Estimating Hidden Numbers That Shouldn't be Hidden

[Dark Holy Elf]

In FEW, damage done was proportional to Str or Mag + weapon might (see this thread). I don't know if it's that simple this time (I've done no testing), but I wouldn't be stunned if it were similar. Str/Mag don't feel that important, and that might be because they're often way, way lower than weapon might. I've had underlevelled characters with good weapons still managing solid damage.

[Burklight]

15 hours ago, Dark Holy Elf said:

In FEW, damage done was proportional to Str or Mag + weapon might (see this thread). I don't know if it's that simple this time (I've done no testing), but I wouldn't be stunned if it were similar. Str/Mag don't feel that important, and that might be because they're often way, way lower than weapon might. I've had underlevelled characters with good weapons still managing solid damage.

I looked over that thread. It's for sure not Str or Mag + Weapon might. Str seems to be significantly more important than weapon might. Absorb Str/Mag are likely one of the BiS skills for those who get them. 

Spoiler

 

Edelgard vs fighter
Iron Axe (+20 atk)	Rusted Axe +20 str
Normal	Normal
42	59
41	57
43	60
41	59
42	56
43	57
42	57
41	56
41	59
42	56
42	60

 

I'll figure out how DO works in a few hours.

[Archeleon]

Spent way too much time on this but at least I found some stuff out regarding damage calculation:

Def works the same way as the prior game, it divides your raw damage. If your attack would do 1000 raw damage and it hits a 10 Def enemy it is reduced to 100. Motivation modifier is applied after the basic damage calculation and my guess is that other modifiers of that nature work the same.

Regarding the following tests...

The first numbers are the median of S1 damage using 30 samples for Mercenary since physical classes seem to have a randomization component to attacks. The numbers in parenthesis are raw damage after removing Motivation bonus and Def/Res reduction. Testing was done as Mercenary and Mage against the 7 Def/5 Res Myrmidon.

The following applies to characters in the Mercenary class only, they don’t work on other classes.

Edelgard – 103 Str

Iron – 160 (974), Steel – 173 (1053), Killer – 186 (1132), Silver – 198 (1205), Brave – 211 (1284)

Shez – 63 Str

Iron – 103 (627), Steel – 116 (706), Killer – 128 (779), Silver – 141 (858), Brave – 154 (937)

Petra – 40 Str

Iron – 70 (426), Steel – 83 (505), Killer – 95 (578), Silver – 108 (657), Brave – 121 (736)

I took the average of the damage deltas between weapon tiers (1041-952, 1113-1041, etc) and got 77.61, then divided it by 20. This gave me a ~3.9 multiplier for Mt. 

If we subtract the theoretical Mt bonus of the Iron Sword (77.61) from the “Iron” damage numbers we get roughly 896.30 for Edelgard,549.35 for Shez and 348.48 for Petra. If we take each value and divide them by their corresponding Str values we get a multiplier of ~8.7 for Str in all three cases.

Using the above factors, I was successfully able to predict the damage Monica with 32 Str and Iron Sword would do in average with S1 as a Mercenary against a Myrmidon with 7 Def (~51).

This also gives insight into just how much damage  +Str skills contribute. Str +10 gives you 87 damage which is ~22 Mt.

A 120 Str character would get 60 Str from a maxed out Absorb Str skill and would gain ~522 damage which is ~133 Mt.

On that note, this means that a natural cap of 120 Str gives 1045 damage (269 Mt).

It also means that the maximum damage bonus you can get from a weapon is 2212 (570Mt~254Str, barring weapons with Reckless Power).

In summary this means that a maxed-out Mercenary gets 2/3rds of their basic damage from weaponry.

 

The following applies to characters in the Mage class only, they don’t work on other classes.

