Heals on Wheels: How to Design When a Class Can’t Be Bad
Of all the standard classes to appear in Fire Emblem, Troubadour might be the single hardest one to screw up. The combination of mounted mobility and staff access offers such potent noncombat utility that you could give a Troubadour 1 HP and 0 in every other combat stat and, while players may not like them, they’d still probably be good enough to use – all it means is you might have to actually buy a Mend staff for once in your life. To make a truly bad Troubadour, you’d need to, I don’t know, have them join not just understatted but also late to the point where your infantry staffers have all promoted, give them reduced mov compared to other mounts so as to have no advantage over said promoted infantry, and make them underleveled enough that you can’t promote them anytime soon to regain that advantage, and stick them with a poor enough base staff rank to be unable to use important staves like Restore in time for when you need them. But surely no one would do that, right?
Oh. Right.
Because the class is so straightforwardly strong, there really isn’t much to say about designing a Troubadour in a vacuum. Unlike Archers or Armor Knights, no one has ever asked “how do I make my Troubadour good”. It’s the only class other than Dancer where the floor is so inherently high as to always be good on class merits alone. Even L’Arachel manages to find her way into deployment sometimes as serviceable last-slot filler. At the same time, it’s hard to make a Troubadour too good either, as the same function that lets them not care about having bad stats also results in them not really caring about having good stats either – at least, not until they promote, which tends to take a while for pure staffers to get to. Now, some hacks – and even some vanilla games, hello Jugdral – do opt to also give their Troubadours or Troubadour-equivalents access to a combat weapon even before they promote. This obviously makes an already-strong class even stronger, and I’ll circle back to it a bit more later, but generally-speaking either their combat is irrelevant and they play the same as normal, or you’re designing a Cavalier with bonus utility and should check out the Cavalier writeup by one of my counterparts.
So, if Troubadours can’t really be bad but also can’t really be broken, what else is there to say about them? A fair bit, actually. And it all starts with a question you’ve probably found yourself asking whenever you look at the deployment screen the map after the Troubadour joins: why the hell would I use the Cleric when the Troubadour exists?
Deployment of the Fittest
This is probably a comparison you’re familiar with. And while there’s a very good point to be made about how it’s perfectly okay, perhaps even healthy for some units to be designed to be used by necessity short-term and replaced by better units later, it’s a comparison many hackers prefer to avoid paralleling so starkly. Most hackers prefer for every unit they design to at least have some argument for deploying them over their competitors, and Troubadour vs Cleric is on its face about as lopsided as it can get. And since stats are so much less important to healers than they are to combat units, it can seem a lot harder to make a Cleric stand out in the face of a Troubadour competitor.
Before I continue, there is one thing I must address. Even when the Troubadour does outclass the Cleric, that does not mean the Cleric won’t be deployed. The role of healer is one that generally doesn’t need that many units to fill it, but it’s quite common for the number you want to still be at least two. Maps that force your army to split into two groups are quite common, especially in hacks, and being able to send one healer with each group is extremely valuable. If you don’t believe me, play Hag in White, where you only have one healer, and compare the progress made by the half of the army Apate goes with to that made by the other half. So, if maps demand it and deployment slots have space for it, the player will bring both the Troubadour and the Cleric even if one is worse than the other, at least in the short term. The problem starts to arise once a third healer joins the army, and a mage gains staves on promo while having way better combat, and the Troubadour always locks in the first dedicated healer slot while the Clerics get forced out.
So, let’s assume you’re one of those hackers who wants your Troubadour and Cleric to be comparably good. Luckily for you, this really isn’t actually that hard or complicated, at least on the surface. Vanilla even gets it right sometimes.
…No, not that screen, we already know that doesn’t matter and it’s not like either of these sets of numbers is good anyway. The other one.
Staff-users don’t care all that much about combat stats, but there’s one value that they care quite a lot about: staff rank. Clarine’s D may not look that far behind Saul’s C, but as anyone who’s played the game can tell you, FE6 weapon ranks grow extraordinarily slowly and that D vs C base makes a world of difference in terms of which of these two units can feasibly make it to Warp rank in time for it to matter. Saul’s base rank lead also gives him immediate access to useful non-healing staves like Barrier and Restore that Clarine can’t click, and often means he’s your first available Physic user. Physic in particular is worth some extra attention in a broader sense, because a Priest or Cleric that can use Physic to heal from farther away effectively cancels out the Troubadour’s movement advantage, though at the cost of spending what is often a more limited resource, and even once the Troubadour’s rank catches up, their usually-lower Magic stat often means their Physic range is noticeably smaller. The early Priest with a Physic staff is a pretty common sight in hacks these days for this reason.
