Who is the worst unit in FE:A ?

[Alastor15243]

Dread Fighter Inigo?

I suppose that's one of the few instances where a faire is useful on a unit with two weapon types of differing pavgis merit, since Dread Fighter has so much less magic than strength.

[Czar_Yoshi]

Yeah, DF Inigo is a bit of a special case. Since his mother's Atk mods are both 0, he'll be underwhelming compared to the other galeboys if you try to specialize so he might as well go mixed. Kellam!Inigo is OK at it, Libra!Inigo is pretty good (I'd call him the best DF behind Stahl!Owain (Chrom is a special case, as always)).

[lordprime]

The worst unit depends on your frame of reference as well as your playstyle and goals.

The worst unit in the maingame is not the same as the worst unit in the postgame.

The worst unit in hard mode is not the same as the worst unit in lunatic mode.

[Czar_Yoshi]

So give the worst units in whatever categories you play most.

[Alastor15243]

Yeah, DF Inigo is a bit of a special case. Since his mother's Atk mods are both 0, he'll be underwhelming compared to the other galeboys if you try to specialize so he might as well go mixed. Kellam!Inigo is OK at it, Libra!Inigo is pretty good (I'd call him the best DF behind Stahl!Owain (Chrom is a special case, as always)).

I'm curious, you've done a lot of challenge runs, I'm assuming you've done no-brave runs. You clearly still have a good impression of vengeance, so how exactly did you get it to work on your no brave challenge run of apo? Because everything I've done indicates that it's inferior to the the other usual options (Luna, Ignis, Luna/Ignis, Luna/Astra, Ignis/Astra, Luna/Aether, Ignis/Aether) in pretty much every way.

Edited January 15, 2015 by Alastor15243

[eclipse]

Ehhh, I'm not the type who drags someone along with the intention of having them sit around in the back... Also, clipsey, I'm sure you're referencing the move bonus, because otherwise, Gaius's pair up bonuses would reek of boring tuna.

Nowi appreciates the Skill/Speed, moreso than the Luck that Lon'qu would provide.

[Czar_Yoshi]

I'm curious, you've done a lot of challenge runs, I'm assuming you've done no-brave runs. You clearly still have a good impression of vengeance, so how exactly did you get it to work on your no brave challenge run of apo? Because everything I've done indicates that it's inferior to the the other usual options (Luna, Ignis, Luna/Ignis, Luna/Astra, Ignis/Astra, Luna/Aether, Ignis/Aether) in pretty much every way.

Unpaired EP setups. When I see a target I want/need it for, the turn prior I drop the Vengeance user in range of a mook, Rescue the support away, let them take damage, and next turn I immediately pair up and attack the Vengeance target, heal with Fortify and have a regular pair take the mook. It does help to have a good deal of Staffbots for this.

In addition, there's a lot of damage flying around anyway because it's very rare to see a 2HKO, so it's pretty common to get 1-turn boosts that last up until healing- in that case it's less about planning where the boosts will happen and more about capitalizing on them when they do.

Oh, and I sometimes skip Rally Def for lower DG rates.

Edited January 15, 2015 by Czar_Yoshi

[Shadow Mir]

Nowi appreciates the Skill/Speed, moreso than the Luck that Lon'qu would provide.

Eh, I'm not convinced, largely because the extra skill just seems inconsequential... and it doesn't make up for the 2 less speed imo.

[Carmine Sword]

Gregor insta-promoted to Bow Knight would most likely outperform Gaius as a support in this situation.

[Czar_Yoshi]

And also hog up a rare and valuable Master Seal at the time when you've only got one and at least three other units who really want to promote (assuming Anna has been unkind).

[Alastor15243]

Unpaired EP setups. When I see a target I want/need it for, the turn prior I drop the Vengeance user in range of a mook, Rescue the support away, let them take damage, and next turn I immediately pair up and attack the Vengeance target, heal with Fortify and have a regular pair take the mook. It does help to have a good deal of Staffbots for this.

...Are we still talking about no braves? Because with no braves that's pretty much every single turn that you need proc damage to kill enemies.

[XeKr]

whoo back to this tangent

AI doesn't play FE well? Even more of a test of the unit's invincibility threshold. It would equalize quite a lot of units were the player just able to kite everything while getting them going, but that would just be more effort over the "easiest" route of zerg rushing/standing still that you then wouldn't be keeping track of.

The only thing this method really has over looking through bases and whatnot is that it produces a hard number that you can use for comparisons rather than a pile of merits/flaws that everyone weighs differently.

How are bases, growths, chapter count individually not hard numbers? How is your proposed number not something people will weigh differently?

