February 27th, 2017, 06:53
Posts: 5,010
Threads: 17
Joined: Aug 2016
So, interesting thing. Started playing my halfling game. Defeated Ariel fairly early - her capital didn't even have 9 units yet (pure life, so no summons, just klackon swordsmen/halberdiers, and 2 heroes).
She had settled 4 cities - I destroyed 3 on my way to the capital - the fourth, her best, turned neutral.
I can't conquer the neutral city. Ariel had cast heroism and endurance on 6 of the halberdiers guarding it - and of those 6, 3 also had holy weapon and holy armor.
So 2 questions arise: why was this city so much better defended than her capital (I actually know the answer to this, so I'm going to rephrase - how do we teach the AI to buff his capital and offensive stacks that way? That's a proper life doomstack! I'm so impressed she did it, just sad she did it to the wrong one...)
And how does a neutral pay the upkeep? I'd be OK if it could maintain whatever upkeep his buildings allowed, but, klackon, they can't maintain that much. (Unless they get difficulty bonuses? Its only extreme though).
February 27th, 2017, 07:07
Posts: 10,463
Threads: 394
Joined: Aug 2015
(February 27th, 2017, 06:53)Nelphine Wrote: So 2 questions arise: why was this city so much better defended than her capital (I actually know the answer to this, so I'm going to rephrase - how do we teach the AI to buff his capital and offensive stacks that way? That's a proper life doomstack! I'm so impressed she did it, just sad she did it to the wrong one...)
And how does a neutral pay the upkeep? I'd be OK if it could maintain whatever upkeep his buildings allowed, but, klackon, they can't maintain that much. (Unless they get difficulty bonuses? Its only extreme though).
The AI already targets units in the capital more. However, before turn ~40 the AI is allowed to push out units from the capital to expand, so these units don't stay there. Simply put the AI is designed with the assumption that no one will attack before turn 40. (or 35 or 50 not sure which is the current number)
The best way to fix this is by lowering this turn count even further.
The Neutral "player" doesn't have a real turn, gold reserve and doesn't pay upkeep whatsoever. Instead, neutral cities have a mechanism for disbanding units if they have too few population to support them (2 units per pop currently). I'm surprised the spells don't disappear when the wizard does that's probably a bug.
February 27th, 2017, 07:38
(This post was last modified: February 27th, 2017, 07:43 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
A further problem, if I remember correctly, is that buffs are cast on highest strategic strength stacks. While the capital is often the best units, they aren't based on strategic strength (this often leads to 9 sprites in a capital, even though there are masses of halberdiers roaming around - a solid mix would often be better) So if the second city had 6 halberdiers, that would be stronger than the halberdier/swordsmen/hero mix in the capital. Then, as buffs were cast (especially heroism), the second city strategic strength would go way up. Also given how many buffs were cast, its quite possible they started long before either city had 6 units, so its quite possible lowering the turn count wouldn't have done anything in this case.
February 27th, 2017, 07:47
Posts: 6,457
Threads: 134
Joined: Aug 2004
That explains why I was able to take out AI fortresses quickly. I suggest lowering it to 30 turns.
February 27th, 2017, 08:58
Posts: 10,463
Threads: 394
Joined: Aug 2015
(February 27th, 2017, 07:47)Catwalk Wrote: That explains why I was able to take out AI fortresses quickly. I suggest lowering it to 30 turns.
Checked, and it's set to turn 35. Maybe 25 would work better? 30 wouldn't make much difference.
Quote:A further problem, if I remember correctly, is that buffs are cast on highest strategic strength stacks.
Not entirely.
It's 100% of the target's value but 25% of the other units in the stack so the best stack is likely to get targeted if it's better by a large margin, but otherwise a weaker stack containing a stronger unit takes priority.
The fortress is worth +100, which is the value of a pretty good unit, considering units are counted at 25% value I'd say enough to outweight 8 halberdiers, so the majority of buffs will go there even if it's only 8 sprites and a singe haberdier - of course they are invalid for most Life buffs anyway, being fantastic units and we assumed the target itself is a halberdier either way.
