Cysx

Disclaimer: First off, some basic things about how exp works. - Unlike with most things in this game, exp gains are rounded up. This is likely to avoid making it possible to gain less than 1 exp. - As you probably know, the exp to level up isn't static in Three Houses, unlike in previous entries. The way the requirements evolve however, largely is: - From Lv 1 to 21, the requirements are multiplied by 1.1 every level, and the result is rounded down. - ie, level 12 would be 100 * (1.1 to the power of 11) = 285,3116... = 285 - Probably in an effort to avoid the numbers getting out of control, that 1.1 multiplier is lowered from Lv 22 onward. - Lv 21->22 itself is a bit weird, with a multiplier of ~1.06545(or roughly +44.05). At the moment I'm not entirely clear on what the game does there. - From 23 onward it's static again, now with a 1.05 multiplier instead. - Thus, the value for, say, level 29 is [(100 *1.1 to the power of 20) + 44.05] * (1.05 to the power of 7) = 1008.6095... = 1008. - Anyway, here's a table from level 1 to 50: - Exp bonuses are made up of two totals: All the statues' bonuses added together, and all the other bonuses added together. - On that note, when you unlock the +10% bonus of a statue, it replaces the 5% one. As a result, the maximum bonus you can get from statues is +10% *4 = +40% - The other available exp bonuses are as follows: - Professor's Guidance/ The lords' P.skills(+20% each) - The Experience Gem(+50%) - As a result of the above, the highest multiplier possible is [1 + (0.1*4)] * [1 + (0.2 + 0.2 + 0.5)] = 2.66. Only the three lords can reach this, since Byleth cannot get his/her own bonus twice. ---------------------------------------------------------------------------------------- Okay, now for what the thread is meant to be about. The formula itself is pretty simple and is the same on Normal, Hard and (from what I can tell)Maddening: (10 + Level -1) * Skill bonus * Bonus exp multiplier ... Yup, that's it. An important thing to understand as a result is that your target doesn't matter at all; it can be lv 1 or 100, that won't affect the result. This is also true for multi-targets spells(such as Fortify), the number of targets is irrelevant. Now as for the "Skill bonus", aka the actual meat of it, it is tied to each support action. Here's the list: Heal: *1 Recover: *1.5 Physic: *1.5 Fortify: *3 Restore: *1.5 Ward: *1.5 Rescue: *2 Warp: *3 Silence: *2 Dance: *2 Resonant White Magic: See Disclaimer Recovery Roar, Impregnable Wall, Sacred Shield, Retribution, Stride, Blessing, Dance of the Goddess: *0/Don't give exp. So for example, a level 15 character healing with Physic and getting a 1.2 bonus from statues would get: (10 + 15 - 1) * 1.5 * 1.2 = 43.2 = 44 points of exp. A level 25 character using Warp, getting a 1.2 bonus from statues and with the experience gem equipped would get: (10 + 25 - 1) * 3 * 1.2 * 1.5 = 183.6 = 184 points of exp. ... And that's actually all there is to it. As usual, my bad if this is old information. Have a nice day! Edits: - Was slightly off on the Lv 21 -> 22 multiplier, leading to an inconsistency on level 31. - Additionally after further testing, turns out exp bonuses don't interact exactly like I wrote down. It should now be accurate.

Ah, okay, the 1/4 odd makes more sense, then. And unfortunately it's really difficult to gauge those two, I'd need to run a set of 5000 attacks to get potent results, and I'll fully admit I don't have the courage for thatXD. Regardless, those are odds low enough that 100% accuracy probably wouldn't help much, it's obvious enough that they're very rare. That's unfortunate! It's not like I've learned nothing but... the way the numbers grow, I just can't make complete sense of them. Doesn't help that the game does not provide a reliable baseline to work off of, since the prologue exclusively gives you units with a 20% exp boost through their personal... Then you have different tiers of enemies giving different bonuses and I can't even tell if beginner tier or intermediate enemies are supposed to be the base unmodified value, considering the former are rarer... Normal to Hard does modify exp but not consistently... it's all a bit of a mess. And now we have Maddening. Yeah, you're right. I just basically came in to post my findings, so that was me letting out a bit of frustration. Though I really need to look into datamining to some extent next time around. Not saying it's easy at all, but I can't help but feel... limited, without it.

