RaIsMyPet Posted November 3, 2020 Share Posted November 3, 2020 (edited) Older version with mistake Spoiler I found a grave error in that whenever you have a multiplier, enemy DEFG is also multiplied (but not HPG). I kind of tried to explain why certain builds are good from a different point of view, yet I fear I need a more careful analysis. Stay tuned while I revise it and see the result. Introduction In Maddening, enemies have crazy stats, especially during timeskip. Many viable strategies become unreliable because of player stats just don't keep up with the enemies. But a handful of units manage to shine in this mode despite all that. So what's happening? Here's a brief analysis of how many physical unit builds compare to enemy scaling, and their long term viability. Conventions By (stat)G we mean growth of that stat, e.g. HPG=HP growth, STRG=strength growth, CRG=crit growth=(DEXG+LUCKG)/2, etc. We assume no-doubling if the attack is not forced double, unless certain unit is usually so fast that can double most enemies reliably (which we will give a range instead of a number), e.g. Petra. Effective Damage growth and Multiplier For example, a vengeance user will have damage growth (DAMG) equal to HPG+STRG, greatly enhancing long term damage viability, since enemy BULKG=DEFG+HPG For another example, a PBV user has a multiplier of 2x to her STRG, while a FIF user has a multiplier of 3x. Regarding Crit We only consider "reliable crit", e.g. a Swift-striker (SSer) with killer lance+ is usually considered not critting as even 50% crit x2 is still no crit 25% of the time, hence not reliable. On the other hand, a War Master with around 80% crit x 2 is only no crit 4% of the time. Factoring how many combats usually happen to one unit, we will consider such a War Master to have x2 with 1 crit, hence a 4x multiplier to STRG Constant Bonus Base stats, powerful weapons, effective damage, etc. Sometimes even if the multiplier is not good enough, a large constant bonus can help keeping up for a while. It helps in that the game is not infinite. Some Other Things I might forgot to mention (add here) Results Base Line Enemy BULKG=HPG+DEFG, so approximately a 2x multiplier is needed. However, since enemy HPG is so large, the multiplier usually needs to be more than 2x. Alternatively, one can consider large constant bonus like Lance of Ruin. List of Long Term Viable Units/Builds Format: Name (multiplier) M!Byleth/Felix (4x - 12x): War Master who can frequently quad and reliable 80 displayed crit, 4x is from 2 hits with 1 crit, and 12x is from quading monsters. Felix has 70% STRG as WM, hence a 280%~840% DAMG, and Byleth is 240%~720%. They outgrow enemies significantly, hence the game gets easier and easier for them. Other WMs are able to reach 4x, but usually cap at 6x since most punching boys can't quad a lot. That's still outgrowing enemies btw. Enemy LUCKG is never high so one can assume CRITG can keep up. FIF (5x - 9x): 3 hits with + 10 crit and likely reliable at least 1 crit, up to 3 crit against monsters. Most punchers break 55% STRG, hence 275% - 495% DAMG Petra (3x - 6x): EP build with BWrath and semi-reliable doubling. WL Petra has 55% STRG, hence 165% - 330% DAMG. Due to being EP, Petra has additional edge in being constantly higher-leveled than your other units. Another "jump" effect is that SPDG matters here. Petra can outgrow enemy Wyverns, keep up with at least enemy heroes, and presumably falcons/WMs as their enemy class bonus for these classes is only 20%, while WL Petra has +10% bonus, and I assume Petra has more base SPDG than generic enemies. Not sure about Swordmasters and Assasins. To clarify I'm talking about SPDG here, not SPD, as I'm assessing long-term viability Claude (3x - 6x (shaky?)): similar to Petra, but he is slower, hence doubling is shaky for him, unless you give him some speed boosters. The difference is small yes, but considering doubling AS is only 4 points margin, one has to keep this in mind at least Hilda (3x - 6x (shaky)): similar to Claude, but with a lot more investment to set up, perhaps even nontrivial favoritism Dimitri (3x): BVBW with at least 70% STRG, 210% DAMG, reliable Swift-Striker (2x): with large constant bonus due to abundance of powerful lances Dedue (~2x): Vengeance with high STRG and HPG, with powerful lance bonus Cyril (~2x): a little weaker Vengeance than Dedue, but should be just enough to keep up with enemy. (add here) List of Unsure/Shaky/Part-term Viable Units/Builds Bernie (~2x): Vengeance with low STRG+HPG, compared to enemy DEFG+HPG (huge). Powerful lance bonus doesn't work well here as it is added only once, unlike twice for Swift-strikers PBV (2x): Leonie and Cyril has middling STRG, and can't use Failnaught so no large constant bonus. Very shaky considering they also have awkward class choice Hunters Volley (2x): similar issue to PBV, but maybe Ignatz can make critting reliable? Brave Weapons (2x): not nearly powerful enough. Perhaps Edelgard can make brave axe work. Not sure. Usually too heavy to quad. Edelgard (1x): but HUGE constant bonus in Raging Storm, and galeforce