Of course, staff rank only matters as much as the staves the player actually gets access to. It’s not enough to give the Cleric a higher staff rank than the Troubadour if the only staves you get for the first 10 chapters are Heal, Mend, and Torch. And while an early Physic can certainly help the Cleric, it may not be enough on its own when all it offers is 15 uses of “equivalent to the Troubadour” healing range with a staff that can’t be replaced until you’re 20 chapters in. For a staff rank-based advantage to actually hold up for more than a few extra chapters, there needs to be a reasonably-steady supply of higher-tier staves with which to use it, and those staves need to offer something meaningfully unique over healers who can only heal.
Lydia’s sleep staff is a prf, but the point stands. Rank-advantaged staves are basically prfs in the short term anyway.
Another option hackers will sometimes employ is giving the infantry staff-user access to a combat weapon, while the Troubadour is stuck with pure utility (or at least, with combat that is far less effective). This is also a fairly straightforward way to distinguish them, and is arguably easier to design around than differentiating on pure staff rank difference. And this is the approach I leaned into when working on chapter 4 of the currently-in-progress telephone hack An Enemy Ambush, spiritual successor to An Unexpected Caller, in which I had to design a second healer to follow up on this:
Irma here is, uh, not very good. Her stats are comparable to the notoriously-bad Serra, and on top of that, at the time, she had a negative skill dealing damage to her after every combat to represent her being a Gaiden cleric. The only things she has going for her are above-average growths (which aren’t enough to keep her fielded long enough to catch up) and her weapon ranks. The one C-rank staff available so far was Restore, for which I didn’t want to spam too many status enemies as early as chapter 4, and her stats meant her combat was so bad as to barely be a factor with the basic Light tome you had to buy for her in preps to even have. Now, I decided fairly quickly that one of the two units I wanted to add would be a second healer, and I also realized that another Priest or Cleric right after Irma joined would be both too similar and likely too much of a direct upgrade. So, instead, I decided to add a unit of the class that’s even more of an upgrade:
Pheros joins the chapter after Irma as a Troubadour with stats better than hers in every way, except, of course, for staff rank. Staff rank would matter somewhat, of course, as Pheros wouldn’t be able to use Irma’s Restore staff for a while if anyone put in status enemies in the next few maps, and I also put a shop on the map selling C-rank Physic staves as an option (though they were somewhat expensive). But more importantly, Irma, while certainly not a good combatant, at least had the option to chip down low-Res enemies safely at range and level up faster by taking kills to let her better growths catch up to Pheros’s intentionally-gimped Magic growth. But that wouldn’t be enough if her combat remained as bad as it was, especially as the power creep among other combat units was already getting out of hand.
Staff rank isn’t the only place a high weapon rank can matter. A Seraphim tome, of which she was the only current user, gave Irma a unique advantage against the exceptionally-dangerous undead on the map, allowing her combat to genuinely contribute while also helping her get a jump-start on catching up in levels. Pheros, on the other hand, did what Troubadours do best: move around with high mobility, trading items and clicking basic Heal. Both units were contributing, not just by necessity of fielding two healers, but by virtue of their own individual advantages and weaknesses.
Ride of the Valkyries
How bad can a prepromote’s stats be while still being a high tier?
Where Troubadour is the best utility class out of all tier 1 classes in the game, its promoted form, Valkyrie, may well be the best class in the game. Period. Not only does it have all the utility of its unpromoted counterpart with the boosted mobility of being promoted, Valkyries add combat on top of that, and most often, this is using the best weapon type in the game: magic. Valkyries are typically only held back by subpar stats, and while this does keep them from completely dominating, it is almost impossible for a Valkyrie to not be one of your best choices to deploy. Now, by the time Valkyries come online, healers’ stats, or lack thereof, do start to matter more. Higher-tier staves like Physic, status staves, Rescue, and Warp have ranges that scale based on the user’s Magic, and the gap in Magic between a Valkyrie and a Bishop or Sage is often large enough for the range difference to meaningfully outstrip the movement difference – especially when the target you need to reach is behind a wall, or you need to warp one of your units from a fixed starting place. On top of that, the obvious potential combat power of a mounted mage tends to also exert downward pressure on the stats Valkyries are allowed to have. Therefore, the logic with prepromote Valkyries is often to lean even more into the same things that make unpromoted Troubadours good.
Unlike unpromoted Troubadours, prepromote Valkyries are often seen with quite good base staff ranks, as they’re designed to be the ultimate out-of-the-box, no-investment-needed pure utility unit at the phase of the game where pure staffers are no longer really a thing. As stats on lategame staffers actually matter now and high staff rank is more widespread, the Valkyrie’s typically-low base Magic stat becomes the main limitation, while also keeping their combat in the “useful when it comes up but never great” range.
And that’s… kind of all there is to say on its face. Like base Troubadour, a bad Valkyrie is still a Valkyrie. Cecilia FE6 is at this point widely-known as a very good unit even if her base stats are horrendous and her base staff rank is only C and her join map has her stuck as a horse in a desert after being bodied by Zephiel in the previous cutscene. It’s just almost never going to be the case that you have a Valkyrie and they’re not worth deploying at least over whoever your 12th-best combat unit is, and the balancing act of keeping them in line is easier now that their stats are a bigger deal.
