XeKr

Unit A with all the resources vs. Unit B with nothing is pretty much the classic tiering argument. From all the recent discussion, I really want to replay FE10. However until then, intuition currently says to me Jill w/ every relevant transfer > Haar > Jill as well. I feel the draft point is a good one (disclaimer: I have little firsthand knowledge about the details). If Jill could reliably cut more turns than Haar, I don’t think she would remain draftable while Haar is banned. Moreover, Jill is never considered free, unlike a bunch of other characters in various chapters. The bexp system (always 3 stat levels if using bexp) and minimum stat up system (no 0 stat levels) means it’s exceedingly complex to obtain truly precise numbers anyway. The fact that the stat levels are mostly independent is enough for me. Tbh I don’t see a clean application of multinomials here. Care to clarify? Moreover, I remember nothing from freshman statistics besides a little Bayesian silliness. It’s certainly good for Jill that she can barely avoid the 3hko threshold from 39 atk, and someone like Nolan with equal resources (better durability with Tarvos) can’t reach the 4hko mark. However, if including 32 and 41 atk tigers and 29 atk Cats, I imagine the durability advantage is not something to be glossed over. Among other points, I'm also still skeptical about her allegedly huge advantage in 3-12 and part 4 as well.

wheeeee discussion! Fair enough. I personally tend to trust the opinion of LTC players over others, as the playstyle lends itself to a better understanding of the game. However, many times, questions such as “how does an average Jill perform in efficient play” arise because no one really has any idea. Oftentimes when developing fast and reliable strats, you find yourself mapping for the absolute lowest turn strategy as well, just to assess its reliability and compare the expected turncount. The consequence is developing these strats is still very time intensive, despite the overall strategy being more lenient. Sometimes it can be immediately obvious that strategies are not reliable (rigging growths/ninja dodges). Often, it’s not a complex analysis to determine precisely how not reliable, but it does take much time to look up the data. Assuming that bexp gave levels as normal (in reality a bit worse) or Jill gets all her levels w/o bexp, 20/1 Jill with no transfers has ~64.7% chance of at least 20 Spd to double the 16 Spd Tigers. She has ~65.6% chance for at least 44 Hp given 2 Seraphs. She has ~68.1% chance of having at least 19 Def, given a Draco. The chance of all 3 occurring is 28.9%. This is exactly why averages are no good. >_> Now there is a Speedwing for Spd (don’t exactly know who else wants it, but it brings the benchmark to 97.8%). One could also relax the numbers slightly to something like 43 hp (89.1%), 18 Def (92.5%) to barely avoid the 3hko from 39 atk Tigers, but it’s still 80.5% to meet all 3 numbers. Assuming the entire chapter strat is predicated on this spd/durability benchmark, that’s already at least 1 expected turn gone. Of course, if Jill can get just a few more levels (having just 6 results in this high variability) or if transfers are assumed, she can do much better. I imagine this is still superior to the alternatives are well. This does also assume that the strategy is nearly 100% reliable given Jill is 3hko’d. I trust you that it is. Of course there are the Cats and 32 Atk Tigers chipping away as well and such. Re #5: Even if this is the optimal strategy, I don’t see Jill getting unique credit for the part 4 turns. Haar can do the part 4 stuff, just giving up the turn in 3-12. The advantage is already included in their pre-part 4 performance. In addition, are you sure Haar doesn’t have a reliability advantage in part 4? Because without Paragon, I imagine Haar has a stat lead (could easily be mistaken). From the earlier analysis, it’s apparent that the reliability constraint can quickly and often negate/mitigate single turn shaving. If neglected, the value of a single turn might be overestimated. Re: philosophy: I see the “context” point as what prevents the tier list from following the single optimal strategy. I would specify by having “in varying non-specific team compositions” in Redwall’s definition. Does Mia save that turn in Titania-free, or even Oscar free contexts (nevermind the fact she probably costs expected turns anyways since you’re losing levels on your flier)? If we assume that needing a specific team composition is bad (or, rather, statistically unlikely in the ensemble) shaving that one turn in that specific context does very little for her overall contribution. Edward is a separate issue. I honestly have no problem if he rises to the top if adopting a rigorous tiering standard, but I see the unfavorable impression that may show and so an “Edward clause” is fine too. People are developing counterarguments as well (“complexity”, lol). Personally I think it’s intuitive not all turns saved are equal. Exactly howso is still debatable. EDIT: My mistake on the Spd benchmark with the Speedwing (88.3% -> 97.8%). Point mostly still stands.

