FE11 H5 0% growths, with commentary (complete)

dondon151 · August 9, 2015

https://www.youtube.com/watch?v=7cWOyvumpsA

Here are in-depth explanations of the topics that Mekkah and I talked about:

FE11-specific mechanics

Reclassing

FE11 shakes up the original FE formula by introducing this totally novel mechanic. Those of you familiar with FE13 should recall Second Seals, which allow a unit to transition horizontally between classes. FE11 reclassing is a more flexible and primitive version of this mechanic; characters can reclass at the beginning of every chapter that has a prep screen without having to use an item, reclassing doesn't reset a units level, and instead of having personalized reclass sets, all characters in FE11 belong to one of two generic reclass sets depending on their starting class.

The inclusion of the reclassing mechanic means that its important now to understand the difference between a unit's personal stats and a class's base stats. When switching classes, the statistical differences between the two classes' base stats (including weapon rank) are applied.

Example: If Jagen were to switch from paladin to dracoknight, he would receive -2 HP, +1 str, +2 def, and -3 res. He would also temporarily lose his sword rank, but gain an E rank in axes. Everything else would remain unchanged. If Jagen then were to switch back to paladin, all of these changes would be reverted.

Through reclassing, Jagen can finally discover his true calling.

In vanilla FE11, total growth rates are also split between personal growth rates and class-specific modifications. Of course, we don't care about that in 0% growths.

All units in FE11 either belong to one of two reclass sets or cannot reclass at all. Classes that cannot reclass (for, I hope, obvious reasons) are lord, thief, manakete, ballistician, and freelancer. All other classes belong to a reclass set designated as set A or set B:

Class set A (unpromoted): pegasus knight (F only), cavalier (M only), archer, myrmidon, mage, curate/cleric
Class set A (promoted): dracoknight, paladin, sniper, swordmaster, sage, bishop

Class set B (unpromoted): knight, mercenary, fighter, hunter, pirate, dark mage
Class set B (promoted): general, hero, warrior, horseman, berserker, sorceror

Classes with a red dot belong to set A; classes with a green dot belong to set B. Also take note of the class limits.

Finally, there is a limit to how many units can be in a given class at once - its not like the game would let the player reclass every single character to dracoknight. For each chapter, the player can have X+1 units of a given class, where X is the maximum number of units who start in that class that the player can have at that point in the game. X includes units that the player didn't recruit and characters from gaiden chapters that the player never went to.

Forging

Forging is not a new mechanic, but FE11's iteration of forging establishes the model for the mechanic in future installments of the series. Whereas FE9's and FE10's forges created a brand new weapon from a limited predetermined selection of weapons for each chapter, FE11's forge modifies the properties of any existing weapon (with some exceptions). For any particular forge, modifications can be made to 4 weapon parameters - MT, hit, crit, and WT - up to 10 stages each; that is, +/- 10 MT or WT, +/- 50 hit, and +/- 30 crit.

In this screenshot, Im forging +6 MT and +10 hit on the Wing Spear.

It becomes quite clear that under this system, forging effective weaponry becomes optimal, because every +1 increment in MT yields a +3 increment in atk. Even on the highest difficulties, a sufficiently strong effective weapon allows its wielder to OHKO a vulnerable target.

Forging any weapon also adds an additional 50% of the weapons base cost on top of the cost calculated from this table. Forging costs are quadratic and increase sharply. Source: Serenes Forest

One major difference between FE11 forging and previous iterations of forging is that FE11s forging cost structure tends to produce very expensive forges. The vast majority of expenditures in this run are on forging. Much like how in FE6 I carefully managed my funds to maximize the amount of Boots purchases, here in FE11 I carefully manage my funds to get the best forges that money can buy.