effect, still the multiplier is too small Non-viable in Long Term Scaling CAs (~1.3x? ~1.33x? I forgot): multiplier too small Standard double (shaky 2x in most cases): worse version of brave combat arts Astra (1.5x): shit Comments and reminders welcome! Edit: Some might say the game is finite hence PBV/HV should be viewed as reliable. But I beg to differ: another difference between 3x and 2x is that the former can afford being STR-screwed for a while, but the latter can't. Archers also have low-middle STRG to start with, hence I would still consider them shaky. Revised version: Introduction In Maddening, enemies have crazy stats, especially during timeskip. Many viable strategies become unreliable because of player stats just don't keep up with the enemies. But a handful of units manage to shine in this mode despite all that. So what's happening? Here's a brief analysis of how many physical unit builds compare to enemy scaling, and their long term viability. Conventions By (stat)G we mean growth of that stat, e.g. HPG=HP growth, STRG=strength growth, CRG=crit growth=(DEXG+LUCKG)/2, etc. We assume no-doubling if the attack is not forced double, unless certain unit is usually so fast that can double most enemies reliably (which we will give a range instead of a number), e.g. Petra. Effective Damage growth and Multiplier For example, a vengeance user will have damage growth (DAMG) equal to HPG+STRG, greatly enhancing long term damage viability, since enemy BULKG=DEFG+HPG For another example, a PBV user has a multiplier of 2x to her STRG, while a FIF user has a multiplier of 3x. Regarding Crit We only consider "reliable crit", e.g. a Swift-striker (SSer) with killer lance+ is usually considered not critting as even 50% crit x2 is still no crit 25% of the time, hence not reliable. On the other hand, a War Master with around 80% crit x 2 is only no crit 4% of the time. Factoring how many combats usually happen to one unit, we will consider such a War Master to have x2 with 1 crit, hence a 4x multiplier to STRG Constant Bonus Base stats, powerful weapons, effective damage, etc. It is affected by multipliers, but not leveling. Enemy Bulk Enemy DEFG is also affected by multiplier, but not HPG. In order to outgrow enemy, we need STRG * M > Enemy-DEFG * M + Enemy-HPG Some Other Things I might forgot to mention (add here) Results Base Line Unfortunately, no unit can keep up with enemy indefinitely. Enemy Paladin has HPG=110%, DEFG=65%. Even WM Felix quading with 100% crit, he is still losing 1 damage every 2 levels. Vengeance is around 2x multiplier, but bypasses enemy DEFG scaling with multiplier, but even Dedue can't keep up with enemy growth. Dimitri is right: eventually I too will fall. What keeps the game alive? All the constant bonuses, obviously, so multipiler is still quite relavant. List of Multipliers Format: Name (multiplier) M!Byleth/Felix (4x - 12x): War Master who can frequently quad and reliable 80 displayed crit, 4x is from 2 hits with 1 crit, and 12x is from quading monsters. Other WMs are able to reach 4x, but usually cap at 6x since most punching boys can't quad a lot. Enemy LUCKG is never high so one can assume CRITG can keep up. FIF (5x - 9x): 3 hits with + 10 crit and likely reliable at least 1 crit, up to 3 crit against monsters. Petra (3x - 6x): EP build with BWrath and semi-reliable doubling. Due to being EP, Petra has additional edge in being constantly higher-leveled than your other units. Another "jump" effect is that SPDG matters here. Petra can keep up with enemy Wyverns, maybe heroes. Claude (3x - 6x (shaky?)): similar to Petra, but he is slower, hence doubling is shaky for him, unless you give him some speed boosters. The difference is small yes, but considering doubling AS is only 4 points margin, one has to keep this in mind at least Hilda (3x - 6x (shaky)): similar to Claude, but with a lot more investment to set up, perhaps even nontrivial favoritism Dimitri (3x): BVBW Swift-Striker (2x): with large constant bonus due to abundance of powerful lances Below things may become shaky endgame Dedue (~2x): Vengeance with high STRG and HPG, bypass Enemy multiplier, Powerful lance bonus doesn't work well here as it is added only once, unlike twice for Swift-strikers Cyril (~2x): a little weaker Vengeance than Dedue, but should be just enough to keep up with enemy. Bernie (~2x): Vengeance with low STRG+HPG PBV (2x): Leonie and Cyril has middling STRG, and can't use Failnaught so no large constant bonus. Hunters Volley (2x): similar issue to PBV, but maybe Ignatz can make critting reliable? Brave Weapons (2x): not nearly powerful enough. Perhaps Edelgard can make brave axe work. Not sure. Usually too heavy to quad. Edelgard (1x): but HUGE constant bonus in Raging Storm, and galeforce effect, still the multiplier is too small Definitely not viable Scaling CAs (~1.3x? ~1.33x? I forgot): multiplier too small Standard double (shaky 2x in most cases): worse version of brave combat arts Astra (1.5x): shit Comments and reminders welcome! Edited November 3, 2020 by RaIsMyPet Link to comment Share on other sites More sharing options...
