Serenes Tiering

[Arcanite]

9 minutes ago, tanuki said:

You missed one category:

Waifu/Husbando

Oboro should have her own separate category

 

@phineas81707 I have one question.

Is this a serious list? Are you trying to make this a thing? Like "The official Serenes Forest Tier List" or something? Cause that'll dictate what I want to say quite a bit heh

[MrSmokestack]

29 minutes ago, Vaximillian said:

144 including not-yet-playable Clarisse.

Which I noted and fixed in my next post. Close enough.

Thanks though.

Edited June 19, 2017 by MrSmokestack

[Bartozio]

1 hour ago, JSND said:

The problem for Elise isn't investment from skill build(which is actually pretty damn light. All you need is Genny) Its the team slot

Her problem mainly comes from lacking the access to the amazing A slot skills that make her look much worse than she seems.

For Elise to actually reach an amazing offensive power level, she need  Wrathful Staff, Hone Cavalry, AND Goad Cavalry. Admitably Elise have a really noticable power jump if you do add Goad on top of Hone, but thats 3 team slot dedicated to enable her.

You say there's 3 team slots dedicated to her, but on a horse emblem team the Hone Cavalry should have been there anyway. Goads are slightly less common, but they still benefit units like Reinhardt enough that they could have been there anyway. This means that you can pretty much dump Elise on most horse emblem teams and you have both a healer and a colorless mage in one package. The more bulky calvalry units like Xander can really appreciate the healing, and if you don't need healing you can just use her to kill more units.

I'm not completly convinced she should be placed in Useful tier, but I do feel she's more useful than Clarine for example, who can only contribute healing with her low attack stat.

[dragonlordsd]

Here is what I think is missing from every tier list, and what I think needs to be taken into account to make something truly useful:

It's not about how good a unit is on paper, it's about what their place is in the meta.

For example, on paper, Reinhardt is merely a good unit. In fact, the gamepress site still only lists him as an "A" tier unit. But spend about ten seconds in the arena (especially without a hard counter) and you'll see a very different picture.

So, a tier list really needs to take into account not just what a unit is on paper, but how it is actually used in game.

For example, I've never seen a single celica in the arena. Period. (tier 17/18)

Because, frankly, as good as celica looks on paper, she's fodder for a vast number of high powered units in the meta right now. Add to that the popularity of red alternative (Ryoma/Ike/Xander) and her slot on a four-man team is easily filled.

[DehNutCase]

2 hours ago, dragonlordsd said:

For example, on paper, Reinhardt is merely a good unit. In fact, the gamepress site still only lists him as an "A" tier unit. But spend about ten seconds in the arena (especially without a hard counter) and you'll see a very different picture.

Actually, on paper Reinhardt is literally the perfect unit.

His bulk is exemplary (4 losses on defense with the -blade, close counter build), his offenses are as good as it gets (he misses 2 units with a L&D -blade build, and gets over 100 wins pretty much no matter what build he's running), and his mobility is the best there is, since he's riding a horse on top of being ranged.

Specialized units can beat him in bulk, offenses, or mobility, but none of them by much, and all of them suffer hugely in some category compared to Reinhardt. (If you beat him in bulk it means your offenses are probably 10% of Reins, if you beat him in offense it means a stiff-breeze can kill you, if you beat him in mobility it means you lose in both offenses and bulk, because you had to take a mobility B-special, or a mobility assist---which costs a turn, too, meaning you don't have a player phase when you use it.)

Edit: And, unlike other units, he tends to have both his A-slot and his B-slot free for utility skills. Vantage for vantage sweeps, breakers for complete overkill, mobility skills, and that's just for the B-slot. For the A-slot he can either beef up his offenses to absurd levels, or take Close Counter and become a unit with a Hector level enemy phase and one of the best player phases in the game.

 

If I were to rate a unit, the main categories would be:

Offensive ability, Defensive ability, Counter-kill ability, Mobility, and Level of Support needed.

Rein is top level in offenses, B tier in defense, C in counter-kill, and A in mobility, with an A in level of support needed. (A high rating in level of support means you need minimal support, and a low rating means you need a lot of support.)

However, if you give him support, his offenses remain top, his defense goes to A (because he now kills people trying to kill him---vantage stronk), his counter-kill goes to A, his mobility stays at A, but his level of support needed drops to C.

