Dark Holy Elf
-
Posts
3,610 -
Joined
-
Last visited
Content Type
Profiles
Forums
Events
Posts posted by Dark Holy Elf
-
-
7 hours ago, Whisky said:
Perhaps I need to try it, but I’m still skeptical. How high are their Hit rates? You say it works reliably so their hits rates need to be close to 100 against the enemies you have them fight.
It seems that Rei’s Lugh support takes 60 turns to reach A. Is that fast?
Nosferatu (70) + A support with Lugh (15) + around 18 skl using your 20/6 figure (36) + around 8 luck (4) = 125. So that should be pretty close to 100 against all but super fast enemies.
I'd consider that pretty fast? This is Binding Blade with its stupidly long maps. Most of the requirements for 5 star tactics are on the order of 20 turns, and if Chapter 21 is what you're worried about, there are 12 full maps between Raigh's joining and it. I don't think it's remotely unreasonable that the twins find time for 5 turns together per map. If you're using Warp to go really fast obviously this is less practical, and in general makes Raigh worse because he also needs time to catch up, which he's not going to get then.
I would describe Raigh as high-investment, high-reward. That doesn't really rate very highly to me in a unit ranking for reasons I've outlined in other threads, but if you wait a few levels before promotion, feed him kills, build his Lugh support, etc., he can certainly do some impressive things in the back half of the game. I just threw him and Lugh onto forests on Murdock's map and watched them destroy everything at basically no risk of dying, it's fun. Compared to physical units their 1-2 range is much better, and enemy res in this game is very low, so their enemy phase has higher potential than e.g. a swordmaster or paladin (even though I would consider both those classes to be better than dark mage overall too, to be clear).
7 hours ago, Whisky said:Killer Axes actually have higher accuracy as long as you’re fighting Lance enemies like Wyverns.
Sure, by 5. Nosferatu beats this by 5 against everything else though. And most of the prominent mono axe (or axe+bow) users lack a reasonable full-hit support, and largely have lower skill than Raigh on top of that.
7 hours ago, Whisky said:If Rei were a Mage I think he’d be better specifically because he’d be able to use Aircalibur after promoting and be much more useful against Wyverns before reaching high enough levels to double them.
To be clear I agree with this; Mage is better than Dark Mage in this game. Fire's high hit is wonderful particularly for lower-res bosses, Aircalibur is nice even if I'm not quite on high as it as you. I just think Nosferatu is a neat niche. I basically agree with what @Eltosian Kadath just posted about the classes/tomes in the game.
7 hours ago, Samu_77 said:I have to agree with Whisky here in that Nos-Tanking seems incredibly scuffed in Binding Blade, especially on Hard Mode. Rei's Luck is generally pretty low which already makes me skeptical when it comes to using to do any kind of tanking
Notice how the people who've actually tried it speak more highly of it than those who haven't.
There's a half-bonus to crit avo from the Lugh support as well. It's not much (+2/+5/+7) but it's easily enough to avoid crit from mercs and thunder mages.
Talking about Raigh getting one-shot by a silver weapon past Chapter 16 borders on being disingenuous. Robes are storebought and obviously a good investment on any combat unit you feel is at risk for this (i.e. if Raigh or any other mage is rolling badly for HP). And on average he shouldn't even need them; even 20/1 Raigh with B Lugh needs 39 atk to be one-shot, and I don't think even wyvern lords have 25 str until Brunya's map.
-
25 minutes ago, Whisky said:
Okay so the problem is, Dark Mages here compete with Mages that deal effective damage to Wyverns. That’s a really nice thing to have. I just don’t see how Dark Mages can keep up with their competition. Ray isn’t great. I’ve seen people talk about Nosferatu tanking, I haven’t tried it myself but that seems very risky and unreliable. Wouldn’t work for Niime either because of how frail she is.
I've done it with both and it definitely works reliably. Niime needed to watch what enemies she faces a bit (enemies with crit etc.), Raigh really didn't, is my memory.
As Cysx notes, Raigh really benefits from his Lugh support, but this isn't hard to set up, quite fast and super beneficial for both since it gives both full hit and full avoid. Niime needs a Seraph Robe or two. The cost of both Nosferatu and Niime's Seraph Robes is fair to bring up, I suppose, but I don't really find money that hard to come by in Binding Blade, at least not by the point of the game we're talking about. I did a challenge run where I promoted every single unit and I didn't need to use the arena to pull it off... my point being that on a playthrough where you're not throwing close to 6 digits of gold down the toilet on that, you really should have lots of money.
The accuracy topping out at 70 hit is definitely a bit of a downside though, for sure. It's a better situation than pure axe-users but that's not saying much. Outside said challenge run I've only ever used Raigh alongside Lugh, which does help him out a fair bit, though even then it's a few maps before he starts really feeling like a contributor. Of course, most other mages require at least some babying in this game, and the upside is neat.
Incidentally I will say I'm not really as concerned with the lack of Aircalibur as you? Promoted Raigh and Niime both double pretty much every wyvern on Normal, and a fair chunk of them on Hard (particularly Raigh, at least if his speed growth cooperates)— blame their fascination with steel lances — and their res is garbage so if they're doubled, they die. Apocalypse can also one-shot them even on Hard IIRC, but that's only relevant at the very end of course.
-
Personally I don't think Dark Knight belongs with dark mages at all, despite the name. Dark mages are largely defined by their access to otherwise exclusive dark magic. The only game where Dark Knight can use dark magic is also the only class where ANY mage class can use dark magic, provided the character in question learns any. I definitely think Dark Knight belongs with Mage Knight and similar. (Obviously it's totally cool for everyone to make their own lists how they want though; very much agree with Zapp on that one.)
1. Awakening: Yeah, Nosferatu is one of the easiest ways to break the game wide open. It's not like the other dark tomes are bad either, with a brave tome and a killer tome.2. Binding Blade: Honestly kinda think they're a bit underrated. Nosferatu is storebought, light, and not horrifically inaccurate, and unlike Awakening it even does healing = full damage, rather than half. It's pretty easy to get a Raigh who just devours entire armies with it once he gets rolling, and Niime can do it easily too if given a storebought robe or two to ward off OHKOs from Silver Lance Wyvern Lords and the like. Sophia sucks but that's definitely a Sophia problem, not a class problem.
3. Fates: Nosferatu isn't nearly as busted in this game thanks to the no-doubling tag, but it's still an incredibly useful tool with the right support, and the Sorcerer class itself has nice stats, including bonuses to the derived stats which are welcome.