To clarify, my point is (also) that the experiment will somewhat give you a parameter to objectively order characters by (though you do propose to count some player-controlled turns, which will then cause variations). But that’s possible for any arbitrary metric, and can be substantially easier to measure/interpret something such as say… base total multiplied by movement, then by chapter availability or something. And more objective, and less subject to rng/sample size.

Effectively the autobattle proposal could be considered (simplified) as how fast/much a “dumb player” (the AI) can complete the game given its propensity to not adapt or improve as humans would (or really do anything strategic). Again, the obvious objection for me is why rank characters, even in part, played purposely stupidly by the FE tactical AI? (not to start on strategic AI implications, like resource allocation, which autobattle doesn’t cover. Or the other stuff you briefly mentioned.)

I can understand how it sort of gets at how “easy” a character makes things, perhaps independent of strategy, etc. Maybe I just like to think that for the most part, tier lists and the ilk implicitly should tend toward the highest level of play (by one’s subjective consideration, or community consensus, blahblah). >_>

[eclipse]

Gregor insta-promoted to Bow Knight would most likely outperform Gaius as a support in this situation.

He's tied up elsewhere (uh. . .no, not literally tied up, he's making Panne super-mobile). Mostly, I'm salty because in my current run, Lon'qu gained eight levels, and is still at base Strength. In other words, all he donates are stats. At least Gaius can damage things on occasion.

[Czar_Yoshi]

...Are we still talking about no braves? Because with no braves that's pretty much every single turn that you need proc damage to kill enemies.

Yup. I don't care about weaker threats (no Braves doesn't affect my defensive capabilities), what I want to get rid of are the things that can 2HKO me or that I can't 2RKO with reasonably bad RNG.

How are bases, growths, chapter count individually not hard numbers? How is your proposed number not something people will weigh differently?

Because it's one number instead of five.

To clarify, my point is (also) that the experiment will somewhat give you a parameter to objectively order characters by (though you do propose to count some player-controlled turns, which will then cause variations). But that’s possible for any arbitrary metric, and can be substantially easier to measure/interpret something such as say… base total multiplied by movement, then by chapter availability or something. And more objective, and less subject to rng/sample size.

But is it more useful? Taking bases/stats into consideration in a vaccuum would overrepresent things like Cavs in harsh terrain, units with high Lck, and units split between physical/magical. Trying to take chapter availability into account as an individual factor would wind up giving undue credit to units who come early and fall off quickly (Anna), while not crediting useful late arrivals (Basilio and his free Rally Str). You have to take everything into account at the same time, and that's something that can't really be done on paper.

Effectively the autobattle proposal could be considered (simplified) as how fast/much a “dumb player” (the AI) can complete the game given its propensity to not adapt or improve as humans would (or really do anything strategic). Again, the obvious objection for me is why rank characters, even in part, played purposely stupidly by the FE tactical AI? (not to start on strategic AI implications, like resource allocation, which autobattle doesn’t cover. Or the other stuff you briefly mentioned.)

Adapting takes work (sure, it's part of normal gameplay and could be fun. So is feeding units. It's still effort). If you want to rank units based on how much of a load they carry- how much using them can reduce your workload- it makes sense to test them using the stupidest thing around.

I can understand how it sort of gets at how “easy” a character makes things, perhaps independent of strategy, etc. Maybe I just like to think that for the most part, tier lists and the ilk implicitly should tend toward the highest level of play (by one’s subjective consideration, or community consensus, blahblah). >_>

They usually do, but in Awakening's case nobody can really agree on what the highest level of play is- normally it would be LTC, but that grades units on a strict pass/fail basis so isn't useful for tiers. So the default is "brisk pace" which is really hard to set up strict guidelines for what counts or not (especially given how easy it is to 1-turn half the game anyway). I suppose I'm trying to do something different.

Anyway, it was just an idea. If you don't like it, that's cool.

[Ownagepuffs]

He's tied up elsewhere (uh. . .no, not literally tied up, he's making Panne super-mobile). Mostly, I'm salty because in my current run, Lon'qu gained eight levels, and is still at base Strength. In other words, all he donates are stats. At least Gaius can damage things on occasion.

A Str screwed LQ and a Spd Screwed Cordy aw man what have you done to the RNG gods.

[amadeuscho]

My Laurent always sucks for the kids. My Ricken always sucks for parents. They don't end up getting used as often for playthroughs.

[XeKr]

But is it more useful? Taking bases/stats into consideration in a vaccuum would overrepresent things like Cavs in harsh terrain, units with high Lck, and units split between physical/magical. Trying to take chapter availability into account as an individual factor would wind up giving undue credit to units who come early and fall off quickly (Anna), while not crediting useful late arrivals (Basilio and his free Rally Str). You have to take everything into account at the same time, and that's something that can't really be done on paper.

