Hey, it’s been a while. The reason for my absence is that I’ve been working my way through the Fire Emblem series, which is quite a timesink.

I’ve been interested in strategy RPGs for a while, but they’ve rarely clicked with me, usually due to an overreliance on RNG that rather defeats the purpose of the “strategy” part. I knew that Fire Emblem was both extremely random (even stat gains are up to chance) and extremely punishing of failure, so I avoided it for a long time. However, the hype over Three Houses (the sixteenth game in the series) convinced me to try it and I quite liked what I found, so I decided to give the whole franchise a shot.

Unfortunately, as with Shin Megami Tensei, I started with the best entry in the series and experienced only disappointment thereafter. I do think there are some notable things to discuss, however.

Before getting into specific games I want to talk about the core gameplay, because it doesn’t significantly change throughout the series and does a lot of… well, let’s say interesting things.

As I just mentioned, the series is extremely dependent on randomness. Damage, at least, is fixed, but all attacks have random miss and critical rates. This is bad and stupid and I hate it. Such a mechanic might — might — be acceptable in a game where there was enough room for error that you could roll with a few unlucky punches. However, with few exceptions, Fire Emblem is not that game. As you can see in the above image, damage is incredibly high relative to characters’ HP, meaning characters frequently face risk of being two- or three-shotted. The attack in this gif had only a 28% chance to hit, yet it landed and took off more than two-thirds of the character’s health. That is an absurd variance in outcomes — either the character is completely unscathed or they’re brought to the brink of death (and likely will die shortly after, as the game loves to throw massive clusters of enemies at you). This isn’t even accounting for critical hits, which inflict a massive triple damage, almost always killing a character outright unless they are supremely tanky. And even that’s no guarantee.

And if your character dies, they’re dead, period. Permadeath is a core mechanic of Fire Emblem, so not only can bad luck hamper your progress in a map, it cripples you for the entire rest of the game by robbing you of a character you’ve invested resources into. That’s fine, you say? You’re willing to waste hours of your life grinding up a replacement character? Sorry, you can’t do that! Barring some of the more modern games, there is no grinding — you are shunted straight from one map to the next, with no stopovers. EXP, money, and items are all strictly finite resources.

This is impossible to balance. As a game designer, you must make assumptions about how lucky the player will be when designing maps and enemies. If you provide enough room for error that players can maneuver out of unlucky rolls, a lucky player experiences no challenge. If you want to make the game hard even for those who play optimally, then there’s no room for error — an unlucky roll will unravel even the best-laid plan, causing players to fail through no fault of their own even if they’re doing everything they’re supposed to do. This, again, could be manageable with unlimited resources allowing the player to recover from any setback, but the combination of extreme chaos with limited resources is just untenable.

What exactly is the player supposed to do under such constraints? Making every move accounting for the worst-case scenario — that all attacks will hit and crit if applicable — is not only impossible (the game simply does not give you enough health and resources for that), it’s a terribly boring way to play. You have to get a little lucky and the game expects you to do so. Nowhere is this more apparent than the recurring decision to give early-game enemies axes, which have high damage but low accuracy — most players will dodge everything and steamroll the enemies, but if you’re unlucky those maps are hell, because your newbie characters die in two hits.

Like random mechanics in more standard RPGs, this is a relic of tabletop and 8-bit games that couldn’t figure out how to make games difficult without making them randomly explode in players’ faces. We can do better now.

I suppose it’s a testament to the rest of the series’ design that it’s remained so enduring in spite of having fundamentally broken gameplay. Certainly, I was willing to play all of them through despite my misgivings (except one, we’ll get to that), and I… suppose I had fun? Mostly? I just wish they were better at testing my wits than my luck.

So what else does the series offer? Probably the most significant feature that distinguishes it from other strategy games is its focus on characters, making it both an RPG and a strategy game. Every unit in your army has a unique name, face, and story, with games from FE6 onwards giving them the ability to form relationships and sideplots with other characters for further development. This is generally considered the primary appeal of the series, and these sideplots are (usually) well-written.

To reinforce this, every unit is also unique in terms of gameplay… well, sort of. See, every character has unique 1) starting stats and 2) stat growth rates, as well as a fixed class (though later games allowed reclassing). Few classes are unique to a single character, so within a group of shared-class characters, the only difference is their stats. Unfortunately, this is, yet again, scuppered by RNG, because stat growths are also random. Every stat has a chance to increase on a level up; the odds are unique to every character, but thanks to the magic of RNG may never be apparent. On average, units will wind up with certain stat distributions, yes, but a unit with great growth rates can be outdone by a unit with poor growth rates just by the luck of the dice. The game does not tell you any of this, by the way, so if you want to know which of the dozens of nearly-identical characters are worth investing in, you’ve no recourse but to look up a guide. This ultimately makes most characters feel samey to the point that they may as well be generic no-names, which goes rather counter to the intended design goals.