4. Blazing Blade: Canas has worse stats than Raigh and Nosferatu weighs a lot... Luna is cool though (albeit you have Athos for the one map it is coolest).
5. Sacred Stones: FE7 nerfed Nosferatu, and now FE8 nerfs Luna to add to it! Also they basically don't get a Sacred Twin. On top of that, both dark mages have crappy combat... but summoning is a unique niche they have. Not sure how to score this.
6. Three Houses: 3H wishes I counted Dark Knight, since I like that! Dark Mage and Dark Bishop are pretty clearly the worst mage classes in the game even before considering the need to go out of the way to obtain their seals; mastering Dark Mage doesn't give Fiendish Blow, and Dark Bishop gives no useful bonuses aside from a Fiendish Blow that doesn't stack with the one you got from mastering Mage. Dark Bishop has a slight niche for Hubert but only because every other Advanced mage class is either not available to him or doesn't offer him anything, but that's not really praise for Dark Bishop itself. (and even then, I personally prefer Paladin for him in this tier).
Abstaining on RD since it's New Game+ only but... yeah it's clearly not great. I'd rank it around Sacred Stones depending on how I feel about summoning on any given day.
-
@JotariYeah, I agree with all of that.
I was just a bit surprised at your original statement, which grouped Aura and Abraxas in with Seraphim and implied they were superior to Nosferatu, just because that's very different than how I would rank these spells. And while I agree that any spell is better than nothing once you have it, I will say that Aura and Abraxas are high on the list of spells I may not even bother learning. Particularly egregious in cases like Annette who gets nothing after C faith until A; I'm not investing 800+ points just to get such a limited-value addition to her spell list. (Though to get back on topic a bit, I suppose I would if I was using Holy Knight Annette, since at that point it would have a notable power lead on Excalibur.)
-
27 minutes ago, Jotari said:
I've never had an issue with Aura. It might lose to Ragnarok in a 1:1 comparison, but you either A) probably don't have Ragnarok or B)Want to use a powerful spell more than 3 times per map. I've specifically used Aura with Byleth even though they also have Ragnarok.
My issues with Nosferatu's not actually that bad hit rate probably comes down to the fact that if I'm in a situation where I'm using Nosferatu then I'm probably using it for a reason and can't really afford the miss. I did genuinely try to give Byleth a Nosferatu build once abusing New Game+ and it turned out very middling.
For me the usual reason I'm using Nosferatu is to one-round an armour with an injured mage and fully heal them as a bonus, saving someone else a healing action. Its hit rate is great against armours. I admit I don't use it against much else.
Aura has some use, but only really in the sense that every spell has some use. It does have reasonably high power, it just has much worse numbers than other similar spells. If it's the only high-powered spell you have (either because it's literally the only one, or you're worried about running out of uses on another) then sure, it has some niche. But if I had to rank every single offensive spell in Three Houses, Aura is likely sitting in last place. (I might be forgetting something; I'm open to a challenge on this statement!)
1 hour ago, Shadow Mir said:The issue here is that because most white magic spells are non-offensive, and the offensive white magic spells mostly suck, any prospective Holy Knight pretty much MUST train reason as well.
This is fair, but if we're using Holy Knight, I think we can assume that we're training Faith more than Reason (else we'd be using Dark Knight). So I'd imagine Reason is probably sitting at C (for the spell earned at that level... interestingly, most suggested Holy Knight candidates learn Thoron) or perhaps C+ for Prowess 3, at least before any incidental extra training from usage. Either way, that's still cheaper than Great Knight.
Also, I'd say Great Knight (and indeed most class builds) is likely training another skill or skills as well. What class do you imagine them going through in Advanced tier? There are several possibilities, but I think it's fair to say the most common choice is Paladin, so that will require some Lance training. And you'll probably want a better ranged option than javelin/hand axe, so you can probably add a bit of bow training too.
-
6 hours ago, Shadow Mir said:
It doesn't help that it arguably needs even more effort to qualify for than Great Knight, which is the poster child for disproportionately high requirements
I don't see the argument. Great Knight requires A / B+ / B+, while Holy Knight requires A / B+ / C. The difference between a B+ and a C (the relevant difference here) is 660 exp. That's well outside the range where other considerations (e.g. what classes the units are spending time in en route to the master class) might tilt things. Great Knight is fairly clearly the most expensive master class to reach.
I think you've often kneejerk overrated the costs of getting a single C-rank skill by Level 30; 300 exp is a small amount by that point in the game. In fact, if one plans to reach >80% certification rate, it is actually cheaper/easier in every way to reach A/B+/C (e.g. for Holy Knight, Dark Knight, or Falcon Knight), than it is to reach A/A (e.g. for War Master or Gremory), because the difference between B+ and A (360 exp) is higher than the difference between C and E (300 exp).
2 hours ago, Jotari said:On the topic, I don't think it's right to say white magic spells are bad. Nosferatu is pretty inaccurate, but balancing Nosferatu is a challenge the series has never quite managed. But Seraphim, Aura and Abraxes are all serviceable if not amazing.
Seraphim is good and Nosferatu has a clear use, albeit a niche one. But Aura and Abraxas are both pretty bad, IMO, especially Aura, which is the least accurate spell in the game in practice. Abraxas is the better of the two, but still not great: comparing it to Ragnarok it has +4 weight, -1 uses, -1 might, and their hit might as well be equal once the relevant Prowess skills are considered.
Definitely agree with @Shanty Pete's 1st Mate that light magic could have used a weight reduction across the board. Drop Abraxas by 6 weight and suddenly it has some edge in that comparison with Ragnarok.
1 hour ago, Jotari said:And if we're talking about adding more white magic, it'd be really cool if Thani was in the game. They let anyone use rapiers, why not toss the ball of armour extinction on a few characters?
I think they might be a bit leery about using that name since it's Micaiah's signature, but I'd still like this quite well. It would be rather similar to Dark Spikes, but since only two characters get Dark Spikes I think expanding this niche is fine. And you could make Thani quite different, i.e. more uses, lighter (let's go for the full RD reference and make it 1 weight, why not), and less power. This would leave it clearly worse than Dark Spikes against enemies too fast for it to double, but absolutely great at shredding them if you are doubling. Set its power well and it could probably kill even Holy/Dark Knights when doubling which Dark Spikes can struggle with.
