Fire Emblem Heroes General Discussion and Links

[mcsilas]

10 minutes ago, Vaximillian said:

Can’t fight mathematics. This is due to how rounding works.

It would be great if it's just -3 to every bane. 

Actually, to be fair @Dayni I just think the boon/bane system is rubbish. It makes finally summoning your favourites annoying when you see a bad bane. Really kills the mood.

I mean sure enough Winter Lissa's HP bane is -4....I guess we have Sacred Seals to fix them but that would mean using up precious resources.

Also it annoys me how refining weapons doesn't completely fix bad Atk/Spd stats. Although I guess that's more incentive to use Def/Res refines. Sigh, I wish I hadn't misclicked and accidentally given Nowi a +Atk refine instead of a +Res one...it would have negated her -4 bane.

[Hilda]

24 minutes ago, mcsilas said:

It would be great if it's just -3 to every bane. 

Actually, to be fair @Dayni I just think the boon/bane system is rubbish. It makes finally summoning your favourites annoying when you see a bad bane. Really kills the mood.

I mean sure enough Winter Lissa's HP bane is -4....I guess we have Sacred Seals to fix them but that would mean using up precious resources.

Also it annoys me how refining weapons doesn't completely fix bad Atk/Spd stats. Although I guess that's more incentive to use Def/Res refines. Sigh, I wish I hadn't misclicked and accidentally given Nowi a +Atk refine instead of a +Res one...it would have negated her -4 bane.

The weapon refinery isnt here to fix bad stats. You use Refinery and skill inheritance to further bolster whats great about that hero. Why would i focus my ressources on a stat that is anyway doomed to do decent, when i can instead focus my ressources on makeing the stats that shine on that character even greater and deadlier?

For example i focused whats great on Clair SPD and RES, so i bolstered her RES up to 54 on enemy phase. Another Example is Beruka, her atk and spd are shit, but her def stat oh boy.
Same goes for Lukas, he has insane DEF and poor res and spd. If you focus on their greatest strenght the heroes become a real threat.

But i do agree that the boon/bane system is rubbish in the sense thats its random, i think a boon bane system should exist, but we as the player should be able to pick ourselfs what the boon and bane of our unit should be. That way it adds another depth to unit building in my opinion.

[Punished Dayni]

22 minutes ago, mcsilas said:

Actually, to be fair @Dayni I just think the boon/bane system is rubbish. It makes finally summoning your favourites annoying when you see a bad bane. Really kills the mood.

I mean sure enough Winter Lissa's HP bane is -4....I guess we have Sacred Seals to fix them but that would mean using up precious resources.

Also it annoys me how refining weapons doesn't completely fix bad Atk/Spd stats. Although I guess that's more incentive to use Def/Res refines. Sigh, I wish I hadn't misclicked and accidentally given Nowi a +Atk refine instead of a +Res one...it would have negated her -4 bane.

Having Boons/Banes in and of itself is pretty fitting of gacha as is, and I could live with it even as free to play. But making any stat 4 can feel arbitrary at times, and for many they don't get a boon to compensate.

Alright, I forgot, after I had checked it. Whoops.

Refining weapons focusing on defences make sense with the intent of reducing offence's OPness. It's not only about fixing boons/banes.

[mcsilas]

9 minutes ago, Hilda said:

The weapon refinery isnt here to fix bad stats. You use Refinery and skill inheritance to further bolster whats great about that hero. Why would i focus my ressources on a stat that is anyway doomed to do decent, when i can instead focus my ressources on makeing the stats that shine on that character even greater and deadlier?

For example i focused whats great on Clair SPD and RES, so i bolstered her RES up to 54 on enemy phase. Another Example is Beruka, her atk and spd are shit, but her def stat oh boy.
Same goes for Lukas, he has insane DEF and poor res and spd. If you focus on their greatest strenght the heroes become a real threat.

But i do agree that the boon/bane system is rubbish in the sense thats its random, i think a boon bane system should exist, but we as the player should be able to pick ourselfs what the boon and bane of our unit should be. That way it adds another depth to unit building in my opinion.

Yeah I'm aware, it just would have been nice I guess :)

I mean at least they're taking steps to try and have different uses for different weapons that are usually overshadowed. Still need to update outclassed prfs like Sanaki's weapon. It's a start at least.

[Poimagic]

For some reason, when I imagine a new year Ryoma, I imagine him in a blue baggy kimono, with a bamboo rod as a Lance, with his description being, "He has somehow turned a piece of bamboo into a deadly weapon."

