XeKr

Heath is quite good, but might actually be better in the case of going really fast (which some people here value). Compare to Fiora, for example, who doesn’t really have time to level at that kind of pace and has much lower Str/Def (Spd also mitigated by promo gains and Con). But in a slower paced run she might have a significant level+supports lead. Anyways, just use him if you like him, most unpromoted units (except like Raven, or a high level Lyn mode unit with possibly statboosters. Oswin too if you like Knights) need a little bit of early training, but they're all pretty usable. I’ll also emphasize this following point because it's really the most important advice (and all you really need to make things easy): Buy Pure Water/Restore, and don’t be afraid of prepromos if/when you’re having trouble.

Firstly note that no reclassing changes things a lot. Parents are not that viable past the midgame, so you'll probably have to baby the children for a bit (unless you want to rig growths and turtle with Rallies). All the children should be viable, as long as their parents were trained to pass stat and skill inheritance. Especially note Lucina, who may be the only character that will significantly damage the final boss (reliably), so don’t neglect her or her parents. Virion is okay in regular Lunatic because he can reclass to Wyvern. If not reclassing, however, he’s not that great. Hard to level at first, little payoff (he’s a parent), and can’t pass really good skills. Bowbreaker is okay for a flying child, but that’s pretty far off in no-grind runs. The best father is probably Gregor because he can pass Sol. Generally Avatar is still a good parent because they can pass Rally Spectrum or Ignis, and Veteran means they can pass higher stats to a early joining child (you don’t have to delay their recruitment to train parents more). Also, because of how Morgan works, it’s the only way to get a Dark Mage or early Manakete (Nah comes in a late and hard paralogue) child. Note that you probably won’t get all the children in no grind runs, because it takes time to grind support points. The earlier you do get them, the easier they are to train, but they have slightly lower 20/20 potential (if not reclassing). And it's certainly worth trying to marry as many pairs as possible, for at least the Paralogue exp. From what I've noticed: tier lists are mostly dead because veterans (especially many that would post re: tier lists, are knowledgeable about the game, and/or play Lunatic) are mostly concerned with efficient play, which typically means beating the game quickly (low-ish turncounts) and reliably (high-ish success rates). Said lists by Redwall/Miikaya ( http://serenesforest.net/forums/index.php?showtopic=40676) and PKL (http://serenesforest.net/forums/index.php?showtopic=42220 ) , are fine for this purpose though ymmv. Discussion typically stagnates on these lists until someone does another somewhat more efficient run of the game and discovers something new/different/interesting they care to argue, which very rarely happens (very few people do so, less of those care, it requires a lot of effort, and generally we’re gotten good at understanding what kinds of characters are good. So just ask.). A few attempts at other tiering metrics ("good combat/durability" or "making things easy") have fizzled out. The arguments that typically spark thousands of posts of discussion/controversy are generally more tiering philosophy related, which people used to get really mad about but no one really cares to argue anymore (literally argued to death). For functional purposes, Avatar, having Veteran, is dominant. Frederick+Sumia is incredible earlygame because of his bases and Sumia giving flight+Spd, though Frederick will fall off. Olivia, Libra, Anna are great for Dance and Rescue Staves, 1 turning a huge number of mid/lategame maps or still very useful otherwise. A trained Chrom or Lucina (pass Veteran) does the final boss reliably. Chrom!Morgan gets Veteran and another good skill (possibly Galeforce). Cordelia can get Rally Speed. Or you can Nosferatu/Sol/be a dragon/Wyvern avoidstack and steamroll. EDIT: for clarity

