Jump to content

Fire Emblem Heroes General Discussion and Links


eclipse

Recommended Posts

11 minutes ago, DehNutCase said:

I tried making some +10s in the summoner simulator once, and, turns out, you need about 2k orbs for a reliable 5* +10.

Yea, I arrived at a similar number when I was calculating how much money I would have to spend if Eirika and Ephraim were the summer units. . . .

 

How did you get 2k orbs?  I got 1150 orbs assuming that you get one everytime you have a 50% chance to get one (which should average).  Then again, probability was the one math class I never had to take in college so it's not my strong suit.

 

The good news is at least it gets more reliable.  If you're just going for 1, it's very unreliable whether or not you will get any but if you're going to go for 10, most likely odds will work out.

Edited by Lushen
Link to comment
Share on other sites

  • Replies 77k
  • Created
  • Last Reply

Top Posters In This Topic

  • Vaximillian

    4980

  • Anacybele

    3374

  • Ice Dragon

    3123

  • Othin

    2728

10 minutes ago, DehNutCase said:

~200 orbs is pretty reliable for getting a focus unit you want

I wish this was the case with me and the Summer banner right now.

Link to comment
Share on other sites

8 minutes ago, Alkaid said:

I wish this was the case with me and the Summer banner right now.

Pretty reliable is mathematician speak for: You're not going to get it a lot of the time.

~200 orbs is the average count, meaning getting it at 200 orbs is close to a 50/50 shot (this, because the average is forced low by the chance of multiple focus pulls very close together, which drives down the average).

 

And past that, you enter the fun world of probability. Where 400 orbs is 75%, and 800 orbs is still only 87.5%, not even 90%, to get the unit you want.

Link to comment
Share on other sites

3 minutes ago, DehNutCase said:

Pretty reliable is mathematician speak for: You're not going to get it a lot of the time.

~200 orbs is the average count, meaning getting it at 200 orbs is close to a 50/50 shot (this, because the average is forced low by the chance of multiple focus pulls very close together, which drives down the average).

 

And past that, you enter the fun world of probability. Where 400 orbs is 75%, and 800 orbs is still only 87.5%, not even 90%, to get the unit you want.

A fun world indeed. I feel the 400 range sneaking up on me already. There is no RNGesus here, only pain and suffering.

Link to comment
Share on other sites

8 minutes ago, DehNutCase said:

Pretty reliable is mathematician speak for: You're not going to get it a lot of the time.

~200 orbs is the average count, meaning getting it at 200 orbs is close to a 50/50 shot (this, because the average is forced low by the chance of multiple focus pulls very close together, which drives down the average).

 

And past that, you enter the fun world of probability. Where 400 orbs is 75%, and 800 orbs is still only 87.5%, not even 90%, to get the unit you want.

Where are you getting those numbers? If you have 200 orbs and use those to get 40 pulls of your color of choice with each pull having a 3% chance of being the unit you want, you have about a 70% chance of getting the unit. (0.97^40 = ~0.3) That's far from reliable, but it's noticeably better than 50/50. And for 800 orbs I'm getting over 99% odds. (0.97^160 = ~0.008)

Link to comment
Share on other sites

23 minutes ago, Alkaid said:

I wish this was the case with me and the Summer banner right now.

If it makes you feel any better, it took getting five off-banner 5*s (Faye, Rebecca, Jeorge, Minerva, Young!Tiki) until I got one of the summer units (Frederick). Even then, Tiki is the one that I want the most out of the four but because she's a focus character that I want, that means I'm doomed to never get her like Bridelia before her. I'm still perfectly content that I got at least one of them and hopefully one of them will find their way to you as well.

Link to comment
Share on other sites

20 minutes ago, Othin said:

Where are you getting those numbers? If you have 200 orbs and use those to get 40 pulls of your color of choice with each pull having a 3% chance of being the unit you want, you have about a 70% chance of getting the unit. (0.97^40 = ~0.3) That's far from reliable, but it's noticeably better than 50/50. And for 800 orbs I'm getting over 99% odds. (0.97^160 = ~0.008)

I napkin math'd it. I tested getting +10 unit and it's about 2k for a +10.