On the topic of seasonals, I'm a little mad that Camilla got one instead of Hinoka. No, I don't like Nohr, so give me some of my Hoshido boys/girls dang it.

[redlight]

At least they'll start giving out seasonal from other games once they've used up everyone in Fates Awakening. Long road ahead till I get Forblaze Eliwood.

[Ice Dragon]

2 hours ago, Dayni said:

Some have it as a balancing mechanism, sure. Rein has one for Defence, likely to prevent someone going for that as the obvious dump stat.

It's not there as a balancing mechanism at all.

It's caused by rounding because growth rates for 5-star characters are in steps of 5.7% over 39 levels (rounded down twice during the calculation) and the nature modifier results in ±1 to the level-1 stat and one step up or down in the growth rate.

You'll notice that 3-star characters never have ±4 nature modifiers because their growth rates are in steps of exactly 5%. 2-star and 1-star characters have ±2 nature modifiers mixed in instead of ±4 modifiers because their growth rates are in steps of 4.65% and 4.3%, respectively, which result in occasional smaller steps instead of occasional bigger steps (though you'll never see them in-game because no 2-star or 1-star characters have non-neutral natures).

Edited December 30, 2017 by Ice Dragon

[Humanoid]

I wonder whether if they had their time again, the devs would just go with boons without banes. That way it'd feel like a nice little bonus on top of the base unit, rather than a penalty, while still retaining the purpose of the system (that is, cynically luring them into "fixing" their unit's IVs via spending). A delta of 3 points is much more palatable than 6, which is huge in terms of effectiveness. Of course, 5 different possibilities makes it easier so technically it's a financial loss, but encouraging people to gamble on a 1-in-21 chance is kind of beyond the pale in the first place.

The rest of the game would be tuned around it of course, but the psychological effect would be rather different. It seems crazy to me that people could end up being disappointed at getting the unit they want, but that's how the current setup works. There'd still be degrees of satisfaction with any system that has any variance in stats, but hopefully the net effect is that the feeling would range from "moderately happy" through to "ecstatic".

Unfortunately it'd be extremely messy to try to retroactively implement due to the existence of neutral units. Letting people choose their boon would suck for those who settled for the second best boon and ditched a neutral unit. Doing it randomly could potentially suck for people who chose the neutral one. Oh well, something for a sequel.

[Punished Dayni]

6 minutes ago, Ice Dragon said:

It's not there as a balancing mechanism at all.

It's caused by rounding because growth rates for 5-star characters are in steps of 5.7% over 39 levels (rounded down twice during the calculation) and the nature modifier results in ±1 to the level-1 stat and one step up or down in the growth rate.

You'll notice that 3-star characters never have ±4 nature modifiers because their growth rates are in steps of exactly 5%. 2-star and 1-star characters have ±2 nature modifiers mixed in instead of ±4 modifiers because their growth rates are in steps of 4.65% and 4.3%, respectively, which result in occasional smaller steps instead of occasional bigger steps (though you'll never see them in-game because no 2-star or 1-star characters have non-neutral natures).

I then have a question: What kind of mechanism could be used to compensate for this? Could the game check for than then compensate for that point given/taken at some point in the character's levels? Or is the system for growths something that could not be modified as is?

[Ice Dragon]

5 minutes ago, Dayni said:

I then have a question: What kind of mechanism could be used to compensate for this? Could the game check for than then compensate for that point given/taken at some point in the character's levels? Or is the system for growths something that could not be modified as is?

Be careful of your question. "Can something be done?" Yes. Something can always be done. "Is it reasonable to do something about it?" Not really. The effort to implement this is not worth the gain. I have several theories on how level-up gains are determined, but nothing concrete (because I've been lazy in moving my data from screenshots to an actual spreadsheet), but none of those theoretical systems would handle adding or removing (if we want to be truly fair) an arbitrary level up point in a particularly well-behaved manner that doesn't leave the code looking like a mess and a pain in the ass to maintain. (This coming from my job as a software developer where tacked on code to handle exceptional cases is always a pain to deal with.)

[Humanoid]

Stats could easily be handcrafted per unit as each is added to the game if the devs wanted to do so. Indeed it's not unlikely that the values are just stored as fixed values per level per IV in the game files instead of being calculated on the fly, but this is speculative (at any rate, dataminers don't have access to it). In such a system it'd just be a matter of someone manually changing the value by 1 in a table.