-
9 hours ago, Jotari said:
Thinking on it, what could have been done to make Holy Knight better? Okay, white magic uses x2 is the obvious choice, but that sort of trips over Bishop and Gremory's niche. Assuming the idea of Holy Knight is to be "offensive white magic" in a game with very little white magic. Maybe it could have got the OG Slayer skill as its mastery that lets all its attacks (or even just it's magic/white magic) deal effective damage to beast units. That would be pretty cool, though might eliminate Seraphim's unique use, which is the main reason to use it as is.
Hm, two proposals:
White Tome Exerptise: +10 damage with offensive white magic, and double its number of uses
Monster Slayer: Offensive white magic ignores the "Magic Immunity" barrier effect and deals effective damage to beasts. [Seraphim is still useful even with this, because it has a lot more might than Nosferatu.]
Either of these replacing White Tomefaire would essentially keep the class's niche in the same place as White Tomefaire is trying to have it, but actually make it work. I also like the idea of just giving the class Seraphim. While we're at it, we could give Dark Bishop an actual good spell like Mire or Banshee or Luna to tempt people that direction.
21 hours ago, Zapp Branniglenn said:I disagree! Not to derail the discussion, but if you swapped Catherine's base class from Swordmaster to Mercenary her stats would be noticeably worse at every recruitable chapter.
In Whisky's defence, I think he was talking about Swordmaster compared to other Advanced classes, or at least that's how I interpreted the comment. I certainly agree that Swordmaster is much better than an Intermediate class assuming equal availability, and this is indeed why Catherine is good! So in the context you're using, you're of course 100% correct. But in the context I believe Whisky was using, he's correct as well. Catherine would be even better if she started as many other physical advanced classes. Assassin would give her the same stats but now she has +1 move and a free C in bows. Sniper would make her completely outclass Shamir (outside the lack of bow combat arts, I suppose, but joining earlier with better growths more than makes up for this). War Cleric would just let her do ridiculous damage (Nimble Combo and Fistfaire off that strength in Chapter 5?!) and fast-track her to a great dodgetank build, plus 6 move and free C+ faith. Wyvern Rider needs no explanation. etc.
-
35 minutes ago, Shanty Pete's 1st Mate said:
Actually, let's imagine a 0% growths Sylvain. How does he do?
[...]
That's cleanly one-rounding chapter 13's Pallardo, along with essentially any generic on that same map. He can one-round basically any non-Armored, non-Monster enemy from that point onward. Even if he gets screwed on his growths, Swift Strikes renders Sylvain a player-phase nuke, who can then Canto away to safety.
Of course, if you don't train his Lance rank up to A, or have him master Brigand, or give him a strong battalion, then sure he can turn out disappointing. But then, that's the player's own mistakes, not a consequence of bad RNG.
Yeah I was trying to be charitable to Mir there. To be fair, I don't think it's completely unreasonable for Sylvain to skip Brigand (maybe you just want Sylvain to spend all that time in Cavalier, it is quite a bit a better-performing class so if a player expects to have trouble with Intermediate-tier maps). And I think it's quite reasonable to skip Str+2; it's a tough call between that and Reposition for him IMO. But this is getting into nickel-and-diming. And there are even sources of damage beyond what you mentioned for Sylvain, e.g. there are relic lances with 3/5/8 more might than Silver+, he can also benefit from up to two +might supports from Felix and Ingrid*, and finally there's his personal ability to consider in a pinch*. So obviously overall I agree with you: Sylvain's quite good, and mounted classes (either cavalry or wyvern) are unambiguously where he shines best.
*While on the subject of these two methods of having allies power up Sylvain, it's worth mentioning that Felix/Ingrid/[random female ally] will be able to provide this support more easily — wait for it — if they have Canto! Canto's good.
-
37 minutes ago, Shadow Mir said:
I worked with Sylvain in Azure Moon. I made him a Cavalier, then a Paladin. He fell off really fucking HARD after the timeskip, and he didn't impress me before the timeskip either. His failure to impress, aside from earning him a grudge being held against him, is gonna give me a lot of pause about working with Ferdinand when I go back to 3H... anyway, high movement is nice, but you damn well better have something going for you besides that. It alone doesn’t make a class good. Case in point: Holy Knight.
Sylvain, like any other unit, is certainly capable of falling off if he gets poor stat gains. So I'm not disbelieving your experience, to be clear.
I'm not sure why that's Paladin's fault, though. Let's say you'd made Sylvain a Swordmaster instead. Even setting aside that it's a significantly more difficult class for him to get to, do you think this would have improved him? Because I certainly don't. He'd have been slightly faster (due to both class modifiers and later, growths), but since Sylvain has Swift Strikes, more speed isn't of amazing value for him offensively (it's still useful, but not highly valuable). But he'd have either 1 less move or 3 less move, no canter, and a -faire on a weaker weapon type. This seems like such an obvious losing trade to me.
-
25 minutes ago, Shadow Mir said:
Many of those other games didn't have terrain that slows down cavalry units and ONLY cavalry units.
This would be a relevant thing to bring up if dismount did not exist. It does exist, so your point is invalid. There is literally no terrain in 3H that will result in a Swordmaster being able to move more squares than a Holy Knight. I can't even think of any situations where they can even move the same number of squares, though it's possible one might exist.
28 minutes ago, Shadow Mir said:Also, you're forgetting another piece of the puzzle here; 3H has combat arts for moving other units around.
These actions are certainly useful, no question. They add 1 or 2 points to a unit's effective mobility for a turn. So, basically, if you use Smite/Reposition on a Swordmaster, they can match a Holy Knight for mobility for one turn (roughly, at least. The repositioned swordmaster can handle terrain better, but doesn't have canter). Of course, the mounted unit still has two big (albeit non-overlapping) advantages: (a) they do not need such a support action for that turn, freeing up their ally to target someone else or use another action entirely, or (b) you can still target the mounted unit with reposition/smite too if you want, giving them even more effective mobility (and this mobility can be used to both advance or retreat after an attack, as desired).
I'm really not sure why you're so set on dying on this hill of "infantry are better than cavalry, actually". More move is always going to result in a better class unless the less mobile class has some convincing advantages to make up for that. Some infantry classes have such convincing advantages, such as Sniper or Gremory. I don't think it's particularly contraversial to say Swordmaster does not.