Dominant strategies are a thing in game theory, assuming the players are rational and solely want to win/maximize utility or w/e. Fun, however, is hugely subjective and so there is no general dominant strategy (and you’re playing an AI…). This is especially relevant since you seem to be arguing players in general will deliberately and irrationally adopt what may be un-fun strategies (to them), because they’re optimal. An interesting claim… (to be clear, some players enjoy grinding or the otherwise “optimal” strats, and this doesn’t apply to them. You seem to be implying grinding is not fun, or the consequence is not fun, yet people will do it anyway. People do a lot of stuff, for a lot of reasons.) You appear to be aware that developers are primarily concerned with some audience that will buy their game; one manner of dealing with this is to add options to appeal to more people. However, there is no reason for you, or anyone, to disparage options outside your preference, if your preferred option is still present. Let me pose a question to you: Are all difficulty modes besides Normal/Casual (take other FE games, if you want) irrelevant because of “principle of dominant strategy”? Why would anyone play different difficulty modes if they can destroy everything easily in Normal/Casual? It’s there, it’s easy, so gamers should all be playing it, no? Also the point has been made, but I'll reiterate: are all characters besides Marcus/other prepromos irrelevant in FE7? Are all characters besides Seth/other prepromos irrelevant in FE8? Etc. They do steamroll the game and newer players would realize it faster if there wasn't such a long-standing "exp-hog" stigma. The existence of dominant options, in a strategic sense, does not necessarily diminish other options, particularly in other senses and particularly in single player. Also applies for casual vs. hardcore options. Re: FE8 grinding. Very few people (here) take that criticism seriously, to my knowledge. I (like to) think we try and call out people with the whole “my taste/way to play > your taste/way to play” rants. Besides, I’m seeing quite a bit of Sacred Stones love nowadays. And to be blunt, while the subjective points are your opinion, your assessment of Lunatic (in your review) is ignorant, more than anything. You don’t need Frederick to solo a quarter of the game. You don’t need to grind. Moreover, Risen encounters both are actually harder than story chapters, and scale in difficulty faster. Other grinding methods outside DLC are hugely nerfed in Lunatic(+). Veteran is indeed broken, though this ties back into the earlier points (it’s as broken as you want it. We can draw an analogy to a few of the better Jagens). Even still, it (or Avatar, or Nosferatu, or w/e) is not necessary in any capacity to reliably beat Lunatic. EDIT: Re: World map complaint. Fair enough, though note there’s tons and tons of ways (i.e. why can you arena for 100 turns when Roy is supposed to be saving Ostia and Lilina? Or when defending Ninian from the Black Fang) to break immersion in FE. Many other RPGs are also infamous for tons of sidequests, many opening just before the final dungeon/boss. It depends on one’s personal suspension of disbelief, and the generally favorable reaction to C10 (despite the dissonant “ooh-shiny-new-paralogue” sentiment) suggests it doesn’t bother most people. maybe that's b/c of the music tho.

1. No one cares about Swagger in the single player. It’s relevant in PvP (and only recently, which does suggest something…) 2. As has been mentioned, the Lunatic(+) “luck” myth is kinda silly. Yeah, insofar as FE is innately an RNG-based game, but it’s really not so much more than previous FE games. Do we just conveniently forget saving Jaffar in FE7, or FE6 C7 in general, etc? tbh Lunatic is probably less “luck” dependent than that stuff (even if you want to save everyone); very reliable clears are possible for all chapters. There’s something to be said about how variable Lunatic+ is, from just barely above Lunatic to really crazy in difficulty. Sometimes it may feel easier to reset the skill distribution than to figure something out. But again, if you try, very rarely is it impossible. So in the end it comes down to player skill if you’re dedicated (or patience, if you do choose that route).

I think you’ll find you’re in the majority opinion, but a few of the people who have investigated Lunatic+ the most have a different opinion. (incidentally, I personally find it tedious too...) EDIT: Also, I'm fairly certain most people don't like the movement range thingy in FE10. Just making you count doesn't add too, too much.