At any rate, I think it's reasonable to say that it's not something that's left like that because it's technically difficult in any way, it's simply that the devs see no reason to "fix" it. After all, it's just another thing players will go out of their way to "optimise" by way of spending money, even though the variance is borne of accident rather than intent.

It does make me wonder though, about whether the game was originally designed around most teams *not* consisting of full 5-star units, and not balanced around level 40 play. A lot of the mechanics around the stats, particularly the trainee/veteran modifiers, only make sense in a game where the focus is on the mid-levels.

Edited December 30, 2017 by Humanoid

[Ice Dragon]

7 minutes ago, Humanoid said:

Stats could easily be handcrafted per unit as each is added to the game if the devs wanted to do so. Indeed it's not unlikely that the values are just stored as fixed values per level per IV in the game files instead of being calculated on the fly, but this is speculative (at any rate, dataminers don't have access to it). In such a system it'd just be a matter of someone manually changing the value by 1 in the table.

It's actually more likely that each growth rate value has a few "tracks" (potentially programmatically generated), and each unit just picks a specific set of tracks that match its growth rates. Software developers are lazy people. There is no reason to do something by hand 15,000 times (200 characters × 5 stats × 5 rarities × 3 nature modifiers) (You want to try?) when it can be done for you by a computer.

Telling the computer to add 1 somewhere on the track only if the nature modifier is negative and only if the nature modifier being negative results in -4 instead of -3 is a fruitless endeavor that unnecessarily complicates things.

 

14 minutes ago, Humanoid said:

At any rate, I think it's reasonable to say that it's not something that's left like that because it's technically difficult in any way, it's simply that the devs see no reason to "fix" it.

Doing special handling for exceptional one-off cases where there is no need to do so is dumb.

It's not technically difficult. It's that doing so incurs technical debt in the fact that other and future developers will need to figure out what the fuck you did there and why the fuck you had to do it and why the fuck you implemented it in that way when they go through that segment of code.

[Punished Dayni]

17 minutes ago, Ice Dragon said:

Be careful of your question. "Can something be done?" Yes. Something can always be done. "Is it reasonable to do something about it?" Not really. The effort to implement this is not worth the gain. I have several theories on how level-up gains are determined, but nothing concrete (because I've been lazy in moving my data from screenshots to an actual spreadsheet), but none of those theoretical systems would handle adding or removing (if we want to be truly fair) an arbitrary level up point in a particularly well-behaved manner that doesn't leave the code looking like a mess and a pain in the ass to maintain. (This coming from my job as a software developer where tacked on code to handle exceptional cases is always a pain to deal with.)

I was wondering, as a bit of a code idiot despite programming being a substantial part of my postgrad. It's not like exceptions being a pain in the ass is new, but I was wondering if such exceptions could be handled.

Again, it's a small thing, but it's pretty irksome imo. Any implementation kind of requires working with everything else, so I get the problem.

[Humanoid]

I'm very much a lazy software developer too, and deal with a lot of hardcoded data. The "track" system you propose is still just a straightforward table lookup, just with a many-to-one relationship. In either case, hand-tweaking the data is not a big deal - highlighting any cases where you have a difference of four instead of three is hardly sophisticated data analysis.

As for the "need" to do it - well, that's their decision on whether a marginal gain in customer satisfaction is worth it. Saying they're no need to do it is not the same as saying there's no benefit in doing it. That said, the beancounters might not even care about something as low level as this so it's possible the any decision was taken at a lower level.

[Ice Dragon]

6 minutes ago, Humanoid said:

In either case, hand-tweaking the data is not a big deal - highlighting any cases where you have a difference of four instead of three is hardly sophisticated data analysis.

Actually, it is a big deal. Every track on the lower side of a 4-point step would need to have two versions made. One for the case where the growth rate occurred from a neutral or positively modified stat and one for the case where the growth rate occurred from a negatively modified stat.

If the tracks are generated programmatically, this just results in the creation of additional branching logic that complicates things unnecessarily. And unnecessary complication is just lost time down the road.

 

9 minutes ago, Humanoid said:

and deal with a lot of hardcoded data.

Then someone is probably doing something wrong.

[Vaximillian]

This got me thinking: do people complain about +4 boons too?

[Humanoid]

Nah, I just deal with a lot of arbitrary payment rates set by legislation/bureaucracy and irregularly indexed.

