Dugy / Legend_of_the_Invincibles

An add-on campaign for the Battle for Wesnoth game
GNU General Public License v3.0
40 stars 22 forks source link

Rebalance of Elves for 1.18 #598

Open Discontinuum opened 1 year ago

Discontinuum commented 1 year ago

There is a huge rebalance of units undergoing in the current dev version of Wesnoth. In particular, there are sweeping changes regarding Elves https://github.com/wesnoth/wesnoth/commit/091b094685f381e6cd4861b3f55ec33916dbacc5

The most notable of them are changes to the high level Elvish mages, Sylph now has 16-5 damage of faerie fire, Shyde has 14-3 thorns. That makes LotI's 4 and 5 extra levels be weaker than the new mainline units from which they advance. So we'll have to rebalance things to account for that.

Regarding arcane there are also discussions about decreasing its variation, so Ud may have less vuln (-30% or -20%), that is not merged yet.

1.18 is not going out tomorrow but I'll still create this issue in advance, so it won't come as a surprise

Discontinuum commented 1 year ago

By the way, new liminal that we account for in this issue https://github.com/Dugy/Legend_of_the_Invincibles/issues/355 is fixed, so liminal now reacts to overillumination and overdarkening, just a notice

Dugy commented 1 year ago

This could be relatively difficult to deal with, because max level LotI units also need to be balanced against each other, as some unit types are almost never used (they are not very well balanced). Elvish Seer, the lvl5 addition to the Sylph line, is already overly popular, and adding some extra damage would only make her more popular.

I already wanted to add some new abilities to underused units and remove some numeric upgrades on overly popular ones, so this will have to be taken into consideration. Though, I would have to get myself to focus on LotI instead of other stuff.

white-haired-uncle commented 1 year ago

If you have to bump up the Seer so that it is still more powerful than the Sylph, for example, how about raising the XP, both to get to Seer and the base for advancements?

Seems like that's almost necessary anyway, since a more powerful Sylph is likely to advance faster.

I've always thought the first AMLA for any line was too cheap anyway. For some units it costs hundreds of XP to advance to that final level, and then they can get multiple AMLAs on the next turn (okay, only by killing BZB or something, but still).

Dugy commented 1 year ago

Well, the Seer has to be more powerful than the Sylph, but her advancements have to be weakened somehow. Increasing the XP is one possibility, but I was thinking more about removing the advancements that increase the number of attacks.

It's totally intentional that Beelzebub gives absurd reward. He is there to make finishing scenarios early actually rewarding. The extra gold was not particularly worth it, and it was more beneficial to wait around the enemy keeps, killing recruits for experience and items, and finish the leaders in the last turn. This required some amount of masochism to maximise reward, and I needed something that would be even more rewarding for finishing early. And Beelzebub serves this purpose well.

Discontinuum commented 7 months ago

I think we need really to start thinking on how to handle this whole elves rebalancing thing. Also arcane has changed: humans now have 10%, undead less vuln (either -20 or -30, don't remember exactly)

Dugy commented 7 months ago

I can think about several things that can be done:

What do you think?

Discontinuum commented 7 months ago

Well, I think the first problem we should deal with (sorry if I didn't start with it right away so it got a bit confused) is that newer 3-4 lvls of core Wesnoth became stronger than LotI's extra levels. Like the Sylph https://units.wesnoth.org/1.17/mainline/en_US/Elvish%20Sylph.html she has 16-5, stronger than Seer's 11-6 (afaik). https://units.wesnoth.org/1.17/mainline/en_US/Elvish%20Shyde.html Shyde has higher damage as well. Elvish Champion got +10 accuracy and a bit less damage but I guess that's easy to deal with, just add +10 acc to Juggernaut as well. Sharpshooter afaik is stronger as well (12-5, almost as LotI's 4 lvl with 13-5)

Dugy commented 7 months ago

Aha, so that would require a significant rebalance. Simply increasing damage of the final units would make them way too strong for the enemies that are currently present.

Discontinuum commented 7 months ago

I wonder how we even approach such a daunting rebalance, where to start?

Dugy commented 7 months ago

I don't know. I can keep throwing ideas, but I need some reality check.

It's possible to simply upgrade all over max level units to have the same damage as the improved units below them, but always make sure they have some extra ability, weapon special or other improvement over them. Because first AMLAs tend to be cheap, they will quickly build up improvements making them stronger than the previous unit. And items useful for these units could be made weaker, but I can't think of any items that are particularly useful to elves.

But, how many unit types are actually affected? I haven't seen many changes that were so significant. Only elves?

white-haired-uncle commented 7 months ago

I don't make much use of elves, but I know some use them heavily. It sure seems to me that if advanced elves become significantly stronger they will be much, much too powerful.

One clear, if unconventional, option would be to not accept the mainline changes. Just replace the new units with ones that match the old ones. I don't know how much work this would be but I'm sure it's a lot less than anything else. The obvious objection is that it doesn't match mainline, but then exactly what part of LotI does?