That said, if you want your prepromote Valkyrie to join a bit earlier in the game, there is also an interesting middle ground, one best highlighted by a romhack unit who isn’t literally a Valkyrie in name but is effectively one by all mechanics: AUC Selena.
FE8 Mage Knight and Valkyrie are literally the same class with different names.
Selena here is a prepromote Mage Knight who you can recruit as one of your two picks from the first interlude, which comes between chapters 5 and 6. Note her weapon ranks. Despite being a prepromote, Selena’s base stats are barely any better than her unpromoted competitors, but they are still better enough in conjunction with her 8 mov and magic access that she can function as a sort of secondary Jagen combat-wise. That A rank in Anima also lets her use the extremely-useful Bolting tome she comes with, something that will have value even as her direct combat falls off. But the main way in which Selena can maintain her relevance long-term is through building her noticeably-poor base staff rank. While E rank staves isn’t really much of a drawback at her join time, it does mean she has some work to do if she wants to use better staves later on, especially since, as a prepromote, there’s no promo gain down the line to bump it up. But on the flipside, if she is able to build up some staff rank, an 8 mov staff-user, as established, is always good, even if her stats fall off. She strikes a unique balance of short-term secondary Jagen and long-term weapon rank training project, while still leaving room for competing staffers to have a huge lead on her staff access and for competing combat units to quickly outstrip her raw bases.
The subject of FE8 Mage Knight also brings me to a topic that, while objectively rather minor and not especially popular as a mechanic among hacks, I always find fascinating: split promotion design. FE8’s split promotion for troubadours, while not as blatantly-lopsided as FE8’s version of Paladin vs Great Knight, is infamously one of the most boring choices in the game. The only differences between promoting L’Arachel to Mage Knight vs Valkyrie are a single point of Res and a choice between two weapon types where the difference barely matters. (Ok, there’s also a hidden EXP bonus on Valkyrie, but it certainly isn’t stated anywhere and may not even have been intended. And without being able to see it, you can’t really use it to make choices.) While differentiating the promotion gains a bit more would certainly help, it still doesn’t fix the problem of the classes being so fundamentally similar. They’re both promoted cavalry with staff access and a magic type, and in both vanilla and the majority of hacks, which subtype of magic you use doesn’t tend to matter very often.
Seriously, why bother?
So, what’s a more interesting choice you can make here? Well, later vanilla games seem to have learned nothing from FE8’s Paladin vs Great Knight, as Awakening gives us the absolutely laughable Valkyrie vs War Cleric split, and Fates presents the only barely better Strategist vs Butler. Wow, I sure love taking 6 mov and a physical weapon type over 8 mov and magic access on my unit with better Magic than Strength! And what really bothers me is that these games actually had the perfect promoted class to use as Troubadour’s alternate promotion: staff-wielding Falcoknights. A flying staff-user is the only thing better than a mounted one, but being stuck with physical melee lances instead of 1-2 magic tomes on your magic-focused Troubadour is a significant decrease in the combat potential the promotion provides. This promo split is what I used for Troubadours in my first ever hacking project, and unlike any of the vanilla Troubadour splits, it actually works! A tradeoff between going even harder into pure utility vs adding competent combat to their toolkit is exactly what a split promo for a utility class like Troubadour wants. This exact split of course isn’t the only way to do it, but it’s a pretty good example of what split promo choices for a class like this should present.
The Big Picture
As I was drafting this writeup, my good friend lowres, who happens to be the one writing the closely-related Cleric counterpart, pointed out something very insightful:
honestly i think troubadours lend themselves better to like, map design discussion rather than class design, as strange as that sounds. in my experience the best feeling troubadours (alta SP, that one from guard of avelon) are ones where you’re doing a lot of both healing and rescuing, and one isn’t just way better than the other
GBA Rescue (not the staff, the mechanic) is one of the most interesting parts of the base engine we hackers all use, and it’s the truly unique utility tool that Troubadour’s movement type enables that isn’t just doing a Cleric’s job either better or worse. When used well, it can generate very powerful movement advantages – pulling other units back to safety, getting slower-moving powerhouses to the front line, delivering the lord to the seize point ahead of schedule, etc.. It’s a very versatile mechanic with a lot of room for skill expression, and one that Troubadours make especially good use of as enablers on turns where their healing is either not needed or less valuable than aiding in a Rescue-Drop. But, as lowres points out, how impactful it actually is depends on the surrounding map design. And admittedly, my own map design is not always very kind to creative use of Rescue – I tend to make denser maps that lean on taxing the player’s action economy, leaving less room to contribute multiple units to proactive Rescue setups. Troubadours in such an environment are still good! They’re still Troubadours! But when they end up spending most of their turns on staff use without room for rescue, they do lose a bit of what makes them truly unique over other staffers. So if you want to lean a bit more into this aspect, you might want to have some maps with a bit more open space so that Rescue strats can really shine.