Basically, his spread is A/B/C/A/A or A/A/A/A/C by my evaluation, and that's more or less unbeatable. (And it helps that his 'support units' can be other Reinhardts.)

 

Double Edit: The biggest problem with tier lists is that the people making tier lists have no idea how to rate units. Setsuna had been consistently above the TKJ trio in the wiki for quite a while, despite being: a, worse than the TKJ trio in her only good set, and b, being unable to run any other sets the trio could run. (Just imagine Setsuna with Close Counter, for example.) It gets even worse when you remember that her L&D Brave+ set costs more than Klein's, because Klein has a brave-bow and she doesn't, yet she was somehow 2 tiers higher for a month or two.

Edited June 19, 2017 by DehNutCase

[MrSmokestack]

33 minutes ago, DehNutCase said:

The biggest problem with tier lists is that the people making tier lists have no idea how to rate units. Setsuna has been consistently above the TKJ trio in the wiki for quite a while,

This too. With how quickly the meta moves and the sheer number of units that need to be analyzed initially as well as later on when more are added, you'll find stuff like this that manages to slip through the cracks and stay there.

This is why I don't like tier lists. People refer to them like Gospel and don't bother doing calculations themselves.

[Troykv]

I like to do experiments in my own; but I also like these things.

I love to see the general opinion about X or y unit and maybe use it to give me and an idea of which units are commom or/and troublesome.

[Ice Dragon]

2 hours ago, dragonlordsd said:

Here is what I think is missing from every tier list, and what I think needs to be taken into account to make something truly useful:

It's not about how good a unit is on paper, it's about what their place is in the meta.

For example, on paper, Reinhardt is merely a good unit. In fact, the gamepress site still only lists him as an "A" tier unit. But spend about ten seconds in the arena (especially without a hard counter) and you'll see a very different picture.

So, a tier list really needs to take into account not just what a unit is on paper, but how it is actually used in game.

For example, I've never seen a single celica in the arena. Period. (tier 17/18)

Because, frankly, as good as celica looks on paper, she's fodder for a vast number of high powered units in the meta right now. Add to that the popularity of red alternative (Ryoma/Ike/Xander) and her slot on a four-man team is easily filled.

This is an offense tier list, not a defense tier list. How a character performs on a defense team doesn't affect how the same character performs on an offense team.

Celica, for example, has the exact same offensive stats as Tharja and Katarina, which means she should place at least as high as they would due to being able to run the exact same set with the exact same effectiveness. Their defensive stats don't matter much because all they really have to do is one-hit kill Gronnblade+ users and tank a counterattack from Hector (which, interestingly enough, Katarina can't do consistently). I say "at least as high" because Celica also has access to the uninheritable Ragnarok, which Tharja and Katarina don't have access to, that behaves similarly to a self-powered Raudhrblade+ (+6/5 without buffs compared to +0/0, +9/9 with +3/4/0/2 compared to +12/4).

 

4 minutes ago, MrSmokestack said:

This too. With how quickly the meta moves and the sheer number of units that need to be analyzed initially as well as later on when more are added, you'll find stuff like this that manages to slip through the cracks and stay there.

This is why I don't like tier lists. People refer to them like Gospel and don't bother doing calculations themselves.

Tier lists work well when the number of knowledgeable people working on them is high enough to review the entire list with each new introduction in the meta. The tier list also needs to be absolutely transparent as to why each character is where it is (and be able to defend its decisions), and the authors need to actually bother to keep the thing maintained.

I do think we have enough people here who actually know what they're doing to pull one off, though, but that really depends on how much work people care to put into it.

[TheTuckingFypo]

27 minutes ago, Ice Dragon said:

This is an offense tier list, not a defense tier list. How a character performs on a defense team doesn't affect how the same character performs on an offense team.

Celica, for example, has the exact same offensive stats as Tharja and Katarina, which means she should place at least as high as they would due to being able to run the exact same set with the exact same effectiveness. Their defensive stats don't matter much because all they really have to do is one-hit kill Gronnblade+ users and tank a counterattack from Hector (which, interestingly enough, Katarina can't do consistently). I say "at least as high" because Celica also has access to the uninheritable Ragnarok, which Tharja and Katarina don't have access to, that behaves similarly to a self-powered Raudhrblade+ (+6/5 without buffs compared to +0/0, +9/9 with +3/4/0/2 compared to +12/4).