Very few can be bothered. Except Redwall. Sorry. <_< Though actually it’s just a quick single line command in Excel/Matlab/R. I imagine it’s fairly trivial to program a comprehensive database too (for stats at least), but who has time for that. >_> The hard part imo is most hardcore/LTC players rig growths/crits and such or are otherwise OCD about turns, so there remain gaps in our understanding of fast and reliable strats. Uhh, you’re proving my exact point about old ideas we used to (and clearly some still) hold? >_>. Why does it matter that Jill requires more “effort” if she gives a better payoff? The argument is that Jill uniquely saves (way) more turns overall whereas Haar can be mostly (but obviously not completely) replicated by Titania and others. Anyways, Celes is the champion of Jill being the best. I said earlier I wasn’t convinced of the reliability.

I’m really behind on reading your write-ups but regardless, really nice work. There aren’t many H4 runs documented at all, nevermind ones dedicated about low expected turns and complete (I never finished my own run either. >_>). Well deserved kudos to you. A bit curious about any other H3/H4 tier list insights (that challenge the conventions) you might have from your experiences. How would you roughly order the god/toptier?

No one (I think) doubts that Jill is very good. The original question was, however, is she the best in FE10? As in better than Haar? Even without transfers? She's pretty clearly the best in the extreme of ignoring reliability (and with transfers), as Celes' Jill centric run is significantly faster than other documented runs. From what I understand, Celes also tries to optimize reliability, when possible, though he emphasizes turncounts first. That's still new and fairly compelling evidence imo, and worth assessing. Handwaving and saying that “she gets lots of levels” is not compelling evidence btw. In my eyes though, her positioning depends on what stat benchmarks she needs to reliably perform her unique turn-shaving strats . Plus how statistically likely it is to obtain those stats. blahblahexpectedturncountetcetc. Admittedly, I seem to be one of few that subscribes to this specific line of thought. I don’t see any issue with the point that some are bringing up where she eats all the resources or takes a lot of "work". It doesn’t matter if Jill (when in play) takes all the resources as long as turncount/reliability are not sacrificed elsewhere. Notably, reliability nearly always is indeed sacrificed elsewhere, as more levels for robust stats are always useful. But we might assume Jill's unique turnshaving outweighs that. Assuming more arbitrary allocations (the classic “competing” for resources argument) tends to result in messy, unproductive and oftentime circular arguments on who “deserves” or is assumed certain resources. I always thought this thing was a “false” idea of the past, but apparently not. From the recent responses, it doesn't seem like many people are over this resource distribution issue and convinced she is the best at all. >_>.

From what I remember (I was there near the beginning of FE10 tiering on gfaqs), Jill was comparable to Nolan (upper mid or something?). Mainly mobility vs. stats. Ofc this was before stacking bexp in fliers became understood as a dominant strategy. And we spent more time debunking exp-stealing arguments than anything else. Haar was always understood to be near the best (in consensus competing with Ike and Reyson) though I always thought he was toppity top. Once tiering went into a more "efficiency" context, iirc his dominance was more accepted. Basically this. A stringent reliability constraint introduces many implications that haven't been well discussed yet (to my knowledge). In actuality, even averages typically aren't enough (it's...about >50% reliable?). Binomial distributions tell us far more important information. Jill also gets messed up by the FE10 bexp 3 stat system. This reliability concern has other wide ranging implications such as, if dumping a million bexp into FE9 Mia so she shaves that turn, I imagine that demolishes the reliability of Marcia/Jill reaching certain stat thresholds (unless FE9 enemies are weaker than I remember). It becomes a debatable tradeoff. Expected turncount ftw? btw no way I'm getting into the "efficiency" of soft reseting arguments.