-
4 hours ago, Wuzzy said:
Reforge sounds good on paper but I find it better as an enemy skill than a player skill
I forgot Royal Knight entirely in my ratings which isn't a good sign. Agreed that it's toward the bottom for reasons you said.
Having just watched a map in which the player used enemy!Mauvier's Reforge to do free damage to him while killing his allies, I'm not certain I agree that Reforge is useful for enemies either. The most notable thing about the Break status is that it allows the attack which inflicts it to not be countered, which Reforge does nothing to stop. Healing the effect for subsequent attacks feels very very niche. I guess I'm saying that I think Reforge is a pretty useless skill.
Cavalry is also indeed a pretty weak typing (outside the +1 move of course). I'd say the most notable things about it is that it allows for 100% bonded shield against other cavalry, as well as bonus damage with Twin Strike which can be the difference between killing a fell wyrm and not. There are probably some others.
-
1 hour ago, samthedigital said:
I am free to rank units based on how much time they save in a speedrun if I like though since I am allowed to use any criteria that I want. For example I could say that Sigurd and Seliph save the most time out of anyone relative to the other units in their cast and they are therefore the best lords. That being said I didn't say that I was interested in ranking units this way; my original post here had little to do with a speedrunning mindset. You might find that a speedrunner wouldn't be interested ranking units that way either. The point was to show that there is an objective best use of a resource in a given type of playthrough.
I agree with your opening sentence, of course.
That said I disagree with the last, I think? At least if you mean "any given type of playthrough" and not just that specific one. (Apologies if I misunderstood.) Sure, a speedrun always (or almost always?) has an objective best use of a resource, because it has extremely well-defined criteria: there is one particular variable you're trying to get as low as possible (maybe with a slight second variable of reliability depending on if we're doing a one-time race or trying for a world record). I would argue that most other types of playthroughs, and most methods people use for rating units/classes/etc., use far vaguer criteria than that; there are basically always multiple variables to consider and their balance is not strictly defined. Now of course there are times when I still feel there is a clear choice for a given resource, but there are others where I think the choice is unclear. (As an easy example since it's on my mind, I do not consider "which unit (or units) get Emblem Lyn" to be a clear choice in the slightest, and I can't imagine I ever will.)
And of course in regards to discussions which take on a more debate-like quality: even when a resource use is clear by my standards, I have to accept that other people will use different standards than me, and thus might come to a different conclusion about the rightful use of resources, so I generally find it more convincing to make arguments not based on my own assumptions of resource allocation when possible, and I feel similarly when reading arguments made by others.
9 hours ago, Whisky said:Slight disagreement with this part. Maybe it’s just the wording. But new strategies for speedruns are discovered years after games came out, the world records for many games are relatively recent records. I think there’s almost always room for improvement, it’s not just solve the problem and you’re done. Sometimes units have surprising uses in LTCs that most people wouldn’t think of. Speedrunners or LTCers need to be open minded to new ideas and strategies being discovered.
Yeah, that's totally fair! Certainly I agree with everything you say here, so perhaps I did misword things. I'll admit I don't pay much attention to speedruns so it's very possible I overestimated in my head how fixed the broad strategies are. (Obviously world records can still improve as people make various small optimizations, both in terms of small strategy adjustments and execution.)
-
23 minutes ago, Shanty Pete's 1st Mate said:
And Tormod gets forgetten again...
*Dark Holy Elf gets an Ace Attorney-style damage animation*
Good point! Though Tormod is even more underlevelled than Mist, doesn't have canter, and requires some grinding to reach C staves.
26 minutes ago, Shanty Pete's 1st Mate said:Eh... Swordmasters with an at-least tolerable Magic stat should be one-rounding Armor Knights with the Levin Sword. And in its forged form, it goes up to 3-range, making it a wash with a 6-move dismounted Holy Knight. Hell, running units like Dorothea, Manuela, or Lysithea as Swordmasters could be fun this way. Although they'd do the same job better in Assassin.
I was admittedly thinking only of "traditional physical swordmaster candidates" (like Felix, Catherine, etc.) with no Fiendish Blow or Warlock certification, for whom Levin Sword isn't strong enough to one-round, typically. Though in my defence, given his hatred of all things that might be interpreted as a build which is both physical and magical in any way, I'd bet that's who Mir was thinking of too.
And yeah I enjoyed using swordmaster Lysithea on an early run (unironically liked it more than any physical swordmaster I've ever tried, solid 3 range really eases the downside of 5 move) before I figured out you might as well just use assassin instead.
-
Unshockingly I basically agree with Whisky about everything here, but I still want to pile on about the key points:
8 hours ago, Shadow Mir said:What about Crimson Flower's last two maps? And the Valley of Torment? And Gronder Redux? The Sealed Forest? The forest map used in Marianne's paralogue (as well as Petra and Bernadetta's)?
That's a lot of maps where Holy Knight can dismount to 6 move and still have more than swordmaster.
8 hours ago, Shadow Mir said:Also, Swordmaster has better speed than Holy Knight (4 vs negative 1),
It's just dishonest to use Holy Knight's mounted speed modifier when even dismounted they are more mobile than swordmaster in all situations, so the gap is just 3. Like yeah maybe there will be one enemy type which swordmaster will double with a bow that Holy Knight will miss, but it likely won't be one they can kill since bow damage with a middling str mod and no Bowfaire just isn't very strong. Meanwhile there's at least one enemy type we know holy knights will kill at range which swordmasters don't: armour knights. They're clearly superior ranged combatants as such.
8 hours ago, Shadow Mir said:You want a class that has literally nothing going for it?! War Monk/Cleric is a thing.
You really want to compare the class that is only useful because Catherine starts in it (but she should eventually get out of) to one of the classes she is recommended to go to? War Cleric has +1 move (and ignores certain terrain penalties), spell access, and Fistfaire on Swordmaster. Oh and Brawl Avoid +20 is way, way better than Astra as far as mastery skills go. Every fist-based character should strongly consider at least making a stop in the class, if DLC is an option.
With that out of the way, rankings!
SpoilerI assume we're going to do Mage Knights / Dark Knights at a later date? Anyway, here we go:
1. Three Houses Valkyrie: I'm not sure if it belongs here or with Mage Knights / Dark Knights, despite its name. Anyway it's extremely good. Range+1, 6 move, +4 magic mod, and canter is just an incredible package for advanced tier. Uncanny Blow is great too. It also reduces the training for Dark Knight (due to all the free riding exp you get). A strong consideration for any female magic unit. Even better for Lysithea and Hapi since it's the only class in the tier which has any effect on the effectiveness of dark magic.
