Jump to content

thelee

Members
  • Posts

    4155
  • Joined

  • Last visited

  • Days Won

    9

Everything posted by thelee

  1. wow, impressive script, shai hulud. this is inspiring me to create a script that's just focused on consumables that i can load up on all my party members. (i pretty much always have one script slot empty) can't run into the AI consumable bug if the AI is always consuming when I need it to anyway
  2. one thing i learned from taking doing physics and stats classes is that it's very important to be clear what your target units of measurement or that you are accurately selecting the correct metric. everything else will flow from that. the correct metric: the RETURNS you get from investing in dex or attack speed is how much effective damage you do over time (generally per second, since DPS is the normal way gamers talk about this stuff). Not recovery time. Tracking recovery time will give you a curve, because it's the denominator. it's not the metric, just a part of the metric. the correct units of measurement: what you should do is invert your attack speed, essentially, because that converts "seconds per attack" into "attacks per second" (which then is a straight-forward shot to "damage per second"). the original relationship is inverted (1/x), and an inversion will look somewhat loggy or polynomial, but it's neither, which is why neither a log transform or polynomial fit will work. there's a graph part way down this page which is a perfect line: https://gamefaqs.gamespot.com/pc/227477-pillars-of-eternity-ii-deadfire/faqs/76599/action-speed-is-linear-returns edit: what bosmer said about relative differences is right, i'm speaking more generally about how those relative differences end up manifesting, mathematically, such that it becomes basically indisputable that action speed has linear returns.
  3. don't quote me on this, but when I was messing around with magran's challenge run (abandoned, it's not worth it at all lol), i found i was somehow able to avoid the potion-drinking bug when my Eder's AI script had a potion drinking action at the bottom of the list (even with other potions). i haven't investigated it thoroughly, but maybe there's a way out of the potion-drinking issue. omg that's hilarious. must be like how AI script can ignore hidden health (like from berserker rage) and trigger health status conditions. i need to do that on my scripts, would make final stand much more useful (frequently i miss the opportunity to use it)
  4. two reasons, both related: one - the game tries to preserve the current % of health whenever max health is adjusted (e.g. from CON changing up and down from afflictions and inspirations). two - more intuitively, if you somehow permanently reduce an enemy's CON, then you objectively need to do less damage to kill it. effectively, you've given yourself a damage boost. an extreme example is if you somehow reduced an enemy's health from 1000 to 200. You only need to do 200 damage to kill it. You've effectively 5x-ed your damage, because you effectively did 1000 damage compared to before the health reduction. the first point I mentioned above is relevant because this would all be for nothing if, when a con adjustment wears off, the enemy gains back the entirety of the missing health. instead, the game continues to try to preserve the relative share of health. So if you somehow reduced an enemy's health from 1000 to 200, but only managed 100, when the con effect wears off, the game puts the enemy's current health back to 500 health (goes from 100/200 to 500/1000), not 900 (i.e. 100/200 and then +800 current health when their max health goes back to 1000). the specific 12.8% multiplicative damage bonus number was derived from the specific numbers you shared from dorudugan's health pool above (vanilla numbers). edit: like i said and you've noticed, there's some weirdness going on with dorudugan's numbers specifically for some unknown reason. but for more reasonable fights (even against enemies with ~1000+ health), any error should be overwhelmed by the general advantage of doing more damage (a ~1% difference in current health only becomes noticable at dorudugan's massive ~16000 hp health pool), and the effect is more powerful the lower the enemy's starting CON is.
