April 3rd, 2019, 14:32
(This post was last modified: April 3rd, 2019, 14:35 by Seravy.)
Posts: 10,463
Threads: 394
Joined: Aug 2015
Quote:If the death knight attacked my bezerkers, while I would do a lot of damage, I don't get to use MY thrown
Except even under the new rules you could use your thrown. The change goes both ways, if the AI can retaliate with their breath so can you with thrown.
We've already discussed thrown, if the AI was smarter, that would at best result in a standstill (allowing you to Holy Word/Star Fires the Death Knights until they all die).
Even if your berserkers only move 3 without Endurance, that means the Death Knights need to stop 4 tiles away. You can then move 2 backwards instead of forwards to restore the original 6 distance and we're back to where we started. At best the edge of the screen stops this movement after many wasted turns, at worst it keeps going until you're out of casting skill and retreat exhausted.
What you seek doesn't exist - if both players play thrown/breath in the smart way it still doesn't allow the thrown unit to hit first unless it's over twice as fast as the target (or the battlefield is so crowded with units there is nowhere to move but in those cases they get to use their breath anyway, you can't cut down that many on turn 1...). Not only is that extremely unlikely, Endurance makes sure it outright never happens.
I feel we are going in circles - you should admit this is an unsolvable problem, and instead tell me if you think nerfing thrown (not for this reason but in general) is necessary on levelups and lionheart, or not. Because it might not be relevant to cutting down those very rares, but it's very relevant to Draconians and Gladiators cutting down enemy war bears and halberdiers for example.
April 3rd, 2019, 15:11
(This post was last modified: April 3rd, 2019, 15:16 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
When things are moving speed 4+ the edge of the map is VERY relevant. It takes 3 turns for drakes to catch something hiding in the far corner.
Yes, you could star fires death knights to death. This would take on average... 12 casts for one unit? If you can already cast that many combat spells, then this is already outside of the abusive stage of treasure hunting. That would be absolutely fine as a tactic without hurting game balance. But death knights are speed 5. That's fast enough it's not realistic to back them into a stalemate without extreme overland cost as well. And while drakes are only speed 4, they take even more spells to kill. In both cases, game balance is even more well preserved.
No I don't think thrown should be made stronger. No I don't think thrown needs to be nerfed, except against units of speed 4 and 5 who walk right into it. (And speed 3 on occassion; chimera before I have prayer and can't afford endurance maybe?)
In early game, I don't really care. Flying/speed 4+ archers, and effectively invulnerable due to relatively high armor units (which could include healing) are much bigger problems than the weak thrown and breath that exist that early.
In fact my problem is the same category - defense is too binary. Fling archers are a problem due to not taking damage. 20 armor heroes are a problem due to not taking damage. Healing a buffed swordsmen is a problem due to not taking enough damage. AI not using first strike and thrown is a problem because my bezerkers don't take enough damage.
In ALL of these cases, the offensive ability only matters inasmuch as you can do significant damage. Sprites were the most extreme example as people might literally take 12 OVERLAND turns to conquer a neutral city. The offense never makes something broken. The lack of damage to human units, that is what makes it broken.
Posts: 5,010
Threads: 17
Joined: Aug 2016
In case I editted too much and it was missed, I did a major post changing edit on my last post.
Posts: 10,463
Threads: 394
Joined: Aug 2015
Quote:When things are moving speed 4+ the edge of the map is VERY relevant.
They aren't moving at speed 4, that's kinda the whole point. If they do you can attack them on your turn because they got too close...
You refuse to accept reality, there is nothing else I can say at this point.
You first want the enemy to move (their speed-your speed) to stay out of range then claim they move the full amount. No they don't, the Great Drake vs Berserkers case would mean moving 1 tile a turn. Death Knights get to move 2. Add Endurance to the Bersekers and Drakes can't move anymore, Death Knights move 1...
April 3rd, 2019, 20:13
(This post was last modified: April 3rd, 2019, 20:27 by Nelphine.)