Tier lists work well when the number of knowledgeable people working on them is high enough to review the entire list with each new introduction in the meta. The tier list also needs to be absolutely transparent as to why each character is where it is (and be able to defend its decisions), and the authors need to actually bother to keep the thing maintained.

I do think we have enough people here who actually know what they're doing to pull one off, though, but that really depends on how much work people care to put into it.

Thank you for this. You read my mind O_O 

I think having one tier list is a jumbled mess. We need at least 4 for this game; Offensive Non-SI, Offensive SI, Defensive Non-SI, and Defensive SI.

Hell, I think we may even need more in the future. (Tempest Trials, I'm looking at you T_T)

[MrSmokestack]

4 minutes ago, TheTuckingFypo said:

I think having one tier list is a jumbled mess. We need at least 4 for this game; Offensive Non-SI, Offensive SI, Defensive Non-SI, and Defensive SI.

Hell, I think we may even need more in the future. (Tempest Trials, I'm looking at you T_T)

And with that, we have dived headfirst into a logistical nightmare.

Non-SI Tier Lists are moot because you can't realistically expect someone to run only a Hero's vanilla set for Arena Offense or Defense. All of the "No SI" discussion would exist in a vacuum since it doesn't reflect the actual game, and hence has no practical use.

@Ice Dragon I just don't see what yet another tier list would accomplish when other sources exist. There's always contacting the individuals who run said tier lists and offering expertise to them, instead. I just think the idea's been overdone.

[XRay]

25 minutes ago, Ice Dragon said:

This is an offense tier list, not a defense tier list. How a character performs on a defense team doesn't affect how the same character performs on an offense team.

Celica, for example, has the exact same offensive stats as Tharja and Katarina, which means she should place at least as high as they would due to being able to run the exact same set with the exact same effectiveness. Their defensive stats don't matter much because all they really have to do is one-hit kill Gronnblade+ users and tank a counterattack from Hector (which, interestingly enough, Katarina can't do consistently). I say "at least as high" because Celica also has access to the uninheritable Ragnarok, which Tharja and Katarina don't have access to, that behaves similarly to a self-powered Raudhrblade+ (+6/5 without buffs compared to +0/0, +9/9 with +3/4/0/2 compared to +12/4).

 

Tier lists work well when the number of knowledgeable people working on them is high enough to review the entire list with each new introduction in the meta. The tier list also needs to be absolutely transparent as to why each character is where it is (and be able to defend its decisions), and the authors need to actually bother to keep the thing maintained.

I do think we have enough people here who actually know what they're doing to pull one off, though, but that really depends on how much work people care to put into it.

1

Is it possible to reduce the number of required people? If we can automate the build process, we would eliminate the biggest time sink trying to figure out the most optimal skill set for every character. I am no programmer, but I imagine it should be possible to let the computer figure out the best nature and skill combination that can kill the most neutral, non-S.I. units for every character on player phase. I do understand that it will take a lot of time and effort though.

Once we have everyone's Win:Loss:Draw ratio, it would be very easy to move units up and down the tier list.

We can also do one for most kills on enemy phase, since quite a few characters perform better on enemy phase than player phase.

[Ice Dragon]

10 minutes ago, TheTuckingFypo said:

Hell, I think we may even need more in the future. (Tempest Trials, I'm looking at you T_T)

Tempest Trials have two problems: they don't last very long (only 2 weeks) and tiering is highly dependent on the map set. For example, fliers have two huge advantages in the current map set: being able to traverse mountains for Story 4-3 and getting around the damned tree in Story 8-2.

 

1 minute ago, MrSmokestack said:

I just don't see what yet another tier list would accomplish when other sources exist. There's always contacting the individuals who run said tier lists and offering expertise to them, instead. I just think the idea's been overdone.

Maintainability and open discussion. One thing that Smogon does exactly right is have its tier list maintained in a forum thread instead of a standalone web page. That way, it's possible to field questions and argue placement somewhere where everyone else can see.

[TheTuckingFypo]

4 minutes ago, Ice Dragon said:

Tempest Trials have two problems: they don't last very long (only 2 weeks) and tiering is highly dependent on the map set. For example, fliers have two huge advantages in the current map set: being able to traverse mountains for Story 4-3 and getting around the damned tree in Story 8-2.