Magekiller is the niche of the lucky unit that gains 2-3 levels from staffspam every chapter, and gets 500 bexp, a forged Steel Sword and +4/5 Atk/Def supports. Or something. <_<

FE9 Mia = saves a turn with bexp+shove. No mobility. No 1-2 range. FE10 Mia: Doesn't one round anything with any reliability, especially not at 1-2 range. The "counterculture" phenomenon is also interesting especially since FE13 has Frederick (who does fall off), Donnel (who, with a little investment, does severely outclass most maingame characters statwise), and internal levels (early promotion will noticeably gimp maingame stats). This Jeigan/Est/etc stuff can be played around by actual strategy/tactics, but many other players who take that kind of "expert" advice blindly might have some issues. I remember when Serra/Erk and even Mist got hype as combat units, Colm > Seth for a time, and FE10 Paladins/Fliers sucked. >_> A lot of these things also go in cycles. Titania is pitiful Jeigan -> is Titania tier -> doesn't fly Edward has the god stats -> blocks axes with his face -> pretty nice, saves dem turns There’s a few nitpicky points that have stood for a long time like Titania/Haar being the best of their respective games when it’s quite arguable (personally not completely convinced of the reliability of the other options though). I don’t think anyone actually cares to argue it though. Either one understands the underlying principles or doesn’t (or does and still doesn't care) and I think we’re no longer all young enough to rage incessantly about it. And buyable Rescue is still underhyped. Not even mentioned yet, pfft

Can we once have a positive topic, especially one that's supposed to be celebrate a nomination for an award? Seriously, FE9 is top 1-2 FE in my eyes and I like the whole series, but I always seem to be defending FE13. <_< Map design and objective variety are imo mostly irrelevant to FE9's (lolollolmarciasoloetc) gameplay "depth" vs. FE13 . I'll let others note specifically how one could invent explanations for deep and complex map design in FE13, particularly in Lunatic. And there's quite a bit of obvious stuff that can be concocted if so desired, especially if some of that rationale already mentioned for map design is "good" (cam's rebuttal is quite sufficient). I'd almost argue that FE13 being way more polished in aesthetics is more important wrt FE map design. at least for me.. What's actually relevant is Canto, Rescue and Shove vs. Pair Up, btw. In addition, I'll note that FE9 has a reputation for being a joke in difficulty and the one FE9 difficulty praised for the challenge, Maniac, is a slug fest in itself in many respects (no one does earlygame damage but Titania, silly enemy density, among other stuff). Regardless of your intention and the level of irony you're going for... -_-

I dunno, I find Lucina's cute side in her supports/dlc convos quite amusing since she's otherwise so "I-challenge-my-fate" serious. >_> Haven't been following this thread, but... Yep. Though, I actually like Cherche too so I don't mind if she wins. ^_^

The worst it actually gets in early Lunatic is like 5% Gamble crits. Hit rates on those can be greatly diminished by equipping swords as well. Getting Avatar Def/Spd screwed just means Frederick needs to be used a bit more. Admittedly, for many first-time Lunatic players, I'd still recommend resetting in that circumstance. All Lunatic chapters can be done reliably with the appropriate strategy (and it doesn’t necessarily have to be lolAvatarsolo). Chapter 1 is one where you “could” kite and use the forest terrain to retreat (in the case of Avatar stat screwed and don’t want to risk a Hammer hit), but it’s often easier to reset and just tank on a fort (as it’s probably >90% reliable). Chapter 2 is one where you could do some minor kiting with Stahl/Sully and microing who’s equipped or not (for Dual Strikes, counterattacks etc) but again it’s often easier to just reset until successful (as again it’s probably >90% success rate if tanking on the mountain/fort). Chapter 5 is definitely also a harder one as the combination of saving Maribelle/Ricken and holding off the reinforcements (especially the Wyverns) can be quite challenging until figured out. (Hint: tank more on forts) I wouldn't worry about getting others exp in Chapter 2. You can kite all the top enemies in Chapter 3 around in a circle. Paralogue 1 is one with many stationary enemies. Those 2 chapters can get some other units competent (Lon'qu, Sully, etc), and there's an easier stretch later from C7 to C11, including some more Paralogues.

What I like to do is use some core group (say Chrom/Avatar, Lon’qu/Panne, Gregor/Cherche), and their children (Morgan, Lucina, Yarne, Gerome), and a few utility people (Lissa, Libra, Anna, Olivia, Cordelia), which tends to add up to full deployment. There's plenty of exp (if well allocated) to facilitate training such an army. Even without DLCs, skirmishes, etc.

imo training Lon'qu or Panne is much easier because Avatar and Chrom are probably competent by that time (instead of just Frederick) so it's easier to feed kills to people. Lon'qu/Panne have serviceable bases (can actually double and take/dodge a hit) without diluting your early exp. Wyvern skills/stat spread combined with their speedy growth spread has some nice synergy too. But yeah, anyone is usable, you just have to be smart about feeding kills and allocating exp. EDIT: You'll be fine as long as you train some non-Frederick units.

As people have said, you way overused Frederick. Even if you did beat C6, you won’t beat the mid and lategame without training high potential/growth units. You can still beat C6 with 1 strong unit by having Frederick hold the middle chokepoint. Marth can hold the right enemies pretty easily by herself. The rest of your team (Avatar, Chrom, Lon’qu, Panne) can deal with the few left enemies. If you don’t want to restart (and don’t want to Frederick solo until Nosferatu is buyable), I’d recommend trying the aforementioned strat, then training your other characters.. You actually still have some time to train people, but get it done before C12 where there’s a minor difficulty spike. I’d recommend Veteran characters (Avatar, his/her children), Panne/Lon’qu (reclassed to Wyvern) and Tiki if/when you get her (or if you get Nowi!Morgan or something, that’ll make things easy as well). If you restart, try +Def Avatar for the easiest time.

Are you all sure non-Panne/Veteran!units can do the lategame reliably after blitzing the midgame? There are forged bows/Rexcaliburs and such. Plus, the comparison is against Frederick, who isn't that great in the optimal HM clears either (best pair up doe?). I still tend to agree; HM (typically) doesn't warrant overkill stats since basically 90% of the cast can reach tanky, onerounding steamroller status quite quickly through Pair Up/Tonics. She competes against those with better Supports, better Weapon Ranks, and do not require a SS. Unless, again, someone can show her stat lead actually matters. I could see a debate (not me though. >_>) Well, when a character has the best availability (all of the game), by far gets the best stats in every way (Veteran), by far gets skills/promo gains fastest (Veteran), access to best mobility (flight, Galeforce/Deliverer), and is in a game that encourages low-manning (Pair Up, abundance of bosskills, buyable Rescue)....yeah... Avatar (more specifically Veteran) obsoletes the cast. No paralogues doesn't cause the problem, per-say, though it does somewhat exacerbate it because exp is even scarcer and therefore more valuable.

If you want to keep the game as easy as possible, play Normal/Casual mode. Besides, at least try to train Lucina so she doesn't have to reset a bunch against the final boss. >_>

I think the general point is that +Def is better as enemy density increases, which is very relevant to Avatar-centric clears of the early chapters (where she faces way more than 2 attacks and healing is limited). Which is really the easiest way to beat Lunatic. No Avatar will fight in Prologue well (alternatively, nearly all Avatars are about equal if using the water thingy), while +Def (iirc even when a bit below average) can fight decently and essentially solo stuff starting from C1. I remember trying that with +Hp Avatar once and dying. >_> Among more fast paced clears, +Def and +Hp are probably about equal since Avatar is being babied until C5-6 or so. +Hp may even be better as helping against Mages does matter. Then again, a flimsier class Avatar (like Pegasus) needs as much Def as she can manage and Elixers/Concoctions are fairly limited early. I ran into some durability issues even with Wyvern Avatar in C5-C8. Some of that was definitely due to magic though. Versus +Spd or +Mag, they allow slightly faster snowballing at the cost of a harder time meeting certain stat benchmarks for bosses/lategame. +Spd matters somewhat as slightly below average Spd will get Avatars doubled by Myrms and such in C1 and C2. But it’s not really a big deal if you know what you’re doing, as Frederick can tank just a bit more to deal with that. Plus, +Spd will start doubling significantly faster, adding a ton of offense (very useful when it's the Frederick show on one front and Avatar/Lon'qu/Chrom/etc are desperately trying to clean up a few enemies). I remember my first Lunatic run was +Spd and it was really fun b/c enemies were really scary, instead of just fodder for Avatar to facetank through. <_<