Weapon rank bonuses and weapon triangle

FE11 is the first entry in the series to introduce stratified weapon rank bonuses. While FE7 and FE8 both feature an S rank bonus of +5 hit and +5 crit, FE11 brings this a step further by providing stratified bonuses at C, B, and A rank (there is no S rank in this game). Weapon rank bonuses depend on the weapon type in question:

Swords gain +1 atk at ranks C, B, and A
Lances, bows, and tomes gain +1 atk at ranks C and A and +5 hit at rank B
Axes gain +5 hit at ranks C, B, and A

Example: Jagen has a B rank in lances and currently has the Silver Lance equipped. B rank lances confers a weapon rank bonus of +1 atk, +5 hit. Jagen has 21 atk (7 str + 13 MT + 1 bonus) and 105 hit (10 skl + 90 hit + 5 bonus).

Jagens old eyes stare into your soul and see all.

The weapon triangle also makes a return, except this time its a bit more convoluted. Weapon triangle modifications are also based on the attackers weapon rank:

Attacker has E or D rank in selected weapon type: +5 hit to self, -5 hit to enemy
Attacker has C or B rank in selected weapon type: +10 hit to self, -10 hit to enemy
Attacker has A rank in selected weapon type: +1 atk, +10 hit to self, -1 atk, -10 hit to enemy

A unit who has weapon triangle disadvantage also forfeits all of his weapon rank bonuses in that round of combat.

The upshot here is that weapon rank bonuses provide a way to level up combat parameters in a context without random growth rates. On H5, weapon triangle advantage tends to be offensively inconsequential, since most player units have low weapon ranks, but because all enemies have A ranks, the defensive application of nullifying enemy weapon rank bonuses makes weapon triangle advantage tactically important in some circumstances.

More esoteric stuff

Even if you're familiar with FE11 mechanics, you may not know these important details.

DS FE random number generator (RNG)

So first of all, the RNG isn't truly random. But you probably know that already. That's a limitation of computing.

My understanding of the DS FE RNG is limited. What I do know, however, is that the RNG is seeded based on the system clock. Theoretically, this means that if one were to start a chapter with the system clock set to a particular time, he could repeat the same sequence of actions with that RNG seed and expect the same outcome every time. RNG abuse for this game isn't as overt as it is for the GBA FE games, but at least in this instance it doesn't have to be.

Displayed hit and true hit

Refer to the appropriate section in my FE6 playthrough introduction. Everything said there also applies to this game.

Hard mode differences

Unlike previous games in the series, FE11 offers 5 degrees of hard mode difficulty on top of the normal mode. These individual hard mode difficulties all have their own name, but no one remembers them, and the community prefers to truncate them to H1 through H5.

In FE11, there are 3 key differences that distinguish each difficulty:

HM enemies get a sizable stat bonus compared to their NM counterparts.
HM enemies have a minimum weapon rank regardless of what weapon they have equipped, whereas NM enemies always have the minimum necessary weapon rank. H5 enemies will always have A in all of their weapon ranks.
HM enemies tend to have better weaponry; H5 enemies progress quickly from iron > steel > silver > brave, and weapons that don't follow that progression may be forged later in the game.

Above is a screenshot of a chapter 2 pirate in NM. Below is a screenshot of the exact same pirate in H5. Take note of huge bonuses to HP, str, skl, and spd, but no bonuses to def or res. Also take note of the H5 pirate's A weapon rank, which confers +15 hit.

Like previous games in the series, HM bonuses appear to be generated by applying extra hidden level ups to all enemy units. The quantity of extra levels applied in H5 depends upon the chapter (information courtesy of Gryz):

Chapters 1-17 and gaidens: 15 extra levels
Chapters 18-22 and 24: 20 extra levels
Chapters 23 and endgame: 25 extra levels

However, HM bonuses are not applied to def or res, and certain enemies and bosses follow different rules. The resulting effect is that enemies are extremely potent on offense, but relatively less bulky on defense than they would otherwise be.

About this playthrough

This playthrough operates under the following conditions:

Player units have 0% growths in all stats.
It recruits all possible units and ends with them alive (there are 52 player units in a playthrough without gaiden chapters).
It aims for the lowest possible turncount.
It aims for a low play time on the game clock.

Videos of each chapter will be uploaded, with audio commentary from Mekkah and me. An explanatory write-up will accompany chapters to express my thoughts that cant be fit into the commentary.