 

Maintainability and open discussion. One thing that Smogon does exactly right is have its tier list maintained in a forum thread instead of a standalone web page. That way, it's possible to field questions and argue placement somewhere where everyone else can see.

Fair enough. 

I nominate @Ice Dragon and the StrategistsThree to come up with builds and create an official SF tier list. 
(I'll help, but I doubt I'd be very useful outside of that Azama build I thought up.)

[dragonlordsd]

19 minutes ago, Ice Dragon said:

Tempest Trials have two problems: they don't last very long (only 2 weeks) and tiering is highly dependent on the map set. For example, fliers have two huge advantages in the current map set: being able to traverse mountains for Story 4-3 and getting around the damned tree in Story 8-2.

This.

There are so many different exploits for tempest trials that aren't as valuable normally. Fliers, as Ice Dragon points out, have a huge advantage in several of the maps.

Furthermore, Veronica always spawns with two cavalry and an armor, making anti-cavalry mages generally better for clearing out her support. I wouldn't normally run Ursula in the arena, but she's huge in Tempest Trials.

Going back to what I said originally though, builds are way too situational. It's too hard, I think, to evaluate what something could be as opposed to how it is actually used.

Setsuna is a great example of this. Even if her statline is better than someone like Kleins' (which I don't really think it is), who's going to put in the effort optimize her outside of a few hardcores? And even if someone does bother to fully flesh out her potential, it's not going to matter much when a Hector just parks himself on a defense tile.

Edited June 19, 2017 by dragonlordsd

[Arcanite]

38 minutes ago, TheTuckingFypo said:

Fair enough. 

I nominate @Ice Dragon and the StrategistsThree to come up with builds and create an official SF tier list. 

Well if this seriously seriously seriously seriously becomes a thing, I'd be willing to help out!

As long as no one questions our abilities XD There may be a few other people to include as well though 

[Ice Dragon]

Just now, XRay said:

Is it possible to reduce the number of required people? If we can automate the build process, we would eliminate the biggest time sink trying to figure out the most optimal skill set for every character. I am no programmer, but I imagine it should be possible to let the computer figure out the best nature and skill combination that can kill the most neutral, non-S.I. units for every character on player phase. I do understand that it will take a lot of time and effort though.

Once we have everyone's Win:Loss:Draw ratio, it would be very easy to move units up and down the tier list.

We can also do one for most kills on enemy phase, since quite a few characters perform better on enemy phase than player phase.

W-L-D ratio is a good starting point, but that's not representative or robust enough to for a tier list. There are things that need to be taken into consideration that the W-L-D ratio doesn't tell you:

• By how much does the unit survive each encounter? Units with Desperation, for example, only need to survive one encounter, and can then go virtually forever without taking any more damage. Sophia and Nowi on the opposite end of the spectrum get by simply by taking no damage to begin with.
• How important are the match-ups that aren't won? Losing to Harold or Merric, for example, should be weighted less than losing to Hector or Spring Camilla.
• Are the match-ups that aren't won match-ups that can be covered by a likely teammate? A red unit that fails to kill Hector, for example, is less likely to have a teammate that can cover Hector than a green unit that fails to kill Hector.
• If you fail to kill an opponent in one round of combat, what are the repercussions? Wings of Mercy is a notable threat for failed kills on enemy phase. Vantage is a notable threat for failed kills in general.
• What support options does the unit have? Celica commonly runs Renewal and can therefore be used for healing. Nowi and Fae typically take zero damage from combat and have massive HP pools that can also be given to other teammates with Reciprocal Aid. Eirika and Ephraim have uninheritable weapons that give Hone Atk 2 without taking up their passive C slot, allowing them to provide buff combinations that others cannot.
• Can the unit get to where they need to be? Hector and Effie do a lot of damage, but that doesn't mean anything if they can't reach what they want to reach.
• How easy is it to build a team using the character? Reinhardt needs no team support. Olwen needs easy-to-get team support since cavalry teams typically already have the buff she wants. Infantry Litrblade+ users need more team support unless they don't need the added coverage. Sophia really needs a dancer to actually get kills on axe users. Sheena and Gwendolyn really want to be on a team with at least three armors to stack buffs.

[TheTuckingFypo]

10 minutes ago, Arcanite said:

Well if this seriously seriously seriously seriously becomes a thing, I'd be willing to help out!