So we have the fact that I do think it’s more useful (or rather as I specifically said for a reason, easier to measure/interpret, plus more objective/less rng), while you do not, while others also presumably may think differently; this implies the obvious, the metrics have inherently subjective value anyways. Your proposal, mine, or any others will de/emphasize various mechanics that people find important or not. I think in principle for this specific proposal that it’s a lot of work to appropriately design, run, and to be statistically sound. And (imo) it doesn’t reveal anything substantially new enough as a result.

If it wasn’t clear, I don’t actually think we should rank characters that way I proposed either, I was specifically pointing out the arbitrariness.

If it matters, I’ve always favored something like how much a character impacts the reliability and speed of beating the game, optimized as best we as players can manage. Optimization in this manner, which requires substantial strategy and tactics, then encompasses the various aspects of a character in a contextual way.

Adapting takes work (sure, it's part of normal gameplay and could be fun. So is feeding units. It's still effort). If you want to rank units based on how much of a load they carry- how much using them can reduce your workload- it makes sense to test them using the stupidest thing around.

I suppose I place less value in how “easy” a character makes things, since it generally all feels very easy if you just want to beat the game.

This line of argument is a traditional one though, in FE and other communities, however I think the common complaint is how characters are not being used to their potential (or the “right way”). Arguing deliberately from ignorance is an awkward position.

They usually do, but in Awakening's case nobody can really agree on what the highest level of play is- normally it would be LTC, but that grades units on a strict pass/fail basis so isn't useful for tiers. So the default is "brisk pace" which is really hard to set up strict guidelines for what counts or not (especially given how easy it is to 1-turn half the game anyway). I suppose I'm trying to do something different.

Well this is more a philosophical issue for FE (and really ranking systems for single-player games) in general, not just Awakening.

Anyway, it was just an idea. If you don't like it, that's cool.

Just discussing the implications thereof. I presumed that since you brought it up on this public forum, it’s subject to discussion/criticism/praise. >_>

Edited January 16, 2015 by XeKr

[Czar_Yoshi]

Just discussing the implications thereof. I presumed that since you brought it up on this public forum, it’s subject to discussion/criticism/praise. >_>

Pretty much.

Using a character "the right way" in Awakening, though... Are you talking about trying to make every character as good as they can be at one specific thing, or finding each character's best use?

[XeKr]

Well that’s what all the arguments are about, right? <_<

It’s more like a very common objection people have, that the character isn’t being used correctly ("my way is the one true way", etc). Usually this means they are/n’t considered to be given some resource like (enough) exp, a statbooster, reclass, or forged weapon, etc that substantially changes their performance.

Edited January 16, 2015 by XeKr

[Shadow Knight]

Maribelle. Can't kill enemies to level up for a second seal. Lissa is a much more efficient healer, and even as a support she only increases luck. Pretty useless unless you're willing to work with her. A lot

[eclipse]

Maribelle, at base. . .can be used as a rescuebot to one-turn a certain chapter featuring a tree, stupid reinforcements, and a boss whose name also appears in the Soul Caliber series. That's more than a lot of other base units can say. ;/

[Czar_Yoshi]

Maribelle. Can't kill enemies to level up for a second seal. Lissa is a much more efficient healer, and even as a support she only increases luck. Pretty useless unless you're willing to work with her. A lot

I look at it as "can level up without killing enemies", which is a pretty awesome ability to have if exp is tight.

And no, Troubadours increase Mag, Res and Spd. Not Lck.

Edited January 16, 2015 by Czar_Yoshi

[isetrh]

One other unit who seems sort of bad is Tiki. She shows up at a time when you've likely already recruited some kids, who have much better growth potential than her. Plus, she's ridiculously hard to recruit. PEMN, but Tiki was harder for me to recruit on Normal than Donnel was on Lunatic+.

[Interceptor]

Ricken is the worst for me, slightly edging out Donnel.

Most units have some sort of niche or saving grace that allows them to contribute, but Ricken is just all sorts of blah for everything that I need from a Fire Emblem character. He's weak, his class set is weird, and he shows up at an awkward time.

[Ownagepuffs]

At the very least Ricken shows up with Wyverns and he can do somewhat notable damage with Elwind. Or Robin can take his Elwind. He can hit res for respectable chip and he can be a pair up bot for your staffers with his free +mag as well. The dude has more going for him than Donny I'd say (even if it isn't much).

Donny is literally a net loss on your team.

Edited January 16, 2015 by Ownagepuffs