Basically pick a god and pray in the truest sense. ^_^

Popping in here b/c I'm vaguely curious the ultimate limit of your patience how low is possible with rigging. 1. I still think +Spd b/c doubling can happen that much earlier. There's also possibility of doubling easier w/o +Spd pair ups, which is more +Atk and such anyways. +Mag or +Def feel (much) more reliable overall but w/e. Basically if something could potentially shave turns, it's +Spd, but if it's not faster, then the others are more reliable. 2. Doubt it, but admittedly I didn't try at all to go up the right side. So...maybe (with sacrifices)? The boss should be killable with the Silver Lance regardless (beware running out of uses though). That Hammer was also needed for a theorectical 2 turn of C12 (Libra and Cherche both need one), but there might be another way. Rig moar dual strikez. 3. I wouldn't bother with Panne honestly (her stats are too low until promotion which takes quite a while, her weapon rank sucks, and the C8 Second Seal costs a turn too I think). I'd feed those kills to Chrom (off the top of my head: could maybe help children bases, help Grima kill slightly, makes C12 2 turn much more reliable /w stronger S Rank Rapier Dual Strikes, etc) or Avatar (Galeforce sooner) or Cordelia (another Rescue, Rally Speed). It's possible with exceptional exp optimization that she'll provide useful contributions in the routs (or rout, as she'll probably only matter in C15), but atm I think it's unlikely. misc. C9 can probably be 3 turned, but probably killing Tharja. I doubt Sumia will be anything but a ferry for Frederick. Grimakill is indeed much easier if rigging Aether/noPavise/etc. lolC2havefun. >_>

It's not so much "rules", but that lowmanning is generally the dominant strategy for LTC in FE. And moreso in this game with Veteran/Pair Up. <_< Chapter 5 is one of the harder chapters (some would say it's the last truly "hard" chapter in Lunatic), so great job getting past it. As long as you delay promotion (reclass first) to get as much out of the fast unpromoted leveling as possible, you'll be fine. Just be prepared for the enemy promoted enemies (which show up en masse Chapter 17). For no-grind, I'd recommend using Frederick as little as possible since his stats are pretty terrible after the earlygame.

Distributing exp evenly often goes against the tenets of efficiency/LTC. >_>. Anyways, in general, there exist highly reliable and faster clears of the earlier chapters. For example I'd consider the standard for Prologue between 4-6 turns and Chapter 1 around 4-5 turns depending on just how safe you want to play. Taking more turns beyond that won't increase reliability by much. At this pace, Avatar can still snowball in time to become competent around Chapter 5-6 as Frederick begins to slow down. Trying the train others to be competent will take way longer and will likely not payoff in turns down the line. I'd highly recommend against doing LTC on a first Lunatic run. The gap between Frederick/Avatar/Veteran!Children and the others is just too massive and if you'd dedicated about relatively low turns, you'll end up skipping a lot of somewhat interesting mid/lategame/paralogue content through Rescue/Galeforce or just simply bumrushing the bosskill chapters. But for those who do want to LTC or do efficient play, some tips would include leaning on Frederick/Sumia while feeding Avatar as many kills as possible. Then lean on Avatar while feeding his/her children as many kills as possible. Pick up some movement skills early while Veteran/stats can still carry you, then go for self-healing stuff (Sol/Nosferatu) as enemies becomes ridiculous with haxxed forges and whatnot. Every character has a niche, whether chip/ferrying/Pair Up. Specifically have a long-term plan for the final boss or it's nearly impossible. There's some other tips/nuances and mapping detailed in Redwall's or my efficiency logs.

Donnel is fine for this archetype. I would argue he is the archetype done mostly right. His combat is quite impressive (relative to the non-Veteran characters) if investment is put in and there's plenty of time to do so. If there was no grinding/postgame in this particular game, or his caps/class set were better (doesn't really matter imo, but people seem to think so), and Donnel was a cute loli, I'd imagine he'd be much more popular among the general fanbase. The archetype is inevitably bad for efficient play since the very defining parameters (which typically include low bases, high growths) are mostly antithetical to fast and reliable clears. You're "supposed" to take turns to train them. And ain't nobody got time for that. >_> I'm not sure what is meant by "fixing" them as they can work well in their niche. But for Ests to be genuinely impressive, they'd have to be specifically designed that way such as having exceptionally high move/flying or something (say if Donnel was the only character who could get Galeforce as a level 10 Villager).

Basically yeah, statistically the effects are small. Where Rightful King has the most effect, activation rates are low and inherently unreliable. For higher activation rates, it has diminishing returns and matters less and less. But because of circumstance, it does tend to help out (as Morgan will likely pick up some proc skills, whether Ignis/Sol/Luna/Vengeance/Armsthrift). Probably overhyped, but it's a nice small boost to reliability. And there are niche Streetpass uses.

