Silly

Sharing this here. This is how base stats are calculated (found by testing multiple units): If the unit is promoted, use (Level+18) instead of (Level-1). You can find Personal Bases, Class Bases, and Personal Growth rates in plenty of places, including VincentASM's helpful spreadsheet. Notably, units do not benefit from class growths in the above formula, only their own personal growths. As a result, units with a high join level will have slightly lower than expected stats, unless their personal bases make up for this. (For example, Mauvier has really high personal bases, which means his base stats are good despite joining at such a high level). Also this game rounds unit stats instead of flooring them like 3H does.

Raging Storm fills a similar niche to brave weapons. It's a tool you bust out on player phase when you need to kill an otherwise too-strong enemy (or in the case of Raging Storm, possibly multiple enemies). Accuracy is lower than something like Iron weapons, but that's okay because there are both plenty of ways to patch accuracy and also because these weapons can deal with important enemies that Iron weapons cannot actually kill. These tools obviously have an important niche. Nobody is really arguing that brave weapons aren't good in this game, and Raging Storm can function basically identically (use it for the first attack and then kill on the second attack with a different weapon), but also has additional uses, such as killing two different enemies in the same turn or using it to smack the final boss five times in one turn.

I kind of actually think there is, to be honest. Mages are just nowhere near as good as the best physical units in terms of killing enemies in a quick and efficient manner. They're slower, have less movement, and are more vulnerable to attacks (and therefore cannot be positioned as flexibly) compared to a good frontliner. Your mage generally makes up for their worse combat with their extra utility, but there is only "so much" utility that you need. For example, one healer is really important, and maybe a second healer is nice to have, but is training a third or a fourth healer that much better than just having two? You can definitely field 4 mages if you want to and still beat the game completely fine, but I think it's far from "needed". In my opinion the upper limit for mages before they start hitting significant diminishing returns is like 2 or 3 (barring very specific strategies).

This is mostly a balance decision. Defense is a stat that "snowballs" on itself, in the sense that the more defense a unit has, the more valuable additional points of defense become. For a brief example, say we have a unit with 20 HP, and are being attacked by an enemy with 15 attack. This is what our bulk looks like at varying levels of defense: At low levels of defense, gaining additional points doesn't really change all that much. Sometimes it lets you survive one extra hit, but often times it doesn't. However, as player defenses increase relative to enemy attack power, you see that each additional point of defense translates to more and more actual bulk. In our example, a 12 defense unit gaining a single point of defense is the equivalent of it surviving THREE additional attacks, which is much more significant than 1 defense was on the lower end of the spectrum. For this reason, I can see it making sense for developers to make sure defense growths are not super high, and for them to be especially wary of giving units that already have very high base defenses high growths as well. Such a situation could lead to defense snowballing out of control. The other two most important stats, strength and speed, do not really suffer from this problem, and are therefore safer to boost growths in. Strength actually scales in reverse, it doesn't take very much extra strength to turn a 5HKO into a 4HKO, but turning a 2HKO into a OHKO takes a lot of extra strength. Speed is a very binary stat that doesn't really scale at all. Either you have enough speed to double/not get doubled, or you don't, and any additional points of speed outside of those two specific breakpoints are not very significant.

Note that Raging Storm has 70 base hit (there is basically zero reason to use Aymr outside of the combat art, because wasting even a single attack means you have one less use of Raging Storm before you need to repair). If we assume weapon prowess level 3 immediately post-timeskip, then you have 83 hit before stats are factored in. This is roughly the same as other powerful weapons that you would be using around this point. For example, look at brave weapons. With weapon prowess factored in... Thunderbrand has 77 hit. Unforged Brave Lance/Bow have 80 hit. Unforged Brave Sword has 82 hit. (I'm assume you would rather get decent use out of your brave weapons before forging them, because their repair material is so rare. Also, no combat arts because using a combat art with a brave weapon is big dumb.) Counting other weapons, with weapon prowess... Silver Bow has 85 hit. Failnaught and Areadbhar have 85 hit. Lance of Ruin is at 75 hit. A lot of other good Swords/Bows/Lances sit at roughly 90 hit. Note that you can boost these numbers with combat arts (for example Swift Strikes Lance of Ruin is 85 hit), but that could come at the cost of being able to double (and therefore ORKO) from most combat arts, as most combat arts don't boost might enough to compensate for not doubling. Also keep in mind that the Bows will suffer -30 hit per tile beyond the second on Maddening. The point is that accuracy on Raging Storm is not significantly lower than accuracy on other important weapons. More powerful weapons are generally less accurate, so if you don't need that Brave Bow to kill then you should use that Iron Bow instead because it's more reliable (70 vs 95 hit). But if you need that extra damage, feel free to pull out that Brave Lance, or Swift Strikes Lance of Ruin, or Raging Storm Aymr. You lose a bit of hit, but it's not so much that you're likely to miss. You are still probably very likely to hit (especially on 2RN, and even more so once Divine Pulse is factored in).