2. Fates: Really good for Jakob/Felicia, since they get Inspiration at Level 15 and that skill rules. (Other characters have to wait an extra 20 levels effectively.) That said it's a nice class for anyone with staves and a magic stat, mobile and solid offence.
3. Genealogy: Horse with staves very good. They don't get all staves but they get enough. Nanna can even have pretty good combat (and always has Charm). I might be underrating them a bit even?
4. Binding Blade/Blazing Blade: They're pretty similar in these games, their class having a big leg-up on the unmounted variant. They get boosted exp as well. Particular shout-out for Clarine just being weirdly one of the best walls in the game.
5. Path of Radiance: Mist will only have offence if you favour her with Arms Scrolls (though it actually gets pretty good then), but it's fine. Get her to 10 and promote her, she's the only staff user until Elincia who doesn't have garbage move and being hit with status staves is about the only thing that can slow down the incredible mounted PoR army, so I appreciate Mist being able to reach people to use Restore.
6. Three Houses Holy Knight: It's been said. I think the class has very little use, at best an off-brand choice for Marianne/Bernie/Ingrid (but all would prefer other options) because Dark Knight overshadows it so badly. But if you're forced to use it, it'll be fine, because mounted mage is still very good it turns out.
7. Sacred Stones: Compared to the other GBA games they lost 1 move. Valkyrie's a decent choice for Natasha, but going Bishop for slayer is valid too. L'Arachel sadly starts rather weak; not so weak she can't get rolling but I'm not sure +1 move is worth it.
8. Radiant Dawn: Mist took a hit in this game because (a) there's more competition for staff users, with Calill / Bastian gaining staves, Micaiah / Laura / Pelleas being added to the roster, and Elincia being buffed and gaining availability, and (b) it's no longer possible to early-promote her; she needs to wait for her own late special snowflake promotion item or reach 21 naturally. Oh and she lost what little offensive potential she had. Still not terrible, she does outmove Rhys for instance, but not great.
Abstain on Awakening; it seems decent enough on paper but I've not used it much, just easier to use falconknights for staves and dark mages for magical offence.
-
9 hours ago, samthedigital said:
It's not subjective, but maybe I'm misinterpreting what you're saying. If I tiered units based on how much time they save over the course of a speedrun then the best way to allocate resources would the one that results in the fastest time
I think that's the difference. I'm not talking about (or particularly interested in talking about) tiering units based on a speedrun. I will readily concede that such a discussion has objective criteria, but it's not one I'm interested in. At that point you reduce discussions to a solvable problem; someone will figure out the strategy that achieves the fastest time and that's that. (Said someone will also be a person who actually attempts or cares deeply about speedruns, and neither of those descriptors applies to me, or I suspect to many other people in this thread.)
Generally in threads discussing unit or class worth like this one, there isn't such clear criteria. We're just generally talking about what we find most effective overall. That can mean different things for different people, of course, hence the subjectivity.
9 hours ago, samthedigital said:I don't have as stringent a rating system as a speedrun that I mentioned above, but the problem then is how to compare units when not as many things are guaranteed.
I certainly think comparisons are still possible, as years of rich discussion threads show. What's less possible is achieving 100% agreement among all participants, but I think that limitation is perfectly fine, and even desirable.
-
On 10/24/2023 at 9:14 AM, samthedigital said:
The issue in my mind is that it seems like an impossible task to come up with an objective aggregate rating based on every single playstyle.
Oh, to be clear, I definitely agree; I wasn't proposing coming up with such a ranking. I was just observing that, given a bunch of people with differing playstyles weighing in on a thread like this, that I would expect Bunet to be the lowest-rated unit. That certainly doesn't mean individual people can't voice why they think Bunet should be higher. (I'm not even trying to weigh in on Bunet vs. Jean or Anna myself; I don't know how i feel personally.)
-
18 hours ago, samthedigital said:
You could do this (and rate units based on how well they carry), but it's also possible to rate units in the context where Seliph gets most of the resources. This is how I rate them.
That's fair. I would only argue that in such a context, we can't actually rate Seliph himself.
18 hours ago, samthedigital said:Yeah, it's a philosophical thing, and I don't penalize a unit when they are the best use of a given resource. Seadall is a bit of an awkward example since he's a dancer, but I would note that it would be relevant to me if he can't be that most important unit and that he isn't able to make use of heavy investment.
I think you either have to penalize them, or just be very careful about comparing them to other people with said resource, because who is the "best" recipient of a resource will always be subjective, and locking it in traps you in circular arguments.
18 hours ago, samthedigital said:Excluding boots though off the top of my head Kliff, Robin, and Jill in FE9/10 would be units that I consider to be "good investment targets" because they don't really do that well at base, but they do quickly snowball and become better than the rest of the army (even if other units in the army got the same resources instead) because of things that are unique to them with investment
I definitely agree with you here; those units are good investment targets. To me, the parenthetical part of your statement is the proof: we can compare them to other units getting those same resources and observe they still come out ahead. "Jill with a lot of boosters" is better than "a myrmidon with a lot of boosters" because they both kill everything and don't die, but Jill does it with canto, flight, and +2 move, etc.
15 hours ago, Cysx said:The main thing arks back to personal experience. Because people do use stat boosters and then have to discuss the games as if they didn't, and nobody seems to see a problem there. This premise leads to the very circular logic you're denouncing here(but with a spice of unit bias added right in), except now it's even worse because no one will want to recognize it's happening. I like to say this a lot but we didn't design the games, if stat boosters are blatantly more effective in certain places, I don't see much of a point pretending it's not the case.
Of course, discussion where we just use all the best tools and ignore the rest isn't as interesting, I agree, but that's why we now consider universal investment and what characters can individually do with that. I don't think it's impossible to work with boosters in a comparable way, but that has to be figured out, and it cannot if we just keep pretending there's nothing to even fix. I guess this means we want the same thing, but the core of my issue is that if I'm understanding correctly you seem to think it's a thing already, and boy do I really not think it is, and we're not even halfway there.
Hm, well.
Regarding personal experience, the beautiful thing about having lots of people weigh in is we have a huge amount of personal experience to work with. People play without stat boosters all the time, or just give them preferentially to favourites. Combining the experiences of lots of people and a bit of theorycraft, I do think we can figure out how a lot of units perform at different levels of investment pretty well. That doesn't mean we're perfect at it and certainly things get missed, but we can get pretty good. And to the extent that we're not better at it, I'm not as bothered by this as you seem to be. Ultimately these discussions are just thought experiments to stretch our brains.