It’s worth noting that this is one of the few times where it can be better to not Pair Up as you may need the additional units/player phase offense to reliably kill everything in time. Also use Frederick’s player phase Silver Lance attack wisely, as it's a big chunk of damage that avoids Counter (if finishing off a unit).

Surprisingly not trivial chapters to do very reliably while maximizing exp/support points. Though I got a bit lazy in the end. <_< [spoiler=Chapter 13: 1/59] Hard to reliably get the Silver Axe without using up Olivia on Avatar, which means Panne is giving up kills. I prefer giving her more exp so no Silver Axe. Henry chips a southern Fighter with Elfire. Maribelle Rescues Tharja down a bit, who chips a different southern Fighter with Wind. Panne with Frederick kills a Fighter, then is Danced for and kills the other one. Requires 18 Spd, which she has with a Tonic and Frederick Pair Up. I wanted to try to build their support if possible, but obviously other Spd pairs could be used if necessary. Base Griffon!Cherche with a Spd Tonic doubles and weakens a left Fighter. Cordelia and then Lissa chain Rescue Chrom down, who then gets that kill. All the hit rates are very high, easily >90% disp hit so even more true hit. Libra Rescues Avatar with pocket Kellam up. She onerounds the boss cleanly with a Steel Axe, no worries about Counter. 51 atk is needed to oneround cleanly. Kellam, Tonics, and Skills give 8 att, Killer Axe (if necessary) gives 13 after weapon rank bonus, which means just 30 Str is needed or 19 str growth procs in around 36 levels. Should be very easy to meet the benchmark as Wyvern gives ridiculous Str growth, plus there are the Energy Drops available and forging if necessary. For reference, here’s Lucina’s bases. Name Class Level Hp Str Mag Skl Spd Lck Def Res Statboosters Lucina Cavelier 10/01.00 36 18 5 17 16 17 16 6 Avatar is 10/16/14.00. Panne is 10/06.15. [spoiler=Paralogue 12: 1/60] I didn’t try too hard, but I think Olivia is necessary to 1 turn this, meaning she can’t be used for getting Lucina an extra kill. Sumia ferries Gaius to the nearby Bullion chest. Panne with Lon’qu kills a Mage. Frederick chips a Myrm. Lucina with Chrom kills the Myrm. Lissa, Cordelia spam Rescue. Libra moves above Panne, Rescuing Olivia up. Avatar with Kellam flies over to Morgan and recruit him. Switch and Transfer Avatar to Morgan, who moves up, Switches and Avatar kills a Myrm. Olivia Dances for them, and Avatar flies all the way up to the boss. My Avatar cleanly onerounds the boss with a Killer Axe. In general, 54 atk is needed. Morgan, Tonic, Skills give 6. Killer Axe gives 13. Which leaves 35 Str needed. Forge and Energy Drops could be used as needed. Or could do something like rely on a Crit or Morgan’s Celica’s Gale, which is around 80% likely. 34 Spd is needed to double and Morgan/Tonics give 5. So 29 Spd, or 19/37 procs. I’m sure some Rescue shenanigans can get Avatar up there with a +Spd pairing, though it could hinder the exp gain of the others. Or could just probably delay this Paralogue if really necessary (if going for a Lucina-centric second half). Note that if passing Galeforce, Morgan’s recruitment will have to be delayed slightly. Something similar like Redwall’s run would probably be good, where taking an extra turn somewhere gets Galeforce before C15. Dark Flier obviously has no trouble 1 turning this. Avatar is 10/16/14.63. Panne is 10/06.78. Lucina is 10/02.00. [spoiler=Chapter 14: 1/61] Lucina takes Naga’s Tear. The main other candidate is Avatar who may want the Spd (she should be on pace for enough, but you never know). Though I think Lucina wants the stats now plus favoritism. Something was blocking the Mag/Res tonic shop, so I had to play around with some positionings to get it all done reliably. Notably, Lucina needed more Res to survive 2 mage attacks, so Libra needed to be available for Ward. Lissa and Cordelia have super-small Rescue range too, which restricts positionings greatly. It is surprisingly tricky with these limitations and the limited deployment slots to reliably get Lucina 2 kills, plus Panne and