Dark Holy Elf Posted November 5, 2020 Share Posted November 5, 2020 (edited) I like this list, even though I have some disagreements about viability. I think you underrate braves a bit, especially brave axe from a wyvern. 2x multipliers with high strength kill things even at the end of the game. I'm not sure why you think Swift Strikes deserves a higher tier than brave weapons... even with the Lance of Ruin (24 might), Swift Strikes only beats out a Brave Axe+ by 6 might when used from a wyvern. Characters with high strength like Dedue, Felix, and certainly Edelgard/Dimitri/Byleth, can compensate for that with higher str than the Swift Strikes users. Unless you go Paladin I guess, but then our brave axe-user has a ridiculous advantage in mobility. You're also underestimating their ability to quad in the hands of faster characters. You mentioned that you feel MByleth and Felix can frequently quad. Well, Petra has 9 more speed than Felix lategame with Darting Blow, and probably about 12-13 more than MByleth (spotting him a level lead). She needs just 25 str to match Felix if fully weighed down by the Brave Axe. Get her higher str or set Weight-3 or give her a lighter brave (or Thunderbrand, Inexhaustible) and she can open up a pretty big lead, as can Leonie and Ingrid... even Hilda and Edelgard end up comparable. (From the above two paragraphs, you can see why it's so important to have either very high str, or very high spd, in this game. Without the former you struggle to 2HKO with braves or HV, without the latter you can't quad except against armours.) For monsters, quad them with the brave weapon they're weak to (almost every monster is weak to one of sword, lance, axe, or bow) and watch them fall to pieces. The fastest monsters are usually flying and die to Brave Bow / Inexhaustible / HV / PBV. In Leonie's case, yeah her peak PBV damage is a bit behind peak Swift Strikes damage (although FWIW, she definitely can use Failnaught, 10 backlash damage is acceptable for a player phase attack... still, it should be pointed out that's only on 2 routes, and quite late on one), but her big advantage over the Swift Strikes users is she's waaay faster than them, so she can often score natural doubles (at range>1) or quad. And even when she can't, she's not that far behind them if she has Failnaught or if someone else already has the Lance of Ruin. Standard doubles definitely have a use, mostly on axe-users, who achieve higher atk than Swift Strikes/HV/PBV (and also are available for units with higher str... e.g. Dimitri, Edelgard, Byleth, Felix; the other high-str types have issues doubling). So they're speed-reliant, but you can really see the difference sometimes. Doubling with Freikugel gives you numbers you won't get out of a brave combat art, and even Silver Axe+ can post huge numbers. Edited November 5, 2020 by Dark Holy Elf Link to comment Share on other sites More sharing options...
RaIsMyPet Posted November 5, 2020 Author Share Posted November 5, 2020 What you said is fair. I started looking at this list with the motivation to find units that will never fall off, and obviously you want high multiplier to do this. But in the end, no unit can really do this due to the “base growth rates” between player and enemy are too disproportionally large, and sometimes multiplier just makes units to fall off faster. Therefore the effect of multiplier is to make the constant part to become larger thus delaying unit falling off. So piling up multiplier is definitely useful provided the unit can do decent base damage. In the end I realized it’s still upto some subtle interplay between constant part and growth part. I know many units can do many things like Petra/Leonie definitely can quad with brave weapons in certain cases, but my goal is not that, rather it’s to find asymptotically what to expect from a unit in the long run in a more RNG/situation-independent manner. Link to comment Share on other sites More sharing options...
Quang Cox Posted August 18, 2021 Share Posted August 18, 2021 I like this list too, very helpful tips. Link to comment Share on other sites More sharing options...
Recommended Posts