Alternatively, as I've mentioned before, serious cripple their AMLAs. The new L3 mainline units would still throw off the game balance, but at least the L4+ LotI units wouldn't be stupidly powerful. Raising the XP cost is the simplest, but weakening the AMLAs themselves would be an alternative/addition.

An option that addresses both the now overpowered L3/L4+ units would be to take away some of their gear. For example, their head gets so big they can no longer wear helmets or amulets (stupid example, it's the idea that counts). Of course, I'll just load up mine before they advance...

Messing with items seems problematic to me since it will affect other lines in ways that we won't think of. But an example of items particularly useful to elves are anything that gives slow, and to a much lesser extent incinerate and poison (grypon riders with lots of MP, hit and run, and slow (and possibly distant attack?) are immensely useful against packs of powerful demons for example).

Another thought is to remove the ability to recruit elves. You pick up some, Corruption and Doomsday Omens come to mind, but that's all you get. C3S1 would need to be updated, for starters. Or limit the scenarios where you can recall them (limiting their opportunity to advance). Or something else, the point being that if the units become too powerful one potential option is just to limit the amount of them you can have. This one is tricky, I tend to play a core of maybe 10 really powerful units, so putting a cap on advances elves would require, for me, a pretty darn low cap.

Dugy commented 7 months ago

I'd prefer to accept the mainline changes because LotI was always adding things rather than replacing things.

Raising exp cost for the strongest of elves seems like an option, they would end up with less AMLA upgrades. Preventing them from using some specific piece of equipment (like helmets) seems okay too.

I don't think that limiting recruitment is a good way of doing this, many players notice some units are exclusive to a specific occasion and will take care to develop those units, and making more units exclusive would only take away some quality of life.

Also, not all elvish units were improved that much, Elvish Marshal is the same, Elvish Champion is only slightly better (and he wasn't great before), Elvish Avenger and Elvish Sharpshooter were only slightly improved, Elvish Shyde was never good offensively, only Elvish Sylph got a massive boost (and undead became more resistant to her attack).

So how about just focusing on Elvish Seer and assume the others' changes are insignificant?

Discontinuum commented 7 months ago

I would review and post a list of changes and my opinion on their significance for us later. Elvish Sylph/Seer indeed is our top priority, its buff was really immense. I think Shyde and Sharpshooter we should take care of as well, cause they became on par or slightly better than LotI 4 lvls for them.

And we need to think carefully about arcane, what changed with arcane in short is that units who were vulnerable to it became less vulnerable and units who were resistant to it became less resistant to it. I'll post most details here later

I'd prefer to accept the mainline changes because LotI was always adding things rather than replacing things.

Rejecting mainline changes was on my mind as well but I've expected your answer so probably we need to try what we can do first.

Discontinuum commented 7 months ago

A possible idea (but I don't quite like it) is to prohibit lvl 3 Elvish Enchantress in LotI to advance to lvl 4 standard Elvish Sylph and instead make LotI 4 lvl Elvish Sylph, with the same sprite but with LotI amlas. But I'm not that happy with eliminating a lvl 5 branch and the problem of Lethalia being weaker than Elvish Sylph is still present (not a problem maybe? Dunno)

white-haired-uncle commented 7 months ago

Random thoughts, not suggestions:

In Arctic Wasteland, I can recruit Elvish Sorceresses. If the Slyph is going to become much more powerful, maybe I shouldn't be able to skip right to one level below it. At least make me work up to Slyph/Seer from Shaman.

Give demons a random trait triple damage against the Elf Sorceress line. This addresses not only the balance between sides, but also within the player side (make Seers much more powerful and I'm likely to concentrate my army on them, but also make them more vulnerable..). Of course, this leaves out a lot of scenarios.

I don't think I've ever seen a Corrupted >=Elvish Sorceress.

Give the Seer a disadvantage. I've seen somewhere units like an Ancient Mage or an Ancient Lich where the unit might get more attack strength, but lose MP and become more vulnerable (such as vs impact for the Lich, IIC) due to the heavy cost of wielding such powerful magic. Make the Seer take poison every turn she is not on veggies, or slowed when not on solid ground, or whatever (I'm not a fan of these particular examples, it just an idea).

Dugy commented 7 months ago

A possible idea (but I don't quite like it) is to prohibit lvl 3 Elvish Enchantress in LotI to advance to lvl 4 standard Elvish Sylph and instead make LotI 4 lvl Elvish Sylph, with the same sprite but with LotI amlas.

This has led me to an idea that seems cool. Since Enchantress no longer advances to Sylph unless explicitly enabled (the advancement line used to normally continue to level 4 at the time LotI was written), it would be possible to give the Enchantress two possible advancements, Sylph and Seer, both level 4. Sylph could continue having superb high damage faerie fire, allowing Seer to keep various special abilities she has right now.

