Test games played

[Seravy]

Found the problem, Orihalcon is removed when checking for dispel targets but after the check for "no buffs" not before. So it doesn't add the the priority as a buff but still enables targeting. Should be simple to fix by moving that line.

[Nelphine]

nice! even still, seems odd that they would cast it. if it doesn't add to priority, is the invisible hero causing the spell to be cast, then can't be targetted, but then it can target the orihalacron so it tries instead of cancelling and casting something else?

[Seravy]

Casting priority is based on the unit with the highest number of enchantments plus whatever other modifiers. If there is no target, the spell will not be picked like every other combat spell. This priority is not very detailed - more enchantments = more priority and that's all.
Targeting priority is based on unit value, and the enchantments on it. If nothing has at least 0 priority then no target is picked. Otherwise the unit with highest targeting priority that is also a legal target is the chosen target. This priority is detailed - some enchantments are worth a lot more priority than others, depending on the situation.

I'm unsure if this is good this way or not. At first sight it makes zero sense - casting priority should be based on the thing that actually gets dispelled. But that might be wrong. Maybe there is a unit with high targeting priority that pulls the dispel onto itself (for example a sky drake with only a few, but high value buffs), but the AI really needs to dispel that hero with 12 buffs on it instead. Since they can't do it immediately, they'll dispel the dragon first and then when they can finally target the hero, do that next. This is not ideal, but better than "oh, I would target that sky drake but it only has 2 enchantments so I rather just cast something else. ". Basically it allows the AI to dispel things in the correct order but still have the priority based on the most buffed thing even if that's not the first choice.

As for why the difference in the formulas : Targeting priority assumes we are already casting Dispel Magic and need to hit the thing that would provide us with the most likelyhood to win the battle. Casting priority on the other hand needs to decide whether we want to use Dispel Magic at all - or use disintegrate or doom bolt or prayer instead. These are not the same. For example even if hitting the thing with elemental armor is the best choice when we have an army of magic ranged attackers, and that target should receive a lot of priority (more than the thing with 5 other unrelated buffs), that doesn't mean it should get high priority over other spells - casting doom bolt on that thing is better than dispelling it. In fact dispelling it is often a very bad move as the enemy can recast the spell next turn, unless the unit died.

Unfortunately invisibility breaks this - the AI can't target the one that is making them use dispel magic even if it would have the highest target priority. However we can't exclude invisible units from the casting priority - the AI might be able to see some of them anyway due to position or true sight.

PS : Dispel Magic casting priority btw is capped, so it won't ever outprioritize legit great moves such as casting flame strike against an army of 9. But if the AI only has medicore moves available, it might end up dispelling instead and failing to target the invisible. (However this likely means it has no invisibility counter, otherwise that would have a large priority boost and would be used before dispel.)

[Nelphine]

My lunatic game continued: Jafar's ally (Tlaloc, 9 death, 4 nature, cult leader, peaceful perfectionist) finally declared war on me when I started spell blasting his Final Waves. It does take him a turn to cast Final Wave, but I sometimes have more important things to do, so they are getting through. This is wrecking havoc with my spearmen garrisons all across Myrror. However, it can't hurt the Veteran Orihalacron Hammerhands guarding all the towers.

Jafar has been banished 5 times. I've picked up time stop, spell ward, suppress magic, mass invisisibility, plus a few rare sorcery spells. However, I can't cast any of those, as I'm busy making sure my doomstacks stay functional (8 great drakes with buffs plus the bard). Jafar actually doesn't scare me.. Tlaloc has as much casting skill as he does, but Tlaloc has doomstacks made up of buffed death knights + buffed gorgons. These things are killer AND have high enough strategic strength to actually attack my tower garrisons, which obviously can't hope to fight that. (I took the time to disjunction survival instinct; on Tlalocs doomstacks it was just too much.)

So my doomstack is running around handling incoming Tlaloc doomstacks, while my older weaker doomstacks continue to attack Jafar.. who conveniently, is getting hurt by the Final Waves I'm missing.

Tlaloc did manage to actually conquer one of my myrran cities for one round, but my doomstack got it back and sealed the tower. He has seismic mastery (actually all of us have seismic mastery cast, due to spell binding) though, so he already has vision on all of my cities. I am SO glad he's perfectionist, or I'd have to deal with mass city curses as well. (Which would do.. very little. I've got dozens of cities on trade goods, so they could all switch to food production, and the only food I need is for unrest garrisons + the towers, and I have more than enough cities to provide food for the towers and just sacrifice all my unrest garrisons if I needed to. Except for Pestilence. That would be awful. Even warp node would just be annoying - most of my power comes from mass crystals on dwarven cities. Between my early game raise volcanoes + transmutes, I think I have literally over a hundred of them.)