Edelgard – 85 Mag

Iron – 130 (565), Steel – 142 (617), Killer – 154 (669), Silver – 165 (717), Brave – 177 (769)

Hubert – 54 Mag

Iron – 87 (378), Steel – 98 (426), Killer – 110 (478), Silver – 122 (530), Brave – 134 (582)

Shez – 43 Mag

Iron – 71 (309), Steel – 83 (361), Killer – 95 (413), Silver – 107 (465), Brave – 119 (517)

This is when I realized I didn’t have it all figured out. Trying to use the Mercenary multipliers to figure out Mage damage was impossible. This leads me to believe that each class has their own Mt and Str/Mag multipliers. I followed the same method as before to see if I could find numbers that would work and came up with this: Mage Mt Multiplier = ~2.6, Mage Mag Multiplier = ~6.

I think I am going to test other classes to see if I can figure out all the multipliers but from what I can tell the formula looks something like:

Damage = ((Str or Mag * Stat Multiplier) + (Mt * Mt Multiplier)/Der or Res)*Crit Mod*Sum of other Mods

 

 

 

 

 

 

 

 

 

Edited July 20, 2022 by Archeleon

[Burklight]

1 hour ago, Archeleon said:

Spent way too much time on this but at least I found some stuff out regarding damage calculation:

Def works the same way as the prior game, it divides your raw damage. If your attack would do 1000 raw damage and it hits a 10 Def enemy it is reduced to 100. Motivation modifier is applied after the basic damage calculation and my guess is that other modifiers of that nature work the same.

Regarding the following tests...

The first numbers are the median of S1 damage using 30 samples for Mercenary since physical classes seem to have a randomization component to attacks. The numbers in parenthesis are raw damage after removing Motivation bonus and Def/Res reduction. Testing was done as Mercenary and Mage against the 7 Def/5 Res Myrmidon.

The following applies to characters in the Mercenary class only, they don’t work on other classes.

Edelgard – 103 Str

Iron – 160 (974), Steel – 173 (1053), Killer – 186 (1132), Silver – 198 (1205), Brave – 211 (1284)

Shez – 63 Str

Iron – 103 (627), Steel – 116 (706), Killer – 128 (779), Silver – 141 (858), Brave – 154 (937)

Petra – 40 Str

Iron – 70 (426), Steel – 83 (505), Killer – 95 (578), Silver – 108 (657), Brave – 121 (736)

I took the average of the damage deltas between weapon tiers (1041-952, 1113-1041, etc) and got 77.61, then divided it by 20. This gave me a ~3.9 multiplier for Mt. 

If we subtract the theoretical Mt bonus of the Iron Sword (77.61) from the “Iron” damage numbers we get roughly 896.30 for Edelgard,549.35 for Shez and 348.48 for Petra. If we take each value and divide them by their corresponding Str values we get a multiplier of ~8.7 for Str in all three cases.

Using the above factors, I was successfully able to predict the damage Monica with 32 Str and Iron Sword would do in average with S1 as a Mercenary against a Myrmidon with 7 Def (~51).

This also gives insight into just how much damage  +Str skills contribute. Str +10 gives you 87 damage which is ~22 Mt.

A 120 Str character would get 60 Str from a maxed out Absorb Str skill and would gain ~522 damage which is ~133 Mt.

On that note, this means that a natural cap of 120 Str gives 1045 damage (269 Mt).

It also means that the maximum damage bonus you can get from a weapon is 2212 (570Mt~254Str)

In summary this means that a maxed-out Mercenary gets 2/3rds of their basic damage from weaponry.

 

The following applies to characters in the Mage class only, they don’t work on other classes.

Edelgard – 85 Mag

Iron – 130 (565), Steel – 142 (617), Killer – 154 (669), Silver – 165 (717), Brave – 177 (769)

Hubert – 54 Mag

Iron – 87 (378), Steel – 98 (426), Killer – 110 (478), Silver – 122 (530), Brave – 134 (582)

Shez – 43 Mag

Iron – 71 (309), Steel – 83 (361), Killer – 95 (413), Silver – 107 (465), Brave – 119 (517)

This is when I realized I didn’t have it all figured out. Trying to use the Mercenary multipliers to figure out Mage damage was impossible. This leads me to believe that each class has their own Mt and Str/Mag multipliers. I followed the same method as before to see if I could find numbers that would work and came up with this: Mage Mt Multiplier = ~2.6, Mage Mag Multiplier = ~6.

