Posts: 10,463
Threads: 394
Joined: Aug 2015
There are no Nature buffs for normal units because it's not necessary. Nature is strong in summoning, so what you want to buff is usually a summoned creature. (Or nothing at all, summoning more can sometimes be better) And you have Land Linking for doing that.
The other reason is redundancy - we don't want Nature to be too similar to Life.
But in reality Nature has plenty of buffs. Water Walking, Land Linking (pathfinding on normal units), Resist Elements, Transmute, later Elemental Armor, Iron Skin, Regeneration. Compare that to the buffs you get for Death...
Posts: 4
Threads: 0
Joined: Jul 2018
(July 18th, 2018, 21:10)Nelphine Wrote: Having spells that buff, directly competes with summons; so it weakens the realm as a whole.
Oh, so that's what's going on, ok. Thanks.
Posts: 10,463
Threads: 394
Joined: Aug 2015
I have spent some time thinking about the AI Transmute problems.
Shared tiles : this would be good to recognize but it's not particularly important. Even if the AI transmutes an ore shared by an enemy city and later destroys it by a raise volcano spell, only a small amount of MP is wasted. Also, the other spells can't recognize shared tiles either, and even if they could, deciding whether to want a shared adamantium or not would likely be hard for the AI. So overall, we can't improve it enough to be worth it.
Multiple ores on a city : this is actually intentional I believe. If there is only one adamantium and it is corrupted, troops produced until purifying completes lose their adamantium weapons. Having a few normal units in a stack of full adamantium weakens the stack overall. With multiple adamantium, the AI is less vulnerable to this. Turning anything to coal might be worth it, but it's questionable - crystals make magical power which is more important for the AI than extra troops, and the only way to reach coal is through crystals. It's probably better to leave gold as is than transmuting it three times to reach coal, especially as gold is versatile while coal does nothing if the city ends up producing buildings.
Transmuting ally or enemy ores : This would need to work differently for human and nonhuman targets, otherwise it can't be scaled by difficulty. However making an ally not help the human but help other AI is really unfair and goes beyond the usual "AI prefers AI in diplomacy" design. The AI selectively making adamantium for your enemies and not for you even when allied to both would be far too obvious. Furthermore, an AI trying to help by giving you an adamantium might be annoying if you preferred to keep the gold. Likewise, transmuting with the intention of removing adamantium would also need to be selective (avoid human on low, avoid ai on high difficulty) which is also bad.
Overall, I believe Transmute is better left unchanged.
July 21st, 2018, 15:43
(This post was last modified: July 21st, 2018, 15:50 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
In which case my rebuttal is: if there are no other wizards with chaos books, they should not make multiple adamantium.
However, in my opinion even if there are wizards with chaos books, they should not make multiple adamantium - if something gets corrupted or volcanoed, THEN turn the next one into adamantium. The amount of power or gold lost (particularly on dwarven cities!) is absolutely ludicrous, and whoever has chaos books may not be at war with that so for years.
Nature also inherently prevents corruption, so that spell is more likely to not even stop the adanantium long. Volcanoes would make it unusable for their enemy and might not even be used.
July 21st, 2018, 15:46
(This post was last modified: July 21st, 2018, 15:47 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
For shared tiles, to me, the more important sharing of tiles is when those tiles are shared by 2 of the Wizards OWN cities. If so, prefer to turn that one into adamantium (to avoid wasting time with multiple transmutes to get both cities producing adamantium.)
For tiles shared with another wizard, prefer not to turn that into adamantium (unless no other ore is available of course) to avoid giving an enemy adamantium.
Posts: 10,463
Threads: 394
Joined: Aug 2015
Transmute isn't an on-demand spell for the AI. The time between losing the adamantium and casting another Transmute can be any amount of turns at random.
The AI won't bother transmuting crystals (3 steps to adamantium) so it'll never lose power that way. It might lose gold or troop production capacity, but the amount should be insignificant, except maybe on Rich maps (which we are planning to tone down anyway). Even then on high enough difficulty (where we would want to apply this), the AI likely has enough production to instantly build most troops without needing the coal. Since there can't be more than one unit produced a turn, it makes no real difference. (Yes, they might get some extra buildings after the troop production from the leftover unused hammers, but abusing coal to build buildings faster is not really something the AI should know about. The human can't do it anyway, as they don't have enough production to build 100+ cost troops and have hammers left over every turn.)
Overall, it's not worth the effort, it would amount to like 0.01% of the AI's economy.
Since the AI turns all tiles into adamantium anyway, doing the shared tile first makes a minimal difference. It's definitely better, but minimally so. Things like location (proximity of enemy forces, how much troops are needed in that particular area) would have a much higher impact and we aren't even going to bother with that.
Also, if it's a shared tile, as targeting is done on a per-city basis, those tiles have a better than normal chance to be selected (best city is targeted first, so double chance to be best city, PLUS if it's an area with two AI cities, it's more likely to be in the center of the empire where the AI has their most developed cities that get targeted first.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
Again, in that case, if there's no other wizards with chaos books, don't do multiples of adamantium.
However, if you are going to tone down rich, that would make it much less of an issue. On fair you occasionally see 3+ adamantiums, but rich you can see 6+ and it just irks me.
Posts: 10,463
Threads: 394
Joined: Aug 2015
I can't find the thread where we discussed the problem of Mineral setting and ways to improve it. Any idea where to look?
Posts: 5,010
Threads: 17
Joined: Aug 2016
I don't remember where... We were discussing density and how many ores to place in a given number of tiles, and the map was split into areas... I want to say 5x5? And the problem is that you could have all the ore from one square in a corner right next to all the other from another square an end up with unbelievably good cities.
Posts: 10,463
Threads: 394
Joined: Aug 2015
|