Anecdotes are not very useful. People have had their children killed. It’s not surprising some feel so strongly about this. Regardless, the brief point I wanted to make is the use of sources. It might help you and also future discussions in this forum lololol >_>. Some of the ones you (BANRYU) are dismissing are from peer-reviewed journals, often pretty good ones in their field. Take, http://pediatrics.aappublications.org/content/88/1/55.abstract. It’s an old paper, certainly, but was published in Pediatrics by prominent MDs from Penn. (You should be able to get the full text if you’re attending a university, and possibly otherwise. It has references, incidentally). It presents data directly from The Children's Hospital of Philadelphia. Meanwhile, you freely accept what media personalities (From what I can tell, Cesar Millan is this.) or bloggers tell you on the internet. Other articles cited (http://www.ncbi.nlm.nih.gov/pubmed/19302402 ) are okay, but from a lower impact factor journal, and they use an online survey for their data. Take what you will from that. Some of these things take seconds to check. Reading obviously biased viewpoints (Animal Humane Society, a “dog-bites.com”) are very useful to an extent, as long as you remember they’re biased and verify the sources yourself, for some of the more controversial and interesting points. Note, even ostensibly scientific press releases are not immune to hype/bias/sensationalism/etc, so it’s often better to go to the actual manuscript they cite and still remember to take that with a grain of salt too (can be very difficult to work through without a specific background, admittedly). Re: the point at hand. From what I’ve read, the statistics are clear, but I freely admit I didn’t read that carefully (or that much).

Interesting, I was just looking at a few boss stats and got a different impression. It seems absurdly hard (more than I remember/expected) to kill them efficiently without crits or possibly multiple bosskillers. Outside Rutger, Fir is arguably one of the best (2nd best?) at facilitating that with the high spd/crit/sword hit combination (and it’s still certainly very, very arguable). The real question is, in what context is Gonzales actually relevant? >_> Also, don’t exactly know what you mean by “super LTC”, but my general impression is that people make proposals based on their understanding of efficient play, then others accept, ignore, or argue it until some vague consensus is reached. etc

Comparison against bosses? Stat benchmarks? If Gonzales 2hkos (+Halberd/Brave or w/e) and Fir 5hkos or something…(hit rate maybe still swings this). Some claims are being put out there, “look at how great 10/1 Fir is”, with little contextual information atm. There’s a lot of superfluous info in traditional arguments, but also a lot of relevant data (that I would like to see) left out, mostly wrt probabilities. Of meeting benchmarks, executing a successful strategy, etc. It also seems like there's other data that people are interested in, that you or I may not be. This is still ostensibly a community list? >_> At any rate, (I forgot) apparently most mounts can’t rescuedrop chain Gonzales, which certainly looks to be problematic…

wait wtf are we seriously defending Fir now? It’s a good point that was brought up earlier, Fir can only rise if you accept Rutger is that good (top tier, maybe the 2nd best). Someone just needs to show how fast she levels, or otherwise convince people. I’m probably more inclined to keep Fir a bit lower atm, until such evidence is presented. I also do think she’s (clearly) better than Gonzales, but I’m curious as to a detailed counterargument. I would imagine it's something like Str lead (need less hits) vs. Skl/Spd lead (hits more often). That’s really useful for improving reliability though (predicated on how much better a status/siege rod Fa is relative to other fillers, which I’m not too familiar with.) Plus blasting dragons, etc.

I’ll throw out some numbers to see what people think. From my (very quick) calculations, level 10 Rutger (Espinosa achieved this by 8x. So the situation is you’re deciding whether to promote or not, and perhaps whether to switch to Fir or not) has BINOMDIST(5,6,0.7,1) = 88.2% chance to get at least 1 Str proc, BINOMDIST(4,6,0.7,1) = 58.0% chance to get at least 2 Str (around his average), BINOMDIST(3,6,0.7,1) = 25.6% chance to get at least 3 Str. Transparency if the calculation is not intuitive (well, it’s not to me…) and so people can verify my terrible/rushed math. My general interpretation is similar to yours. In the cases Rutger has below average stats, his lead (Weapon rank+potential promotion for Str and Crit) is most likely such that it’s still worth sticking it out with him (switch to Percival, etc later), over slowing down to train Fir. Though I suppose it’s possible you’re in that 42% chance of 9 or 10 Str Rutger who is level 10 (or 12% of 9 Str Rutger), and looking at 9 Str Fir who is level 1, and maybe it becomes debatable for some players. It does require some leniency/flexibility in strategy, admittedly, especially since Fir still needs to reach promotion/higher Sword rank. The main use of Fir, imo, would be in no-Rutger teams. (This may be relevant to Gonzales’ use as well, so I’m still certainly interested in a Fir vs. Gonzales analysis.) swords ftw

Garret doesn’t take a promotion item, has up to like 70 crit with existent Str, can take a few hits and peak/waterwalk, and base Skill/Luck comparable to Percival (Axes still suck, etc). Clearly not great, but I could see some uses. (I could see Fir being better though) I’d be fine with dropping some of the worse Mid characters down. Some others aren’t that bad (Sue, Chad, Ellen, even Fir), just outclassed. Bartre, Fa, etc also have their uses. Again, to be fair, Lilina’s “nuking” starts at 13 att, 87 hit or so (marginally improving per level). Nuking might apply better to what enemies do to her. >_> EDIT: Mainly nitpickying here, but Geese can promote and Brave stuff with some crit. Or something.

The Donnel comparison has been made before, but like Donnel, they're usable with some investment. Anyway, it typically depends on the growth potential of units you choose and your ability at efficient exp allocation and reclassing. In general, a larger team probably means several children, maybe a few with Veteran/Aptitude, and perhaps Manaketes who can all be effective with less levels or have otherwise accelerated leveling. Panne is also a character with 2nd gen-like growth potential, during the main game. If it’s your first Lunatic run or so, or using more of the 1st gen, you should consider using a smaller, more consolidated team with more filler utility units (Libra, Anna, Olivia, etc). Perhaps Chrom+Avatar, their children, then another family or 2 (or something like Tiki/Say’ri). Another thing to keep in mind is the usage of uneven combat pairs, where one does most of the fighting and the other is mostly for Pair Up bonuses. A bit weaker bonuses and Dual Strikes, but more combat units overall and you can distribute exp accordingly. Avatar with pocket Chrom and Nowi with pocket Gregor are the commonly cited examples, though I’m partial to Wyvern!Panne with pocket Assassin!Lon’qu too (or Wyvern!Lon’qu with Taguel!Panne). And if you do use children, the parents are often good Pair Up bots for them. For the most part that’s good advice, though I’ll note I trained at least 5 usable lategame units by Chapter 21 or so (all children near 20/20 in a No Second Seal challenge) and none had Veteran. In a more typical run, I probably could have added another child (and had some Veteran/Aptitude children), maybe Tiki/Say’ri, then Flavia/Basilio later and continued training Avatar (if he could reclass) for more units. At some point you’re running out of slots, due to how Pair Up works and because it’s very nice to have some Rescue Staves, Rally users (especially Spectrum and Speed), and Olivia. It does depend a bit on however much Paralogue exp you manage as well. More children is more chapters is more exp.

I mean, 12/1 Raigh has something like 95 hit, 10 AS with Nosferatu, which doubles….nothing in C13 (at a glance). Even if he somehow got a few levels he’s still isn’t doubling most things. And still bad hit, too. Does he really compare that favorably to Hugh? Mid seems like a stretch. Also, not really going to defend Geese here, as axe hit rates vs. existent enemy luck don’t mesh too well in whatever marginally usable context one might imagine. But to be fair, he’s already quite low.