As the capital is larger, it's unlikely to have more halberdiers somewhere else, unless they came from the capital itself and got pulled in to defend the new city.
February 27th, 2017, 09:39
(This post was last modified: February 27th, 2017, 09:56 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
Wait, a halberdier (klackon) has a defensive rating of 528 at level 1 or 2 (I don't remember exactly when halberdiers get armor 5).
If its level 3 its defensive rating is 792.
Without even accounting for the offensive rating, how does the capital +100 remotely compare to that, even with the 25%?
I would expect at least +500-750 if you wanted it to be roughly equal to 1/4 of a strong unit. If the strength is a direct comparison to its offensive and defensive ratings.
Can we also add half the fortress bonus to the strongest stack that is not in any city or node? That way its offensive stack has a better chance than non capital garrison's?
February 27th, 2017, 09:47
Posts: 10,463
Threads: 394
Joined: Aug 2015
(February 27th, 2017, 09:39)Nelphine Wrote: Wait, a halberdier (klackon) has a defensive rating of 528 at level 1 or 2 (I don't remember exactly when halberdiers get armor 5).
If its level 3 its defensive rating is 792.
Without even accounting for the offensive rating, how does the capital +100 remotely compare to that, even with the 25%?
I would expect at least +500-750 if you wanted it to be roughly equal to 1/4 of a strong unit. If the strength is a direct comparison to its offensive and defensive ratings.
Unit rating is Defense Rating * Offense rating /8192 capped at 1024. So a unit with 800 offense and defense rates 78 on that scale. (and that's a very above average halberdier)
February 27th, 2017, 11:00
(This post was last modified: February 27th, 2017, 11:05 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
Ah OK. I would have to assume it buffed the first halberdier in the capital, then sent it out, then the city ended up with one or two more units than the capital, so it buffed a second halberdier there (2 heroism halberdiers would be 93 rating between them with that formula, enough to match the capitals bonus.
Level 3 Hammerhands though could certainly be higher than the 100 the capital gives. I'd suggest putting in 250 for capital (that's still weaker than my doomstack halberdiers or very rares, but about on the level of strong rare summons), and 100 for the strongest stack not in a city or a node.
Or if you want to prioritize offense, 150 for capital and 125 for the other stack.
February 27th, 2017, 11:19
Posts: 10,463
Threads: 394
Joined: Aug 2015
Actually, it's +50 in the fortress, I checked again.
Quote:(2 heroism halberdiers would be 93 rating between them with that formula, enough to match the capitals bonus.
This isn't true : as I said units in the stack are only considered at 25% value. On the target itself is at full value. So you need 4 of those extra halbderiers to outweight the bonus, assuming there IS a halberdier to target in both cities. If there isn't, that's a different story but you said there was.
Maybe a bug? Can I have this save file you are talking about?
The full list of fixed bonus amounts :
+50 in fortress
+50 for all invisible units in the stack when targeting invisibility
+50 for all regen units in the stack when targeting regeneration
+2000 for heroes (this is definitely enough)
+3000 for water walk on settlers (enough)
+2500 for Land Linking to target a fantastic unit (enough)
+16 for each buff on the unit for Spell Lock
+75 for Flight on settlers
+20 for each movement of the target for Wind Walking
+10 for each armor for Immolation
+3000 for Planar Travel on Settlers
+520 for Focus Magic on touch attacker
+120 for Focus Magic on casters
and that's all, other modifiers are "valid or invalid target" types.
February 27th, 2017, 12:06
(This post was last modified: February 27th, 2017, 12:57 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
What about wraith form on settlers?
Also, if heroes are that high why would halberdiers get them at all if there are any heroes in play? (Which there was, and they didn't have all 4 buffs each)
But I definitely think +50 is too low for fortress. The problem with halberdiers (or hammerhands) is that level 3 mithril hammerhands for instance have a rating of ~250-300. A summon like a storm giant only has ~150-200. The fortress bonus won't make up for that which means the storm giant won't get buffed.
I'm using this as an example because a primary sorcery wizard prefers to pick up warlord; and if dwarf, a war college and mithril isn't unusual for an off city, but stork giants aren't unusual to see filling a capital.
|