. Welp, I just wasted my time then, for the most part. It's really nice to have all those odds, though, I only got the combat ones on my own. Though there is one thing that doesn't match about the Crest of flames; in 500 attacks, the preventing counters and +5 damage effects activated 12 and 16 times respectively, and never separately from the healing, or together. The healing activated 103 times, which matches... but as a result, the other two effects should be much, much rarer than 1/4. Also an additional tidbit, the crest of Indech cannot activate off of brave effects at all, from what I've tested. Outside of that, I have a lot of raw exp data that I'm having trouble extracting a formula from, anything you've seen within the game's data that could help, by any chance?

That's weird. Got it on every route, with only Ashe, only Catherine, and both.

Can confirm. ... er, that it isn't an exception. Simply having Ashe unlocks it, but you miss out on the battalion tied to Catherine... assuming there is one worth anything to begin with.

The point being, you should still have enough uses unless you spam them; her chapters aren't even demonic beasts heavy. There's also the fact that Kotaro is the chapter boss while Petra can be ignored(more or less easily depending on the route/chapter), but most importantly... just use gambits man. Low uses doesn't mean they magically don't exist, and no character uses them as their main source of damage, not even Hubert(who cannot face Petra anyway).

That's twice per character though(well, not always admittedly), which you have a dozen of. The difference with Kotaro is that you had to deal with his high dodge, but here, you don't.

Phew, okay, that one will be shorter. Ashe is wyvern lord. He's good at axes and needs the damage, and chest opening utility without having to hoard chest keys of check every map is nice(as the game very much does not provide free ones often, it stops doing that quite early in fact). Ingrid...? I read Ignatz. Well that's easier then! Wyvern lord! You'd think Falcon knight because of her relic but it's impractical to only use that, and she otherwise really likes the better damage from axes and a +3 base. Lorenz I'm not sure, I guess wyvern lord by default just because he can? He's very comparable to Sylvain as a DK, with better damage(don't forget about that personal), but some speed issues, and no ranged healing. And since he doesn't get to be a super paladin... In general his speed is in this weird place where it can't really be fixed since it always barely reaches the best bases for his tier(WL notwithstanding, that's too much work), so you kinda just have to deal with it. That's partially why he's probably better on a wyvern overall, even though that's really boring for an hybrid unit. I guess he could find some success in a levin sword build on faster classes? Idk.

This is not necessarily true(or rather, fair), both have an easy time into wyvern(one strength, no weakness) and Caspar's personal is legitimately nice to have in many situations, since axes like the extra accuracy. Raphael gets his speed fixed by WL's base+mod and has easy access to weight-3; he otherwise brings potent durability and great damage(gets a +mt support from Ignatz, too, although it's not that relevant here since they're from the same route). Definitely not essentials and won't always shine on the path to wyvern, but once there they're not underwhelming, they'll clear you the game just fine. I think I've ran my mouth about Annette way more than I should have already, and I'd never pretend she's an essential recruit anyway. See I agree with the rates, but got some pretty bad surprises when I tried to rely on his tanking; in the end he otherwise has pretty mediocre natural durability(magical especially), meaning in many cases it amounts to tossing a coin. Which is not that big a deal considering DP and all, but I don't know, I think it's nice to be reliable in that department, and he's kinda not. And +6 weight on top of weapon is pretty bad still, especially if it's axes; one way or another, it cuts into his offense. I just don't use the thing anymore, it's cute but I don't think it really makes him a better unit outside of speedster classes. He's still amazing however, especially early on. Gets stuff dead better than most anyone else. I will say that he has a bit of a tougher time getting into wyvern lord than most physical units due to being neutral in everything, and a weakness in authority is always a bit of a problem. But overall he probably will not disappoint, his stats are just too good. You know, I kind of have to say guys, in the end considering the tools we have I don't see the big deal. She's a foot unit stuck at one range with manageable charisma, two gambits(that can be made safe through canto, or range) and she's generally out.

I think even neutral is almost impossible. The +2 bonus every time you tutor and to a lesser extent +4 from weekly studies really add up, in the end it's probably around a 800 points difference by the end of the game? Something like that? Honestly, my experience is: do not go for it on a fresh playthrough. It's not worth the trouble in the slightest and you'll get it way too late for it to matter; S+ is for new game +.

