Redwall

Since you bolded perfect play, let me point out that even people (for example, Interceptor) who adopt the approach used in most of the tier lists use the term. Perhaps we need to define that term, as well. I haven't played the Tellius games, so pardon any errors I make specific to those games. My point in bringing the Titania example was to offer a context that I thought had non-trivial mutual dependence between them; since that is not the case, just change Titania to a character who does, and consider many such contexts as well as those in which the mutual dependence is negligible. If, weighted across these contexts, Jill's absence increases turn-counts moreso than the absence of any other single character, then I would consider Jill the highest-ranked character in these contexts. In response to the Mia thing, everyone capable of shoving Mia can simply receive shared credit. Finally, note that I was arguing under the assumption of a hypothetical tier list concerned with turns and reliability as the end goal, not one that adopts the approach used by Red Fox and others (enforcing an efficient playstyle, but not with the end goal of turns and reliability). Which is why I offered my own definition of "best unit" for consideration. You'll have to be clearer, since although I've showered my monitor with lemon juice and X-rays, I haven't uncovered any hidden messages. While I was writing my previous post, my only guess was that you were addressing Celes' point #6, which argues that some of Haar's contributions are not unique.

:< There's an opportunity cost, yes, but if it remains the best thing to do in a variety of contexts (for example, one may eschew Titania), then obviously you're going to give her those boosters/BEXP. Since the tier lists assume perfect play, there's no issue, save for the need to make some semantic definition of "best unit," which for the purposes of LTC-like tier lists I would personally take to mean "cuts the most turns in the ensemble of all possible contexts (females-only run, horses-only run, etc.) assuming perfect play." I don't see how this is relevant to any of the recent points brought up.

Hey now, he could be worse.

That is kind of interesting. Either a Steel or Iron Bow forge, or an Arms Scroll to get someone like Palla to use unforged Silver Bows, is necessary for a four-turn clear of C17, though AFAIK; the only reason I got a five-turn clear in my playthrough is that my Avatar and Luke were both below average in certain stats at that point. The bow forge does help throughout in all the stages with Dracos. Not sure if it's the optimal thing to do, but it's nice to have. Yes, I think that's preferable; my whole point in bringing up that Bullion (L) example was to argue that the Bullion (L) is strictly superior to the Speedwing since you can, if you wanted to (you probably shouldn't, for the reasons you mention), replicate the Speedwing's value. Yes, anyone can post.

if the awakening board sucks, then fftf swallows (just look at the influx of moved topics)

I like characters who save turns. I dislike characters who cost turns.

I see no intent to start any shipping wars. If people flame each other in arguments, then we should be calling them out on the personal attacks, not condemning them for having arguments.

Yeah, RP is assumed in all my calculations.