There is an integer overflow glitch present in Azure Moon Ch 22 on Maddening Mode. From what I can tell, in Three Houses, healing from forts works by setting a unit's HP equal to the following formula: This is pretty straightfoward, and produces the intended result in almost every use case. For example, if a unit has 18/20 HP and is standing on a heal tile that heals 30% of their HP per turn, then their HP after healing would be AfterHealingHP = MIN(18 + 20 * 0.3, 20) = MIN(24, 20) = 20. The problem with this implementation, though, is that HP in Three Houses is stored as an unsigned 8 bit integer, which means its maximum value is only 255. Any number greater than 255 overflows and wraps around to 0. On Maddening, Azure Moon Ch 22 Edelgard has 199 HP on her last HP bar, and is standing on a heal tile which heals her for 30% of her maximum HP per turn. If you chip her for a very small amount of damage, then when the game goes to heal her, you get the following result. AfterHealingHP = MIN(198 + 199 * 0.3, 199) = MIN(198 + 60, 199) = MIN(258, 199) = MIN(2, 199) = 2 You can test this yourself by putting Impregnable Wall on a unit and having them chip Edelgard's last HP bar for 1 damage. When the game goes to heal her, her HP bar should drop to a very small number.

Raging Storm is pretty important on CF route if you actually just do the objective for each map instead of "kill every last one of them". If you just rout all the maps then it's not that useful, but if you're mostly focused on boss kills and utilize it to grab multiple boss kills in a short time, then it's actually an invaluable tool. All the CF post-timeskip maps are Kill Boss maps, and Edelgard having Raging Storm means that she is the best choice for your primary combat unit (aka the one that is eating the majority of your stat boosters), since a strong Edelgard can kill multiple bosses in one turn, whereas another unit might only be able to kill one. For example, this is an example of Ch17 CF LTC, which makes very important use of Raging Storm to kill multiple bosses in one turn. The footage is of HM, because there isn't any good recordings of MM LTC, but the strategy for MM is fairly similar.

I don't understand the draw of the bad units on a route being slightly better or worse than other bad units. Ideally all of them are being replaced or relegated to a support role anyways. It doesn't matter whether a unit is the worst or the second worst in the game when regardless of what happens I am probably just going to use them for a few chapters and then stick them on the bench. ---------------------------------------------------------------------------------- BE route has Edelgard, Petra, and Ferdinand as attractive long-term units. Everyone else is either more of a support unit (or mage) or benchworthy. GD route has Claude, Hilda, and Leonie as good units. Both routes also have Sylvain for free, which puts you at 5 good combat units (including Byleth) already. Add in some support units (healers and someone to carry support gambits such as stride) and you don't even need to add that many cross recruits to round out your team.