As long as no one questions our abilities XD There may be a few other people to include as well though 

Yeah! We can do this! I'll let @Ice Dragon and @MrSmokestack do the calculations though. I can't math. >_<

[MaskedAmpharos]

I mean I can help with calculations and general theorycrafting if necessary, but to be quite honest, I don't really like the idea of a tier list in general. 

There are too many different contexts and scenarios to consider for each character (many of which @Ice Dragon, @MrSmokestack, and others have brought up already) to make a fully accurate tier list, and I also fundamentally don't like the idea of a tier list because it discourages experimentation and just makes more people aim for "high tier" characters and ignore lower tiered ones.

Not to mention this game is entirely PVE (even Arena, since you're fighting the AI and not actual players), and you can beat just about any AI-controlled team with just about any unit. You don't need a Lucina with full SI to take down Hector when Laslow or Stahl can do it just fine even if they aren't "high tier". 

Also, tier lists by nature focus on how a unit performs as an individual, but Heroes is a team-based game. It doesn't matter that Sophia is shit at taking out anything but greens and colorless (if you're running Raven) because that's all she ever needs to take out. You have 3 other slots dedicated to filling in the gaps. Tier lists like these put too much emphasis on independent units that can take out as many different units on their own as possible, but that kind of mindset is inherently counterintuitive to the nature of the game. 

tldr - Tier lists don't work well with Heroes because they're inherently biased towards a unit's individual merit, which matters much less in a team-based game. Additionally, even a list based on individual merit has too many factors to consider to be truly accurate. 

[Alkaid]

I think it speaks for itself that most knowledgeable players just disregard all the tier lists elsewhere.

Ever since we got SI, this game reminds me more of Pokemon than other gacha games. While it takes much more work, I feel like something akin to Smogon's Pokedex for competitive Pokemon would apply to this game better. Every Pokemon has a write up, suggested sets, teammates and counters. That makes it an excellent resource regardless if you play in their self-made meta or not. However, Pokemon only has to deal with big changes once every couple years and not a regular introduction of new characters/skills into the meta like Heroes. So keeping a similar resource up to date would no doubt be a nightmare.

Still, with how SI, natures and teammates factor in (cav buffs, etc.) I feel like a standard tier list is really hard to apply to this game while maintaining much accuracy.

Edited June 19, 2017 by Alkaid

[Ice Dragon]

2 minutes ago, MaskedAmpharos said:

tldr - Tier lists don't work well with Heroes because they're inherently biased towards a unit's individual merit, which matters much less in a team-based game. Additionally, even a list based on individual merit has too many factors to consider to be truly accurate. 

You can say the exact same thing about Pokémon where every party member is still a member of a 6-member party. Tier lists still work, and Pokémon has far more build variance than Fire Emblem Heroes has. Any mechanic that Fire Emblem Heroes shares with Pokémon is a mechanic that can therefore be accounted for in a tier list.

As such, there is only one complex mechanic that Heroes has that (singles) Pokémon does not have, and that is the presence of non-self-targeting buffs (movement range, attack range, and things like that are not complex by any stretch of the imagination). This can, however, be accounted for in a similar manner as Mega Evolutions in Pokémon where each notable combination of buffs is a separate entry on the tier list.

For cavalry, fliers, and dragons, I'd expect two entries, one with no buffs (i.e. used on an otherwise all-infantry team composition) and one with +6/6/0/0 (+0/0/6/6 for dragons) (i.e. used with at least one other unit of its type). Additional buffs, including +6/6/6/6, +3/4/0/2, +4 to a single stat, and additional buffs from Rally assist skills can be wrapped up in analysis details because they don't change viability by a large margin. Furthermore, +4 to a single stat is already somewhat expected and should already be accounted for in a tier list.

[JSND Alter Dragon Boner]

 I mean even pokemon have an objective non usage based tier list that most players would agree with. Ask anyone about the best pokemon atm and people would tell you about Arcanine, Tapu Fini, Kartana and Celesteela as the cream of the crop.

Pokemon do have a widely used non self targeting buff. Thats basically how Stealth Rock really works. A lot of analysis in pokemon do have notes that say things along the lines of "this is an OHKO against X after Stealth Rock damage"

I kinda want to bring up before that 4/0 or 6/6 or 0/0/6/6 buff state could be a part of tiering but this had an awkward effect on some unit(Setsuna jumps a tier with buff comes to mind)

[XRay]

@Ice Dragon WLD ratio does not tell us every thing, but we can use more numbers and ratios.