I think I am going to test other classes to see if I can figure out all the multipliers but from what I can tell the formula looks something like:

Damage = ((Str or Mag * Stat Multiplier) + (Mt * Mt Multiplier)/Der or Res)*Other modifiers

That's some really good work. That would also explain some of the findings in my previous post about Str being significantly more important than might. I also think the formula you have is likely spot on for normal attacks.

My only immediate question from your findings are "It also means that the maximum damage bonus you can get from a weapon is 2212 (570Mt~254Str)" 

That's based off the assumption that you cant get higher weapon might than 570, yes? Because you absolutely can with reckless power. 

And the other question is how do you know it it's class based and not attack based? It makes sense that S1 would have a different modifier than S2, right? Unless you used it to predict multiple strong attacks in the same class and had consistent results.

 

[Archeleon]

48 minutes ago, Burklight said:

That's some really good work. That would also explain some of the findings in my previous post about Str being significantly more important than might. I also think the formula you have is likely spot on for normal attacks.

My only immediate question from your findings are "It also means that the maximum damage bonus you can get from a weapon is 2212 (570Mt~254Str)" 

That's based off the assumption that you cant get higher weapon might than 570, yes? Because you absolutely can with reckless power. 

And the other question is how do you know it it's class based and not attack based? It makes sense that S1 would have a different modifier than S2, right? Unless you used it to predict multiple strong attacks in the same class and had consistent results.

Right, that is assuming you don't have things like Reckless Power on the weapon as that is the highest Mt you can achieve without relying on drop luck.

Regarding class vs attack, that is mostly from prior observations Ive made while grinding classes on Edelgard and Shez. It seemed like some classes just did more damage for some reason and I couldnt figure out why. I'm not sure what the relation is between different attacks chains yet but I will look into that at some point as well. I am hoping it is just an additional modifier that is added to the damage calculation based on how far along you are on the chain.

 

 

[Archeleon]

Here are the numbers for sword users. It seems like the Mercenary and Thief lines use their own modifiers each, with their Master Classes having a small bonus to Mt Mod. Though Assassin and Trickster might merit a bit more testing.

Myrmidon - Mt Mod = 4(3.88), Str Mod = 9(8.78)

Mercenary - Mt Mod = 4(3.88), Str Mod = 9(8.71)

Swordmaster - Mt Mod = 4(3.63), Str Mod = 9(9)

Mortal Savant - Mt Mod = 5(4.87), Str Mod = 9(8.60)

Thief - Mt Mod = 3(3.35), Str Mod = 8(7.85)

Assassin - Mt Mod = 3(3.42), Str Mod = 8(8.22)

Trickster - Mt Mod = 3.5(3.5), Str Mod = 8(7.90)

Edited July 17, 2022 by Archeleon

[Archeleon]

It looks like United Front is about an 8% increase per nearby ally.

Also, they have to be very close, it's not enough to be in the same stronghold. You can easily outrange the effect in the Mock Battle map. Pairing up guarantees you get the bonus though.

Edited July 20, 2022 by Archeleon

[Burklight]

I don't have numbers to support this, was eyeballing it. Pretty sure DO adds half of the off stat. So for physical hits, it would be (str + (mag*0.5)). It's for sure less than 100%, looks like about half to me. Either way, it would make it one of the better abilities for scaling damage based on @Archeleon's findings because in a scenario where you have maxed str and magic, it's basically another fully stacked Absorb Str/Mag.

[Archeleon]

Information regarding how Spd affects CA's CDs. 

Data below:

Spoiler

 

	41					68					97
	CD	CD-Spd	Delta	%	%perSpd	CD	CD-Spd	Delta	%	%perSpd	CD	CD-Spd	Delta	%	%perSpd
Lightning Axe	10	7.09	2.91	29%	0.7%	10	5.95	4.05	41%	0.6%	10	5.08	4.92	49%	0.5%
Drill Arrow	15	10.64	4.36	29%	0.7%	15	8.93	6.07	40%	0.6%	15	7.61	7.39	49%	0.5%
Armored Strike	20	14.18	5.82	29%	0.7%	20	11.9	8.10	41%	0.6%	20	10.15	9.85	49%	0.5%
War Strike	30	21.28	8.72	29%	0.7%	30	17.86	12.14	40%	0.6%	30	15.23	14.77	49%	0.5%

 

Conclusions reached:

Speed affects all levels of CAs and Spells at the same rate, regardless of Might and Level. Spells are affected at different rates, seems about double the effect.

