August 14th, 2018, 14:07
(This post was last modified: August 14th, 2018, 14:50 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
Right, but it doesn't 'just' turn and kill enemy units. My numbers already account for both turning and killing other units, and killing the main target, and dispels, and potentially killing a melee unit that has to handle it.
With all of that accounted for, confusion averages 2 deaths per successful cast, whereas possession averages 3.25. account for save penalty difference, and possession is now at ~3.9 (less than twice as good as confusion, which is about right considering tier and cost difference between the two spells).
So where in my numbers did I go so wrong, that you are getting the difference between confusion and possession to be 2.5 times stronger than the difference I'm getting? I simply don't see it.
And, you haven't accounted for why, regardless of the gap between confusion and possession, confusion can be 2-5 times better than firebolt, when it costs less than twice as much.
No, direct damage isn't more reliable, until the target has high enough resistance to make confusion much worse. Against halberdiers, confusion is better than firebolt, even if the halberdiers have 8, maybe 9 resistance. (Such as halberdiers with resist magic, or orihalcron dark elves), even with the higher cost of confusion.
The idea that you can piecemeal targets with direct damage is true.. but is only relevant if the target has lots of figures (where direct damage is the weakest due to armor counting multiple times) AND the target had a ranged attack (but ranged units generally have higher resistance than melee anyway, so they aren't the ones you're concerned with in this whole topic!)
Posts: 10,495
Threads: 395
Joined: Aug 2015
Quote: confusion averages 2 deaths per successful cast, whereas possession averages 3.25.
Confusion gets you those kills several turns in the future, 75% of the time. Possession gets them immediately.
Thus confusion does not help end a battle quickly without significant losses on your side, (enemy gets to cast more spells, their units attack more times etc) while Possession does, in fact it's so effective a single casting cuts a 9 stack down to "now Wave of Despair can kill the rest is spammed" size and two uses cuts it down to "Wave of Despair ends battle in 1-2 more turns".
Meanwhile Confusion might actually prevent you from ending the battle since it's an enemy unit that has to die but you can't kill it until it turns back to the enemy side, assuming you couldn't double KO it (which is often the case as the unit does not move together with the rest of the army due to the spell's effect)
Your numbers aren't wrong but don't calculate what makes the spell good : IMMEDIATE impact on the enemy army. Things dying 4 turns in the future, yeah that's a thing that will happen but no one cares about. Not when 2 great wyrms, 3 minotaurs and 4 beastmen magicians are knocking on their city walls. Then you need results and need them immediately - confusion would be a guaranteed loss of that battle while Possession is a guaranteed win about...95% of the time? (assuming your garrison is...pretty worn down since this is the fifth attack coming in these past 3-4 turns)
And no I'm not saying Confusion is bad - it still takes out one enemy unit immediately from action and MIGHT do it to 2-4 at a 25% chance now or 75% later. But that's vastly inferior to always taking out 2-4 units instantly. And there is a catch - that one unit is only prevented from attacking you. It STILL counts towards enemy army side and the battle won't end while it's there.
August 14th, 2018, 16:25
(This post was last modified: August 14th, 2018, 16:38 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
Except my numbers ONLY count the deaths from confusion if it acts on the first turn. That's my 25%, it's already accounted for. If you include deaths that units causes in later turns, my numbers would go up, but my assumption is that for both confusion and possession, the unit targeted by the spell only gets one turn to act, the turn the spell is cast
So in both my comparison to possession AND my comparison to firebolt, I'm only counting the immediate benefits (the exception to this is the targeted unit. I've assumed it always dies, except for the 25% dispel chance. Which, yes, means on occassion that unit does get to do a bit more damage to the casting wizard. However, I think that cancels out the fact that I haven't accounted for it being able to do damage in future rounds. But it doesn't affect the amount of damage the spell causes.)
Strictly speaking, 25% (or 50%?) Of the time, the targetted unit of confusion can't even be attacked on the first turn, so except for dispel, they'll still get that 25% of wiping out a unit of two on future turns. However, I've chosen not to include this as I assume the targetted wizard will move out of attack range. But strictly, that's not actually that likely, and the confused unit will probably always (exvept odispels) get to kill at least one other unit; so confusion will be affected LESS than possession if the unit can't attack on the same turn as being targeted.
