Retort balance, hopefully last time

[Nelphine]

I actually think 'reverts to' is amazing, from a flavor standpoint. I don't particularly think losing death immunity will make much difference - you still get them too fast and can make an effective mini stack out of them. If black sleep works on them, at worst, your mini stack needs 1 more unit in it. More likely, you already sent enough extra (like 3 vs 1 swordsmen) that it won't matter anyway.
However, not being able to Regen vs zombies in particular and ghouls, that might make a difference. Makes those two units much more dangerous, and zombies are very hard to avoid.
Conversely, although I appear to be the only one who cares, this will make ghouls stronger, by giving them another unit to create while treasure hunting. 

Overall, I think it would be a good change. Even just being targetable by drain life will make things miserable for wolves.

[Seravy]

The question is, do we want wolves to be weak against one more realm? I'm afraid that would push them below master difficulty quality. I draw the line for that around 2 "not very good against" realms, make it 3 and I won't use the unit anymore. The chance of not running into either of the 3 realms is below 20% per enemy wizard. If the odds are so bad, even rushing Night Stalkers and Shadow Demons seems more viable.

I agree sleep would make no difference, and failing to regenerate from pretty much anything the death wizard can use, would.

...but this works in both directions, it's a huge nerf to AI death wizards, even if they manage to build up a good stack of wolves, the human will just say "sure but you can't regenerate those". Specifically zombies, the AI won't be able to avoid, while the human will easily do so by not moving the wolf close enough to the enemy half of the battlefield while the AI still has skill. If it isn't done by surprise, the wolves can easily distribute killing the zombie so that none of them gets over 50% undead damage from it (nor do they die in the first place for that to be relevant).

Of course in reality it doesn't matter - Blood Lust allows the human to beat the death wizard anyway (although slightly later than usual - but the first few wolves can be used on neutrals, nodes, etc or a nondeath wizard while Blood Lust is being researched, so it's not a major loss of time) while the AI won't be able to do that (AI can't push research for that spell.) which means letting everyone have the death immunity by default helps the AI more.

Haven't we gone through all this discussion already? I'm keeping the death immunity because the humans can take better advantage of the lack of it, and yes, undead wolves were the largest unwanted side effect of that, but being able to kill the AI's main strategy too easily is a problem even if you don't convert them to undead - that just makes it worse.

The thing is, the AI will always play werewolves if they are mono death - but the human might be playing anything. So any change that makes wolves weaker for the AI is really bad. Any change that makes wolves bad for the human with minimal impact on AI, is good. Regen 0, no illusion immunity, no poison immunity are such changes - the AI doesn't know how to abuse regen, the human will rarely attack a death wizard using sorcery magic in the early game (that's a suicide move) and the human rarely uses large stacks of nagas - even if they did the cost makes it much less effective for them. The only place where the human might benefit from the existing changes is when using spiders against wolves - that's unfortunate but can't be helped. Pure Nature is kinda on the weak side anyway compared to death even though both are early realms, so it's not particularly bad for the game overall.

[Bahgtru]

Seravy you really like writing  

Weakness, you can't make it good enough to beat a high HP unit without either changing it completely (as it targets attack, not defence or HP) or making it broken in the hands of the player against any other kind of unit. I'm proposing something far simpler...

The question is, do we want wolves to be weak against one more realm? I'm afraid that would push them below master difficulty quality. I draw the line for that around 2 "not very good against" realms, make it 3 and I won't use the unit anymore. The chance of not running into either of the 3 realms is below 20% per enemy wizard. If the odds are so bad, even rushing Night Stalkers and Shadow Demons seems more viable.

I agree sleep would make no difference, and failing to regenerate from pretty much anything the death wizard can use, would.

...but this works in both directions, it's a huge nerf to AI death wizards, even if they manage to build up a good stack of wolves, the human will just say "sure but you can't regenerate those". Specifically zombies, the AI won't be able to avoid, while the human will easily do so by not moving the wolf close enough to the enemy half of the battlefield while the AI still has skill. If it isn't done by surprise, the wolves can easily distribute killing the zombie so that none of them gets over 50% undead damage from it (nor do they die in the first place for that to be relevant).

Are you kidding? Who ever cares for losing fights in order to kill a werewolf with non-recoverable damage??? That's not a thing, it costs you more than the AI when you account for the bonus.

Come on Seravy, remember my test game. As I've proven humans can take much better advantage from the immunity (doing it at turn 00-30) than from its lack. Instead, its lack is not so bad for the AI that has more skill and mana and can use BS/ghouls/zombies more. Players don't sacrifice ghouls to prevent AI wolves from regenerating, that just doesn't happen, you're donning your AI hat too tightly. Players simply choose and win the fights in tactical.

