliminate this multiplier and use a hardcoded effect of 25% chance per difficulty level to gain the additional growth.

On the other hand, I used this table slot to implement a new AI advantage multiplier : the overland casting cost multiplier. This scales down the overland cost of spellcasting based on difficulty for the AI, reducing the amount of skill and mana spent, effectively multiplying the amount of spells they can cast. I believe the AI had a disadvantage in this department because they do not use any strategy to cast spells, they just select at random, based on priorities of each spell, yet the amount of skill available for them is limited. Their skill is obviously higher than what the human player has, due to the additional power income, but power does not translate into skill in a linear way. They also build faster, so from the new Amplifying Tower building they can get extra skill, but this is again a quite limited resource because you cannot have more than one per city, while they do build them faster, they won't get more on the long term. On the other hand, a direct discount to casting costs does help this issue, and does so without providing excessive combat casting skill to the AI. Additional overland casting capacity will both help the AI fill more offensive armies with higher rarity fantastic creatures, most of which are rarely seen outside their fortress garrison, and will make them more of a threat when casting curses against the player.
On the other hand, this does make disenchanting stuff quite a lot less effective, simply because they will be recast much sooner, and the costs of dispelling will not be any cheaper, unless another AI is trying to dispel.

I'm considering to the the multiplier 100% for easy and normal (no change), 80% for hard (25% more overland casting power), 66% for hard (50% more), and 50 for impossible (twice as much spells).

I would like to hear some opinions about this because it's quite a major change.
Reply

(March 26th, 2016, 14:56)Seravy Wrote: I'm considering to the the multiplier 100% for easy and normal (no change), 80% for hard (25% more overland casting power), 66% for hard (50% more), and 50 for impossible (twice as much spells).

I would like to hear some opinions about this because it's quite a major change.

I'm fully in favor of it. I think it's a great idea and excited to see it in action, though I'm not sure whether 50% for impossible is overkill.

Reply

I'm playing a game now on Extreme with the 50% cost multiplier and it's working well, I see a lot more fantastic units and curses now.

This might however be bad for Spell Blast. Since the reduction is done by effectively granting the first X percentage of progress in the spell for free, blasting will be tremendously expensive on higher difficulty, on top of having a rather low timeframe to do it (as spells complete faster).
Some changes to Spell Blast might be necessary but not sure what.
Reply

Rather than giving the AI an advantage on casting all overland spells, why not just reduce the cost of mid-tier summons and curses? It's not like those are spells that are much used by players either.
Reply

Seravy has proposed further improvements to summoning units and some overland curses and enchantments. Simply put, they are much more cost-effective than the original game. If the costs decrease even more, it'd break the balance of the game.

Reply

Well, you can probably deal with the spell blast system by instead giving overland-only casting skill. Caster heroes already add overland-only casting skill, so it's clearly possible.
Reply

(March 28th, 2016, 19:15)Anthony Wrote: Well, you can probably deal with the spell blast system by instead giving overland-only casting skill. Caster heroes already add overland-only casting skill, so it's clearly possible.

That could have been an easier way to do it I guess. I went with editing spell blast to count the reduction into the total price if the target is an AI. So it only requires the actual spent mana as always.

That solution would mean the AI still needs the mana to fuel the extra casting skill, while in this one, the bonus applies to both (skill and mana), not sure which is better.
Reply



Forum Jump: