Random Notes on Damage Mechanics in Granblue Fantasy: Relink
Preface
This is probably a rather unorganized and incomplete write-up, but at least I would like to get some information out there. I enjoy testing things a lot, especially useless things, and when it comes to Relink, there are quite a few topics that piqued my interest during my initial playthrough.
The problem with (testing) this game
What troubles me the most is the fact that Mastery Tree upgrades are mandatory. This, in many cases, makes controlled testing unnecessarily complicated with upgraded characters. The game also does not show the total value for important stats such as critical damage, damage cap, etc., forcing us to count each Mastery node by ourselves.
Therefore, if I missed anything important or made any errors, feel free to correct me. Sharing any interesting data or topics is also very welcome.
Damage Basics
Relink takes a lot of inspiration from the original mobile game, so a few mechanics have been adopted with similar calculations, but not everything is the same.
Damage is scaled based on ATK stat. Each action/skill has an assigned ratio and a damage cap.
The only modifier known to be applied after damage cap is elemental advantage (1.2x multiplier).
Base critical damage without any Mastery is 20% (1.2x multiplier).
DEF stat might exist, but so far I have not found any enemy (or character) with more than 1 DEF.
As stated in Lyria's Journal, bosses have part toughness/weakness and critical hits ignore part toughness.
Ancient/elemental dragons have 1.2x weakness on their heads (1.5x after broken) and 0.8x toughness on their legs.
Gallanza (ガランツァ) and Id (イド) seem to take 1.2x damage during certain attack animations.
Overdrive state seems to reduce damage done to the boss by 10% on the first time, and 30% on the second time.
Damage variance seems to be in a range of ±5%.
I believe Maygi has already done more extensive research than me on damage calculation and offensive Sigils overall, so I would like to suggest checking out her current and upcoming resources.
Sigils (ジーン)
Sigils (ジーン) still require more extensive testing to confirm how they stack, though most of them are not used very often. Generally, they are multiplicative with at least a few confirmed exceptions.
Critical Damage is additive to Critical Hit Damage Up from Mastery nodes.
Exploiter (弱点攻撃) is additive to Back Attack Damage Up from Mastery nodes.
Life on the line (捨て身), Combo Booster (コンボボーナス), Power Hungry (修羅), Less Is More (裸一貫) are in the same additive group (they are additive to each other).
Concentrated Fire (集中砲火), Overdrive Assassin (オーバードライブ特効), Break Assassin (ブレイク特攻) have been confirmed to be outside of this group. Other Sigils still need more confirmation.
Garrison (堅守) seems to max out at roughly 20% HP remaining.
Status Effects
One interesting aspect of Relink is that the same buffs/debuffs cannot stack. The one with more potent percentage gets priority and cannot be overwritten by weaker ones.
ATK/DEF
ATK↑/↓ effects are additive to each other.
DEF↑/↓ effects are additive to each other.
If the result is a damage reduction, it is applied as × (1 - ratio).
Otherwise (damage amplification) it is applied as × (1 + ratio).
Damage Cut is multiplicative to DEF↑/↓.
Phantasmagoria is additive to ATK↑/↓ but multiplicative to DEF↑/↓.
Here are some values unstated in the game that I have personally tested (still a very small list I know):
Cagliostro
Phantasmagoria: ATK↑ 30%, DEF↑ 30%
As noted earlier, it is additive to ATK↑/↓ but multiplicative to DEF↑/↓.
Evyl Blackwyrn (黒竜イーヴィル)
Abyssal Glory: DEF↑ 50%
ATK↓: 20%
DEF↓: 50%
Burn/Poison
Burn and Poison do non-lethal damage (cannot kill) over time based on a percentage of max HP. The percentage amount varies by source. Here are what I have confirmed so far:
Infernadile/フレイダイル (The Next Hot Topic/斬新なファッションには斬新な素材から - Very Hard): 2%
Vrazarek Firewyrm/火竜ヴレッザレク (Blazing Trial/紅蓮 - Proud): 3.2%
Cobra (Chapter 6 Part 2: Relics in the Sand/砂漠に眠る遺跡)
Action: 1.2%
Hard: 2%
Ultimate: 3.2%
I still need to figure out DoT from player to bosses. It is likely similar but the percentage amounts are unknown at the moment.
Another not very useful thing to know is that DoT damage is calculated from internal HP value including decimals (not the displayed HP value which is rounded to the nearest integer), while the actual DoT damage displayed and applied seems to be rounded down.
More interesting topics?
There are some more tidbits that I left out because I wanted to gather more data, or because I believe other people can test and present them better.
In the future, I will probably test more things and look into more interesting topics (which might be useless for gameplay, who knows), but for now, I would like to get back to the grind and have fun.
I hope this has been an interesting read, if not helpful for those out there trying to calculate things.