  5. CON inspiration and CON affliction are both just two sides of the same coin. But that thread was specifically about how those conspire with effects that add a set amount of health (mainly those amulets that grant +health), and how the game doesn't handle the math right. I actually don't remember which way the interaction goes, it's pretty obscure (but had an application in letting you infinitely restore health in e.g. woedica's challenge). But the Dorudugan thing shouldn't be the same thing here, I doubt he/she/it has a specific health buff. Given the closely-aligned numbers involved in Shai Hulud's post outside BPM (87.4% of health to 88.7% of health, back down to 88.3% of health), I half suspect maybe it's some funky rounding with the difficulty health boost (which, thanks to this thread I now realize is a multiplicative bonus on the enemy's total health), which we only really notice here because of the massive numbers involved with dorudugan's health pool. edit: i speculate that it has something to do with the difficulty adjustment mostly because back in that thread i think Boeroer had specific numbers and MaxQuest had an equation that matched the game to a T, so I don't think CON adjustments alone causes the health healing weirdness. edit 2: also want to underline that this concern is probably mostly for solo, since if you have reasonable uptime and reasonable party DPS, even with dorudugan healing some tiny % of health every time the CON effect wears off, it'll get overwhelmed by what is a 12.8% multiplicative party damage bonus against it by virtue of good uptime on the con affliction.
  6. i don't think there's a default way in the AI scripting system to control the range something is used. and rage doesn't take a target, so that especially i don't think you have control over. i think your best bet to avoid wasted action is just to maneuver serafen manually next to enemies.
  7. in the past i've heard that playing at midnight can help, because it reduces commoners and those are the ones that can trigger the bug. maybe give that a try?
  8. sorry to necro this, but any chance of incorporating whatever fix @Elric Galad did for the animal companion attack inconsistency into the community patch?
  9. i guess an open question is - what CRPG has done summons generally very well across the board? i'm not sure i have a good answer to that. I've been trying Solasta of late, and even with extremely rigid adherence to 5e SRD, there are some way too good summons and a lot of garbage, because the tabletop game itself doesn't seem to have put a lot of consideration into making summons work well, even though it has stronger restrictions than Deadire on how it copy-pastes monsters (challenge rating?)
  10. it's funny though because incarnates suffer both from being underpowered, but also because spiritual ally is essentially overpowered. like arguably in a vacuum the incarnate summons (except maybe wael) still are servicable use of your action economy in keeping enemies occupied and dealing a bit of damage. but b.c. of the same design oversights, spiritual ally comes with infinite uses of healing, which is ordinarily limited by an AI script, but which you can spam endlessly. compared to that, even Rymrgand is a hard sell, since they both compete for the "summon slot" that you're allowed to have.
  11. maybe i'm missing something, but are you not able to do the Strand of Favor cheese? that should give you plenty of invisibility (You can rapidly pick it up and drop it back into your inventory slot, and every single instance of this gives you a compounding 10% increase in duration. Also works with that armor that has the +10% duration enchantment (the gambeson))
  12. Ive been off social media for lent, so sorry for the necro, but say whaaaaat?? This is a vanilla game issue? seems like this should also be addressed in the community patch bug fixes
  13. alright i think you've convinced me on energized v swift. it also means you have a little bit more work to benefit from the interrupts (even if it's just marked prey and stalker's link) whereas swift would have been an easy pickup for anyone.
  14. I guess Bonded Fury granting Energized + Swift would solve the problem for Maia for a non-pet maia I think Heart Seeker is a great capstone (esp bpm version) BTW does BPM do anything with Twinned Shots? It always struck me as underwhelming.
  15. you ninja-ed me on post-timing, but I disagree that this is bad. Maia's pet is too good. Maia's subclass is too good. Swift on a pet would be a payoff for SC Rangers that gives them an option that can compete with Maia's pet. As it is, it's frequently hard not to feel like a mainchar or merc ranger+pet setup is a bit of a trap compared to just rolling with Maia. Maia would still get an incredible pet for 100% of the game, and could multiclass and still have an incredible pet. Swift on Bonded Fury would just let mercs/mainchars have a similar lategame option in their toolkit if they want to.