Chapter list:

Turncounts:

01   8 turns   06   6 turns   11   3 turns   16   5 turns   21   1 turn
02   5 turns   07   4 turns   12   2 turns   17   2 turns   22   1 turn
03   6 turns   08   5 turns   13   2 turns   18   7 turns   23   1 turn
04   5 turns   09   4 turns   14   6 turns   19   1 turn    24   1 turn
05   4 turns   10   7 turns   15   2 turns   20   3 turns   F    1 turn

Total turncount: 92 turns

Edited December 28, 2016 by dondon151

Ownagepuffs · August 9, 2015

I am so stoked for this oh god.

Espinosa · August 9, 2015

So does playing this with 0% growths (entailing that class growths are reduced to 0 as well) not also cause the enemy level-ups to be less potent?

Carmine Sword · August 9, 2015

The hype is real. I can't wait for this to start in earnest.

Dark Sage · August 9, 2015

So does playing this with 0% growths (entailing that class growths are reduced to 0 as well) not also cause the enemy level-ups to be less potent?

iirc, enemies use different class growths than players do so they should be the same as normal.

Topazd · August 9, 2015

~~Don't know if I should comment on that 'gaiden' pronunciation~~

Looking forward to it!

Edited August 10, 2015 by Topazd

dondon151 · August 9, 2015

iirc, enemies use different class growths than players do so they should be the same as normal.

this is correct

sim · August 9, 2015

awesome! cannot wait for this series to unfold

AlexMH3 · August 10, 2015

I'm hyped. Though I am worried about whether or not I should watch because I never played the Marth games.

dondon151 · August 10, 2015

I'm hyped. Though I am worried about whether or not I should watch because I never played the Marth games.

at least you can be relatively certain that i won't spoil any of the plot lol

Carmine Sword · August 10, 2015

~~People still play Fire Emblem for the plot?~~

MartyTheDemonSlayer · August 10, 2015

Hype!

kirbymastah · August 10, 2015

at least you can be relatively certain that i won't spoil any of the plot lol

"plot" "FE11"

( ͡° ͜ʖ ͡°)

looking forward to this though, especially as someone who's really unfamiliar with this game. ~~I need to stop being lazy and go play FE12~~

Also wtf i didn't know that's how weapon ranks worked in this game lol, interesting

Edited August 10, 2015 by kirbymastah

dondon151 · August 12, 2015

https://www.youtube.com/watch?v=ngO0ianRpP8

Chapter 1 is completed in 8 turns.

Chapter information

New characters

Marth is the main character. I can somewhat confidently assert that, especially in this context, this iteration of Marth is the worst lord in the series. Whereas even bad lords such as Roy fight from time to time, Marth ends up fighting a total of twice throughout the entire run, and he spends much of it getting OHKO’d, doubled, or both.

Because FE11 doesn’t have the rescue mechanic, Marth ends up spending most of his time moving from point A to point B without being able to take a detour, so even in a typical run, his use is limited. He also doesn’t promote and his lategame prf weapon can be skipped with little consequence.

Personal bases: 0 HP | 0 str | -1 mag | 1 skl | 5 spd | 9 luk | 2 def | 0 res | lances 30

It took awhile after FE11’s release for the community to realize it, but Caeda is the best combat unit in the game. You wouldn’t be able to discern this from her base stats; her utility lies almost entirely in her prf weapon, the Wing Spear. The Wing Spear is like an 8 MT Rapier, and forging a good chunk of +MT on it lets Caeda OHKO most important enemies and maim a large subset of bosses. Among effective weaponry that aren’t bows, it also has the highest base hit, the lowest cost per use, and the highest base durability.

Caeda’s 5 personal base spd is on the high end relative to the rest of the cast, and this helps her to double general bosses and avoid being doubled late in the game. Of course, she also flies and promotes to the class tied for both the highest base str and the highest mov.

In contrast to the biblical story, Cain doesn’t murder Abel (and they are probably not brothers), but the red cavalier would be envious of his green counterpart. Cain and Abel inaugurate the cavalier duo archetype that would become a series fixture, but in this game, their base stats and growths are very similar. The major difference between them is in their base weapon ranks.