And just to be clear, I'm not disagreeing with your assessment that stat boosters are blatantly more effective on some units than others. I just disagree with using that information to then compare units at different levels of investment.
-
9 hours ago, Shanty Pete's 1st Mate said:
There - I think that 10 points, distributed this way, would make a not-the-only-Seizer Roy about equivalent in value to "vanilla Roy". I evalyated based on NM, since that's the one I know best (and the one with more consistent enemy stats), but such a boost would obviously improve his HM performance as well. This is all very "touchy-feely", and I acknowledge that playing the game with said changes applied might cause me to "tweak" the dial a bit.
Interesting stuff! To me, of course, your new Roy is a clear upgrade on the old one, but I liked seeing you lay out your thought process.
9 hours ago, Shanty Pete's 1st Mate said:Well, this is another question of one's point of view. Is an "unrescuable" Roy a substantially worse unit than vanilla Roy? Or, is it the "Rescuers", who lose some of their functionality, all becoming slightly worse? I think there's a case to be made either way. Anyway, I do agree that such a change would make Roy a better Boots candidate.
My response to your first question is "both". Being unable to be targeted by a helpful action is a weak point for a unit; if Alan could be targeted by Dance but Lance could not be, I think this would unquestionably be a valid thing to bring up in Alan's favour during a comparison of Alan and Lance as units. And conversely, every unit who becomes untargetable by an action makes the action in question (and those who use it) worse. If "can't target Lance with Dance" was a property of Larum, but Elphin could target both just fine, that would be a point in Elphin's favour when comparing the performers.
There are non-hypothetical versions of this too, of course; e.g. Hawkeye can not be rescued by promoted paladins or fliers, which is a point in favour of his competition that is (correctly, IMO) brought up in conversations. And similarly, Vaida's ability to target fewer people with Rescue than other FE7 fliers is a point against her.
But obviously we broadly agree; I just thought your tangent was an interesting one.
10 hours ago, gnip said:I do think that it's a bit silly to just assume that these units get these stat boosters, and just be compared as such vs. other characters without this kind of help. "Can use stat boosters well" is just another of those unit qualities that are hard to quantify, like "staff utility", "rescue utility", "thief utilty" and all that.
I definitely agree with you. It is absolutely fair to observe that some characters make better uses of stat boosters than others; that's an important conversation to have (arguably more important than which unit is better, even!). And, all other things being equal, I would even agree it is a slight point in favour or the unit who makes better use of them (so I misspoke when I earlier said the two topics were "entirely" unrelated).
That said, there is, fortunately, a very elegant solution: whenever you assume a unit receives favouritism such as a stat booster, you always compare him or her to other units receiving equal (as much as possible) favouritism. If Miredy really does make better use of Boots than [insert other unit here], it should be possible to show that in a Miredy vs. [other unit] comparison where each is assumed to receive the boots. That doesn't mean we can then turn around and compare Miredy-with-Boots to others without, though, even if the general consensus ends up being that Miredy "should" get them.
9 hours ago, samthedigital said:This is something I'm conflicted about. I assume that Seliph gets a bunch of resources because it's overwhelmingly the fastest and most efficient way to beat FE4. I could even compare him to other characters given a similar level of resources and the result would be the same. The problem is that if I compare other units in this way then it wouldn't be an accurate representation of how the characters perform in a practical setting.
I don't see a problem, perhaps because I disagree about what counts as a practical setting. If I wanted to give all my resources to Scathach instead, I can, and I can still win. More realistically, if I spread all the resources out in essentially random ways, I can easily do so and still win, as evidenced by the fact that I suspect this is in fact closer to how most people actually play FE4. You can't argue these aren't "practical" ways to play the game. You can certainly argue they're not "optimum", but I would shy away from only comparing characters in optimum situations (even if we can agree on what optimum means, which isn't a given in a single-player game), because it easily leads to circuluar logic. Taken to its extreme you end up being unable to talk about "less optimum" units at all. "Wil is not a good unit to invest in therefore we try to avoid giving him exp therefore Wil ends up underlevelled and terrible and therefore Wil sucks" is not a valid argument.
In general I feel that the value of a unit is, in broad terms, equal to their performance relative to the investment required to achieve that performance. e.g. I think Seadall is the best unit in Engage because he doesn't take investment away from other units the way top combat units do, and even if someone could prove that [insert unit here] is the "objective" best recipient of a ton of favouritism that would make them more important to an "optimized" run than Seadall, I don't think that would make said unit nearly as good as Seadall. But this is a philosophical thing and I can see disagreeing.
-
I agree with @Jotari. The weird thing about class bases in games with no reclass possibilities is... well, they're invisible and undetctable. Like, if I hack the game and up Lord and Master Lord's stats by 2, but lower Roy's personal stats by 2, the game is unchanged from an end user perspective. So I'm a bit uncomfortable with ranking classes based on their stats in such games, for all that I can see the case to do so.
In fact this is largely why I basically just didn't comment on the lord classes in this very thread. Because lords can't reclass in most games (outside Awakning/Fates, and 3H kinda), and they have a sample size of 1 per class (unlike e.g. cavalier vs. myrmidon, where you can compare an aggregate and infer some statistical trends about the class), it feels like there's not much to discuss when talking about the class itself, in a way which distinguishes it from the individual lords as units.
On 10/25/2023 at 5:05 AM, Shanty Pete's 1st Mate said:So, if I'm rating units within the context of FE6, I would broadly side with position 2. As you say, Seizing is essential. I do think there's some wight to give it, say in deciding stat boosters - Roy is a good candidate for defensive boosters and the Boots, because he Seizes, but also because his death is a "Game Over" condition.
Oh, I definitely agree Roy / other seize lords are a good candidate for boots. Notably, though, I think "is a good candidate for Boots / other resources" is an entirely unrelated conversation to how good I think a unit is. (Though, to be fair, I've noticed not everyone feels similarly... different opinions on Seliph are pretty revealing here, for instance.)
This is easiest to illustrate with a thought experiment: what if we made Roy untargetable by Rescue (as well as repositioning staves such as Warp)? What if we lowered his move? Both of these changes make Roy an unquestionably worse unit: you would never trade the Roy we have for this new one, given a choice. But they also make him a better candidate for the boots.