Haven't we gone through all this discussion already? I'm keeping the death immunity because the humans can take better advantage of the lack of it, and yes, undead wolves were the largest unwanted side effect of that, but being able to kill the AI's main strategy too easily is a problem even if you don't convert them to undead - that just makes it worse.

No, actually you kept the immunity only to prevent undead WW abuse. That's why you swapped between sorcery and death, not this. "Reverts" fixes that.
As players don't care about AI regenerating units as much as theirs - they normally tend to win fights - the immunity is not important for AIs.

The thing is, the AI will always play werewolves if they are mono death - but the human might be playing anything. So any change that makes wolves weaker for the AI is really bad. Any change that makes wolves bad for the human with minimal impact on AI, is good. Regen 0, no illusion immunity, no poison immunity are such changes - the AI doesn't know how to abuse regen, the human will rarely attack a death wizard using sorcery magic in the early game (that's a suicide move) and the human rarely uses large stacks of nagas - even if they did the cost makes it much less effective for them. The only place where the human might benefit from the existing changes is when using spiders against wolves - that's unfortunate but can't be helped. Pure Nature is kinda on the weak side anyway compared to death even though both are early realms, so it's not particularly bad for the game overall.

Substituting death immu for "reverts" then is exactly what you describe as good: a change that makes wolves somewhat weaker for the player without bloodlust, while giving minimal penalties to the AI, and at the same time reducing player annoyance when fighting against death AIs thanks to some undead being created (without brokenness).

• player using WW: slowed down against death AI, fixing the problem I've exposed.
  
• AI using WW: no change with non-death player, death player doesn't really use non-recoverable damage the same way the AI does (sacrificing units to reduce the stack) so it's no change there either
  
• player against WW: no change unless the player is death, then the player enjoys some garrison-level undead if they don't use black sleep (which forces a choice, and so is a good mechanic) (yes, I've grown fond of the black sleep = no undead rule)
  
• AI against death AI WW: death AI: no change as strategic combat isn't really affected I think? Minor issue anyway.
  
• AI against player WW: does not get WTFPWN'd until the player uses bloodlust on all their doom-stack WWs, by which time the player is supposed to beat the first AI anyway.
  

Note that last "all": the AI is very good at targeting the player's wolves that still have the weakness in case of a mixed bloodlust/non bl stack, forcing the player to buff all or none of the WWs. 

In definitive, it's a very AI friendly change and your resistance to it is starting to look like "not invented here" to be frank. It gives players the option to use bloodlust - and so good gaming - when they meet death, like they are already supposed to do with sorcery. And its flavour would be great, especially if you managed to save the WW's original unit - although that's a nice-to-have, as random or AI-controlled race swordsmen would be good enough. Finally it lets us play with a very limited quantity of undead WWs coming from lairs, which is very fun and rewards dungeoneering.

[Bahgtru]

I actually think 'reverts to' is amazing, from a flavor standpoint. I don't particularly think losing death immunity will make much difference - you still get them too fast and can make an effective mini stack out of them. If black sleep works on them, at worst, your mini stack needs 1 more unit in it. More likely, you already sent enough extra (like 3 vs 1 swordsmen) that it won't matter anyway.
However, not being able to Regen vs zombies in particular and ghouls, that might make a difference. Makes those two units much more dangerous, and zombies are very hard to avoid.
Conversely, although I appear to be the only one who cares, this will make ghouls stronger, by giving them another unit to create while treasure hunting. 

Overall, I think it would be a good change. Even just being targetable by drain life will make things miserable for wolves.

Thanks for the support. Exactly. The wolves would still be a great unit, especially in dungeons, and I for one have started to use them and would definitely still use them. It would just not be the draconian archer level of abuse that it is (against death AI) anymore.

On regeneration: you just need to avoid losing the unit when its damage is mostly non-recoverable, I've learned to do that when playing against exorcism with resist magic (an exorcism normally still kills 3-4 of the figures and is non-recoverable). Remember that it was lunatic, I lost 3-4 units then figured it out: all you have to do is move the unit back, it's not so difficult. So this change really stimulates good play.

[Seravy]

Weakness, you can't make it good enough to beat a high HP unit without either changing it completely (as it targets attack, not defence or HP) or making it broken in the hands of the player against any other kind of unit. I'm proposing something far simpler...