Discontinuum commented 7 months ago

Sylph could continue having superb high damage faerie fire

Which amlas would she have? Also, what to do with any possible standard Sylph in the code, just replace it with LotIfied one?

Dugy commented 7 months ago

Which amlas would she have?

Making her ranged attacks more powerful, but not much to improve her melee. The main problem with Seer is, as far as I know, that she can have both good melee and ranged, so she can be powerful at offense but also versatile enough not to be very vulnerable to melee enemies. With this change the exceptional offense would be passed to Sylph, while the versatility would be kept by the Seer.

Also, what to do with any possible standard Sylph in the code, just replace it with LotIfied one?

The advancement from Enchantress to Sylph has to be explicitly enabled, so this could be removed and replaced by a LotIfied Sylph. For backwards compatibility, a normal Sylph would upgrade to this new Sylph.

Discontinuum commented 7 months ago

Regarding most sweeping balance changes, arcane resistance stuff:

Smallfoot, armoredfoot, elusivefoot, mounted, fly, smallfly, lightfly, deepsea, mountainfoot, gurefoot, rodentfoot, lizard, dunefoot, duneelusivefoot, dunearmoredfoot, dunehorse, dunearmoredhorse (basically humans, ghouls, gryphons, saurians) - arcane resistance changed from 20% to 10%. Treefolk - arcane resistance changed from -30% to -20%. Undeadfoot - arcane resistance changed from -50% to -20%. Undeadfly - arcane resistance changed from -40% to -20%. Drakefly, drakeglide, drakeglide2, drakefoot - arcane resistance changed from -30% to -10%. Paladin - arcane resistance changed from 60% to 30%. White mage - arcane resistance changed from 40% to 30%. Mage of Light - arcane resistance changed from 60% to 50%. Lich - arcane resistance changed from -40% to -20%.

About units, elves:

Elvish Outrider - ranged damage changed from 8 to 11, melee damage changed from 7 to 8, hp changed from 57 to 60, cost changed from 43 to 58. Elvish Shyde - hp changed from 46 to 51, melee damage changed from 6 to 7, ranged slow (entangle) damage changed from 6 to 8, ranged magical (thorns) damage changed from 8 to 14, cost changed from 52 to 69. Elvish Enchantress - cost changed from 55 to 70, ranged slow damage changed from 5 to 7, ranged magical damage changed from 9 to 13. Elvish Sylph - hp changed from 60 to 68, ranged slow damage changed from 6 to 7, ranged magical damage changed from 10 to 16, cost changed from 67 to 161.

And in theory those undead might require some attention:

Necrophage - melee damage changed from 7 to 9, cost changed from 27 to 23, xp changed from 120 to 61. Ghast - melee damage changed from 10 to 12, cost changed from 43 to 50. Necromancer - ranged cold damage changed from 17 to 19, ranged arcane damage changed from 12 to 16, hp changed from 70 to 76, impact resistance changed from 0% to 10%, cost changed from 50 to 90.

Costs of many high level units changed, it may affect some scenario where you can recruit lvl 2 or lvl 3 but that doesn't seem to be a top priority issue, also many xp requirements changed, but usually not too much

Dugy commented 7 months ago

I have changed Seer into a level 4 unit, and made a special Sylph with damage-focused advancements. What do you think about it? It's the commit 0e13c5f.

konecnyjakub commented 5 months ago

FYI elves are getting a bit nerfed for 1.19/1.20: wesnoth/wesnoth@72d2f45. There was a significant backlash against the changes so they partially rolled it back.

Dugy commented 5 months ago

Thanks for telling. I can change their base damage once 1.20 comes out.

Anyway, I think the difference isn't very significant, it's more important for maxlevel LotI units to compete with other maxlevel LotI units, smooth progression from previous level isn't so important (unless it's a downgrade). And at this point, I have no idea what the Seer/Sylph split will cause.

Discontinuum commented 5 months ago

I've had not too much time for LotI unfortunately, but as far as I'm aware, the Sylph thing was done. We probably still have to think about how arcane change will affect demons (especially considering their crazy shadow attacks). Basically almost all undead's vulnerabilities to arcane were reduced (arcane does less damage to undead) while humans' 20% were lowered to 10% (as well as White Mage and Paladin's ones). So in general arcane now does more damage to those, who were resistant in the past and less damage to those highly vulnerable to it

Dugy commented 5 months ago

Lawful human units are already overly popular, while undead are quite unpopular, so this could actually help make less used units more popular.

Discontinuum commented 2 months ago

833

Although Elvish Champion has +10% accuracy but LotI Juggernaut doesn't. Although new Juggernaut, as far as I could see him, is very nice and has cool amlas

Dugy commented 2 months ago

I gave him the 10% accuracy and slightly reduced his damage to compensate in f9b2fb9.