Abel’s D lances allows him to throw Javelins at base and puts him within reach of C lances after several chapters. Cain’s D swords allows him to use only Steel Swords at base. Additionally, enemies reach 10 AS fairly early in this game, so Cain soon becomes unsafe attacking enemies unless he kills them or attacks a bow user.

In a game where knights have half the movement of paladins, Draug isn’t a very happy camper. On H5, enemies have high enough atk that Draug is much more proficient at dying than he is at stubbornly choking a point. Finally, Draug doesn’t get access to any other classes with a base lance rank.

Personal bases: 0 HP | 0 str | 0 mag | 5 skl | 1 spd | 1 luk | 1 def | 0 res | swords 30, lances 105

Jagen is the OG badass veteran paladin. Relative to his successors, Jagen’s base stats actually look quite bad; they’re barely higher than those of Cain and Abel! However, small statistical differences matter a lot in the earlygame, and more importantly, Jagen is the only user of the Silver Lance for the first 4 chapters of the game. His per-hit damage output is nearly double that of most other characters.

Jagen’s class set includes dracoknight, which preserves his lance rank. The lance weapon type contains the Ridersbane, which is the most useful effective weapon that’s not the Wing Spear. Essentially, these two qualities ensure Jagen’s longevity.

Gordin’s base stats are sooo bad, mostly because the archer class itself has atrocious base stats. However, Gordin is a godsend in chapter 1. He’s the only ranged attacker who doesn’t attack at WTD with a weapon that already has only 70 hit, and avoiding a counter from the strong enemies is paramount.

Wrys is the only staff user available until chapter 3. Due to the complete dearth of Vulneraries in this game, Wrys’s healing ends up being extremely important, even though he only heals 9 HP per Heal use.

Mechanics notes

The enemy AI, in review

While in FE6 we reserved discussion of the enemy AI to the latter half of the game, here in FE11 it becomes necessary to explain some AI behavior starting as early as chapter 1. In such a difficult game where the enemies hit like trucks (and rarely miss), we have to dig deep in our bag of tricks. Good AI manipulation can render a seemingly impossible scenario easily beatable.

Classically, Fire Emblem AI weighs several major variables in determining what unit to attack:

How much damage it can do on average (e.g., dmg * hit)
How much HP the target will be left with after being hit
Whether the target dies after being hit
Whether the target can counter attack
How much damage the target can counter on average (e.g., dmg * hit)
Whether the attacker has a displayed critical rate
Whether the attacker has weapon triangle advantage
What is the target’s deployment slot (tiebreaker)

This is not an exhaustive list, but it takes into consideration the variables that one would expect an enemy AI to use.

I do not know the mathematics behind FE11’s targeting priority formula (though Gryz might), so all that follows is borne from empirical observation. It seems that of the above variables, FE11’s enemy AI weighs heavily:

How much damage it can do on average (e.g., dmg * hit)
Whether the target dies after being hit
Whether the attacker has a displayed critical rate
Whether the attacker has weapon triangle advantage

It’s difficult to tell if all of these are independent variables or if, for example, that weapon triangle only has an effect because it significantly affects average damage. It may also be that there are other variables considered but weighed less heavily.

Regardless, the takeaway here is that enemies don’t seem to care about how much damage they take, so they will gladly attack armed targets instead of unarmed targets or high-atk targets instead of low-atk targets if the circumstances are right. They also seem to prioritize maximizing damage over leaving a target lower on HP. Finally, enemies are very much inclined to attack targets who face non-zero crit and/or weapon triangle disadvantage.

Examples:

An enemy pirate attacks Jagen (on fort) instead of Draug (on plain).

In this example:

Enemy pirate does 6 HP dmg @ 87 hit to Jagen and 5 HP dmg @ 100 hit to Draug.
Jagen counters for 15 HP dmg @ 84 hit while Draug counters for 8 HP damage @ 77 hit.
Jagen starts with 14 HP while Draug starts with 20 HP.
Both Jagen and Draug face 1 crit.
Both Jagen and Draug have weapon triangle disadvantage.

Given that the enemy pirate does, on average, about the same amount of damage to Jagen and Draug, clearly it doesn’t care about the damage on counter attack. There may be a deployment slot tiebreaker in play here, but it’s hard to tell.