  16. i think this is just way too good. i think you're overemphasizing the pet. if you need to have an ability this overtuned at tier 8 to make a pet viable, then what you're basically saying is "pet isn't viable." which i don't agree with. imo, it should be reasonable to have a pet build and go w/out a talent without major cost, but you're basically making a skill that would be a trap for pet builds to not take. (for this reason I don't particularly like Vicious and Resilent companion, they are basically talent taxes for rangers, at least on PotD) the problem we're trying to address, if I were to summarize, is simply that Bonded Fury as it is, isn't a particularly great payoff for single-class rangers, especially since a party of 5 (or a multiclass) will frequently be able to cover several tier 2 inspirations. We don't have to go all the way to all that to make it useful. I honestly believe as I mentioned upthread that simply just granting a medium-duration Swift even is Good Enough to make it worthwhile. Tenacious+Swift if you want a bit more (and Tenacious is a bit more "Fury"-ish). (Energized + Swift would be pretty good, but might be busted on lower difficulties where crits are easy to come by with a pet.) Yeah, it kinda breaks the original theming of Bonded Fury, but I don't think it was a particularly compelling theming to begin with. It would give pet builds a lot more survivability and utility, and would make single-class pet rangers less of a trap compared to just including Maia into your party (again, her bird basically is the best all-around pet and there's nothing much the mainchar can do about it). And it would not be a talent tax for pet builds, because you still have relative alternatives (Master's Call).
  17. still a mini combo! not everything has to be hermetic to a solo character. though the big problem with this stuff is that at least on PotD enemy health way outscales your character, so relying on returning some of the damage you receive in any meaningful way does not seem like a great strategy. at least with stuff like riposte, you are avoiding damage at the same time.
  18. ime it's far more likely for the pet to be stationary for extended stretches of time, so it'd be mostly "free" resource gen. i like the idea of a mobility reward, because the ranger has some mobility pay-off (shot on the run mostly) and it encourages an atypical playstyle. i'm not sure how that would actually play out in terms of implementation though
  19. in terms of heal companion buff, have you considered giving it an instant cast and no recovery? there's still an unpleasant action economy to heal companion, and Play Dead has instant cast. at least then "Heal Companion" has value in that it doesn't cost you action economy to use to use more pre-emptively. I haven't played aroudn with it in BPM mode, but when heal companion is available, enemies tend to be able to take a pet from healthy to dead pretty quickly, so you want to use it pre-emptively, but also not necessarily waste cast-time (even down to .5s BPM) and recovery time just to do that in case there's no real danger coming up on the pet. adding resource recovery to bonded fury seems too good, imo. it would basically be an auto-pick in any build. i'm circling back to an earlier point - if bonded fury had a narrower set of tier 3 abilities, that would be extremely good for a pet-support. Swift in particular (whatever other T3 is an exercise to the reader). Maia has the best pet in the game because her bird can face tank megabosses because it's immune to engagement, a single-class ranger focusing on their pet should get a similar payoff and not need to rely on an exhortation.
  20. if like, you get a couple skellies in one hit you get two free full attacks on the enemy? (or is it one cleave per attack. doesn't really come up often, so curious)
  21. probably not what you're looking for, but "everything bad" i find to be really hilariously fun challenge. everyone's dead, you get cursed by basically every god, etc. more seriously, the non-"everything bad" pre-built legacies generally get you everything in the game, whereas if you import or do your own custom history you could easily choose to resolve quests and story points in ways where you deny yourself unique stuff. the one exception that i can think of is the Devil of Caroc plate, only some of the pre-built histories will get you that, and i'm only pretty certain of "everything good" option being one of them. (there may be another) edit: you won't be able to get "everything" using a pre-built history, because the berath option (fair and balanced) is the only way to get the unique consumable "berath's bell", but also impossible to get devil of caroc breastplate. similarly i don't think there's any other way to get hylea's consumable without benefiting her in the legacy, which is mutually exclusive with berath's bell. can be important for e.g. a no-rest or low-rest run where the hylea's consumable is very very good.
×
×
  • Create New...