Posts: 5,010
Threads: 17
Joined: Aug 2016
no.
assume the AI has one higher speed than I do. I move backwards enough that the AI is moving forward at 1 tile per turn. Then I have at most 6 turns before they back me into a corner and attack me; except if I have 9 units, then some of my units are in fact 2 full squares closer than that. That makes it 4 turns before they can attack me. That's 8 full turns before I can expect to kill ONE with my vaunted star fires spell, which means they slaughter me.
But wait! what if I have both ranged and first strike units? Then insert the formula to see which one of them is more dangerous, under the proposed number of turns it would take them to catch up to me. They may charge forward first anyway, but in that case, they're doing so because it's more dangerous to stay still and soak up fire than to charge forward.
Now assume the AI has the same speed as me. Then the AI backs into the corner. If I want to attack, then I have to move into it's attack range first. Star fires wins! Except it takes so many turns of star fire to kill one, that my previous argument prevails. If you have enough casting skill that you CAN kill a very rare with such a spell, then it's completely reasonable for you to be hunting very rares in the first place, and you wouldn't need to rely on this strategy, and therefore, many other realms and units could accomplish the same thing.
Plus, even against the same speed units, the AI could at the very least ensure that the human can't attack TWICE before they can retaliate. That's sometimes not very relevant (most of my examples have been about 2 bezerkers killing a sky drake, in which case, neither bezerker ever attacks a second time, unless the sky drake gets very lucky and survives both bezerkers initial attacks with only a few hp left), but against other examples (your orc hordes with a medium number of buffs for instance, or my very original beastmen halberdier, against hydra), that might be much more relevant.
My point about speed 4 being the relevant point is that for most ranged units (which don't move speed 3, they move speed 2), you'll only be backing away one square at a time, so that speed 4 means that within 3 turns, they'll catch you.
Speed 4 absolutely matters, but right now, it HURTS the AI to have speed 4 or 5 if they're against strong melee units and either side has first strike/thrown- they aren't as powerful as slower units (*cough hydra*), but they simply move forward into slaughtering range of the human melee units.
Your comments about other units, like orc hordes, are absolutely relevant here. If your hordes CAN attack them in the first place, for instance, via flight, then the speed 4 great drake is WORSE OFF than the speed 2 hydra - against both a bad player AND a good player, the drake charges into attack range, then the orc horde attack and the drake isn't allowed to use it's breath weapon. With a hydra against a good player, the human baits the hydra into coming into range, then attacks and the hydra isn't allowed to use it's breath weapon - BUT against a bad player, the human also charges, and then the hydra DOES get to attack with it's breath weapon. So the speed 4 on the drake is somehow worse than the speed 2 on the hydra.
Yet we price the speed 4 and 5 units as if that speed is good (because it IS good against ranged, or when neither side has first strike/thrown). Which brings the human to figure out loopholes, such as this one.
The map is absolutely small enough for it to matter. We absolutely can calculate if it's better to move forward to avoid ranged attacks (you could even put in estimated spell damage, although that would be tricky if we don't want to assume the AI simply automatically know the human spell list and casting skill) and when it's better to move forward, and even when it's better to move back and stalemate into a corner.
That's why my whole position this entire time has been about AI not being able to use first strike/thrown and not being able to fight first strike/thrown. I know i've recently refined that to primarily be about fast speed units, but strictly, that's because in most play, anything speed 3 or less, is probably susceptible to most ranged/spells the human will have anyway, so it's not worth worrying about those cases, because the calculation will just say charge forward because stalling will make you take too much damage.
It's also why my entire position here is NOT about thrown or life units being overpowered. They are strong - not overpowered. If the drake were allowed to attack, then it would absolutely crush my bezerkers. But right now, I don't even need to waste time on endurance (which is GREAT because it means sorcery opponents won't dispelling wave my doom stacks), because I don't need the bonus movement in combat for anything. When looking at multiple stacks of bezerkers, saving 70 overland casting skill per unit, 630 casting skill per stack, is absolutely relevant to getting the snowball rolling.
