Fire Emblem Heroes General Discussion and Links

[property of nuvelle]

Hey, how long does the double SP last? I could probably find out myself, but I'll ask here first.

1 hour ago, Lushen said:

Anyone else praying that we get some orbs for the glitch?  This one is very minor, and affects even less people than the previous one.  But...still... Need more orbs.

Oh? What's the glitch this time?

[MrSmokestack]

1 minute ago, KongDude said:

Oh? What's the glitch this time?

Ploy Skills. If you use them and return to the title screen, they stop working unless you restart the app.

There's a message on the announcements board for it.

Edited June 30, 2017 by MrSmokestack

[Troykv]

3 hours ago, SatsumaFSoysoy said:

Maybe that HP Def 2 is worth keeping on just for the 4 HP, instead of something like Fury.

When a man with no shirt takes magic better than you, you probably should rethink your life. Perhaps that bib is too flammable.

-HP is lol. As a melee unit she can't really deal with even lower HP. I guess you'll have to have her be some sort of glass cannon, maybe with a Brave Lance, because she can't really make use of her good mixed bulk with just 32 HP.

 

3 hours ago, AbsoluteZer0Nova said:

That's good. +Atk is always good and because it's -HP Fury would work great with her and it goes nicely with Deft Harpoon+'s ability. I would give her Renewal so that Deft Harpoon+'s ability stays in tact and just have someone on your team with Ardent Sacrifice just in case as back up. She would have during combat, 54 Atk, 39 Spd, 33 Def and 34 Res. 

Eh...   I'm not sure XD 

[property of nuvelle]

1 minute ago, MrSmokestack said:

Ploy Skills. If you use them and return to the title screen, they stop working unless you restart the app.

Huh. I don't have any units with ploy skills so I would have never noticed. Well let's hope orbs comes out of it then. 

[Othin]

2 hours ago, Ice Dragon said:

Eh? Luke was easily (and by far) the least likely character to be dropped to 4-star rarity because of Panic Ploy. I'm not sure why anyone would have expected Luke to be one of the two dropped to 4-star.

 

Katarina was by far the least likely character to get dropped, because she was the head of the banner and the first one on the trailer. People were also making a lot of fuss about Firesweep and Wo Dao because they'd only been on 5*s so far. I can see the logic behind Panic Ploy, but it struck me as the least obvious concern.

[Lushen]

The update drops 7/5 while daily orbs for summer banner end 7/6.  This pretty much confirms we will not get an extension for our daily orbs and they will go away for a time.  Maybe they'll come back for Tempest.

[Ice Dragon]

18 minutes ago, Othin said:

Katarina was by far the least likely character to get dropped, because she was the head of the banner and the first one on the trailer.

And I'm still not convinced that that argument is not just people looking for patterns to put their minds at ease. All of the first characters on their respective trailers have other compelling reasons why they wouldn't be dropped to 4-star.

And preempting anyone bringing it up, 7 for 7 means little, statistically.

Edited June 30, 2017 by Ice Dragon

[Othin]

1 minute ago, Ice Dragon said:

And I'm still not convinced that that argument is not just people looking for patterns to put their minds at ease.

And preempting anyone bringing it up, 7 for 7 means little, statistically.

If you flip 7 coins, the odds that they are all heads is 1/128 and the odds that they are all tails is 1/128, so the odds that they are all the same one way or the other is 1/64. If you're talking about a statistical perspective, 1 in 64 is well within the typical 1 in 20 test for statistical significance. That's not to say appearing at the start of a trailer is necessarily 100% odds, but it's certainly evidence in favor of a character staying at 5*, and it's backed up by more than anything else you say is a predictor, like weapons.

[Lushen]

9 minutes ago, Othin said:

If you flip 7 coins, the odds that they are all heads is 1/128 and the odds that they are all tails is 1/128, so the odds that they are all the same one way or the other is 1/64. If you're talking about a statistical perspective, 1 in 64 is well within the typical 1 in 20 test for statistical significance. That's not to say appearing at the start of a trailer is necessarily 100% odds, but it's certainly evidence in favor of a character staying at 5*, and it's backed up by more than anything else you say is a predictor, like weapons.