Made a testing error but now I know that the Magic Mastery abilities reduce Spell CDs by 25%.

Speed does not seem to reduce cooldown linearly but rather diminishes in effectiveness the higher you get in Spd but not so much that you'd want to avoid stacking it. This is likely the case to avoid situations in which very high Spd characters would potentially reduce CDs to 0.

What might be worth figuring out is cooldown breakpoints for each category (IE Medium Spell vs Large CA); something to work on.

Edited July 20, 2022 by Archeleon

[ZanaLyrander]

Has anyone done any testing of the weapon attributes? I feel like those are gonna be a nightmare to test given the sheer number of vague ones not to mention each of them have three different levels.

Like, for example, Boost Combat Arts/Magic lvl 3, does anyone know how much of a damage boost that conveys? It makes evaluating weapons difficult when I'm not really sure how much of an advantage each attribute actually gives you.

[Archeleon]

23 minutes ago, ZanaLyrander said:

Has anyone done any testing of the weapon attributes? I feel like those are gonna be a nightmare to test given the sheer number of vague ones not to mention each of them have three different levels.

Like, for example, Boost Combat Arts/Magic lvl 3, does anyone know how much of a damage boost that conveys? It makes evaluating weapons difficult when I'm not really sure how much of an advantage each attribute actually gives you.

The hardest part about testing this would be getting three weapons with that attribute, one per level, and then hopefully buying another from the armorer without attributes. 

Then it's just a matter of using a single-strike CA/Spell (Wrath Strike/Fire) against an enemy type in an MB and recording the results. 

[Burklight]

18 minutes ago, Archeleon said:

The hardest part about testing this would be getting three weapons with that attribute, one per level, and then hopefully buying another from the armorer without attributes. 

Then it's just a matter of using a single-strike CA/Spell (Wrath Strike/Fire) against an enemy type in an MB and recording the results. 

I think we can skip figuring out what level 1 and 2 look like. The purpose would be to figure out what a BIS weapon would look like, and there's no way a level 2 attribute would end up being BiS. I'm also fairly confident that weapon attributes are significantly less important than attack given some of the early testing I did. If I had to speculate why, I would say that probably all damage modifiers are added before multiplying instead of each one actually multiplying. So the difference between say, 150% extra damage or 180% extra damage is less meaningful than 50 attack. If you've dealt with any PoE damage multipliers, think "increase" and not "more."

[ZanaLyrander]

3 minutes ago, Burklight said:

I think we can skip figuring out what level 1 and 2 look like. The purpose would be to figure out what a BIS weapon would look like, and there's no way a level 2 attribute would end up being BiS.

You're probably right, assuming no bugs in the coding, which, given what has already been discovered here in this thread about Apex abilities, may be a mistaken assumption. Certainly there's no way a level 1 or level 2 attribute should be BiS.

[Archeleon]

3 minutes ago, Burklight said:

I think we can skip figuring out what level 1 and 2 look like. The purpose would be to figure out what a BIS weapon would look like, and there's no way a level 2 attribute would end up being BiS. I'm also fairly confident that weapon attributes are significantly less important than attack given some of the early testing I did. If I had to speculate why, I would say that probably all damage modifiers are added before multiplying instead of each one actually multiplying. So the difference between say, 150% extra damage or 180% extra damage is less meaningful than 50 attack. If you've dealt with any PoE damage multipliers, think "increase" and not "more."

Given the damage formula we think the game uses, it sounds like you are saying that multipliers are additive and not multiplicative (I think I got the terminology right). 
Something like this:
100 damage (from basic Str/MAg+Mt formula)

with 50% multiplier = 150. 
with 30% multiplier = 130.

with 50% multiplier and 30% multiplier  (multiplicative) = 195

with 50% multiplier and 30% multiplier (additive) = 180 

Something else to look into.