And again, I'd be perfectly fine with AI only caring on this at master or lunatic. It's a specific trait, that only matters when the human and ai ranged/spell offensive capability is low, and one side has the first strike/thrown, and the ai has high speed (usually 4+, but occassionally 3). That's specific enough to never come up on the lower difficulties; but it's common enough in treasure hunting as to allow ridiculous snowballing, making it ideal for high difficulty strategies.
P.S. the gaze change to go at the front of all abilities, on either offense or defense, (but NOT to be doubled by haste) would probably be a good one, regardless of the rest of this entire discussion.
Posts: 10,463
Threads: 394
Joined: Aug 2015
If the AI works like that you don' bring anything but melee (or thrown if it doesn't trigger on normal melee but considering your example cases, it should. Melee kills the drakes not thrown.) units. That should be obvious, anything that doesn't trigger the new behavior can't be abused so you avoid doing that.
Second you don't bring many units. Why would you? They are not going to get a chance to attack anything. 1 is plenty.
Third, you most likely won't get cornered. Yes, you'll usually move 1 but if things get risky you can take a turn moving 3 to get away from the corner as needed. I'm pretty sure a good player can come up with a movement pattern that's safe.
Fourth, since the feature triggers on high melee enemy+ AI unit having breath (it has to, melee is killing the AI and preventing the use of their breath, not thrown), and a high melee unit can still be ranged, you can in theory still shoot at the AI while backing up, or if ranged disables the feature then force the AI to go too close to your berserker. All it takes is having Focus Magic on it.
Either way it ultimately means the AI still doesn't get the chance to hit you with their breath attacks. It just means instead of being able to melee it, you are forced to stall and use spells. Which isn't the direction I'd like to take.
Of course this is just theory - in practice I can't implement such an AI system anyway.
So while being pointless in theory, it's also impossible to do. Overall, a solution to this problem doesn't exist, that's my point.
Quote: If the drake were allowed to attack, then it would absolutely crush my bezerkers.
Instead of chasing after an illusionary solution, how about a real one? Drakes are Chaos and Sorcery. So their owner can potentially cast Mystic Surge, or Haste to make them fast enough to attack on turn one. Of course they won't always have the spell, but that's something at least. The AI already gains extra priority to cast Haste on high cost units, but we can add new conditions like even more priority if that unit also has a breath attack. If that's not good enough (the AI likely has something higher priority for the first turn like a global) we can consider making Haste available overland.
As for Mystic Surge, it is probably not very usable for this purpose because casting it means missing out on a flame strike, apocalypse or blazing march.
Ofc lairs don't cast spells. For lairs the only thing we can do is reduce the treasure point value of the monsters, if you feel they are too easy to kill. It's not like lairs cause losses for players, they almost never do if the player plays is smart, in fact they can wait until they have something that can beat the lair without losses, they are not forced to take the loss. So if you say the drakes are too easy and can be killed too early, that means they need a lower treasure budget.
These are solutions possible to implement and worth talking about. Improving the AI isn't.
Posts: 441
Threads: 4
Joined: Apr 2018
Maybe part of the issue is that thrown doubles the effectiveness of buffs. Breath doesn't. I feel that all buffs should apply only to one of melee, thrown, breath (or ranged if that happens) after this discussion.
I'm not saying that this solves it but the fact that super-buffed units are OP is not helped by doubling their effectiveness.
April 4th, 2019, 10:34
(This post was last modified: April 4th, 2019, 11:58 by Seravy.)
Posts: 10,463
Threads: 394
Joined: Aug 2015
Yes, that's a reasonable observation. While units typically can't use their melee and ranged attack effectively at the same time so raising both isn't a big problem, thrown and breath does double the effectiveness of attack buffs if they apply to it and melee at the same time.
In reality the effect is even more - thrown and breath on normal units starts at low values so any gain to it is much more significant relatively.