2k/11 ~= 2k /10 = 200.

 

I assumed 50/50 because a great deal of the expected value of low odds gambles like this is locked behind 'chain wins,' where wins come close together. This is another napkin math thing.

 

Also: Pulls aren't 3% of getting the unit you want per pull. Characters are generated on the summon orbs first, rather than after you pick a color, this affects pull odds. A, it affects odds of getting the color you want at all, b, it makes colors with large numbers of foci have high 5* chances but lower chances of pulling the focus you want, and vice versa, and c, colors without foci have less than 3% 5* odds because math, although c is irrelevant in this scenario.

 

Edit: This means: Colors with 2 foci tend to have ~10% 5* odds, colors with 1 foci tends to have ~7% 5* odds, reds tends to have more than other colors in terms of 5* odds no matter what, due to higher number of 5* characters, but also lower % chance of getting your focus character, this time due to # of characters in general.

Edited by DehNutCase
Link to comment
Share on other sites

8 minutes ago, DehNutCase said:

I napkin math'd it. I tested getting +10 unit and it's about 2k for a +10.

2k/11 ~= 2k /10 = 200.

 

I assumed 50/50 because a great deal of the expected value of low odds gambles like this is locked behind 'chain wins,' where wins come close together. This is another napkin math thing.

 

Also: Pulls aren't 3% of getting the unit you want per pull. Characters are generated on the summon orbs first, rather than after you pick a color, this affects pull odds. A, it affects odds of getting the color you want at all, b, it makes colors with large numbers of foci have high 5* chances but lower chances of pulling the focus you want, and vice versa, and c, colors without foci have less than 3% 5* odds because math, although c is irrelevant in this scenario.

 

Edit: This means: Colors with 2 foci tend to have ~10% 5* odds, colors with 1 foci tends to have ~7% 5* odds, reds tends to have more than other colors in terms of 5* odds no matter what, due to higher number of 5* characters, but also lower % chance of getting your focus character, this time due to # of characters in general.

Yeah it's not exact, struck me as close enough on average though? Curious how it compares to actual probabilities.

Link to comment
Share on other sites

33 minutes ago, Othin said:

Yeah it's not exact, struck me as close enough on average though? Curious how it compares to actual probabilities.

Chance should be close to:

20 orbs = 5 * [color chance] * [focus chance per color orb].1

 

\[ color chance]  is a function of [units color] / [units]2, while [focus chance per color orb] is derived from the formula {1 = [3* color]/[3*] * [3* chance] + [4* color]/[4*] * [4* chance] + [5* color]/[5*]  * [5* chance] + [focus color]/[focus] * [focus chance] }

Edit: Note that something like [3* color]/[3*] * [3* chance] is a ratio, not a probability. The ratios of each could be silly numbers like 200 for [3* color]/[3*] * [3* chance], 500 for the 4*, and 1 for the 5*. You have to divide or multiply everything so it equals 1 to make it a probability, but this formula is ugly enough already.

 

1.This formula should handle the 5 orb cost of 'total luck failure,' where you don't get a focus color at all, pretty well, but even this formula doesn't come close to approximating the efficiency of sniping. I have a feeling 'total luck failure' is a big part of the costs, just like how chain 5* foci is a big part of average pull returns.

2.actual formula a lot more complicated than this, due to the fact that 3*s are more common than 4*s, and 4*s more common than 5*s.

 

Edit: The big thing about my formula is that: It handles the fact that it's a lot easier to get green foci than other colors, and that it's a lot harder to get red foci than other colors, as well as handling the existence of multiple foci. The math looks so asinine that I think it'd be faster just to run simulations, though.