• By how much does the unit survive each encounter? Units with Desperation, for example, only need to survive one encounter, and can then go virtually forever without taking any more damage. Sophia and Nowi on the opposite end of the spectrum get by simply by taking no damage to begin with

Spoiler

 

Maybe we can judge a character using a few different different ratios?

Player Phase Maximum Win Ratio
+Atk/+Spd; Brave/Blade/Poison Dagger, Damage Special, L&D/DB, Breaker(100%HP)/Desp.(75%HP)
Breakers usually gives the most kills, but I am pretty sure I came across at least one that has Desp.@75%HP that got the most kills. I think it is a mage with a Blade tome build, do not remember who though.

Enemy Phase Maximum Win Ratio
+Atk; Silver/Killer/Wo Dao/Raven/Poison Dagger, Damage Special, Counter/Fury, QR

Desperation Health Range Ratio
+Atk/+Spd; Brave/Blade/Poison Dagger, Damage Special, L&D, Desp, 100%HP
This will track four numbers, encounters that end in 100%HP, 100%>X%HP>75%, 75%>=X%HP>0HP, and 0HP. The second number tells us how likely it is for a character to take damage and not be in Desperation range, so -HP may be preferable over -Def or -Res to make the second number smaller.

Player Phase Desperation Maximum Win Ratio
+Atk/+Spd; Brave/Blade/Poison Dagger, Damage Special, L&D, Desp.(75%HP)

 

 

• How important are the match-ups that aren't won? Losing to Harold or Merric, for example, should be weighted less than losing to Hector or Spring Camilla.

Spoiler

 

I think a sample of 30 T20 players' first three matches that pop up is a good indicator of what units to include and what units to exclude. That will be a total of 90 teams and a total of 360 unit slots. I think that should be a decent sample size. However, if the same team from the same player pops up more than once though, I am not sure if we should reroll until we get someone different.

We can then take the top 50% (or whatever percentage is appropriate) of units that appear the most frequently and include them in a weekly or monthly list. If we have enough resources to do weekly lists, we can track the meta better and combine the last 2 to 4 weekly lists into a current meta list. We can use the current meta list as a benchmark for how well a character performs against them.

If we do not have enough T20 players, I think using lower tiers would be okay. We can always increase or decrease our sample size as resources allow.

 

 

• Are the match-ups that aren't won match-ups that can be covered by a likely teammate? A red unit that fails to kill Hector, for example, is less likely to have a teammate that can cover Hector than a green unit that fails to kill Hector.

Maybe we should revise the benchmark and judge characters on how they perform against neutral (or another nature), Counter/Fury, QR units? Most units will not have that build, but it does give us an idea on how a character performs against the worst case scenario.

• If you fail to kill an opponent in one round of combat, what are the repercussions? Wings of Mercy is a notable threat for failed kills on enemy phase. Vantage is a notable threat for failed kills in general.

I think this might be outside the scope of a tier list and more about tactics and maneuvering.

• What support options does the unit have? Celica commonly runs Renewal and can therefore be used for healing. Nowi and Fae typically take zero damage from combat and have massive HP pools that can also be given to other teammates with Reciprocal Aid. Eirika and Ephraim have uninheritable weapons that give Hone Atk 2 without taking up their passive C slot, allowing them to provide buff combinations that others cannot.

Characters with utility builds not based purely on combat are hard to quantify. They should be judged on a case by case basis, and debated on how they contribute to a team. For example, dancers/singer have some of the worst combat potential, but their ability to renew allies more than makes up for their lackluster strength.

• Can the unit get to where they need to be? Hector and Effie do a lot of damage, but that doesn't mean anything if they can't reach what they want to reach.

Unlike utility builds, I think movement is pretty clear on how they impact a character's performance and tier. I do not think this is quantifiable, but I do not think we need to.

• How easy is it to build a team using the character? Reinhardt needs no team support. Olwen needs easy-to-get team support since cavalry teams typically already have the buff she wants. Infantry Litrblade+ users need more team support unless they don't need the added coverage. Sophia really needs a dancer to actually get kills on axe users. Sheena and Gwendolyn really want to be on a team with at least three armors to stack buffs.