This isn't true - the value of the unit is the product of its attack and defensive power. Halving the attack halves the unit, regardless of how much hp it had.
This is fairly easy to realize - drop the attack to zero and the unit has zero value. Even if it has a hundred hit points, if it's dealing zero damage, it isn't doing anything to win the battle (aside from stalling for time which isn't relevant in what we discussed - neither party has enough casting skill to care).

That's not a thing, it costs you more than the AI when you account for the bonus.

1. You attack with a random spearmen, summon some zombies, kill wolves. Definitely worth it. Do it 2-3 times and that unstoppable doomstack of 9 wolves is gone.

2. In a less ideal world, that stack is attacking your city. You summon the zombies, take advantage of the walls, kill half the stack. The stack is now not a threat, the next time it tries to attack anything, you beat it.  You lose a city, unfortunate, but does happen sometimes. Otherwise the stack would regenerate, raze your city, go and raze the next and keep doing that until you have no more cities.

As I've proven humans can take much better advantage from the immunity (doing it at turn 00-30) than from its lack.

Try to be on the receiving end of werewolves otherwise you can't make this assumption.
Also, taking advantage of your OWN wolves is an intended feature. Taking advantage of the enemy wolves, is NOT.

On regeneration: you just need to avoid losing the unit when its damage is mostly non-recoverable,

That's what I have been talking about. The human can do that, the AI cannot.

Seravy you really like writing

I don't like it at all but you are really persistent and I'm not a fan of being a tyrant who just says "no" without explaining why. While 90 times out of 100 that would be the better thing to do to save time, in the remaining 10 cases some unexpected argument or idea shows up that changes things and turns it into a valuable discussion. It also helps me actually know why I made a choice which reinforces it so I don't change my mind later.

Either way, things have to be done in the correct order. First we need to have enough data to be able to judge the situation. Play the same thing 5 more times, and if you lose fewer than 2 out of that, then I'll accept we have a problem - until then this discussion is a complete waste of time no matter how good or bad ideas are posted. (if other people do the same, that works, too. I definitely won't, I'm sick of playing wolves already. We have entire realms and races I haven't played in years, those are a bit more urgent.)
Alternately, win one as quickly as Sapher did - victory screen in 1406-7 - such a landslide victory is enough proof by itself - unless, and this is important, unless it's caused by a different game component, such as a retort (Spellweaver?), or excessive luck (Wraiths spell in a node in 1402?) etc.).

But hey, there is one case when I can say "no" without explaining myself. When something is on the "rejected suggestions" list already. So lets add reverting werewolves to it now. It's clearly worse for the AI than for the human, it's hard to implement (don't think there is much space where undead are raised, although I'm not 100% sure), and it even makes ghouls better (which I know aren't OP but the last thing we want is making them better anyway) while serving no practical purpose at all - as a werewolf nerf it's ineffective because the human can get around it : by blood lust or not letting the unit that took zombie damage die, whichever is more convenient for him. That's more than enough to add it to that list.

Also we are completely offtopic, this is the retort thread. This belongs to the werewolf or Death threads. Doesn't matter I guess, I'll link the rejected suggestion list here so we'll find it later. But if you do post more data on werewolves to advance the discussion after playing more games, do it in the "test games" or "death realm" threads.

Probably has been mentioned but might as well repeat for future reference, the current ratio of "easy to beat" and "hard to beat" realms is good, so moving death to the "hard to beat" group would make this ratio worse. Also, while we want to steer the player away from killing Sorcery early (it's way too beneficial to do that so the "cost" should be higher on it), we have no reason to do the same for Death (you'd be strong against Death all game anyway if you are playing Death yourself, so it's not a very tempting target.)

[Bahgtru]

Definitely not invented here syndrome.

[Nelphine]

Seravy I think the biggest problem with your argument is this simply isn't true:

1. You attack with a random spearmen, summon some zombies, kill wolves. Definitely worth it. Do it 2-3 times and that unstoppable doomstack of 9 wolves is gone.

The AI doesn't USE doomstacks of werewolves. They aren't intercontinental, and they don't get targetted by wraithform often enough. Instead, the AI penny packets them around, 2 in this garrison, 3 in that garrison, 2 more coming and joining a few ghouls and halberdiers, and another 4 all out individually. None of those are dangerous, regardless of whether they have death immunity.

[Seravy]

Definitely not invented here syndrome.

Because you were offering to write and insert the code and it does exactly what I wanted to do myself.
Oh wait...it does a completely different thing than I wanted and I have to do it myself? How could that be?

Let's stop this here I say.