Ice Dragon

Not physically possible to have more than one positive stat.

Currently running Ryoma [+HP, -Res], +4 Robin (M) [+Spd, -Res], +9 Julia [=], Takumi [+HP, -Spd]. Ryoma can kill 2 Hectors and damage the third before going down, but that might be able to go up to 3 Hectors and damaging the fourth with Julia's Breath of Life to heal him on each turn (Julia lands the finishing blow on each Hector). If Ryoma managed to damage the last Hector before going down, Julia or Takumi can easily finish him off. Robin (M) can comfortably take out 2 Takumis without dying. Ryoma can comfortably take out 1 Takumi without dying. Takumi and Julia have no trouble taking out the last Takumi. The only possible trouble is which map each of these is on and whether I'll be able to split them up enough.

All weapons have their own sprite, and the sprites in this game are modular based on rips of the graphics assets in the game. There is no issue at all with swapping out weapons for other weapons from a graphics standpoint.

At the minimum, I'm assuming the following: Weapon skills (i.e. weapons) can only be transferred between characters of the same weapon type. Exclusive weapon skills (e.g. Falchion) cannot be transferred at all. Assist skills on colorless staff units can only be transferred to other colorless staff units. Dance and Sing cannot be transferred. Effective damage protection skills (e.g. Svalinn Shield) can only be transferred between characters of the same movement type. I'm hoping there are restrictions to certain other extremely powerful skills. As much as I would love to be able to do it, things like Lilina with Swordbreaker would be pretty insane, even without touching the counterattack skills. Perhaps only allow a character to receive a maximum of one skill (not like that would affect Lilina with Swordbreaker at all)?

The Japanese event notice has this as the second bullet point in the bottom section: Which translates as, "After this period is over, the new heroes introduced in this summoning event will continue to appear." I don't have the English text for this notice due to my language settings.

I don't see anything that hints at such. The game runs on single global server, so events and updates hit all regions at the same time. The Japanese news feeds and articles have time stamps that line up perfectly with the updates in the West.

Stats are not equivalent to each other, so losing or gaining those points in specific stats can make a large difference. For example, [+Spd] Robin (M) cannot be double attacked by Takumi without buffs or debuffs active, whereas [=Spd] Robin (M) will be double attacked by [+Spd] Takumi and [-Spd] Robin (M) will be double attacked by any Takumi that isn't [-Spd]. Similarly, [=Atk] Ryoma will never fail to kill Takumi in 2 hits, again without buffs or debuffs active and without healing, whereas [-Atk] Ryoma will fail to kill certain Takumi variants in 2 hits. Ryoma is also guaranteed to avoid being killed by Takumi in 2 hits with certain natures regardless of Takumi's nature. Yes, it is entirely possible to get a particular nature on a character that makes them suck. (Whether the particular example is one of them is up for debate; there's no need to bother with specifics here.) Furthermore, because certain stats have a +4 or -4 on a positive or negative nature, it's entirely possible that one defensive combination dies whereas another survives with 1 HP in a particular match-up simply by trading Def and HP modifiers, for example.

Cuan, Trabant, Altenna, and Areone all should have access to legendary lances that could help balance out the current dominance of red sword lords. There's also Ishtar who, if implemented, would likely be blue and carry Thorhammer. If we get Camus sometime soon as a Grand Hero Battle reward, it's possible he may have Gradivus available as his 5-star weapon (I don't recall what he had during the story, so correct me if he had some generic 5-star weapon), which I would assume would behave like a lance Falchion (10 HP regeneration every 3 turns). Light magic is currently split between blue (Aura) and green (Naga, Light Breath). Both of the green weapons are related to the divine dragons, so it's likely that "standard" light magic will end up blue, and Micaiah would probably end up here. I'd be pleasantly surprised if Thani is implemented with its effectiveness against cavalry and armor still intact. As for dealing with Takumi, Michalis is a potential future Grand Hero Battle reward, and he comes with flying mobility and Iote's Shield. I'd imagine he's no slouch in terms of Def either. It's also almost a certaintly that we'll get a larger selection of characters with access to weapons and skills effective against archers, such as the Raven magic series, the Poison Dagger, and the Bowbreaker skill.

Tharja's Raudhrblade+ weapon adds all of her buffs a second time as Atk. This includes Atk buffs. I've been using her in the training tower with Sharena, who has Rally Attack (+4 Atk) and Fortify Defense (+4 Def) for a total of +12 Atk and +4 Def, and she demolishes literally everything. You just need to watch out for enemies capable of inflicting Panic, but that goes without saying for any buff-centric team composition. Lilina has the highest Atk in the game among mages at neutral 53 (equal to Chrom even though Bolganone+ has 3 less Atk than Falchion), though her speed is rather low. I haven't used her yet, but I assume she's used as if she were a red Julia. There aren't many enemies in the arena that you'd be able to kill in a single hit, but she should have no problem softening up enemy reds and greens for your front line to clean up.