[Burklight]

43 minutes ago, Archeleon said:

Given the damage formula we think the game uses, it sounds like you are saying that multipliers are additive and not multiplicative (I think I got the terminology right). 
Something like this:
100 damage (from basic Str/MAg+Mt formula)

with 50% multiplier = 150. 
with 30% multiplier = 130.

with 50% multiplier and 30% multiplier  (multiplicative) = 195

with 50% multiplier and 30% multiplier (additive) = 180 

Something else to look into.

That accurately describes what I think it's doing most of the time. There might be some edge cases. It wouldn't surprise me if some of them were always multiplying and not adding, but I suspect most of them are adding.

[Archeleon]

Can confirm that the majority of multipliers are added together before being factored into the baseline damage.

Testing details below:

Spoiler

 

Used Wrath Strike to get numbers.

Kept track of the Minor Crest of Seiros effect (MS/60%), Sword Prowess(SP/30%), Wild Abandon(WA/50%) and Offensive Tactics(OT/20%).

Baseline Damage	1805
MS+SP (Actual)	3328	MS+SP (x 60% x 30%)	3754	MS+SP (x 60% + 30%)	3430
MS+WA (Actual)	3845	MS+WA (x 60% x 50%)	4332	MS+WA (x 60% + 50%)	3791
MS+SP+WA (Actual)	4270	MS+SP+WA (x 60% x 30% x 50%)	5632	MS+SP+WA (x 60% + 30% + 50%)	4332
SP+WA (Actual)	3251	SP+WA (x 30% x 50%)	3520	SP+WA (x 30% + 50%)	3249
OT+SP+WA (Actual)	3701	OT+SP+WA (x 20% x 30% x 50%)	4224	OT+SP+WA (x 20% + 30% + 50%)	3610
MS+WA+SP+OT (Actual)	4692	MS+WA+SP+OT (x 60% x 50% x 30% x 20%)	6758	MS+WA+SP+OT (x 60% + 50% + 30% + 20%)	4693

 

 

 

Edited July 19, 2022 by Archeleon

[Pie Burritos]

A shame about Apex skills. I'll start dequipping them right now.

Do we have a spreadsheet where we can see the results of the testing and see a short and digestible version of what each skill does?
Also, I'm very interested in seeing what multipliers are added together, because having sources that multiply instead would obviously be the best for increasing damage. I seem to vividly remember a flat 1.5 increase to damage once I equipped Impossible Feat, even with all my other damage boosting skills equipped. Maybe it was just a coincidence, though. That being said, it seems like a lot of work to see which skills belong to which addition group because if we want to be a thorough as possible, every skill combination has to be tested. 

[Archeleon]

26 minutes ago, Pie Burritos said:

A shame about Apex skills. I'll start dequipping them right now.

Do we have a spreadsheet where we can see the results of the testing and see a short and digestible version of what each skill does?
Also, I'm very interested in seeing what multipliers are added together, because having sources that multiply instead would obviously be the best for increasing damage. I seem to vividly remember a flat 1.5 increase to damage once I equipped Impossible Feat, even with all my other damage boosting skills equipped. Maybe it was just a coincidence, though. That being said, it seems like a lot of work to see which skills belong to which addition group because if we want to be a thorough as possible, every skill combination has to be tested. 

Every skill would have to be tested to know for certain. So far though it seems they all are added together before being factored into the standard damage equation. I don't have access to Impossible Feat yet but it would be easy to test as we know Prowess is one of the abilities that follows the addition rule. Assuming IF does in fact give a 50% increase then using it with Prowess should show an 80% increase and not a 50% x 30% increase.

[Pie Burritos]

I got off my lazy bum and tested it, using a 120 str myrm shez with an unforged iron sword. The short version is that yes, it's additive, and the 1.5 was just a coincidence. Impossible Feat provides a 100% damage increase. I also tested it with Axe Buster Lvl.1 since it was more convenient for fast testing since I could alternate between hitting the training myrm and training fighter.

The full test results are below.