Just don’t overuse Frederick too much and you should be fine. The hardest part is over (though beware Chapter 5, as mentioned). Also re: your post in the other topic which is relevant here... It’s up to you. You can have a small team (or even just Avatar) beat everything (Frederick will fall off hard), or you can use a larger team. You might need to spend some time babying other units (feeding them kills, etc) to get them to fight competently, but they’re fine once over that hump. It can be rewarding to see your weaklings eventually fighting very well after trained up (especially no-grinding runs). If you like larger teams, I recommend getting Rally Spectrum so your Avatar can support everyone.

iirc it was theorized that a trained Lugh/Lilina can obtain the Anima rank to Bolting stuff (Lugh far more likely?). So yeah, it depends on how fast you think they level. If all they do is chip (in different parts of the game), then maybe they should be ranked closer. No idea what Ray does.

All good points, I agree with you. I was mainly noting that while recent discussion has proceeded quite reasonably so far, there exist some glimmering points of contention (Zeiss, Chad, Wolt, Killer usage, etc), that have traditionally suggested some fundamental clash of understanding/philosophy (remains to be seen if people care to argue them in detail). And tbh, I'm a bit surprised Allen/Lance's severe drop is as controversial as it's been (meaning, not very), given they used to be top of top. EDIT: @Polydeuces (below): People are just discussing whatever points they find interesting it seems, as a lot of stuff has been thrown into the ring. Right now, it's mostly the Snipers, I think? I'm essentially positive this is vanilla mechanics, but shifting toward the new translations for names.

Understandable. One of the advantages imo is a slightly larger and perhaps more interesting discussion space. We can talk about what Geese could potentially do, instead of trivially nothing. I’m not so sure what others think, but I find unreliable strats quite inelegant (especially in other games, but because RNG is fixed here…) or at least so much as really slow but really reliable strats. Though for the most part, most people will draw the line somewhere. Exactly where is quite subjective and with little consensus. Therefore my view is it’s preferable have to have a statistical model naturally draw the line, in a sense. In some cases the reliable strats are identical except with more cautious play vs. the boss. In other cases, there remain existing enemies that you don’t deal with because you get there/kill boss/dance/seize so fast. Some variant strategy may be preferable. To clarify, for the most part it requires an extra turn or so, but not always, particularly as mentioned in FE6. In an (perhaps feeble) attempt to prevent further philosophy tangents, the simplest solution is to have CM clarify what he wants for his list (or debate so with him).

I always thought we were assuming/pretending that the RNG wasn’t fixed (Colonel M's recent comments also suggest this). Though it doesn’t bother me too, too much (either way) I can already imagine the potential absurdities (that I’m sure some people will bring up) For me, even if you can’t necessarily Killer crit every time, the chance to crit still noticeably improves reliability (and facilitates a faster progression in general. Sometimes much much faster). In addition, Rutger in particular has perhaps the best combination of Speed + high hit + effective/crit weaponry, so he’s by far the best bosskiller even when taking into account the actual probability of his criticals. So basically, atm the current rankings look reasonable to me, despite the fact that I probably prefer far more reliable clears than some others.

Lissa w/ Miriel and Rescue is the easiest way, or you could also have one of Ricken/Maribelle Pair into Frederick and the other Pair into Sumia on the cliff. Lon'qu or another good sword unit can hold the left. There's a few other semi-aggressive positionings that reliably work (push toward the forts), even without a strong Avatar. Turtling in the corner can also work, but it can be tricky to hold off the Wyverns properly and you may have to think more turns ahead.

Most of the relevant points in this discussion re: thieving utility were made just a few pages (though years) ago. Page 96ish. Baldrick also mentioned the crux of the issue earlier, whether treasure has inherent value beyond their use to improve the efficiency of clears (bearing in mind the investment to obtain it). I don’t think it does; it’s as arbitrary as slowing down to train and get pointlessly higher/maxed stats. Considering the low hit rate nature of FE6, personally I’m very accepting of slightly slower but far more consistent clears. Still, I wouldn’t say the majority of that early treasure Chad gets is a big deal. iirc the most valuable thing is the C7 Barrier, given a highly stringent 7 turn (other potential considerations here are the reliability, the reinforcements, and buyable keys in that chapter that might get chain traded over. dondon also theorized a potential 6 turn with growths, but to my knowledge no one has shown it, you skip more items, and the consistency is probably terrible.)