However, this will really come down to whether I can actually get my doomstack to start killing things. I'm casting meteor swarm (and hope to cast doomsday shortly, but I'll need to dedicate 2 full turns of casting skill to it, which isn't easy to do); I had held off casting those two to try to avoid warring with Tlaloc before Jafar was defeated, but no such luck. (Of course I don't have Armaggedon. Which sucks; for me, I consider Armageddon by far the best of all chaos very rares. I also missed Hydras. I also missed warp reality.) Hopefully that will help mess up both of the AI enough that my doomstack will have a moment to start killing cities. (By the way, almost every Jafar city has chaos ward on it. That's why I'd be running around with hammerhands and steam cannons prior to the war with Tlaloc.)


For reference, its June 1415. I've been at war with Jafar since.. early 1413.

Edit: I'm actively allowing Jafar to cast time stop, just to use up some of his 25000 mana (after 2 years of full scale war, losing half his empire, and having earthquakes smash his cities regularly, he still has that much mana).

Edit edit: Oh crap, the timer on that is bugged.. oh hell.

[Nelphine]

Banished him before he completed time stop (actually took 2 rounds!). Got Magic Immunity.

GG.

Great Drake + Magic Immunity + Haste (+Spell Lock + Endurance +Land Linking +Chaos Surge +Focus Magic, but none of those really matter) + Healing, it doesn't matter what my enemy has. Actually, against the buffed death knights, the extra's matter, but only a little. And with divine order, I have 9 of those great drakes (going around in groups of 2, killing everything.)

[zitro1987]

Continuing my expert high elf 10chaos specialist omniscient on fairly small land, 3 AI

Year= 1416
Powerbase = 712
Base research = about 400
Skill = 157(157)
Towns = 20 (I sometimes lose a few, then conquer them later ... myrran towns starting to develop ok)

Biggest Successes: fire storm, flame strike, immolation gargoyles + fire giants (mainly for cracking a hydra tower in a fairly big continent I settled and rushed magic markets), doom bats + warp lightning (for sneaking on transport galleys)

Biggest challenges: main opponent also having flame strike, extended periods with war against many resulted in nice lost/razed cities and all dead heroes. Main opponent has an endless supply of stag beetles, fairly difficult to beat with chaos other than spamming flame strike.

2 AI wizards left. I heavily debilitated the death wizard without heavy losses (mainly firestorming their magician-defended garrison*) and got 100% usage of a small cluster of islands with 10 or so arcanus cities. The main chaos opponent I struggled and lost a bit due to his fire-storming and flame-striking (before I researched the latter) After a few dips during the war against allied wizards and giving some offers for some peace, I regrouped, got a bunch of summons and let the doom mastery really shine with towns pumping units (mainly pegasi), then I reconquered/maintained that small cluster of islands. Some of my nodes got taken away and there's random stacks of stag beetles all over. Fortunately, the main opponent is at war with me and the other wizard, spreading himself a bit thinly.
*Why is 'magician' such a popular defense unit? why not griffin?


Graph: I'm rising more rapidly than my main opponent (who is still 1st place, primarily due to having more than twice the army strength I have). Other wizard is mediocre, about half of me and not rising in past few years (due to war against main opponent), and the death wizard which was about 20% of me finally got defeated by the main opponent.




MomTweaker experiments (for better or worse):
*Lightning Bolt costing 20 but having 5-6 less strength is an interest experiment allowing regular magicians to spam it on first turn (like pscionic blast or petrify).
*Fireball having a bit more power, cost and research (by cost, being 21-25) 'feels' powerful as a wizard casting choice (at least equivalent to lightning bolt) early to mid-game and forces magicians to not be so dependent on this spell. Of course this would mean fire giant costing 15-20 more to remain balanced.
*Warp Lightning having a bit more power, cost and research prevents the brutal 'efreet' usage of spell and grants chaos the ultimate general 'direct damage' spell ... but the drawback is a 40 caster efreet surprisingly lacking in high-cost spell choices.

[Nelphine]

Griffin's have enough health to survive a round of attacking. Magicians do not .therefore, magicians on defense and Griffin's on offense. (High speed flying means Griffin's also make good doomstacks.)

[zitro1987]

If I banish a channeled wizard, does the wizard suffer the 5X combat mana penalty or not?
*I doubt i'll be able to, but just wondering.

[Seravy]

No, Chanellers only suffer a 2x penalty instead of the normal 4x.

[Seravy]

Finished the Orc artificer game. Nothing special, as always with Sorcery, once I survived the early game it' was an easy victory. Spell Binding recharge seems to work correctly this time.
New treasure system seems to be good as well - found 6 picks worth of stuff, but I was playing heavy treasure hunter strategy - heroes are good at that.
Orcs were ok but their advantages felt a bit lacking. Growth was not bad but the unit maintenance reduction, idk. I think we should push that a little further.
This concludes Orc testing, next race will be one of : Gnoll, Barbarian, Nomad. However I'll be taking some break before that.