It takes like half an hour tops to hit level 20 in the tower. Do that twice. Then just toss orbs at the stamina restore until she's maxed out. Took about 6 hours and about 15 orbs to max her out.

My luckiest streaks for Lucina with the Legendary Heroes focus were 4 in 14 pulls (#564, #570, #574, #577, about $40), followed by 6 in 78 pulls (#1372, #1393, #1414, #1432, #1444, #1449, about $200). And I was trying for Tiki at the time. The streaks were almost certainly related to the fact that there were 3 red focus characters in the set, which significantly skews the odds towards focus characters when only pulling one character. If we ever have 3 of the same color focus in a single set again, you can probably manage to get a +10 with about $400-$500 or by being really, really lucky.

Confirming level 40 stats. Exactly as I predicted. I don't have her Defiant Res 3 unlocked yet.

Le'see... Linde [+Res] has 30 Res. Robin (F) has 46 Atk with weapon triangle advantage, so 16 damage per hit. Linde has 32/35/39 HP, so she'll only die if she's [+Res, -HP]. Linde [=Res] has 27 Res, so Robin (F) will deal 19 damage per hit, which will kill Linde as long as she isn't [=Res, +HP]. Robin (F) kills Linde [-Res] regardless of other stats.

Happens when you pull as much as I do. Robin (F) is guaranteed to be perfectly neutral, just like Narcian and all of the daily 1- and 2-star characters. Unless you're meaning the target characters.

Got Robin (F) to 5-star. Her base stats at all rarities are identical to Robin (M)'s. I'm going to guess that her final stats at level 40 5-star are going to be identical to Robin (M)'s also. If that's true, you're looking at 40 HP 39 Atk 29 Spd 29 Def 22 Res with Groenn Wolf equipped. We'll see.

Not staff-users since they don't really use the same weapon upgrade system as the other weapon types. If I recall correctly, staves get their second weapon at 3 stars and their third staff at 4 stars, so the only benefits of getting a 5-star staff-user is the 1 or 2 extra points to each stat and any 5-star skill unlocks.

A 4-star at level 40 is approximately 2 points lower in every stat compared to a 5-star at level 40. It takes 5 merges to gain 2 points in each stat. The 5-star weapon is 3-5 points of Atk higher than the 4-star weapon it upgrades from. You won't be able to make up for the missed Atk (you can get at most 2 more points of Atk from merges), but additional merges can eventually give you 2 points in every other stat. You obviously won't be able to equivalently make up for the loss of 5-star only skills. Every 5-star character has a new weapon because every 4-star weapon has a 5-star version with a "+" appended to the end of its name.

If anyone wants to contribute, I don't mind, but I'm not going to go out of my way to ask. As far as I'm concerned, what really matters are the level 1 stats, the level 40 stats, and any patterns I can find between them because in the end, the most important thing is "what are my stats going to look like with a maxed out character?" when given the level 1 stats from a fresh pull, and this much is not account-dependent.

Do I have to correct people every time this gets mentioned? Is the thread titled "How stats work in this game" not descriptive enough? There's no need to level up a unit to level 40 to check its stats. Looking at its stats at level 1 will tell you exactly what its stats at level 40 are. Growths are not random. They are erratic. There is a huge difference. Because growths are erratic, a character's stats at level 20 are not indicative of their stats at level 40. Unless you consult a chart that has the character's stats at all levels, there's no good way to determine a character's final stats except at level 1 (where deviations are exactly +/-1) and level 40 (where you can see the final stats in front of you).

I've seen it done by generic and named enemies in the training tower. The unit simply hops across the map during its movement. There's no reason to believe Anna's weapon effect is any different.

I'm pretty sure when you select Anna, you'll see blue squares next to all of your other units and can simply drag her to one of them. I wouldn't know from first-hand experience because my Anna ends up straight-up dying rather than being left with low HP, but that's what I'd assume considering how the AI uses the skill.

It's likely caused by rounding of some sort rather than a deliberate decision. After all, multiplying or dividing by 39, the total number of level ups, does not commonly result in nice, clean numbers. On the other hand, I'm a bit hard-pressed to find an actual set of numbers and calculation system where rounding error results in these particular values having rounding error in that direction.

You don't want to know. (And if you do, you can check my spreadsheet in my research thread. Every line in the data pool is one of my own pulls.)