Not really.  There's another case in which one thing can cause two things to happen.  That does not mean that those two things are related, however.  Katarina is also a very major character in her game, as are all the ones that have stayed 5*.  It could be said she appeared first because she was the most significant and she is still a 5* because she is the most significant.  If this is true, 100/100 would not be enough to prove they are related, but rather that they are likely to both occur.

So yea, I suppose 7/7 improves the ODDS of a them both happening, but it does not mean there is an actual correlation between them.  Rather, they could both have a common relational factor.

Edited June 30, 2017 by Lushen

[Othin]

If the most prominent character appears first in the trailer and the most prominent character stays 5* exclusive, the character who appears first in the trailer stays 5* exclusive even if there isn't any direct connection. It's a reasonable predictor.

[DehNutCase]

42 minutes ago, Othin said:

If you flip 7 coins, the odds that they are all heads is 1/128 and the odds that they are all tails is 1/128, so the odds that they are all the same one way or the other is 1/64. If you're talking about a statistical perspective, 1 in 64 is well within the typical 1 in 20 test for statistical significance. That's not to say appearing at the start of a trailer is necessarily 100% odds, but it's certainly evidence in favor of a character staying at 5*, and it's backed up by more than anything else you say is a predictor, like weapons.

If you flip 7 coins, the odds that they are HTHTHTH is also 1/128. Same for HHTTHTH. Etc. Etc.

Edited June 30, 2017 by DehNutCase

[Othin]

1 minute ago, DehNutCase said:

If you flip 7 coins, the odds that they are HTHTHTH is also 1/128. Same for HHTTHTH. Etc. Etc.

I am aware. I'm not sure what your point is.

If you flipped a coin seven times and it came up heads every time, you would probably assume that it is weighted to come up heads all or at least most of the time, and if you flipped it again, you would also be pretty sure that it would come up heads. If it came up heads twice, then tails twice, then heads once, then tails once, then heads once, you would probably not assume that that is the result of the coin having any sort of pattern in its results and you would not think you could predict the results of future flips.

[DehNutCase]

6 minutes ago, Othin said:

I am aware. I'm not sure what your point is.

If you flipped a coin seven times and it came up heads every time, you would probably assume that it is weighted to come up heads all or at least most of the time, and if you flipped it again, you would also be pretty sure that it would come up heads. If it came up heads twice, then tails twice, then heads once, then tails once, then heads once, you would probably not assume that that is the result of the coin having any sort of pattern in its results and you would not think you could predict the results of future flips.

My point is that: The only reason a pattern like HHHHHHH or TTTTTTT points more to a 'rigged' coin than HTHTHTH or the like, is the fact that it's more likely to rig a coin to show all heads and all tails to start with. It's actually just as much evidence for a rigged coin as any other order of results* (because any order can be rigged---indeed, a rigged coin probably would be rigged to show a non-standard order that isn't all of one, or at least you'd use it sparingly. Something like using a rigged coin that shows heads 4 times in a row out of every 50 flips or so is undetectable via any low-powered test, because random noise drowns it out**).

That is, the reason the coin is more likely to be rigged when all heads or all tails happens is due to the priors saying it's more likely that a coin is rigged to be all heads or all tails than because you saw all heads and all tails.

*It's also because people suck at math and statistics, but nevermind that.

**High-powered tests would say a lot of innocent people are cheating.

 

Back to the main point of banners:

I don't actually watch the trailers that much, but, if the first character shown is always 5* after the banners end, that's still not evidence for the first cahracter always being 5*. This is because being the 1st character is also correlated with having a unique skill (Wo Dao+ for Karel, Ragnarok and Distant Counter for Celica, Fortify Dragon for Ninian, Heavy Blade for Ike, etc.) The 'unique new skill' qualification, itself heavily correlated with staying 5*, is also correlated for being the 1st in a banner:

This is because the 1st unit is the unit most people notice, so there's good reason to put a new character with new, unique skills there. As IceDragon says, the people who stayed 5* while being the first in a given banner have other reasons other than being first to stay 5*, and, indeed, those other reasons are also reasons to be first in a banner in a first place. Until a unit that isn't particularly likely to be 5*d shows up in the 1st position, and stays 5* afterwards, there's really not much evidence for being 1st meaning anything.

[Othin]

You're getting lost in the analogy. It's plausible that IS would make the lead unit in the trailer always 5*. It's not plausible that they would set up some weird pattern to confuse people.