Yeah, I am not sure how to quantify this either.

Edited June 20, 2017 by XRay
Added spoilers to make the post shorter.

[MrSmokestack]

1 hour ago, Ice Dragon said:

You can say the exact same thing about Pokémon where every party member is still a member of a 6-member party. Tier lists still work, and Pokémon has far more build variance than Fire Emblem Heroes has. Any mechanic that Fire Emblem Heroes shares with Pokémon is a mechanic that can therefore be accounted for in a tier list.

As such, there is only one complex mechanic that Heroes has that (singles) Pokémon does not have, and that is the presence of non-self-targeting buffs (movement range, attack range, and things like that are not complex by any stretch of the imagination). This can, however, be accounted for in a similar manner as Mega Evolutions in Pokémon where each notable combination of buffs is a separate entry on the tier list.

Pokemon vs Fire Emblem Heroes is like apples vs oranges. No, broccoli vs bread.

One is a movement-based top-down game where you have perfect information on both sides, with nothing left to chance. The other is not; you can make a perfectly valid move--probably the only one at your disposal--and still lose for no reason other than the RNG just didn't agree with you.

In Pokemon, you pit 1 or 2 creatures against each other in any given instance, and much of the challenge comes from predicting what the opponent might bring out next, or what moves they will pick--which makes battles more dynamic, especially in the days before Team Preview. In Heroes, all 8 units are on the board at the same time and leave no factors hidden from consideration. Matches can be played out in roughly the same way; Bait / Kill --> Dance / Position out. If you run all the calcs beforehand you will know exactly what will happen in each game, every single time, without fail. How can an AI that can be played into performing what you want it to be anywhere near as smart as an actual person who is evaluating your every move, turn by turn? Good players, even just people in general, can know when they're being set up, but computers can't--least of all the one that you play against in Heroes, at any rate.

"But heroes can have different skills just like pokemon can have different moves!" This is a fair point. But do I really need to adjust my game plan if I go up against a Triangle Adept Julia as opposed to a Soren running the same set? What if that same Soren runs Gronnblade instead, a la Nino? Still no. Unit colors simplify most matchups and make minute differences in damage irrelevant. The former wouldn't threaten my team any more than the latter would, assuming my team already has an answer to Adept or -Blade Greens.

Even if we ignore hit checks on moves that don't have base 100 accuracy, there is still RNG involved in using attacks. Moves in Pokemon don't deal a flat amount of damage; rather, they will inflict a value of damage within a range of possible values. In other words, get a high roll and you can OHKO something that would have lived otherwise; get a bad roll, and that wall is hanging on by 5-11 HP just to use Recover again. And that's before taking critical hits into consideration. But I digress.

Never mind you're playing against a highly exploitable, basic AI  in one and a real, live, breathing human being in the other. PvE =/= PvP, at all.

And unlike Pokemon, Heroes undergoes "meta-shifts" every two weeks rather than a year. In literally two years we will have as many units as Pokemon has, well, pokemon. Not exactly, but Heroes's cast is growing at a much faster rate than Pokemon's. Assuming we get four new units every two weeks, after two years the roster will have more than doubled before taking seasonal and Hero Battle units into consideration. In addition, some focus banners like to have more than four units, as Blazing Shadows can attest to.

Spoiler

Maths for people who care:

144 current

4 Heroes every 2 weeks --> 208 units after 2 52-week years

208 + 144 = 352

 

My point is, the two games aren't comparable in the slightest--your analogy distorts the big picture. Implying that the metagames for both are equally easy or difficult to maintain is simply the most egregious one.

Even if they might share some mechanics in common--buffs, "types" or otherwise--Fire Emblem Heroes is not Pokemon. So don't make it Pokemon.

[Ice Dragon]

21 minutes ago, JSND said:

I mean even pokemon have an objective non usage based tier list that most players would agree with.

That's exactly what I just said.

 

26 minutes ago, JSND said:

Pokemon do have a widely used non self targeting buff. Thats basically how Stealth Rock really works. A lot of analysis in pokemon do have notes that say things along the lines of "this is an OHKO against X after Stealth Rock damage"

Rocks, however, are so widely used that it's practically a given to have active at any given time. They are taken into account in analyses and match-ups because you can rely on them being there, and a standard team composition has a member that sets rocks.