You definitely want have have at least two ranged units. The fact that most units are limited to 2 movement means both you and your opponent are capable of using impassable terrain to very easily force free attacks against melee units. [=Atk] Ryoma kills all variants of Takumi in 2 hits. [+Atk] Takumi kills all variants of Ryoma in 2 hits except [=HP, +Def], but [=Atk] Takumi is only able to kill Ryoma in 2 hits if Ryoma is [-HP] or [-Def]. [+Spd] Ryoma doubles [=Spd] Takumi and cannot be doubled by [+Spd] Takumi even after Threaten Spd. [=Spd] Ryoma is doubled by [+Spd] Takumi only after Threaten Spd. [=Atk] Ryoma will always kill Takumi 1-on-1 as long as Takumi cannot double Ryoma on Takumi's first turn. Let Takumi initiate combat first so that Ryoma is unaffected by Threaten Spd until after the first round of combat. If Ryoma doubled-attacked Takumi, Takumi is already dead. If not, Ryoma is guaranteed to kill Takumi on his next attack. I haven't had time to test it out more thoroughly, but I've been having reasonable success with my current team of Ryoma [+HP, -Res], Robin [+Spd, -Res], Julia [=], and Takumi [+HP, -Spd] (eww). Robin can easily tank a hit from Takumi and retaliate for most of Takumi's health (Takumi has 6-13 HP remaining after one round of combat), making him easy to pick off with anyone else. Takumi is unable to double Robin without buffs. Ryoma's Hone Spd allows Robin to double [-Spd] Takumi. Robin deals considerable damage to most common red units (and outright kills them if he double attacks) and leaves them easily in range of being cleaned up by Takumi or Ryoma. Julia kills Nowi in 1 hit and leaves Robin with single-digit health (1-7 HP, unless Robin is [-HP, =Res] where he dies outright) while taking no damage from the counterattack. Julia leaves Linde with single-digit health (1-8 HP, unless Linde is [-Res, =HP] where she dies outright) and takes at most 8 damage in return. Takumi kills Camilla and Cordelia without risk of counterattack. Takumi can win a trade against any Takumi that is [+Atk] or [+Res] (or [+Spd] with Ryoma's Hone Spd) in any scenario where my other units are busy. Taking pot shots at enemy melee units is also nice. Ryoma is fast enough and strong enough to kill virtually any sword unit 1-on-1 and has no problem running around cleaning up the enemies that Robin and Julia fail to kill in one round. Ryoma takes little damage from Camilla on Camilla's turn and finishes her off on the following turn or can double attack Camilla and kill her outright if he initiates the first round of combat. As mentioned above, Ryoma can also win any 1-on-1 trade against Takumi. The biggest advantage of this team is the fact that every member is capable of counterattacking against a ranged enemy, meaning ranged enemies are usually going to split up predictably by color (or just beat on Ryoma and his abysmal Res) instead of by who on my team can counterattack. On the other hand, only two members are capable of counterattacking against a melee enemy, and only one member is capable of initiating combat at melee range. At the very least, Robin is usually capable of tanking a stray hit from a sword enemy. So far, the biggest challenges I've had to face are Kagero and teams with 3 or more swords. Teams with a large number of swords are troublesome because Julia is usually completely unable to participate. If the last member is a blue unit or a dragon, Julia usually needs to do a suicide run to snipe that unit unless I can somehow separate it from the rest of its team. Kagero causes problems because all of my units are infantry, only 3 of the 4 can initiate combat at range, and Kagero has high Res and often enough Spd to double my mages. I haven't fought against Tharja yet, but I'm pretty sure she'll also be troublesome to take down. Same with Abel, who can probably wreck my entire army if I let him initiate a round of combat.

The Falchion isn't an exception to the rule because it is first and foremost a +16-Atk weapon. By default, it already outclasses a Silver Sword+, which you already rated higher than type-effective weapons one paragraph earlier. In the Arena, the primary use of type-effective weapons and skills is to be prepared for specific common and otherwise potentially difficult enemies or enemy compositions. Their use, however, is limited by which enemies are common to see as well as how good the unit using the weapon or skill is when put against other enemies: The Falchion and Naga cover Tiki and Nowi. Both weapons are also powerful in their own right. The Raven series covers Takumi. Bows cover Camilla and Cordelia. Bows, being physical, ranged weapons, are useful on their own. The Swordbreaker skill covers the plethora of natural 5-star sword units, considering it is common to see two or more on a given team. In the story, they are useful because you can already know the enemy composition. In the training tower, they aren't useful unless you are lucky, but the main point of the training tower usually is to make your units useful elsewhere. Young Tiki, Lucina, and natural 5-star Marth have had the exact same drop rate as each other at any given time with any given gacha set. They are also all the same color, so there is no way to bias your pulls towards one of them by focusing on a single color. By all means, I should have similar numbers of each of them. I have so far pulled: 5 natural 5-star Marth, 8 young Tiki, and 21 Lucina, 6 of which are -HP, +Res. I have pulled more -HP, +Res Lucina than I have pulled natural 5-star Marth. I am beyond confused. I'm pretty sure the RNG just hates players and Desire Sensor is in full swing.