Spoiler

Edit: Added crits to the dataset. It was already confirmed but obviously critical hits are an independent multiplier. Getting lots of them becomes more important the more damage boosting skills we have.

 

Edited July 19, 2022 by Pie Burritos

[Archeleon]

Interesting that Crit is it's own separate multiplier. It does make critical damage boosting skills a lot more powerful than I thought. Shame that Apex is bugged right now.

Also makes Impossible Feat and Lifeforce the highest multipliers I've seen so far. But IF applies to everything unlike LF.

How bad is the backlash from IF?

Edited July 19, 2022 by Archeleon

[ZanaLyrander]

2 minutes ago, Archeleon said:

Interesting that Crit is it's own separate multiplier. It does make critical damage boosting skills a lot more powerful than I thought. Shame that Apex is bugged right now.

Also makes Impossible Feat and Lifeforce the highest multipliers I've seen so far. But IF applies to everything unlike LF.

How bad is the backlash from IF?

In my personal experience, it's pretty painful. I don't know the exact numbers (and the description seems to indicate that it varies based on the might of the attack), but I feel like I lose roughly 1/8 of my max hp when I use Hades.

[Pie Burritos]

The backlash from Impossible Feat seems to tick three times as fast as Renewal recovery, at the same 0.5% rate. If you have a sacred weapon and the corresponding crest, then the backlash is offset a decent amount but it still gets outpaced. Though if you have some form of recovery, like Mortal Savant's elementally infused C7 (or C6? whatever, it's the last combo), then it's easy to keep your health topped up. Other characters like Shez can use Rejuvenation from Grappler to keep their health up since Fluegel/Asura deal so many hits so fast. The health drain is far less severe than Lifeforce.

Edit: TLDR; Impossible Feat is worth it on every character that gets it, but you should have some sort of plan to recover health.

Edited July 19, 2022 by Pie Burritos

[Burklight]

Actually had a busy work day so I'm a little late to this one. I've been running IF on everyone who can use it. If you're playing well, you shouldn't be getting hit anyway and the damage only matters once your hp drops below 25% and then I use an elixir so it goes back to working. The damage doesn't count against your rank, so as long as you don't get one tapped from half hp you're good. If you're paranoid, you could always run Mercedes or Constance for Light Assist and heal that way. 

@Pie Burritos I listed numbers for everything I tested in the first post. They're in the spoiler.

As far as knowing whether a skill adds or multiplies, you won't need to test every combination. Just need to test every skill with one other skill that's in a known category. It's going to be particularly important for weapon abilities, because weapon abilities that add are going to be deceptively not as impactful as they seem like they should. For example, lets say you have a weapon ability that in isolation gives +30% damage, but in the build you're using it on, you already have +300% damage though passive skills. That +30% is going to in practice only add 10%.

[Pie Burritos]

Like you said, it seems highly likely that they're categorized and that for the most part all regular abilities are in the same category of damage boosting, whereas stuff like motivation and crits are in different categories. The next thing to test would be personal abilities. Which I've done with Felix's Lone Wolf! Seems to be additive, which is what I suspected, at a +30% boost for level 3.

 

Spoiler

Obviously my testing was slightly lacking in certain areas but I think it's good enough.

[Sire]

Interesting.

So, if I have this right, most abilities are simply additive. (So, Sword Prowess 3 [+30] and Offensive Tactics [+20] will result in +50% damage.)
-- Getting a Critical Hit is a multiplier, although Apex Weapon is bugged and has a lower multiplier than default.
-- Motivation is a multiplier, ranging from +5 to +15 depending on the level. (I'm not sure if this has been tested yet)

If I'm interpreting this correctly...

(Damage * Abilities) * Motivation * Critical Hit = Total Damage
-- Damage likely has its own formula regarding STR, MAG, MT, and what attack/combat art/magic is used. [Alternatively, perhaps its class based as it was outlined in a previous post]

* * * * *

So, Felix's Lone Wolf is an additive multiplier with only a +30. Guess it's a useless skill since having a Battalion is almost always superior, for having advantage as well as reducing damage taken. I suppose the same can be said for Catherine's personal.

Edited July 20, 2022 by Sire
Motivation has been tested