So again, I never believed that Ryan is flat out worse than Linde overall (quite the contrary) when considering Prologue performance; I simply think he's not as good as Linde in the stages after Ch. 5 or so. I wrote my comments in my H4 thread specifically with these subsequent stages in mind. Throughout this post, I'll assume the tier-list playstyle: H3 with Rainbow Potion and the Drops. I agree here, but I think commonguard only mentioned the Iron Bow forge since H4 makes it harder to otherwise level up Luke. But since I'm assuming H3 in this post (since that's what the tier list assumes), then yeah, Steel is better. I'm talking about the very, very late stages of the game, though--the game after and including Ch. 20. Indeed, if you give him a Speedwing, 15/1 Ryan can double Fire Dragons with decent reliability (BINOMDIST(9,14,0.5,1) gives 91%, without the Speedwing it's about 60%), but come on, even Sirius can double those clowns; these mid-game contributions of Ryan's are not unique, whereas Linde is the most reliable Nos-tank for C15 and the only character AFAIK who can get the Speedwing without forcing you to miss more than one other booster. As I have no firsthand experience with using Ryan in the lategame, I'll adopt the levels Luke had in my playthrough. We'll assume Ryan is a Horseman and gets one Speedwing and the RP throughout, but we'll treat growths using binomial calculations and assuming he has a 50% chance of gaining Spd upon a level-up (though on average Ryan's observed mean will be slightly lower due to the Hunter tree). C18 (Lv. 14/8 after C17): You need 26 AS to double the LHS Paladins as a Horseman. Ryan can take the Starsphere if Marth has the Maturity Drop. Ryan has only a BINOMDIST(8,20,0.5,1) = 25.17% chance of reaching 26 AS even after giving him a RP, Starsphere, and one Speedwing. C19: Your Falcoknight is taking the Starsphere, so Ryan needs to reach 28 AS on his own if he wants to double (which probably isn't necessary if you wipe out Sedgar's Horsemen during a single EP with a +1-Mt or better Poleax). C20 (Lv. 14/10 after C19): You need 27 AS to avoid getting doubled by Hardin. Starsphere is included here. The odds are BINOMDIST(9,22,0.5,1) = 26.17%. Ideally you would have enough to double Heroes, but that benchmark is even further out of reach (requires 30 AS). C20x (Lv. 14/13 after C20): Snipers can have up to 26 AS, so 30 AS is needed to double them (sometimes they'll have 25 AS, and the Berserkers will never have more than 25 AS). Reaching 30 AS only occurs BINOMDIST(7,25,0.5,1) = 2.16% of the time even after accounting for the Speedwing and RP. The sharp drop is explained by the fact that the Starsphere is gone by this point. C21: You need 30 AS to double Berserkers and 28 AS to double Sorcerors. C22: You need 30 AS to double the boss (Swordmaster doesn't have enough move) and clear with only one Again and one Unlock use in four or five turns. You'll have three more Speedwings by this point, though, so I suppose that's a plus. C23: Not an issue for Ryan since he can reclass to Swordmaster, I think. It was not until midway through C23 that Luke capped his level in my playthrough. Even at this capped level, Lv. 14/20 Ryan only reaches 30 AS as a Horseman with RP and one Speedwing BINOMDIST(14,32,0.5,1) = 29.80% of the time. That's bullshit, son. OK, let's look at benchmarks for Linde. I'll assume 70% Spd growth throughout (because of her promotion, Linde actually gets 75% Spd growth as a Sage, but I'll neglect this since I really don't want to do joint-binomial calculations...) C11 (Lv. 13): She needs 17 AS (we'll give her the Scorpio shard and the RP here; I think giving the Avatar the Scorpio is preferable in H4, but since this is H3, we can afford giving someone else the Scorpio) to avoid getting doubled by Flying Dragons on her way to the Speedwing. Lv. 13 Linde (really, this is sandbagging I wasn't using the RP in my playthrough, and Linde has an easier time gaining EXP with the RP than without obviously). Linde has a BINOMDIST(4,9,0.3,1) = 90.11% chance of reaching this threshold given the Scorpio and RP. She doesn't need Shaver; what she needs is Nosferatu on the enemy phase, and a Vulnerary + Physic on the subsequent player phase. C15 (didn't record stats here, but my guess is Lv. 14/5): She needs 24 AS after receiving the Starsphere and RP. If she gets no Speedwings at this level, she'll have BINOMDIST(3,14,0.3,1) = 35.51% chance of doubling. If she gets the C11 Speedwing (which isn't available [w/o giving up more than one other booster] unless you're actually using her seriously), the odds go up to 78.05%. That's bad, of course, but note that I was playing H4; in H3 her level will undoubtedly be higher because she has ~10 Nosferatu uses to snowball in the Drill Grounds beforehand (the other ~14 uses must be saved for Ch. 15). And of course the RP gives her more opportunities to level up than in my playthrough. If you think I'm pulling this out of my ass, check out some of the posts made by KoT and SDS in the tier list and in their playthroughs; they had higher-leveled Lindes than I did throughout. She's not gonna get OHKOed as you suggest, nor will she get worn down as long as she doubles. Oh, and let's compare to Mallesia (who was about Lv. 19/3 at that point), since you brought her up. Mallesia, without any Speedwings, has only a BINOMDIST(5,20,0.4,1) = 12.56% chance, and with one Speedwing, has a 41.59% chance. That's even worse. I'm not doing any handwaving here, so there's no need to accuse me of such. Ryan's just much less likely to reach the Horseman benchmarks than Linde is of reaching her own benchmarks. Ryan isn't the best Horseman, either, since Arms Scrolls aren't seriously contested and since money isn't an issue after C16 or so (assuming you grab the Bullion [L] over the Speedwing; I think you should always do this since you can simply use the money for enough arena fights to gain a Speedwing's worth of Speed anyway). In contrast, Linde is the best Nos-tank for C15 and the best way to get the C11 Speedwing.