August 14th, 2018, 17:43
(This post was last modified: August 14th, 2018, 17:46 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
As a way to complete this: I agree possession is too strong. Averaging 3+ kills on successful cast is incredible, far better than other type of spell except wave of despair and flame strike.
However, confusion averaging 2 kills is also ludicrously powerful - and it's common.
So yes, I think possession should be toned down. But not by a tiny bit by making the save different. I don't think any single target spell should AVERAGE 2 kills. I'd like confusion to average 1 kill, and possession to be somewhere in the 1.5 unit range.
For confusion, you practically meet that simply by making sure it can't act on the same turn the spell is cast. Possession does the same thing. Possession stays better because it can't be dispelled - it's ALWAYS going to kill that unit, whereas confusion won't. And since confusion can't guarantee attacking even if it survives to the second round, possession still stays ahead.
To get a similar set of numbers but still allow for first turn actions, you'd need to set the save to be a BONUS, not a lesser penalty. At +1, possession wouldn't be insane. But that would be worthless.
August 20th, 2018, 10:01
(This post was last modified: August 20th, 2018, 10:03 by Seravy.)
Posts: 10,495
Threads: 395
Joined: Aug 2015
I just had this thought of "Annihilate is not useful enough for a very rare" and started wondering why. For multi target you obviously have other spells, but for single target it does have the best modifier, yet, I'm not using it, why?
Well, (normal) low resistance units are better to hit with Possession. Sure, instead of 100% I get a 60% chance but if it works I killed 2-4 units. Simply a better deal. For higher resistance normal units, they are usually low on health so Wave does better if they are in a swarm, and if it's a mixed army I possess the low resistance unit and that kills the high resistance one.
Damage is irrecoverable so I can't use it to enable Animate Dead.
Against higher end fantastic units, I rather use 2-3 waves than Annihilate at a 10-30% chance albeit, this doesn't really happen anyway in this current game as I mostly face 16 resistance Wyrms.
So it's only really useful against another Death wizard (to bypass immunity) or stacks of very rare creatures with 13-14 resistance. (at 12 or less, Black Prayer+Massacre does better. At 15 or more, Annihilate has a too low chance to be relevant in stopping such a doomstack.)
Anyway, one idea I had was "what if we swapped Possession and Annihilate". Obviously, by changing the save modifiers to match the new tiers better. I wonder if we ever considered this, we might have but I don't remember. Doing it would mean Death loses the ability to kill heroes at -7 save in very rare though, and instead they'll be a threat at uncommons (but probably at a low enough save modifier to not really be one as long as the hero has minimal resistance boost available)
August 20th, 2018, 11:30
(This post was last modified: August 20th, 2018, 11:31 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
My primary issue is confusion - without possession to compare against, confusion simply becomes the best combat spell in common and uncommon. Although we could swap confusion and psionic blast?
Secondary: I think annihilate should be the go to for killing very rares, and massacre shouldn't be. What if we made annihilate ignore all buffs to resistance (so on the unit screen, it only mattered what the white ones are, not the gold; so prayer, resist magic, prayermssters, hero items, node auras, etc, all wouldn't matter)?
August 20th, 2018, 14:37
(This post was last modified: August 20th, 2018, 14:40 by Seravy.)
Posts: 10,495
Threads: 395
Joined: Aug 2015
Quote: What if we made annihilate ignore all buffs to resistance
Then reductions would also have no effect if we take the data directly from base units stats. Otherwise, many of the effects don't add gold icons... it also uses generic "save or die" spell process so it would need to be a special case. Not that it helps since we assumed no buffs when designing the spell.
Massacre isn't really a go to spell to kill them unless you find a doomstack of low resistance very rares. Problem is against high resistance ones, even Annihilate is ineffective. (You won't have that many turns in such a battle to do Black Prayer+Annihilate x3 to kill one creature and then hope you can do another.)
No, if we want it to be a spell to kill very rares, it should be having a better save modifier.