Again though, you dismiss changes as "unnecessary" as a reason to not make them.  I have no issue with people who are happy with the status quo, if it's something that doesn't affect your personal satisfaction with the game, then obviously any effort whatsoever is effort you don't see worth spending. Necessary is a needlessly black and white term, but it is important to establish the differing viewpoints of whether it's *desirable*. This is important because this colours the perception of the effort spent. Is looking at level 40 stats for each rarity for each character a large amount of data? You assert it is, but I'm not so sure.

Even if the stat values were stored individually for each character with no efficiency measures, it's still a pretty reasonable volume data for low-level non-technical staff to check even manually. Even if you had just a couple of interns, and had them review the data in an inefficient format like manually reading something resembling, the wiki pages, it's something that's doable in a fairly short timeframe. Further, the maintenance aspect of the effort would be a simple eyeballing of the eight or so new units they add each month unless they raise the level cap or add additional rarities.

I don't really want to go back and forth any further. It's simply a change that I think would be nice and therefore worth the effort. For those who think the change is pointless, then it's fair enough to believe the effort would be a waste. Both natural and reasonable viewpoints and I'd be happy to leave it at that.

 

P.S. Yeah I don't like +4 boons either. One factor is just OCD, but consider also it means the unlucky people lose out on even more potential than they otherwise might. Imagine if Reinhardt had a potential 8 point swing in attack stat, and thus 16 damage per round purely because of IVs.

[Ice Dragon]

50 minutes ago, Humanoid said:

Imagine if Reinhardt had a potential 8 point swing in attack stat

Physically impossible because no two 4-point steps are ever adjacent to each other. At 4-star and 5-star rarity, any stat with +4 with a positive modifier necessarily has -3 with a negative modifier and vice versa.

[Humanoid]

Yep, acknowledge that, and I used a bad example. Just trying to say how even if a unit had a 4 point bane I wouldn't want a hypothetical 4 point boon to "balance" it, because if anything it'd feel even worse for the majority.

[Ice Dragon]

Training up Lyn was so much fun.

Spoiler

Only took 50-odd turns of Normal and 70-odd turns of Hard to get to level 39.

I might consider this setup for training other ranged units. It seemed to work pretty well (+6 Lachesis [-Def] takes zero damage on Hard from any pegasus knight hit by Atk Ploy).

[Sunwoo]

Goodness gracious. I have really mixed feelings about the 8% banner.

Last time, I got really lucky -- especially for being f2p. Hector, Celica, and Ike -- and Ike was the one I wanted most of all. This time, I didn't spend nearly as much orbs because I do want to save some to try for Takumi, but I really wanted Innes. I also told myself that I would stop pulling whenever something broke my pity rate, whether it was someone I wanted or not.

So, last night, I summoned on a colorless stone and got ... Elise. I'm not exactly that upset, since she's not a seasonal I didn't want or a character I already had. But I really, really wanted Innes. And I can think of at least one other person here who would've wanted that Elise more.

[Hilda]

I have to say Gunnthra and Arvis complement each other very well. Arvis works as a Far ranged debuffer/healer/nuker with ardent Sacrifice, while Gunnthra goes to town on the foes he debuffed. Now if i had a Lute (and give her Atk Ploy Seal) i could put together the Tri-fuckta of debuff mages lol

[Hirotune]

Im just here to ask a simple question. If you give the Wind Blessing to someone else, does the afterimage appear green? or is it a different color?

[Tybrosion]

2 hours ago, Sunwoo said:

So, last night, I summoned on a colorless stone and got ... Elise. I'm not exactly that upset, since she's not a seasonal I didn't want or a character I already had. But I really, really wanted Innes. And I can think of at least one other person here who would've wanted that Elise more.

If it makes you feel better, I haven't gotten Innes in my search for Elise. Granted, I wouldn't even be upset if he does show up since I'd quite like to have him as well but I can't even have that happen. Like the one last month, I've gotten a couple characters I didn't have going into it (SF!Chrom, SF!Lucina and Amelia from this one) so I can't be upset by the end results but I'm already prepared for the disappointment of the one I wanted the most to elude me yet again.

[Sunwoo]

I still feel bad because I know you really wanted her, while I'm not even fond of her. She's +def -HP, so far from ideal. Going to have to give her HP+5 or something, 27 HP is just lol. I think I would've preferred the reverse, +HP -def puts her at 33/16.