Is Leonie that good an example, when her canon master class is much more likely to be bow knight than great knight, and it'd be weird for her to stop at paladin, making her actual main weapon bow, and thus suggesting that archer is a proper route for her even though her timeskip outfit is only on cavalier and paladin... bleh. Anyway, it's pretty solidly Edelgard, I think. Nothing comes close to Raging Storm providing up to 5 more actions in a turn... not even in the whole series maybe. Contrary to the description, it doesn't actually have to hit either I think, it works regardless. Even Dimitri's wrath vantage build with infinite counter range can be beaten with gambits and the firesweep sword art.

His advantage is that he can essentially gets both archer and brigand masteries at once. Make him a brigand, learn death blow, then go sniper. If he has good enough speed, go bow knight, otherwise, stay a sniper and abuse hunter's volley at max range with max accuracy and death blow. He can also abuse less accurate weapons easier than others, so you could build around that, I don't have any specific idea as to how, though. And indeed, mage works too as the biggest weakness of his main spells are their low accuracy(to compensate for high crit most likely), which doesn't matter for him.

Absolutely, you're getting a more accurate average on the very level itself, but then you have a faulty number for everything that comes after; besides, it is your personal choice to interpret a, let's say 14.7 as 14 or 15, to me it translates as 14 and thus it remains accurate; but what I'm really trying to say here is that in a context where you're selectively considering averages to that extent, you'd simply have to note that on level 20 specifically, despite an average of 14.65, the only statistically relevant value is 14. I don't think it's worth getting everything else wrong to avoid having to do that. I guess a compromise could be to put the decimals between parenthesis, to notify of the issue. Promotions in FE7 are indeed tied to promotion gains rather than class bases, so there is a difference in when the stat boosts occur; but they're still flat stat boosts. I have to say I fail to see how the difference you're describing is relevant outside of the problem you've brought up above; again, averages aren't a game mechanic. That being said, it basically largely amounts to Strengthofmorning's decision here, which is one person like you and me. So it wasn't exactly relevant for me to bring up either way.

I was mostly referring to a combination of their HP, Def, and Res. Though I wasn't telling you to throw them in the middle of 10 enemies either.

The issue I have with that approach is, sure you have a solidified value for level 20, but when it comes to level 21 and 22, you're behind averages, at 14 spd instead of 15. What if level 21 was the level of reference here, that we cared about, instead of 43? Wouldn't it matter to get it accurately? And as a result, doesn't it matter to get it accurately, period? From 19 to 22, on average she's supposed to gain 1.05 from growths. With that method, she's gotten 0.7 for no good reason. That's why I used the stat booster example, when taken to its extreme it shows that the very reasoning is flawed. The averages on Serenes work that way as well(which I misremembered); for example, when promoting, Lyn's average Skill jumps from 18.4 to 20.4, and not a flat 20. It does cut decimals on caps though, which I don't really agree with but is even harder to fight against, and irrelevant in Three Houses. - Basically I think I need to debate this with more people, but I don't see any kind of food beyond +spd to really be relevant. This is because the other major stat, strength, gets the ingredient for its main recipe from doing auxiliary battles, which consumes a weekend and essentially equates to grinding. The spd one, however, you get from fishing, which only consumes negligible amounts of money and doesn't have to give you any big advantage as long as you don't sell what you get. Post timeskip, you have more activity points than you need to cook once a weekend, too. So basically, a sketchy speed problem gets fixed to an extent through this. Not in ch22 though, since cooking effects fade after ch21 and you don't get to refresh them. - Honestly, I just went through a playthrough where I tried to get 5 units neutral in riding to +1 move at once, and... it kind of opened my eyes on how much of an investment it really is. Everyone barely reached the minimum rate to get into Dark Knight/Bow knight by 30, too. I'm feeling pretty confident spending between two to four months tutoring something completely unrelated will make a sizeable difference, however, I look forward to your own conclusions if you ever share them.

Yes, it is. Don't remember on which chapter/month, however:x

I can mostly give general advice since I've never played on normal: If you're being overrun by enemies, use gambits to immobilize several at once. Claude's battalion in particular, the Immortal corps, is amazing for that as its gambit covers a huge area. You also don't have to deal with the north enemies on turn 1, there's enough place to the south to retreat, and in general, it might help focusing on only the south side where there's breathing room and no swamp; there's also a demonic beasts summoner there, that you want out of the picture asap. Cavalry bosses can die easily enough to Lysithea. You probably know that already, but your Byleth, Claude, Sylvain and Raphael should be able to act as walls if needed. Hope that helps. Dex is mainly for accuracy when it comes to weapons. It also affects magic accuracy, but less so. It gives you a bit of Crit for both as well. Cha is for Gambit accuracy/avoid first and foremost, and gives a slight boost to gambit damage as well.