Double Edit: Holy hell the brackets are owning my edits, it keeps killing the [chance color] <--- I can't even type the right thing here, wtf---I have in the second line.

 

Triple Edit: Actually, I think the final formula is:

 

20 orbs = 5

(for 5 orbs per full pull) times

{[3* color] / [3*] * [3* chance] + [4* color] / [4*] * [4* chance] + [5* color] / [5*] * [5* chance] + [focus color] / [focus] * [focus chance]}

(for the chance of a given color appearing) times

{[focus color]/[focus] * [focus chance] / [3* color]/[3*] * [3* chance] + [4* color]/[4*] * [4* chance] + [5* color]/[5*]  * [5* chance] + [focus color]/[focus] * [focus chance] }

(for the chance of a focus appearing per pull from a given colored orb).

 

The formula has to be modified slightly for colors with multiple foci, however.

20 orbs = 5 * {[3* color] / [3*] * [3* chance] + [4* color] / [4*] * [4* chance] + [5* color] / [5*] * [5* chance] + [focus color] / [focus] * [focus chance]} * {[focus color]/[focus] * [focus chance] / [3* color]/[3*] * [3* chance] + [4* color]/[4*] * [4* chance] + [5* color]/[5*]  * [5* chance] + [focus color]/[focus] * [focus chance] }

And multiply the above by 1 / [focus color] if the color has multiple foci.

 

 

 

Simulator update:

2327 for 12 S!Robins (12 because the last 3 came within 5 pulls, and I was spamming pulls so couldn't stop in time). Fun fact: Up until 2k orbs, I was pulling almost as many 5* Reinhardts as S!Robins. It also took about 800 orbs for the first S!Robin, feels bad man.

Spoiler

sbq7kcS.png

 

Edited by DehNutCase
Link to comment
Share on other sites

Whats wrong with saying it takes 23 pulls to get 50/50 chance of getting a 5* focus.  This comes from

1 - (1-.03)^23 = 0.5'''

Thus, 23 * 5 orbs / pull = 115 pulls for 5* focus. (50% chance)

This is DRASTICALLY overestimated because I assume you're always spending 5 orbs when roughly half the time you'll get at least two of one color.  Additionally, the 3% chance increases by .25% every 5 pulls which adds up.  Though, theres the odds of getting no orbs of the right color, but I'm going to ignore that because it is somewhat rare and I'll assume it cancels with getting 2 or 3 of the same color.

If you assume these approximations which should come up with a larger value, you still only get 115*11=1265.

Rounding down the aproximations would give an average of around 1000 orbs for a +10

This would require 1000 * 140/75 = $535 USD for a +10 on average.

Edited by Lushen
Link to comment
Share on other sites

4 minutes ago, Lushen said:

Hey that's nothin.  I had a 5% chance only to get the wrong focus unit on the summer banner.

 

 

i went super cashul on that banner tbh

 

since all 3 of possible banner that IS could put up next is a trigger finger for me(only reason i didn't went ham for ultimate waifu Roderick is because after Roderick went away, its either important characters of Sacred Stones or Thracia)

Edited by JSND
Link to comment
Share on other sites

10 minutes ago, GuiltyLove said:

I have a 4% pity rate for the summer banner...

Mine is sitting at 3.5% right now, though I already got the bae Robin.

4 minutes ago, JSND said:

 

 

i went super cashul on that banner tbh

 

since all 3 of possible banner that IS could put up next is a trigger finger for me(only reason i didn't went ham for ultimate waifu Roderick is because after Roderick went away, its either important characters of Sacred Stones or Thracia)

Isn't the next one CYL?

Still no Summer Tiki...

Link to comment
Share on other sites

4 minutes ago, JSND said:

since all 3 of possible banner that is could put up next is a trigger finger for me(only reason i didn't went ham for Roderick is because after Roderick went away, its either important characters of Sacred Stones or Thracia)

I love Roderick but I didn't pull for him because it was very obvious that he would be 4*s after his banner.  And I don't see the point in pulling for people who will turn 4* when I will probably eventually get them naturally and can raise them with feathers.