That said, limiting every single effect to melee only (let's be realistic, most people won't care about a buff that raises only thrown/breath - most units don't have either.) is definitely overall. I think it should be ok to have a few buffs apply to either breath or thrown (probably not both, to make sure the abilities play differently and work better with different strategies) but it should be less than currently.
In particular, I believe Nature and Chaos should be buffing Breath attacks (the latter especially as it can grant Breath to units), while Life should be able to buff thrown. Death and Sorcery has no relevant buffs.
Applying this to our list :
+1 Thrown/Breath/Ranged at levels Regular, Veteran, Ultra-Elite
Should not include thrown and breath on all these levels, at most one of them. Veteran seems ideal - Regular is too easy to get, and Ultra-Elite is only relevant for Life Warlords which seems to be the main culprit behind Nelphine's strategy.
+1 To Hit on Thrown from having Magical Weapons (or Heavenly Light)
+1 To Hit on Thrown from Holy Weapon
+To Hit in general
Good as is. Without boosting the base strength of the attack, adding To Hit only is far less effective. Multiplying the two bonuses is what makes the effect to strong.
+1 or 2 Thrown if Mithril/Adamantium (according to wiki)
Probably should not apply to thrown.
Edit : actually doesn't matter. Barbarians can't use these ores and other races don't have thrown units.
+2 Thrown/Breath/Ranged from Chaos Surge if Chaos Channelled
+1 To Hit, +1 Thrown/Breath/Ranged from Animated
Fine as is, these are very rare spells from realms who don't care about buffing at all by then.
+2 Thrown from Flame Blade
Would prefer this to add Breath but most units that do have breath attacks can't be buffed by it. Also it's not Life so requires extra picks spent to have. Not sure if it needs nerfing.
+3 Thrown from Lion Heart
This is definitely among the most effective ones although it does came later in the game and it is a very expensive buff. If we do care about the problem we most likely should remove it.
+2 Breath from Land Linking if Chaos Chanelled or Undead
I think this one is ok, Nature uncommons are a fairly unlikely thing for a buff based strategy to have.
+3 Breath from Focus Magic (thrown is turned to ranged instead)
Fine as is, never grants melee.
+1 Thrown/Breath/Ranged from Leadership at level 6 or higher
How often do you have a level 6 leadership hero? Even then it only adds 1. Not a threat.
+3 Thrown from Blazing March (but no Breath)
Should be breath instead? Most Chaos creatures have Breath attacks. I thought this actually works for Breath, I'm surprised it does not. Was it nerfed or did I misread the code? Alternately we can make it add a lower amount, or we can make it not work on them at all.
+1 Thrown/Breath/Ranged from Darkness if Undead
Probably good as is. Undead are hard to keep alive so that reduced retaliation damage actually matters in a good way.
Posts: 5,010
Threads: 17
Joined: Aug 2016
I'd actually leave lionheart - it IS late enough that if you're using it, it's probably balanced by the creatures you're fighting, in the way that buffing is actually intended. I've never seen it actually cause anything but win-more situations; the attack portion is never a source of actually winning a battle.
Flame blade I'd be tempted to leave as +3 melee, +2 ranged, no bonus to thrown or breath (specifically because it DOES stack with life buffs on barbarians and draconians who are the races who have the most problems from going with life buffs anyway, and if you're not playing life, then you're probably focusing on your chaos summons, so you aren't those two races in the first place)
Leadership is not a problem, but the Chosen One probably is. However, I don't think I've ever actually needed to use the Chosen One except against life buffed behemoths, so the Chosen One (if a problem at all) is a problem due to general overpowered heroes.
Since I'm already a fan of reducing level bonuses, I'd be fine with that. But, given that I do NOT think thrown is the issue here, I think you should leave it on 2 levels, including ultra-elite. In particular, for non life barbarians, that's still going to be a fairly major impact.
Posts: 441
Threads: 4
Joined: Apr 2018
I agree with all the suggestions including nerfing flame blade.
My only request is to kindly add this in the help text - although I know that that's painful.
|