The original design was for very rares that had twice the current cost, so 10-30% chance to kill them was good. As is, not so much, Wave of Despair, used 3-4 times, can kill one, without rolls, only need to clear away the useless things first. But those 3-4 waves also kill two, or three of the same very rares. (Not four, though. And you might need to cast it 5 times.)
Basically, Annihilate is only better than Wave if the enemy stack is 7+ units making Wave not an option. But then Massacre will kill the weaker things in the stack, or your army does it, reducing it to a smaller size, or if the entire stack is strong units, you aren't getting enough turns to use Annihilate many times to actually deal damage. So there is no room to use Annihilate either way.
Problem is Wave of Despair is designed to kill very rares, and seems to do it better.
However if Annihilate can easily kill very rares, enemy wizards with them stop being a threat entirely. You only really need to face a stack that has more then 3 of them once or twice a game, in all other cases, single target is effective enough (and so is Wave, if you can clear away the trash units, or there aren't any.)
August 20th, 2018, 14:43
(This post was last modified: August 20th, 2018, 14:45 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
Agreed. But yeah that's why I thought annhilte could be ignore immunities ignore buffs. Since it would have to ignore debuffs too, make it -6 or -7 to account for not getting black prayer. Also means you don't need to spend the round casting black prayer first. However, still wouldn't work on archangels, so, hope there's only one and wave it?
But I could also see confusion to uncommon, possession to very rare, annihilate to uncommon.
Edit: no, very rare means city troops are supposed to be meaningless. Possession can't come that late.
Posts: 10,495
Threads: 395
Joined: Aug 2015
Ok so no swapping with possession.
I think no higher save modifier either for normal units, as that makes it strictly better than disintegrate. (If we raise the current -7 to -9 then it's the same 100% chance to kill below 10 resistance but with an additional chance above that Disintegrate doesn't have.)
We also can't do higher save on fantastic units, as the AI will usually only have 1-2 of their high end very rares in a stack and sniping those too easily means the other wizard is not a threat (unless they have combat spells that make them into one. Sorcery does not (if their very rare was killed, there is nothing to haste or make magic immune), Life does not, Nature is semi-ok (call lighting, paired with earth elementals), Life doesn't care to (their normal units are good and archangels don't die anyway even with a higher save modifier), Chaos does, Death does not (besides their own Annihilate, nothing works on your death units.)
So if we do allow Annihilate to kill very rares at a reasonable chance, half the possible enemies cease to be threats in combat (aside from making you run out of MP, or if they already had a fully built 9 very rare doomstack, which can destroy a few cities before annihilate ruins it) and that doesn't seem very acceptable to me.
We also can't do lower cost - otherwise Demons will be able to use it.
So I see no room for improving the spell whatsoever which is bad because as is, Possession+Wave+Massacre pretty much covers all possible battle scenarios (except an enemy Death wizard, where Annihilate is your only option). So maybe we should instead move it down to rare and move something else up? There isn't really anything in Death rare that could be moved up... definitely not the economy/city spells, nor the only creature, nor zombie mastery (yes, it's powerful, but it should be something Death is good at and gets reasonably early). Nor Terror (autocast by Cloud of Shadows so can't be higher tier), nor Wrack (DoT must be earlier in Death than Nature), nor Wave (we specifically designed that to be rare).
Annihilate is still good in specific scenarios (you have Mind Storm, or fight a Death wizard, or missed Massacre/Wave/both) but does seem a little underwhelming even considering that.
August 22nd, 2018, 16:50
(This post was last modified: August 22nd, 2018, 16:52 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
So, math why confusion and possession can't be changed to not act on the first turn after spell is cast?
Specifically when compared to direct damage spells?
Assuming we go that route, then possession no longer makes annihilate feel weak.
Then modify wave of despair to do max damage against 3-5 units instead of always doing more damage individually if there are fewer targets. (Flavor: the despair feeds on the other units, so too few targets actually weakens it. Too many and the group strength fights off the despair.) Probably worth considering even without the annihilate discussion.
Then annihilate is better at solo kills than wave, and flame strike remains better against large stacks than wave.
Also gives added benefit that new players running around with a solo unit, say a favourite hero, don't get 'ambushed' by wave and lose that favourite unit without understanding why.
|