Would you prefer capped enemy Talent levels w higher Power?

[Snarvid]

There's been several posts about the way that 1.6 has a non-trivial tendency to spawn rare+ enemies with absurdly high talent levels (highest I've seen is 50+) that are either a. 1HKO specialists b. are virtually unkillable or c. both. In this post I discuss a possible fix to this system, capping enemy talent levels at some reasonably low level (not necessarily the player's cap of 6.5, but within a small distance of it, perhaps +1 or +2 more per difficulty level) while giving them substantial, difficulty-scaling buffs to power (physical/mind/spell/steam) in a manner analogous to the way enemies receive large bonuses to HP/tier/difficulty level. I think this would help limit the way certain problematic enemy qualities scale with talent level increases, notably their resistances, global speed, respen, and the duration and cooldown of their talents, while not making them essentially immune to all power-based status effects the player would wield (as would be the case if their stats, and thus saves, were buffed instead of their power). It might also allow for the tuning of rank-and-file enemies, who do not currently possess such over-leveled talents, to be somewhat less helpless. Finally, it would give a clear set of expectations for players when choosing a difficulty level - you could easily know ahead of time that on Insane rares got +X to their power, uniques +2X, and bosses +3X, as opposed to playing enemy lottery on every level and hoping not to pull the black dot.

Now: I have no ability to make this change happen, I'm just trying to get a sense of the feelings of the community on this being a one possible way to change the current build for the better. I think one's Insane win ratio in 1.61 depends primarily a. on avoiding bugs and b. on whether or not the game happens to spawn an immortal murderbot rather than player choices being a meaningful predictor of performance, and I'm hoping this sort of conversation could help change it.

GLHF,
Snarvid

e: While I included "it depends on implementation" as a possible answer to the question included in the topic line, this obviously applies to all possible build edits (e.g. they could add +10000 to power and that would clearly not be an appropriate balance tool). I have therefore changed the poll question to determine whether or not respondents would be excited to see this approach tried, rather than asking them to express confidence it is the correct final answer ahead of time. (Also, I clearly have problems being concise, and the polls 100 character limit is bad for how I write!)

[helminthauge]

I agree with the idea of capped talent level, but I don't want enemies to have higher power. That's only making saves even less useful. And TBH, if you want to scale damage, why don't just scale inc_damage?

Plus I'd like to see enemies picking up more talents instead of investing all their points into several talents to some ridiculous level. Currently there're a lot of talents with no npc-ban but seeming never picked randomly.

[richardhawk]

Power boosts are not the way to go because that just makes Saves worthless. (This already happens when an enemy rolls a particularly large number on stat boosting talents anyway)

What the answer is is substantially more difficult to figure out. I don't think the current method of slapping on gigantic TLs is interesting at all, on top of it inevitably breaking all the time. It's just annoying when the enemy taunts you with a talent that does not work at all like its actual design. The most exciting ToME is when the wild variety of normals and elites - with their highly divergent natural statlines, attributes and natural talents - matters, but that only happens on lower levels of play. Is it possible to rebottle that lightning?

[Snarvid]

Fair enough. I think of saves as a non-factor in relevant fights in Insanity already once enemies significantly outstrip player level, but I suppose that should be fixed. Not sure if the answer is smaller boost to power or not boosting power (and going with the above damage boost suggestion, perhaps along with damage shield boosts).

I don’t know how to rebottle the lightning. Most solutions I can think of reduce variety to remove overly strong combos and nonbos, but if you had prebuilt class template combos then over the course of a game they would become very repetitious indeed.

A game where you didn’t know what enemies could do just by looking at them (maybe once you had seen a Fire Drake you would know what it could do, but you couldn’t tell what additional classes they had, or what talents within those classes, until after they had used them within your LOS) might add some of the lightning back, but higher difficulties were balanced around full information transparency and choosing appropriate counterplay so other contributors to difficulty would have to change quite a bit to make sure it felt more like wonder than cheapness.

[visage]

I don't want enemies to have higher power. That's only making saves even less useful. And TBH, if you want to scale damage, why don't just scale inc_damage?