Link to comment
Share on other sites

2 minutes ago, Lushen said:

I love Roderick but I didn't pull for him because it was very obvious that he would be 4*s after his banner.  And I don't see the point in pulling for people who will turn 4* when I will probably eventually get them naturally and can raise them with feathers.

4* is somewhat rare, also if you do get 5* copy,its a step towards +1s

 

Having a good character at+ran is pretty valuable for me so im all for it

Link to comment
Share on other sites

Just now, JSND said:

4* is somewhat rare, also if you do get 5* copy,its a step towards +1s

 

Oh they're rare.  I'm still wondering how I don't have a Kagero yet.  But I imagine a year from now I'll have all the 4*s that have been released up to this point with the freebie orbs.

+ing units is a nice incentive though.

Link to comment
Share on other sites

3 hours ago, Quintessence said:

Whelp, funny because I've gotten tons of Bartres. Best usage of Brash Assault is on Lyn, else other units will have to be tanky enough to eat a counter or exploit WTA to ensure its effectiveness.

Well, by default, Bartre, Hinata, and Seliph are physically tanky with Hinata having both Fury and a Ruby Sword to make him even tankier against the greens he's fighting.

The only other units I can think of who could work Brash Assault well is Chrom and Ephraim and that's mostly because of how hard Chrom hits and Ephraim being a high attack, bulky, but slow unit who happens to be blue. Otherwise, Alfonse, Lyn, and Seliph's legendary weapons synergize better with Defiant Attack or Defense and Brash Assault. I don't think Arthur, Barst, Donnel, Laslow, or Oboro would want to run Brash Assault, especially considering some of them already have good B-skills and Lukas crosses the line of being a bit too tanky to the point where he can end up not taking damage in the first place making Brash Assault useless. Also, Tobin, but... Tobin.

I doubt any of the cavaliers, knights, and fliers who could work Brash Assault would want it, especially if they're on an all-type team. Knights in particular don't want it since they're mainly enemy phase units and Brash Assault is player phase.

3 hours ago, DehNutCase said:

Because one of the requirements of using BA at all is not being 2HKOd (since you'd die on the counter below half in that case), brash units at least guarantee 2 hits if you initiate---initiate, then firesweep hits back on their turn, initiate again.

If a BA user isn't 2HKOing (or 4HKOing with a brave), you picked the wrong brash user, so it shouldn't be too bad.

 

To be honest the bigger problem with Brash is that there's no easy way to get a unit to barely below half. Just imagine if the requirement had been below 70% instead. BA would've been so amazing, since you get into range even while fighting people who 6HKO you or something.

Eh, Firesweep weapons just feel like really annoying things to deal with. They can let you chip away at any unit while killing units you have weapon triangle advantage over with without fear of counterattacks which could pose a problem, especially since to use Brash Assault, you're going to be at least 50% HP.

Y'know, maybe the Defiant skills and Brash Assault should have been 60% to 65% at level 3. 50% is really tight range to work with. Meanwhile, Desperation and Vantage at 75%. Oh, you tickled me? Have fun getting counter killed because you made a terrible mistake. Now I'm imagining the horror of Escape Route and Wings of Mercy activating at 75% HP. If that change were to happen, Anna would need to drop down to 50% to use Noatun's Escape Route 2 since that would be its new range. Combine it with Wings of Mercy 3 and you'd have a better flying Anna than what the current game allows.

Link to comment
Share on other sites

22 minutes ago, Kaden said:

Eh, Firesweep weapons just feel like really annoying things to deal with. They can let you chip away at any unit while killing units you have weapon triangle advantage over with without fear of counterattacks which could pose a problem, especially since to use Brash Assault, you're going to be at least 50% HP.