Morgan each a kill (plus open up a clear path to the boss). Both Lucina and Morgan should have pocket units to take advantage of Veteran. Avatar does not need a pairing and it’s just 10 exp anyways, which does free up a slot. Cordelia should be deployed to get Rescue exp. Gaius needs a helper to get to the Bullion chest. It seems possible to get the Second Seal, but gives up reliability/exp. Besides, I expect Morgan to be a short term contributor for the C15 rout and he’ll be a better fighter promoted rather than reclassed. It could potentially be useful for Paladin Frederick or something, but w/e. Anyways, Cherche chips a Knight on the left. Morgan with Henry kills that Knight. After getting Ward by Libra, Lucina with Chrom goes far left to kill a Mage (Str tonic for clean one-rounding). Cordelia Rescues Olivia a little left, who Dances for Lucina who kills another Mage. Lissa Rescues Gaius to the right a bit, who then grabs the Bullion chest. Panne with Lon’qu kills a Mage on the left ship, which clears the way for Avatar to Hammer the boss (who doesn’t have Pavise). It would probably be ideal for Panne to do this instead, but she requires a few more attacks to get to D rank Axes. There are numerous other objectives that I choose to prioritize less than giving Panne, Lucina, and Morgan as much exp as possible. Among them, building Panne/Frederick support, building Avatar/Lucina support, giving Chrom exp, and some Second Seal chests. Of these, the Avatar/Lucina support is the most worrisome, as I don’t really know when there will be time to build it. There are a lot of small optimization decisions like this that are difficult to tackle. Name Class Level Hp Str Mag Skl Spd Lck Def Res Statboosters Chrom Lord /05.96 24 8 2 10 10 8 10 1 Avatar Griffon Rider 10/16/14.83 53 39 12 30 31 27 31 16 Energy Drop Frederick Great Knight ??/12.76 40 22 2 18 18 9 20 4 Lissa Cleric /05.98 21 3 7 6 6 12 3 5 Maribelle Troubadour /05.20 19 1 5 4 7 5 3 9 Panne Wyvern Rider 10/07.35 39 24 2 17 17 12 18 4 Cordelia Falcon Knight 10/02.02 33 14 4 19 20 11 10 12 Libra War Monk ??/03.99 42 16 18 14 15 12 12 18 Lucina Cavelier 10/04.00 44 21 7 20 19 20 20 9 Naga’s Tear Morgan Tactician /10.85 32 17 12 16 16 14 14 12 Supports: Chrom: S Avatar. C Frederick, C Lucina, C Morgan Avatar: S Chrom, B Stahl. C Lissa. C Kellam, C Gaius, C Lucina, C Morgan Frederick: S Sumia. C Chrom. C Virion. Lissa: C Avatar. Panne: B Kellam, C Lon’qu Lucina: C Avatar, C Chrom, C Morgan Morgan: C Avatar, C Chrom, C Lucina Skills (Stat bonuses from skills and weapons included in the above stats.): Chrom: Dual Strike+ Avatar: Veteran, Solidarity, Strength+2, Tantivy, Deliverer Frederick: Discipline, Outdoor Fighter, Luna Lissa: Miracle Maribelle: Resistance+2 Panne: Even Rhythm, Strength+2 Cordelia: Speed+2, Relief Libra: Miracle, Healtouch Lucina: Dual Strike+, Charm, Aether, Veteran, Discipline Morgan: Veteran, Solidarity, Rightful King, Deliverer Blahblahexpoptimizationblahblah Avatar da gawd. Are we sure that’s –Luck flaw? Lucina pls. I do think I should go to Anna’s paralogue since Lucina/Panne are still kinda weak and don’t have 1-2 range. If they are both competent for C15 (ideally both will be promoted and with 1-2 range), they can start to snowball and get tons more kills/levels without babying (especially during the few chapters I’ll have to 2 turn or more like C17 or C20). It depends on how many turns P4 ends up taking. Or if I can already reliably 3 turn C15 without training I’ll maybe just move on. In addition, there’s still a few other really hard chapters later to keep in mind like C20, C23, C24, which are heavily impacted by the snowball effect. Anyways, the general skeleton plan is Avatar goes Hero after C18/19, Lucina goes GK or Paladin depending on her Spd. Panne goes Griffon. Morgan goes GM then Priest/Sage (or possibly just stay GM for the routs as Celica’s probably does some hefty chip). I’d welcome others’ thoughts on this and everything else. However, this run is now on indefinite Civ 5: BNW hiatus. Thanks for reading what’s done so far; I will definitely be coming back to it …eventually >_>.