On 10/25/2023 at 5:05 AM, Shanty Pete's 1st Mate said:However, in this thread, we're specifically ranking the classes between their apppearances in various games. In that light, I'm ranking "Lords who always Seize" against "Lords who sometimes Seize" against "Lords who never Seize". In that light, I feel compelled to acknowledge that the first is contributing something more than the second, who is contributing more than the third, in at least this one axis. Likewise, visiting Villages is usually to the player's benefit; so, in a game (SD) where the ability to Visit is concentrated into one unit, that unit (Marth) will have an advantage over Lords who don't uniquely possess this power.
Villages are slightly different because they're not essential, so arguably Marth is similar to thieves in other games and deserves a score boost because he is providing material goods you couldn't get without him (much like how Matthew can provide the Silver Card and hence insane amounts of effective gold). This is still hard to score (as discussions of Matthew have proven in the past), but at least I can see scoring it in some way.
Seizing just feels impossible to score. I'm definitely curious how much you weight it (and I mean this non-judgementally to be clear, I'm not trying to use this as a gotcha), e.g. if you consider a hacked version of Binding Blade where anyone can seize (but the game is otherwise unchanged, i.e. Roy is still force-deployed and story-promoted), how much better would Hacked Roy's stats have to be than Current Roy's in order for him to be equally valuable in both versions of the game, in your estimation?
-
On 10/22/2023 at 7:07 PM, Shanty Pete's 1st Mate said:
I find myself half-agreeing, half-disagreeing. I follow your logic, but it leads me to the opposite conclusion. Picture this: every map is "Defeat Boss", but Roy is still a forced deploy. Even if Dieck, or Fir, or Ogier would be better, you're stuck fielding... Roy. In my opinion, that'd be an even worse scenario for him. You're forced to bring him, but he's contributing nothing unique. Why force deploy him, if you don't need him to complete the chapter? It's a wasted slot. At least in the vanilla FE6, he's doing something important.
The issue I have with seizing in this conversation is that I can only really see two logical positions:
1. Seizing is an essential action. You could argue Rutger is "better", but even on hard mode a skilled player can easily beat the game without ever having Rutger take an action (or even recruiting him). Roy's Seize actions, meanwhile, are absolutely essential to clearing the game. They are his contribution and they are unquestionably more valuable than anyone else's. Therefore Roy is the best unit in Binding Blade.
2. Seizing is essential, certainly, but the argument that Roy is automatically the best due to Seize is not very informative/useful. Usually if a player asks "is Roy or Rutger better", they probably are thinking in terms of non-essential contributions (i.e. primarily combat in their case, but also supports and rescue utility, etc.) because the essential nature of seizing is already understood. In this case it makes most sense to just ignore seizing entirely for unit rating/ranking purposes.
(If it's not obvious: I take position 2 because I feel it is more useful, but I think position 1 is logically consistent as well.)
Something in the middle, where Roy gets some finite value assigned to his seizing, just feels unsatisfying to me. How do you decide how much to weight seizing? If someone hacked Binding Blade so anyone could seize (similar to Fates), but Roy got +1 to all stats, would you consider that Roy to be better or worse than the current one, and why?
-
7 hours ago, samthedigital said:
How are you determining this exactly?
I was mostly thinking in comparison to Jean (the next lowest) or Anna: there are of course playstyles where those two will be worse than Bunet. But players who are into investing into growth units heavily will get a fair bit out of them; they do have legitimate advantages over others once they get rolling. I don't think there's any type of player who will view Bunet as anything better than vaguely passable filler for a window after he joins. That's his ceiling, while his floor is "do almost nothing in the desert map, then get benched". I think many players see Bunet join at a point in the game where they already just had to bench a bunch of other units, at least some of which they were getting definite use out of (e.g. someone like Framme), and won't be inclined to bench yet someone else for someone whose only positive point is low-investment 6 move.
The other low-ranked units are folks like Alfred, Etie, and Boucheron, but those characters have a window where they will be used.
-
Just piggybacking off the comments of others, agreeing with some and disagreeing with some:
-Citrinne I agree feels too low, agree that she is better than Celine (and probably everyone else below Celine, though you could argue me on a couple). One of the better characters for the "one-round things with magic" build, and that's a good build. And she has early Canter over her competition, at least post-well.
-Veyle's ranking is roughly fine, I think? She deserves to be above the other lowish-availability characters because she actually brings something new and unique to the team, i.e. dragon typing to carry Byleth/Corrin/others. She's less than a point and a half above them anyway.
-Chloe's big advantage on the midgame Elusia/Solm crew is that she has Canter and they do not. So in part 1 she is an outstanding combat unit, in part 2 she is still a great combat unit and has Canter over those who would otherwise be better, and in part 3 she's... passable. That's a very good overall package. I don't feel strongly on how she compares with the other high-end units but I don't think she's too high by much if she is.
-I think Bunet deserves to have the lowest score in the game taking into account different possible playstyles.
To add one of my own, the thing that jumps out at me is that I think Alfred should be below Boucheron. Boucheron has the niche of early backup, and doesn't compete with Louis + Chloe for his weapons, and has slightly better stats overall too IMO. But it's not a huge gap, and someone could easily weight +1 move more highly than me.
-
More than any other class this one just ends up feeling like a unit rating thread to me. I'm just gonna use the "a lord is a unit whose death always causes a game over while also having very high availability on at least one path through the game" to avoid weird corner cases like Lord Zelgius. And is also provides two reasons to exclude Lucina. Anyway I'm ranking based on the average score I would give lords in each game.1. Three Houses: The lord class sucks but I'm not rating that (fine if anyone else wants to do it! It's just not the list I'm making). Edelgard and Dimitri are the best units in their routes, and Byleth is the best unit in Silver Snow. Claude... is extremely good too. There's a case to be made that these are the four best units in the game, period, and no other game manages that for lords.
2. Fates: Corrin varies from "one of the best" (Birthright, Conquest) to "probably outright best" (Revelation) due to immense flexibility, good stats, and an easy ability to snowball. Corrin isn't the second best lord in the series, to be clear, but there's no other lord in Fates to lower the average score.
3. Genealogy: Sigurd is the best lord in the series. Seliph is... decent? Mediocre before promotion, great after. Yes it's the objective best plan to get his mediocre unpromoted phase over quickly, but this doesn't reflect well on him as a unit IMO. This probably averages out to about here.