Firesweep on the AIs side is free food for teams like mine, who are all out offensive. Counter-killing teams might have issue, but you should be running TA-3 or something anyway. Firesweep on players hands is nice, but it's not Reinhardt or -blade tomes in general, never mind Reinhardt & -blade. Over 100 OHKOs before A-slot, fair and balanced.

Edit: And it's not like BA users can't hit people while at full hp, particularly firesweepers who can't even hit back. A good brash user should be able to 2HKO most units with a brave, anyway.

Edited by DehNutCase
Link to comment
Share on other sites

I think I might just go ahead and give Faye Brave Bow+ and DB3 actually. Maybe Swordbreaker too. B!Cordelia is obviously first priority but Squad Assault is a thing also and I would hate all of the SI I've given to her to go to waste at this point. With just a Brave Bow and DB2 (+ New Moon/Luna) she is performing invaluably in the trials, she's more useful than she ever was with Firesweep. 

Plus, Faye does a decent job at killing things and it's nice to have a unit that doesn't get oneshot like B!Cordelia. As these trials keep coming and people are forced to use units they otherwise would never, I bet some people will end up with some pretty impressive units they may not have expected.

Edited by Zeo
Link to comment
Share on other sites

38 minutes ago, Kaden said:

Well, by default, Bartre, Hinata, and Seliph are physically tanky with Hinata having both Fury and a Ruby Sword to make him even tankier against the greens he's fighting.

The only other units I can think of who could work Brash Assault well is Chrom and Ephraim and that's mostly because of how hard Chrom hits and Ephraim being a high attack, bulky, but slow unit who happens to be blue. Otherwise, Alfonse, Lyn, and Seliph's legendary weapons synergize better with Defiant Attack or Defense and Brash Assault. I don't think Arthur, Barst, Donnel, Laslow, or Oboro would want to run Brash Assault, especially considering some of them already have good B-skills and Lukas crosses the line of being a bit too tanky to the point where he can end up not taking damage in the first place making Brash Assault useless. Also, Tobin, but... Tobin.

I doubt any of the cavaliers, knights, and fliers who could work Brash Assault would want it, especially if they're on an all-type team. Knights in particular don't want it since they're mainly enemy phase units and Brash Assault is player phase.

What about ranged units?

Link to comment
Share on other sites

1 hour ago, DehNutCase said:

Firesweep on the AIs side is free food for teams like mine, who are all out offensive. Counter-killing teams might have issue, but you should be running TA-3 or something anyway. Firesweep on players hands is nice, but it's not Reinhardt or -blade tomes in general, never mind Reinhardt & -blade. Over 100 OHKOs before A-slot, fair and balanced.

Edit: And it's not like BA users can't hit people while at full hp, particularly firesweepers who can't even hit back. A good brash user should be able to 2HKO most units with a brave, anyway.

It's more of the feeling of helplessness and being robbed of your counter kill that annoys me. Arthur would have kicked your villainous butt if it weren't for your meddling Firesweep Lance, Roderick! I'll get you for this on my turn! Archers being a problem in general to my army is a different, but related problem. I really need to level up that +Atk, -Res 3* M!Robin and dump T-Adept 2 on him. Maybe Bowbreaker 3 from a random Setsuna too since it's just fantastic when Gordin and Klein show up bursting through with their Brave Bows on maps where you have to get in range before the AI moves. Stupid colosseum map on the last Tempest Trials.

Anyway, got bored and apparently Faye is the best archer to use Brash Assault. Her extra bulk from being a villager is really helping out. The other contenders are Gordin, Takumi, and Virion where for Takumi, it's probably not worth it compared to just sticking with his default kit. I think KageroChart might be glitching out with Defiant skills, since I remember messing around with Chrom and Brash Assault and I don't think his default Defiant Def 3 was being applied. I checked with Alfonse and Seliph and Tyrfing's effect applied, but not Folkvangr's. So, I added in the Defiant Atk 3 buff manually as +7 bonus to their attack. Hopefully, this is the case and I'm not wrong.