I meant to confirm if this was HM or not. And I dunno, when you have FE6 Normal Efficiency Tier List Rules in bold letters in the op and just a single passing “HM” mention on the first pages… Shanna is fine as is, imo. She facilitates efficient clears tremendously while she’s the only flier, but she’s extremely dependent on having very competent fighters around and she’s surrounded by some heavy competition (unless you want to move some of them to top also, which may crowd things. And top looks pretty good atm to me). Don’t have a strong opinion on the thieves. For the most part I agree that Astor is enough better than Chad while they are around for there to be a sizable gap. It requires substantial effort to feed Chad’s 8 base att self.

Might as well change the names to the new translations, for consistency? Also, this is NM (hard to tell from the first page alone...)? At a glance, the only thing that really bothered me about the older list was Alance, so I support the change. Not sure if they should go further down. What do the Thieves do again, compared to Shin/Tate/etc?

I see people play efficiently without the optimal team or with varying conditions all the time. Drafts, girls only, bow only, recruit all, no sacrificing, etc. They don’t matter? (that’s why I called it a ground rule thing) And people who don’t play efficiently still might care it’s “harder” to recruit Harken (or say…Xavier). We don’t really know, and I don’t think you can assume such a binary distinction. And in the cases (multiple) where lacking a mount is non-trivial (which I agree is non-trivial. As in, not having a mount matters), it doesn’t render the other distinctions between Raven and Harken absolutely irrelevant. The space of stuff that is non-trivial is not all identical (or singular). Could you please clarify? I think I understand what you’re saying but (atm) I legitimately don’t follow your logic in reaching that conclusion. Or answer this in a word, in your opinion, is FE9 Jill (or FE13 Male Avatar) top or bottom tier? I totally promise this is relevant. Harken is expected to cost 3ish turns. Harken2 is expected to cost 0 turns. Harken3 is expected to cost 100 turns. They are otherwise identical. Are they equivalent (in efficient or casual or whatever play)?

“otherwise identical” was supposed to refer to stuff like that too. Interesting though, you think that Karel is the only relevant consideration between Harken, Harken2, and Harken3? And I see waiving recruitment costs as a separate issue from Raven (and Harken) lacking a mount. And anyways, that’s essentially what I was getting at with the earlier statement (reproduced below)… Along with inserting deliberate qualifiers like “(in most teams)” and such. I spent so many words trying to make that distinction. I know some people may actually think that way (i.e. only the optimal is worth anything), but I also specifically mentioned a fluid, non-specific team composition for my own evaluation. Even if one adopts a strict framework like turn-based ones, it’s possible to consider varying contexts, and the ensemble of cases where the toptier mounts aren’t used. Raven, with his stats, is probably one of the best units from his recruitment on, in most teams. Of course so is Harken, but for much shorter time.

Yeah, it’s typically just turncount or reliability (clearly the former in this context). If I want to beat the game efficiently quickly and reliably, I’m not so sure Harken has a worthwhile role (in most teams). If I have to slow down to get Harken, that’s nearly equivalent to slowing down to train [insert growth unit] or slowing down because [some unit] can’t move or kill stuff fast enough. I realize a lot of people waive that cost however, or assume recruiting all, but it feels more self-consistent (to me) if considering it. The simplest thought experiment is to take Harken2 who is otherwise identical and suppose him to be recruited automatically at the end of the chapter regardless and suppose Harken3 was recruited at turn 100. I personally wouldn’t consider them equivalent, but by waiving the cost, one does and introduces some fuzzy ground (and that’s fine. It’s understandable to be more concerned with post-recruitment considerations. It’s a ground rule thing). EDIT: To clarify, I actually consider it more appropriate to mention both cases (i.e. [unit] has some recruitment cost, but if we ignore that he's good/bad b/c of x/y/z), as it's usually fairly easy to do so. and trivial statements are boring.