4. Awakening: Very similar to the above, there's one OP lord and one "pretty good I guess" one. My own personal tiebreaker is that I don't personally consider Robin to be as good as Sigurd.
5. Radiant Dawn: And Ike10 isn't as good as Robin. Micaiah's weird and hard to rate, I consider her decent enough but she's not going to go above Seliph or Chrom so this is the obvious spot.
6. Engage: Alear definitely isn't as good as any one lord from the above games, but they're pretty good and there's nobody to drag them down either.
7. Sacred Stones: Ephraim is certainly very good for a unit who is infantry-locked for half the game. Eirika is okay, held back by a weak start and no 1-2.
8. Path of Radiance: I consider Ike9 to be pretty comparable to Eirika: shaky start, but supports later prop them up to have good durability and good speed, held back by being sword infantry. Ike9's supported durability grows better/faster, but on the other hand he never gets a horse. Anyway, what was my point again? Oh yeah, if Ike9 is comparable to Eirika he's obviously behind avg(Eirika,Ephraim). This is quite a drop actually.
9. Blazing Blade: Well. Lyn is a worse Eirika, and Hector is way worse than Ephraim. And Eliwood is IMO worse than any of them.
10. Binding Blade: Roy bad.
From what I recall Shadow Dragon/New Mystery would be near the bottom and Shadows of Valentia would be pretty high, but I'm not confident enough to say where exactly.
-
Thanks for running this, @Imuabicus der Fertige!
On 10/17/2023 at 11:16 AM, lenticular said:The one thing that really stands out to me here is how much undifferentiated middle there is here. From Céline in 12th down to Rosado in 29th is fully half of all units, but they're only separated by 1.48. I don't know if that's just an artifact of having multiple people giving scores and that driving scores towards the middle or if Engage really does have that many units of very similar quality. Probably a bit of both. Either way, I think it's at least vaguely interesting.
It's definitely interesting. Of your two explanations, I favour your second one, that Engage does indeed have many units of similar quality. The ability to reclass everyone means that, outside some early utility (e.g. their starting class and immediate contributions), characters tend to feel a bit same-y, and ones with comparable overall stats therefore all earn relatively similar ratings.
On 10/19/2023 at 2:07 AM, Bartozio said:I second this. I overall like how the rankings came out, but there's several placings that I feel would at least be interesting to discuss (Goldmary having a considerable higher score than Saphir and Lindon for example).
A gap of 1.24 really isn't that big, even if there are a lot of units who fall between them. Certainly I feel that all my scores have at least a ±1 margin of error, and I suspect it's the same for most others.
That said I do think Goldmary vs Saphir is an easy enough comparison which I do tend to think Goldmary wins. I suppose this might be less true depending on when you do paralogues, but for me, I did quite a lot of them between Chapter 16 and 19, so Goldmary ends up joining 6 internal levels and at least that many maps earlier. Once you adjust for this, it's fair to say Goldmary has significant wins in speed/def/res, while Saphir's wins are in less important areas like HP and Build, and Goldmary has higher availability on top of that. I don't think it's a massive win but it's clear enough IMO. If you do paralogues at a different time you may feel differently.
Ranking each game by class: Dark Mages / Druids / Sorcerors
in General Fire Emblem
Posted
Well keep in mind that Raigh himself has something past 75 avoid when on a forest with Lugh nearby. I don't remember wyvern hit rates but most of their numbers should be super low too. Toss in true hit and even if it were true that a single miss would lead to his death, the odds of it happening (consecutive hit, miss, hit) would be incredibly low. And of course some attacks may not 2HKO (depends on robes I'm sure), and some of these wyverns will be doubled where one miss won't be fatal. etc. I don't remember feeling in danger and I'm pretty sensitive to low-percent chances of death.
For what it's worth, I think this conversation is showing that (at least to me) Raigh is only worth it on Hard if you're investing into him at least fairly heavily. If you're not gonna, just don't bother at all. He has too many benchmarks he wants to hit, and early promotion doesn't really help him hit them. On Normal you can probably get away with lighter investment, of course.
Missing the Wyvern Lords (either by damage or not doubling) is unfortunate but not the end of the world. If you enemy-phase down all the riders and leave the lords on death's door they're an easy mop-up next turn. I also don't think too many setups have a better showing here - Lugh himself is admittedly potentially one (especially if he gets lucky with speed growth, he can get many of the wyverns to zero hit iirc, and Aircalibur is obviously great in this situation. He also doubles them more easily).
(As a point of interest, using some numbers from a project for another site I have lying around, I can say that if you give Raigh around ~9% of the combat exp (i.e. spreading exp equally among a combat team of ~11) from his join map on, he'll hit around Level 20/8 for Murdock's map; most units vary between that and 20/10, with a few obvious exceptions. This was calculated for NM, but I don't believe HM changes much. Obviously this is an abstract ideal; a lot of things can change it in practice (especially completing maps extra fast and leaving a bunch of unharmed enemies) but gives an idea of what's possible depending on different levels of investment.)
Axes aren't bad in FE6, and I bring them up to observe that even in optimum conditions they barely outperform Nosferatu for hit. If you make the conditions less favourable to axes, such as against other weapons, or trying to attack from 2 range, they just plain start losing.
Geese's support situation is dire (59 turns to C with Echidna is his best), and Garret is little better (joins even later, need to save a Lilina support for him and even then it's significantly slower than Lugh/Raigh). Both fall behind Raigh for hit pretty quickly even against lance users (especially on Normal Mode for Garret, though if you want to argue we just don't use Garret on Normal that's fair).
I know this is an aside to your discussion with Jotari, but for what it's worth I definitely wouldn't say they have the same classes, in that in these threads I've usually only had to consider a Hoshido class or a Nohr class, but not both. Hero, Dark Mage, Valkyrie, Wyvern, Berserker, Armour, and Cavalier are Nohr; Swordmaster, Mage, Pegasus Knight are Hoshido. Thief is I think the only class we've talked about where I've seriously considered one from each side (and even then, the Nohr thief uses a bow and the Hoshido ones uses shuriken so they feel very different, if both most likely thieves because they share locktouch. Fighter is the other one which arguably has a representative on both sides, and those two are at least somewhat similar, though one emphasizes offence and one defence... you could argue that Fighter (the Nohr class) is just an unpromoted Berserker, actually, since it doesn't promote to Warrior.