An enemy pirate attacks Abel (on plain) instead of Marth (on village) or Gordin (on plain).

In this example:

Enemy pirate does 9 HP dmg @ 100 hit to Abel, 8 HP dmg @ 89 hit to Marth, and 9 HP dmg @ 93 hit to Gordin.
Abel counters for 7 HP dmg while Marth and Gordin are unarmed.
Abel starts with 20 HP, Marth starts with 10 HP, and Gordin starts with 18 HP.
None of these characters face crit.
Abel has weapon triangle disadvantage.

The enemy pirate has the opportunity to leave Marth at 2 HP, but chooses to attack Abel instead, who is also the only player unit in range that is capable of a counter. We can reasonably conclude that the enemy AI doesn’t care about leaving a target at low HP or avoiding a counter.

(watch general horace tell me that he did this chapter in 7 turns)

Edited August 12, 2015 by dondon151

Espinosa · August 12, 2015

That's a more creative way of doing chapter 1 (hardest chapter in the game?) than I could've imagined. Caeda (I've always pronounced it as "Ki-dah") is already the second unit after Jagen in spite of lacking opportunities to use the Wing Spear effectively.

Edited August 12, 2015 by Espinosa

dondon151 · August 12, 2015

i think chapter 2 is way harder

caeda essentially contributed 3 kills (on top of the 1 hp kill that she got) in chapter 1 because that's how many pirates i ended up not having to face due to her shenanigans

Edited August 12, 2015 by dondon151

Irysa · August 13, 2015

Damn and I thought we'd have to wait till Chapter 4 before Caeda started really pulling weight!

Due to the icky hitrates of this map, how many rerecords were neccessary, since you're not utilising RNG fixing via the clock? Also, would I be correct in assuming the missed Javelin from Draug near the end was to avoid a miss for Jeigan against the boss?

dondon151 · August 13, 2015

Due to the icky hitrates of this map, how many rerecords were neccessary, since you're not utilising RNG fixing via the clock? Also, would I be correct in assuming the missed Javelin from Draug near the end was to avoid a miss for Jeigan against the boss?

yeah.

i am RNG fixing the clock; it's just that i have to keep resetting to find the right time. when you record a DS movie file, you have to specify a date and time from which to start, so it's kind of impossible to not RNG fix the clock.

i don't remember how many times i tried. it was definitely less than 24 because i would only change the hour and i never had to go to another day. you know, overall it's not that bad; the boss strategy has like a 26.5% CoS but the rest of the chapter is fairly reliable except for one mandatory javelin throw from abel and a javelin attack from draug that could be an iron lance stab instead if that pirate proc'd 25 HP instead of 26 HP. i know that i scrapped some takes because jagen would get a crit and it would save a silver lance use and make the strategy appear unreliable because the crit happened.

Edited August 13, 2015 by dondon151

Upside-DownFish · August 13, 2015

I always pronounced caeda's name cay-duh, but to each their own I guess.

Can't wait to see more of this playthrough, especially all the community picked forge names!

JSND Alter Dragon Boner · August 13, 2015

So when are we getting Dracoknight Pope Windell mk 151?

The comments about the AI shits are interesting. I remember using S!Ryan to bait the hits, regardless of his defense and it always works. I guess DSFE in general have the most predictable AI?

dondon151 · August 13, 2015

DSFE targeting priority probably has fewer terms in the equation, which makes it more predictable. i think that FE12 doesn't care about weapon triangle whereas FE11 does.

overall there are still instances of unexpected behavior, but such behavior is easier to characterize here than in GBAFE.

Emperor Hardin · August 13, 2015

i think chapter 2 is way harder

caeda essentially contributed 3 kills (on top of the 1 hp kill that she got) in chapter 1 because that's how many pirates i ended up not having to face due to her shenanigans

Probably because the boss, Gomer, has a Handaxe.

Also wish they'd given Darros better growths.

Paused · August 13, 2015

Also wish they'd given Darros better growths.

All zeros across the board is harsh, but at least he's average compared to everyone else here.