OK. I'll address your latest post also.

Maybe we should start a new topic specifically for Ryan vs. Linde. I think we'll get input from other users as well. I agree Ryan is better, but I disagree sharply on how much better he is.

I can see Ryan being a bit better than Linde overall simply because of the Prologue stages and the earlygame up to Chapter 5 or so. Because of his terrible Spd growth, Ryan makes a terrible endgame Horseman, though, compared to Luke (and possibly Rody). You need a Horseman with capped Spd in order to double the Chapter 22 boss; 28 Spd is required for the Mage Dragons in Ch. 23; 25 Spd before the Starsphere is needed to avoid getting doubled by Hardin; you need either 28 or 29 Spd to double the Sorcerors in Ch. 21 and the Snipers/Berserkers in 20x. Ryan's 50% Spd growth isn't borderline; it is just bad, plain and simple. At important Horseman-y points in my playthrough, my Spd-cursed Luke had been averaging just over 50% observed Spd growth--something that Ryan actually needs to get lucky to achieve, since he's never going to be in a class with above 50% Spd growth and since he's only going to be above average in a given stat half the time--and still needed three Speedwings. You can't assume that everyone's going to be at or above their average every time since it's only going to happen about half the time, nor can you hand-wave that away by saying that someone like Sirius may be Spd-blessed, freeing up a Speedwing. Linde saves a turn in Ch. 15 that Mallesia can't do without multiple Speedwings or lucky level-ups; if we're talking H3, Linde shouldn't need a Speedwing since she can just snowball through the arena with Nosferatu. Also note that Linde doesn't need to OHKO Flying Dragons in Ch. 11; her goal is to get the Speedwing, so she simply needs to equip Nosferatu and have 17 AS to not get doubled by the Flying Dragons.

Someone got 28 turns with a Mercenary MU. The seven-turn clear of P8 is in the tier list. You need Ogma and Est. It's pretty much all-out offense.

I think he means that most of the turns that would be saved in a ~160-turn full-recruitment playthrough would come less from crits and more from things like ninja dodges and rigged growths (to the tune of having a 23-Def Dracoknight by Chapter 4).

Accountability matters. The problem is that dressing "like a skank" isn't immoral per se. If she has any real moral problems, then she should work to rectify those, and accept criticism accordingly. It sounds like she simply doesn't appreciate the criticism of her dress, while still believing that how she dresses is fine.

Yes

In high school, I did Science Bowl, Ocean Science Bowl, Quiz Bowl, and Science Olympiad. Oh, I guess I also played piano (poorly).

Thanks for the feedback, all. The minimum raw TC for H3 is just over 160 (KoT got 165), and I can't imagine the minimum for H4 being much different; both cases would require way more rigging than I did in this run, the goal of which was to minimize (or at least come close to minimizing) the expected turn-count. To that end, I avoided rigging crits unless they occurred on the same turn as a save-point usage or on turn 1, which followed a "real" save anyway. I suppose my C23 clear was arguably worse than a no-Rescue/Again clear performed in six turns since the expected turn-count for my C23 clear was about six turns with one Rescue and one Again use, but eh.

The ability to save lots of turns.

Thanks for making it even more annoying, by the way.

If by "max efficiency" you mean the lowest LTC with no concern for reliability, then apparently you need a lot of arena-fights (check out his Merric). If considering the tier-list playstyle--high-ish reliability, low-ish turn-counts--arena use for the purpose you alluded to (beating the unpromoted Spd cap or similar) only really benefits Catria and Caeda (and even then the benefit is small, moreso if the Rainbow Potion is being used), both of whom need only fight like once or twice each. The good-not-great units like Luke and Sirius benefit from occasional arena topping, but I think most of the funds will be spent on stat-boosters and forges.

Does Shadow Dragon have an in-base arena like FE12 does? In FE12, you can and should do something similar with Catria, allowing her to beat the unpromoted-unit Spd cap pretty easily and stockpile some growths for after her promotion by fighting in the arena as a Cav or Archer and then subsequently reclassing to Pegasus for her to do her thing in the upcoming actual level.

Probably grad stat mech. Either that or this one cooking class I took way back in middle school.

1. Bulbasaur 2. Cyndaquil 3. Torchic 4. Chimchar 5. Never played 6. Probably never will play since I sold my 3DS