Agreed. Ridiculously high stats, talent levels, and power lead to utterly broken combos on enemies and make entire mechanics (saves, application power, ability uptime, defense, etc) completely irrelevant.

[visage]

Given the differences in how weapon damage and non-weapon damage scale, wouldn't nerfing enemy stats while boosting their power dramatically shift the relative dangers of casters versus weapon-users?

[Snarvid]

I don't want enemies to have higher power. That's only making saves even less useful. And TBH, if you want to scale damage, why don't just scale inc_damage?

Agreed. Ridiculously high stats, talent levels, and power lead to utterly broken combos on enemies and make entire mechanics (saves, application power, ability uptime, defense, etc) completely irrelevant.

Well, let’s be clear that boosting power would only further break an already broken mechanic (saves) and would not increase most ability uptime, mess with player application power, or boost most defenses. However, I am happy to trade in power for inc_damage (and in fact, I did not know that such a thing existed because I’m not a coder). This would leave enemy damage shields weak, but would avoid scaling ACT and a few other problematic talents.

Does someone else (helminthague, since it was your idea?) want to create a different poll suggesting this, or shall I? If there’s relative consensus around any particular proposed fix to try I think it would be beneficial for the devs to see it, even if they don’t ultimately implement it.

[Arcvasti]

Very high talent levels being a problem is only really a thing because many talents do not properly implement diminishing returns like they're supposed to. If talents like Freeze that do not diminish properly are fixed to diminish properly, then very high talent levels are not a problem.

[Snarvid]

I don't believe that is true at the levels currently available in game (up to 50 or more), or else the problem is present in way more talents than just outliers like freeze. Flat DR, resist all, respen, and speed are all problematic at very high levels, as are duration near or longer than cooldown for certain powerful talents.

Or put otherwise - I would have a hard time believing that 50 levels of Automated Cloak Tesselation would be fine.

[Arcvasti]

resist all, respen, and speed are all problematic at very high levels, as are duration near or longer than cooldown for certain powerful talents.

Yes, that was what I meant by many talents not properly implementing diminishing returns.

[Snarvid]

According to the wiki, as I don’t code-dive, Freeze simply did not have a log function in its damage, which made it unlike most other things, so it on purpose did not have diminishing returns. ACT uses cTStmD, again according to the wiki. So one of these would be “sometimes Tome doesn’t implement its main talent level scaling technique,” and the other would be “when they are implemented it’s still not enough at high talent levels.”

[visage]

According to the wiki, as I don’t code-dive, Freeze simply did not have a log function in its damage

No, Freeze uses the normal combatTalentSpellDamage. The problem is that it multiplies its damage by its cooldown, and its cooldown increases with talent level.

[GlassGo]

Let's put it like this - if any changes will make game more predictable in a meanin "bland boring none-challenging" I prefer it stays like it was - because at least these "you are f***d up" situations are memorable.

[Nevuk]

Very high talent levels being a problem is only really a thing because many talents do not properly implement diminishing returns like they're supposed to. If talents like Freeze that do not diminish properly are fixed to diminish properly, then very high talent levels are not a problem.

Agreed. Freeze was also just changed in this patch.

The current setup is balanced around NPCs having very high scaling. Them having higher power instead would require more then even a 1.6 level rework of the system (it would make saves literally meaningless for players 100% of the time rather than just more often than not, for instance).

It's much better to just address edge cases as they emerge (disruption shield, freeze) than to redo the whole system.

[Snarvid]

It seems like there are multiple different lines across the relevant threads who are identifying as agreeing but seem to me not to be agreeing:

- some say Freeze and disruption are emblematic edge cases having to do with those specific skills.

- some say broad categories of skills (flat DR skills, resist all boosts, speed, global speed, and respen skills).

In response to my concerns with this second group, Arcvasti assigns them to list of “many talents not properlyimplementing diminishing returns.” If there is rough consensus that these all belong among the edge cases, cool. If not, I don’t see the game you see.