Buffs in Heroes are more complicated than this model. Heroes has multiple types of rocks that work differently from each other. Furthermore, rocks in Heroes are simply not "reliable", not from a gameplay perspective, but from a team building perspective. It's possible to build a perfectly viable team without using the best rocks, and many units play differently without said best rocks, meaning rocks cannot simply be assumed to be present for all calculations.

 

47 minutes ago, XRay said:

Maybe we can judge a character using a few different different ratios?

That doesn't address the reasoning behind that point, but I admit I didn't follow through.

The point I was intending to get at is that a unit should be expected to have to fight multiple rounds of combat. Winning a round of combat with 1 HP remaining is nowhere near the same as winning a round of combat at near full health unless the unit has Desperation.

 

51 minutes ago, XRay said:

I think this might be outside the scope of a tier list and more about tactics and maneuvering.

Leaving any unit you're intended to be used against that commonly runs Vantage alive after one round of combat is a bad thing and should be taken into account.

Any unit intending to counter Xander, Camus, or Ryoma on enemy phase must either finish the job or have enough HP remaining to take the hit from Vantage. Failing to do so should be weighted heavily against a unit in tiering.

 

18 minutes ago, MrSmokestack said:

My point is, the two games aren't comparable in the slightest--your analogy distorts the big picture. Implying that the metagames for both are equally easy or difficult to maintain is simply the most egregious one.

I have not at all implied anything about the ease of the two tasks. My response was in direct response to the statement I quoted, which claimed that tier lists "don't work well" in Heroes due to individual merit mattering less than team composition.

Pokémon has the same emphasis on team composition as Heroes does. You can slap together the top six Pokémon into a team and that team will most certainly not be topping any charts. However, even with the importance of team composition, Pokémon still has a coherent tier list that is objectively defendable.

My argument effectively states that Heroes contains a subset of the mechanics available in Pokémon. Pokémon has more complicated mechanics, but that simply means that if a tier list can be created with something as complicated as Pokémon, it can also be created with something simpler.

As for the actual evaluation of viability, Pokémon essentially collapses down to the results of one-on-one match-ups with limited team intervention (e.g., sneaky pebbles). Who do you reliably counter? Who do you reliably check? Who can you survive a hit from? How you do avoid unfavorable match-ups? These questions are no different than the ones that need to be asked for Heroes and the process for verifying the answers to these questions is similarly no different: You run numbers and see who comes out on top in what circumstances, then you gauge how often said circumstances apply.

[XRay]

4 minutes ago, Ice Dragon said:

That doesn't address the reasoning behind that point, but I admit I didn't follow through.

The point I was intending to get at is that a unit should be expected to have to fight multiple rounds of combat. Winning a round of combat with 1 HP remaining is nowhere near the same as winning a round of combat at near full health unless the unit has Desperation.

 

You could split WLD ratio into more components if you want. For example, perfect win (100%HP), win (100%-75%HP), Desperation win (75%-50%HP), bad win (50%-25%HP), Pyrrhic win (25%-1HP). I think WLD is sufficient in that it nicely sums up a character's performance without it being overly complicated.

How common is multiple rounds of combat though? Top tier characters usually aim for ORKO, and then have another character Assist them out of enemy range. Outside of armors and a few other tanks, I do not think multiple rounds is important enough to warrant too much attention. Multiple rounds maybe more applicable for Enemy Phase characters, but combat at most usually lasts two rounds (enemy phase initiation then player phase counter attack or retreat) and I assume people usually do not make their tanks agro more than one or two units at once. Even when I used to play Hector in the beginning, I always aim to place him against one unit at a time to prevent him from being overwhelmed, unless they were daggers and lances where he will obviously kill them quicker if he agros more.

29 minutes ago, Ice Dragon said:

Leaving any unit you're intended to be used against that commonly runs Vantage alive after one round of combat is a bad thing and should be taken into account.

Any unit intending to counter Xander, Camus, or Ryoma on enemy phase must either finish the job or have enough HP remaining to take the hit from Vantage. Failing to do so should be weighted heavily against a unit in tiering.

 

I think that is more on the player's skill and tactics. I am guilty of not checking my opponents skills sometimes due to laziness and get myself killed from Vantage, but usually, players make sure they kill a Vantage unit before it activates, or make sure their own characters can take the hit if Vantage is active.