CF final chapter has a lot of really poor terrain, and fair number problematic enemy Falcon Knights. The terrain can be mitigated by having a good number of flying units, but you will have to have some way of dealing with the Falcon Knights, unless you did a lot of setup and can one-turn the map. The boss itself is mostly a big HP sponge that you can chew through rather quickly with Raging Storm and Brave weapons. SS final chapter has a really annoying gimmick, in that the majority of enemies have Miracle. However, it is also the easiest map to "accidentally LTC", because the boss starts within 10 tiles of your entire army. It doesn't take much brainpower to realize "hey, I can just use stride and then the majority of my army can dogpile the boss on turn 1". The boss is a slightly stronger version of the CF route's boss. Mostly just a big HP sponge. BL final chapter has a lot of problematic enemies. There are a lot of mages with long range magic, as well as War Masters that have ridiculous stats. Another really annoying gimmick is that as soon as you enter the boss room, four mages with siege tomes will spawn. These are ambush spawns on Maddening, which is a very big problem. If you deal with all of these and get to the boss, then there shouldn't be much of an issue. Like the other bosses, the BL boss is mostly a big HP sponge that gets wrecked by Brave weapons. The GD final chapter doesn't really have any super horrible gimmicks besides the abundance of poor terrain. If you have a lot of fliers you should just be able to clear the map while doing typical flier shenanigans (kill an enemy on PP, canto to a location where you can bait 1-2 enemies on EP, repeat next turn). The boss itself is definitely the strongest statistically, and is therefore the most threatening to engage. However, the boss kind of just stands there waiting for you, so you can take your sweet time routing the map and then setting up a kill on him. Overall, the GD final chapter is probably the easiest because it is the most like a typical map, where you just go around killing stuff until you eventually get to and kill the boss. If you have a lot of fliers then the map gets easier. The SS final chapter can be annoying, but should also be very doable, especially if you can realistically kill the boss in 1 or 2 turns. CF and BL maps are the "hardest" maps, though both have ways of making the map a lot easier. CF endgame is significantly easier if you have a lot of fliers. If you have a lot of ground units then dealing with the Falcon Knights is a nightmare, since they have so much movement advantage over you. BL endgame is significantly easier if you have a strong EP unit with Retribution (either a dodgetank or Vantage/Wrath Dimitri), as that makes dealing with siege tomes infinitely easier. Having some way to reliably ORKO War Masters is also a big plus, as they are actually somewhat problematic if none of your units can ORKO.

I'd like to point out that growth rates don't matter for master classes. By the time you reclass into them you likely only have 10 more levels before the game is over. 15 if you take some extra time to rout maps. A 10% difference in growth on your master class is literally only 1 point of difference by the end of the game, and even less than that in the chapters before the endgame. Also, in terms of support, I do think the best magic class is Bishop if you have Physic/Fortify, and Gremory if you're Lysithea (or Manuela, but she has a little bit of trouble making it there due to her Reason weakness). If you want to do combat, then Dark Knight is the best choice by far, but your combat will probably end up worse than a generic Wyvern Lord.

Mages are mediocre combat units anyways. I think most of their value (compared to generic flier #6) comes from being able to use support magic, which Lorenz is not particularly good at. He's not a horrible unit, since everyone in 3H is at least a fairly good unit with investment. But I don't think he's very good relative to most of the cast either.

By difficult I mostly mean "difficult to complete in a reasonable timeframe". Playing slower definitely makes early maps a lot easier. But if you want to complete something like Ch 3 or Ch 5 (which is especially horrible) without taking a ton of turns then the maps are really annoying (Ch3 has really poor hitrates if you're playing fast due to the abundance of enemies in forests, whereas Ch5 is just a massive pain). On the other hand, later on in the game you have a bunch of tactics that can possibly be employed for increasing the speed and reliability of finishing maps. Currently I think that the first 5 chapters on Maddening are the most "difficult", the later post-timeskip chapters (and Ch13 on some routes) are probably next on the list. Earlier post-timeskip chapters next, and then the easiest chapters in Maddening are likely chapters 6-12. Without proper knowledge and preparation though, I can definitely see the second category (later post-timeskip and Ch13) being harder, since enemy stat benchmarks are pretty high and there is a lot of stuff that you need to deal with that you might not be fully equipped to handle easily.

I assume on any sort of tier list that you don't do auxiliary battles, because that's a bit too close to grinding.

I would say that BE has the worst starting roster. GD has Claude, Hilda, and Leonie with fairly good base stats. You have two units with Tempest Lance at base, as well as multiple ways of increasing damage output (Hilda personal, Leonie personal, Lorenz personal after Ch1). You also get access to Rally Strength and Rally Speed after Ch1, since Ignatz and Raphael learn their rallies at D rank. BE route has Edelgard and Petra with good bases. Ferdinand only has okay bases and is your only unit with Tempest Lance, but his stats are worse than both Leonie and Hilda. You also have no rallies or ways to get more damage out of your units outside of Bernadetta's personal.

It depends on what you value, I guess. An easier early game but harder late game is probably overall easier than vice versa if you know what to expect and are good at planning ahead, but it could be more difficult if you haven't planned well for the dramatic spike in endgame difficulty. -------------------------------------------------------------------------------------------- Also, speaking about BL specifically: BL has the easiest early game due Felix's base offenses and Dedue's base defense. However, the last few chapters are probably "harder" than other routes due to the high number of war masters and magic users (many who have siege tomes). It is also the route with the best support battalions. However, it has the lowest number of "good" flying battalions, which sort of hurts.