By now, every banner character has unique new skills. But nonetheless, some banner characters are higher-profile than others. It's reasonable to say that IS will have the first unit in a trailer be the one they think players will identify as highest-profile, and it's also reasonable to say that the one they think players will identify as highest-profile is one of the ones they'll keep at 5*. The result is that a character being the first one in their trailer is a reliable signal that IS plans to make that character be one of the ones they'll keep at 5*, and the data supports this. The only way it could go wrong is if IS decides to make the first character in the trailer one they don't think is worth keeping at 5*, and that seems very unlikely.

Edited June 30, 2017 by Othin

[DehNutCase]

2 minutes ago, Othin said:

You're getting lost in the analogy. It's plausible that IS would make the lead unit in the trailer always 5*. It's not plausible that they would set up some weird pattern to confuse people.

By now, every banner character has unique new skills. But nonetheless, some banner characters are higher-profile than others. It's reasonable to say that IS will have the first unit in a trailer be the one they think players will identify as highest-profile, and it's also reasonable to say that the one they think players will identify as highest-profile is one of the ones they'll keep at 5*. The result is that a character being the first one in their trailer is a reliable signal that IS plans to make that character be one of the ones they plan to keep at 5*, and the data supports this. The only way it could go wrong is if IS decides to make the first character in the trailer one they don't think is worth keeping at 5*, and that seems very unlikely.

That is essentially what Ice and I are saying, except:

For us, it's, 5* unit > first in trailer.

For you, it's first in trailer > 5* unit.

That is, the cause and effect are flipped. The only way to prove one way or another is when:

 

A, A unit that shouldn't have stayed 5* remained 5* because they're first in trailor. (Unlikely because, under my view, a unit has to be very likely to remain 5* to be first in trailer).

or

B, A unit first in trailer gets demoted to 4* after banner passes. (Relatively unlikely, for the same reason as A, but can happen, if, for some reason, IS stuffed a banner with 4 candidates to remain 5*.)

 

If B happens then it's clear that first in trailer > 5* isn't a real thing. If A then 5* unit > first in trailer isn't a real thing.

[Ice Dragon]

13 minutes ago, Othin said:

You're getting lost in the analogy. It's plausible that IS would make the lead unit in the trailer always 5*. It's not plausible that they would set up some weird pattern to confuse people.

You're taking the explanation out of context. You were saying "if people see a pattern, it's likely that the pattern was intended." DehNutCase is saying "perceiving a pattern isn't the same as a pattern actually existing, and patterns that exist might not be as easily perceived".

Essentially, you're seeing "first character on trailer remains a 5-star", and we're seeing "good skills remains a 5-star, popular character remains a 5-star, popular characters and characters with good skills commonly get to be first character on trailer (for marketing purposes, namely to generate hype), and popular characters often get good skills".

Edited June 30, 2017 by Ice Dragon

[geozeldadude]

quick random question: when you sort your heroes by level, how do they sort heroes who are the same level and rarity? i thought it might be by total stats, but that doesn't seem to be the case, nor does it seem to be the order that they reached level 40. anyone know?

[Othin]

3 minutes ago, DehNutCase said:

That is essentially what Ice and I are saying, except:

For us, it's, 5* unit > first in trailer.

For you, it's first in trailer > 5* unit.

That is, the cause and effect are flipped. The only way to prove one way or another is when:

 

A, A unit that shouldn't have stayed 5* remained 5* because they're first in trailor. (Unlikely because, under my view, a unit has to be very likely to remain 5* to be first in trailer).

or

B, A unit first in trailer gets demoted to 4* after banner passes. (Relatively unlikely, for the same reason as A, but can happen, if, for some reason, IS stuffed a banner with 4 candidates to remain 5*.)

 

If B happens then it's clear that first in trailer > 5* isn't a real thing. If A then 5* unit > first in trailer isn't a real thing.

 

1 minute ago, Ice Dragon said:

You're taking the explanation out of context. You were saying "if people see a pattern, it's likely that the pattern was intended." DehNutCase is saying "perceiving a pattern isn't the same as a pattern actually existing, and patterns that exist might not be as easily perceived".