At this point you may wonder how on Earth the designers can possibly balance a complex strategy game around RNG both in moment-to-moment actions and in determining what level of strength they have over the course of the game, plus permadeath and limited resources. The answer is… they don’t, really. Instead, they design around it such that the character mechanics don’t really matter. A difference of one or two stat points isn’t going to make or break the game (unless that stat is speed). Weapons and items have a much bigger effect on how well you can fight, and those are doled out in a deterministic fashion; additionally, permadeath is mitigated by (in most games) giving you a steady trickle of new characters, most of them quite strong, so that the designers can set a bottom line for your power level in a given map.

These are clever and subtle solutions that allow the games to remain playable in spite of their core mechanics… but the thing is, if you find yourself constantly having to make your game in spite of your core mechanics, your core mechanics are a problem. A good game, like a good story, should have all its parts working in harmony to create a product greater than the sum of its parts. “We wanted to make a character-focused strategy RPG, but we had to make the character stuff irrelevant in order to make the game playable” is the kind of nonsensical design you can only find in these ossified franchises that refuse to reconsider the design choices they made 30 years ago when video games had less computing power than a thumb drive.

Fortunately, the latest games do seem to be inching towards progress — the thirteenth game and onwards feature more deterministic ways to build your characters and even unique personal abilities to give every character a truly unique identity, which are welcome additions. Random and opaque growth rates are still a problem, however, and I just feel like reiterating it took them thirteen games to do even this much.

The other notable element of this is how you recruit new characters, which is… I think well-intentioned but not thought out very well: Occasionally a map will feature a soldier with a unique name and portrait, and that’s your cue to talk to them with the protagonist or a character relevant to them, at which point they will immediately defect and help you slaughter their former allies. If you kill them instead they have a sad death scene to remind you you messed up, unlike the faceless soldiers you can slaughter endlessly without remorse.

On the one hand, I think it’s a nice idea to encourage players to pay attention to who’s in the enemy army and pursue nonviolent options when possible. On the other hand, the very existence of a divide between the privileged named characters who get to be spared and the nameless NPCs who don’t turns this into something of a farce. (One game does actually let you recruit any enemy, but you’re clearly still expected to kill most of them.) I just can’t take these halfhearted attempts to humanize enemies after Spec Ops: The Line and UnderTale actually walked the walk. Sure, giving every single enemy a name and face like UnderTale does, even with some kind of random generator, would probably be too much to — oh wait, they did that for the generic units in Shadow Dragon? Okay never mind, they have no excuse.

The games are full of this cognitive dissonance. The stories are all about how war is terrible and the protagonists almost always state they want to minimize casualties and end the war as soon as possible, but of course the gameplay is all about murdering people for EXP so you can get the coolest weapons to murder even more people in style. War is hell, but also war is really cool. Sometimes I wish they’d give up the pretense and just let us play as the evil empire instead of maintaining the fiction that we are the good imperialists who are killing people for the right reasons.

Lastly, we of course have to talk about how the series handles women. This varies wildly by entry, so I’ll evaluate individual games too, but as a whole I found it… surprisingly not terrible? Even the first game, which came out in 1990, has an impressive number of female characters in combat roles, many of them among the best units in the game. We also get male healers, even from the beginning! While the cast almost always tilts male (of course), they make sure to include plenty of women among the massive casts. Furthermore, the stories almost always present an egalitarian world where no one feels the need to point out WOW HEY IT’S SO WEIRD YOU’RE A WOMAN WHO CAN FIGHT GOOD. (There are exceptions, which are always irritating, but less frequent than I expected.) The female outfits, too, are impressively normal and there’s rarely a significant difference in appearance between male and female versions of a class… Until the series went 3D, of course.

Unfortunately, the series does commit a very grave sin which it has now become infamous for: the 1000-year-old loli. Nearly every game has a prepubescent girl who insists you can’t treat her like a kid because she’s actually a 1000-year-old dragon (subtext: it’s totally cool if you want to bang me)! It did, however, take a while for the designers to go fully mask-off on this; prior to the 3DS games (because of course that’s where everything went wrong), none of them had romantic subplots or skeevy designs, and if you weren’t familiar with lolicon tropes you’d probably find nothing untoward about them. This surprised me given how much people talk about it, but that is probably because the later games were so disgustingly awful that it tainted the entire franchise.

That’s all for now, I think. Next week, I’ll begin with the first games in the series… or their remakes, anyway, because I don’t hate myself.