Can confirm that it does. However, it seems you will only get the battalion tied to characters you have recruited, the one exception being Seteth's.

I actually have very little when it comes to combat exp, I was planning on looking into it next. Healing exp ended up being pretty basic so at least that's covered; I'm always up for making informative threads however, and the exp formula could be a good occasion for a megathread, I just don't have enough "new" to justify it at this point I feel. Though if more people are up we can also figure it out together. Crests activation rates would also be something pretty useful... that would require a ton of testing. You get two chances to save after clearing the final map, one to make a post battle save, which allows you to see the ending/battle data/character endings again, and one that gives you a "cleared" save, which can only be used to start a new game +. As long as you keep the post battle save, you can see the ending and paired endings again, without having to redo the final map. Paired endings aren't otherwise visible anywhere, to my knowledge anyway. Technically the Switch also allows you to make recordings/screenshots of the paired endings. That's a bit more practical imo.

Actually, I'm pretty sure the formula is Res *0.3 rounded down. I just checked with a 24 Res Annette, using a Bolt Axe, she dealt 37 damage vs a 10 Res enemy. Using Lightning Axe raised this to 48, aka 4 + 7; if it was your formula, it would have been 24/4 = 6, so +10 resulting in 47 instead. Just pointing it out, I doubt it will matter much here. I don't really have anything to say about the chapter 6 test anyway; this is the worst part of the game for pure mage Annette imo, so I'm not really surprised. As for the later one, let me spoiler box it for brevity In any case, overall, I am wondering more and more if darting blow isn't the way to go for even magic Annette, myself... But -> mage->peg->DK? Idk, I guess it works on paper? Lightning Axe with steel+ is roughly as strong as Excalibur + tomefaire + Warlock mod(no magic battalion however)... But getting to C+ Axes, D+Lances and E+/D flying will probably make you enter DK a few months later, and not much you can do against targets you don't ORKO. She'll be better in the end for sure, I don't know if it's really that practical though.

I mean, I've made one:

I think as others have said above, while we have quite a few cases of people who got their lives heavily influenced by crests, they're a complete minority if you think about it. Like how many Marianne can there really be? Or Sylvain? Crests are rare. In the grand scheme of things, if not for crests rich families would have been at the top instead(so, nobility, so, the exact same thing). They're problematic, but not the massive menace on society that Edelgard describes. Plus, Rhea does reveal after Miklan's transformation that this kind of things happen, and the church works hard to prevent them, they just arrived too late that time. Which means that they are in fact doing something.

Assuming you are referring to her P.skill here... I'd say it is mostly useful to get class skills before everyone else, and she's the only unit that can realistically pick up two blows between level 10-20. If combat gets your odd unit 3 exp in their weapon of choice, and they fight 10 times in a chapter(which is highly generous for a squishy mage) for 30 exp early on, that's still less than one session of tutoring, which you get 2.5 of plus the same amount of personal studies per month(which also grant around 30 each). Multiplying combat weapon exp by 2, while nice, still doesn't make that value your main source of weapon levels by a long shot, and thus she cannot really build up a massive lead thanks to it, outside of like, authority... or faith if you play slowly. ... things are a bit different if you're spamming auxiliary battles, but in that case nothing matters since you're overleveled anyway.

Oh, my bad, I misunderstood then. ... well to be fair, she does seem to be in a much better state of mind the instant her revenge has been enacted, that's also something the cutscene shows us. Imma let you two settle this one, though. Right, although for some reason, the main thing she did was glorifying Nemesis and co, apparently. I'm still not quite sure why.

She's lost it by that point, to be fair. And she has decent reason to be in that state, considering what Byleth pulled on her, siding with Edelgard after she desecrated the holy tomb, stealing both the sword of the creator and the heart of Sothis as a result; and then there's everything else that happens afterwards. She's kind of extreme to begin with for sure, but I think it's fair to say burning a whole city to the ground isn't something she'd do normally.