Essentially, you're seeing "first character on trailer remains a 5-star", and we're seeing "good skills remains a 5-star, popular characters and characters with good skills commonly get to be first character on trailer (for marketing purposes, namely to generate hype), and popular characters often get good skills".

Both of you are misunderstanding what I'm saying. The fact that I think being first in the trailer is a reliable indicator that a character will stay at 5* does not mean I think being first in the trailer causes a character to be 5*, or that trailer order is decided before which characters drop.

My chain of logic is similar to yours, but with much less independence between the factors.

• IS gives the skills they consider best to the characters they think people will be seeking most.
• The characters that stay at 5* are the ones IS believes people will most want to pull, based on both skills and popularity.
• The first character in the trailer is a character IS considers particularly iconic and well-known, who will get people excited.
• Due to being one of the best-known and iconic characters, the first character in the trailer is one IS thinks people will be particularly likely to want, probably more than any of the others.
• For this reason, IS has already given that character a skill set they consider appealing.
• Between that and their prominence, that character is naturally one of the ones they already picked to stay at 5*.

Because the criteria are so closely tied together, the character first in the trailer is inevitably also one of the ones they consider to be most deserving of staying at 5*. Even if they might not seem deserving of it based on our analysis of their skill sets, we can reasonably assume that the first character in the trailer is also one that IS would have given a skill set they consider particularly good, because IS's views on skill sets are not necessarily the same as ours but IS's views are the only ones that determine any of this. Therefore, saying "despite this character being prominent enough to be the first in the trailer, they probably won't stay 5* because their skill set isn't good enough" doesn't hold up because we can safely assume that whichever character at the start of the trailer is one to which IS gave a skill set that they do consider good enough.

What this all works out to is that it is highly implausible that a character first in the trailer would not be one of the characters IS considers worth keeping at 5*, and therefore when we see a character first in the trailer, we can reasonably infer that that character is one that IS considers worth keeping at 5*. So seeing a character first in a trailer tells us to expect that character to stay at 5*.

[XRay]

4 minutes ago, geozeldadude said:

quick random question: when you sort your heroes by level, how do they sort heroes who are the same level and rarity? i thought it might be by total stats, but that doesn't seem to be the case, nor does it seem to be the order that they reached level 40. anyone know?

Once sorted by level, it is then sorted by:
Rarity (5, 4, 3, 2, 1)
Color (Red, Blue, Green, Colorless)
Weapon (Melee, Breath, Tome; Bow, Dagger, Staff)
Movement (Infantry, Armor, Cavalry, Flier)
Origin

[Roflolxp54]

Oh boy. Squad Assault mode doesn't look fun at all. TWENTY units have to be ready to reliably take it on.

[geozeldadude]

5 minutes ago, XRay said:

Once sorted by level, it is then sorted by:
Rarity (5, 4, 3, 2, 1)
Color (Red, Blue, Green, Colorless)
Weapon (Melee, Breath, Tome; Bow, Dagger, Staff)
Movement (Infantry, Armor, Cavalry, Flier)
Origin

got it. thanks.

[ruruo]

31 minutes ago, Roflolxp54 said:

Oh boy. Squad Assault mode doesn't look fun at all. TWENTY units have to be ready to reliably take it on.

So many ._. I'm going to level up some of my 4* for it. Since they're in the Story section, I'm guessing they would have no time limit to complete. So maybe it'll okay to take the time to build up units. I'm glad that they're doing a big update for the story section. 

[Roflolxp54]

Just now, komasa said:

So many ._. I'm going to level up some of my 4* for it. Since they're in the Story section, I'm guessing they would have no time limit to complete. So maybe it'll okay to take the time to build up units. I'm glad that they're doing a big update for the story section. 

20 units, split into 5 teams.

In Squad Assault, with the information floating around...:

- 5 maps.

- Team cannot be reused after clearing a map until the challenge ends. This means that you can only use 1 team per map.

- All units must survive.

[DLNarshen]

I'm really glad I trained up all my worthwhile 4* units during the game's slow periods.  I think I'm sitting on close to 50 Lv. 40 units right now.  

[NoirCore]

5 hours ago, Lushen said:

Anyone else praying that we get some orbs for the glitch?  This one is very minor, and affects even less people than the previous one.  But...still... Need more orbs.

You appear to have an addiction to orb excuses, my good friend.

I suggest consulting a cleric about this.