Spoiler

Against only ranged units with Fury 3, +Atk, -Spd Faye with Brash Assault 3, Defiant Atk 3, and Brave Bow+ gets 53 wins, 6 losses, and 1 draw. With a Silver Bow+, she gets 40 wins, 15 losses, and 5 draws.

+Atk, -Spd Gordin with Brash Assault 3, Defiant Atk 3, and and a Brave Bow+ gets 50 wins, 9 losses, and 1 draw. Silver Bow+ gets 35 wins, 22 losses, and 3 draws.

+Atk, -Res Takumi with Brash Assault 3, Defiant Atk 3, and Brave Bow+ gets 49 wins, 10 losses, and 1 draws. Fujin Yumi gets to 26 wins, 27 losses, and 1 draw.

+Atk, -Res Virion with Brash Assault, Defiant Atk 3, and Brave Bow+ gets 50 wins, 9 losses, and 1 draw. Silver Bow+ gets 32 wins, 26 losses, and 2 draws.

Against ranged units without Fury 3, +Atk, -Spd Faye with Brave Bow+ would get 56 wins, 4 losses, and 0 draw. Silver Bow+ would get 50 wins, 5 losses, and 5 draws.

+Atk, -Spd Gordin with Brash Assault, Defiant Atk 3, and Brave Bow+ would get 54 wins, 6 losses, and 0 draws. Silver Bow+ would get 44 wins, 14 losses, and 2 draws.

+Atk, -Res Takumi with Brash Assault, Defiant Atk 3, and Brave Bow+ would get 56 wins, 4 losses, and 0 draws. Fujin Yumi would get 42 wins, 18 losses, and 0 draws; -Spd would have 1 more win and 1 less draw.

+Atk, -Spd Virion with Brash Assault, Defiant Atk 3, and Brave Bow+ would get 54 wins, 6 losses, and 0 draws. Silver Bow+ would get 44 wins, 16 losses, and 0 draws; -Spd would have 1 less loss and 1 more draw.

So, yeah, Faye's mixed bulk pulls through letting her survive unlike the others who kind of just get screwed over one way or another. The question is will she be able to survive to get around Brash Assault's range. She has bulk, but it's going to be iffy when her base speed ends up being 21 with -Spd which is really weird when you consider that Gordin has the same base speed as her, but -Spd puts him at 22 rather than 21.

Edit: Flubbed some of Faye's numbers. Still the best, though. I think I had her Noontime at a lower cooldown by accident.

58 minutes ago, Quintessence said:

What about ranged units?

I was just looking into that. Look above for Brash Assault archers or rather, archer. :p

I'm not sure about mages using Brash Assault since it's going to be difficult for them to drop down in damage like that if they're running -raven tomes and T-Adept. That and being able to nuke units in the first place with that setup or with -blade tomes. For the dagger units, the only bulky ones are summer Frederick and Matthew. Felicia's fast and has low attack and Saizo has low HP. That leaves Gaius who has high HP and speed, but average-ish defense and resistance, Jakob who is Mr. Average, Jaffar who's kind of like Matthew if he traded some defense for resistance, and Kagero, the glass cannon. It could work, especially with Poison Dagger, but it's probably not that great.

Edit: Yeah, +Atk, -Res summer Frederick with Brash Assault, Defiant Atk 3 and Poison Dagger+ gets less results than if he was +Spd, -Res and had L&D3 and Poison Dagger+ which would let him cover every unit since Brash Assault requires that your opponent can counter back -- in this case: ranged units-only -- while L&D just boosts your offensives allowing you to do as much damage and double as many units as possible. Seashell would conflict with this. Anyway, if he had Silver Dagger+, the Brash Assault build would also underperform compared to the L&D3 build.

Edited by Kaden
Link to comment
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Recently Browsing   0 members

    • No registered users viewing this page.
×
×
  • Create New...