SSJDennis · August 15, 2015

Good luck DD :) Been looking forward to a next 0% run of yours.

dondon151 · August 16, 2015

https://www.youtube.com/watch?v=rt5nWHfLLnM

Chapter 2 is completed in 5 turns.

Chapter information

New characters

Ogma appears solid on paper, but he gets weighed down by Steel Swords, and 12 spd actually isn’t high enough to reliably double any enemies beyond pirates in chapter 2 and knights. Ogma also can’t reclass to a class with higher movement, and his per-hit damage doesn’t exceed that of Cain. At least he doesn’t run the risk of getting doubled.

Personal bases: 4 HP | 5 str | 0 mag | 5 skl | 2 spd | 5 luk | 3 def | 0 res | axes 30

Among non-gaiden characters, Barst’s personal base str is the highest in the game. Against enemy axe users, Barst has more atk with a Steel Axe than Jagen has with a Silver Lance. Unfortunately, all unpromoted axe users are ground-bound and limited to 6 mov, and Barst’s personal base spd isn’t stellar, so there’s not much benefit to grind him for long-term use.

These dudes are a pair of fighters who are strictly inferior to Barst in almost every way. At least Bord has a higher base axe rank, which allows him to use the Hammer for a period of time before Barst can reach C rank axes. Cord, on the other hand, is stuck using Iron Axes, which means that typically he gets to attack once in a chapter before being too low on HP to do anything else.

The fourth axe user that IS throws at the player in chapter 2 is the weakest and the slowest. Why.

Hunter is a much better class than archer, so Castor is basically a better Gordin. Chip damage with bows is still much more reliable than chip damage via Javelin or Hand Axe.

Mechanics notes

Explaining hard mode bonuses

In the introduction post, I explained that enemies received a certain number of bonus level ups depending on the chapter being played:

Chapters 1-17 and gaidens: 15 extra levels

Chapters 18-22 and 24: 20 extra levels

Chapters 23 and endgame: 25 extra levels

HM bonuses are applied to all stats except for def and res (and luk, for which enemies don’t have a growth rate). The magnitude of the HM bonus for any given stat is simply L * G, where L is the number of HM bonus levels and G is the growth rate of that stat as a decimal.

Example:

Pirate has class growths of 90% HP, 50% str, and 15% spd. All chapter 2 enemies receive 15 HM bonus levels on H5. Pirate’s expected HM bonuses for these stats are:

HP: 15 * 0.9 = 13.5

str: 15 * 0.5 = 7.5

spd: 15 * 0.15 = 2.25

The expected bonuses in these cases are all non-integers. In FE11, this means that the actual HM bonus for each of these stats can either round up or down. The probability P of a stat rounding up is most likely equal to the decimal part of the expected bonus, and the probability of the converse would obviously be 1 - P (so a chapter 2 pirate would have a 75% chance of +2 spd and a 25% chance of +3 spd).

Chapter 2 pirates are all L1, so their NM stats are equivalent to their class base stats. Pirate has base stats of 18 HP, 5 str, and 6 spd. Their expected stats on H5 would then be 31.5 HP, 12.5 str, and 8.25 spd. Is this consistent with in-game observations? See for yourself:

This pirate rounded down his HP but rounded up his str and spd.

Some enemy AI types

Typical Fire Emblem enemies most commonly have one of three types of AI:

Aggressive, where an enemy always moves towards the closest target
Patient, where an enemy doesn’t move unless a target is in attack range
Stationary, where an enemy will never move under any circumstance

Less commonly, some enemies have an AI type where they move towards the closest target only if that target is within a certain distance away from them (and this distance can be outside of that enemy’s attack range). We saw this in chapter 1, actually, with the hunter and one of the pirates that started near the gate.

One more note on thieves: thieves in this game will generally not attack unless player units completely block them from their target. If you give them room to run, they will usually not try to hurt you.

FE11 H5 0% growths, with commentary (complete)

Recommended Posts

Link to comment

Share on other sites

Top Posters In This Topic

Popular Days

Top Posters In This Topic

Popular Days

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Link to comment

Share on other sites

Join the conversation

Recently Browsing 0 members