Want to make a new tier list?

[Moishe Oofnik]

Jill T would be above Haar, and Haar would be above Jill N.

I still think rating by parts is better

but yea Jill T Is like slightly better then haar to warrent the same placing on the tier list

Haar Transfer really doesnt matter

[Chiki]

I haven't looked or considered this in great detail, but here is what I'd do:

Top

Jill (T)

Haar

Jill (N)

High

Titania
Rafiel

Reyson

Ike

Mid

Sothe

Nolan

Leanne

Marcia

Volug

Mia

Janaff

Ulki

Nailah

Oscar

Tanith

Edited March 19, 2015 by Chiki

[Yojinbo]

Because they're in the same game. I don't think your reasoning is sound here because it's possible to compare units in diff games; for example, FE8 Seth is obviously better than FE9 Makalov. No need to deny the truth.

I'm not saying you can't compare them. You can compare anything if you want. It's just not very conclusive to do so for a tier list that aims at accurately reflecting the state of the game as a whole imo. I think the latter is impossible anyway due to the game's nature.

The most accurate way to handle it would be to have multiple lists. Sothe's position in the game is described more accurately if you have him in top tier of a DB Tier list for Part 1, 3-6, 3-12 and 3-13 and in more or less useless tier for Part 4, than by having him in the no man's land of midish tier for the entire game. In the same hand multiple tier list would allow to just compare Haar's and Jill's performances in Part 4, draw the conclusion that Jill is better and then acknowledge that they don't actually have anything to do with each other for the rest of the game and that it'd make a whole lot of sense to leave it at that even though they could be compared for whatever pointless reason.

[Moishe Oofnik]

I haven't looked or considered this in great detail, but here is what I'd do:

Top

Jill (T)

Haar

Jill (N)

High

Titania

Rafiel

Reyson

Ike

Mid

Sothe

Nolan

Leanne

Marcia

Volug

Mia

Janaff

Ulki

Nailah

Oscar

Tanith

Birds VS Mia should be a thing Ill look into it later

But yea seems fair

Instead of naming it Mid

Name its

DragonStomp:

Top:

High:

[Chiki]

I'm not saying you can't compare them. You can compare anything if you want. It's just not very conclusive to do so for a tier list that aims at accurately reflecting the state of the game as a whole imo. I think the latter is impossible anyway due to the game's nature.

The most accurate way to handle it would be to have multiple lists. Sothe's position in the game is described more accurately if you have him in top tier of a DB Tier list for Part 1, 3-6, 3-12 and 3-13 and in more or less useless tier for Part 4, than by having him in the no man's land of midish tier for the entire game. In the same hand multiple tier list would allow to just compare Haar's and Jill's performances in Part 4, draw the conclusion that Jill is better and then acknowledge that they don't actually have anything to do with each other for the rest of the game and that it'd make a whole lot of sense to leave it at that even though they could be compared for whatever pointless reason.

The most accurate way to handle tiering for any game would be to have a separate tier list for each chapter. That way, Edward and Nephenee won't be so high up on the tier list for the whole game simply because they save a lot of turns in one chapter. But we know tier lists have these problems: we just acknowledge them and move on.

[-Cynthia-]

I think Anoulethe explained it pretty well, here:

Got me curious to actually using him for the first time, and he seems pretty salvagable. He's like a sort of Cat, without the awful gauge and with excellent durability.

My problem with Kyza is that you're using a non trivial portion of BEXP and whatnot and the payoff isn't that good. He doesn't improve the 3-4 clear or the 3-5 bosskill, 3-7 doesn't matter. He might contribute to the 3-8 and 3-10 maps because they're routs, but he still has gauge and is 1 range locked. 3-11 he has issues with pitfalls like all non-fliers, he doesn't really have any contribution to 3-E. By Part 4, he's too slow and still has 1 range lock and gauge to deal with- kinda like Part 4 Volug or something. Ultimately, we put this BEXP into him and we get 2 chapters where he might help rout things- it's hard for me to see how this makes him better than like Geoffrey or Tauroneo who have a lot of use without many resources.

[XeKr]

Here's an interesting question that isn't entirely related to the current discussion - since we traditionally attribute items that thieves obtain exclusively via stealing and the like to be items the thieves "get" us, and some tierlists also give points to units for recruting others (although outside of the FE12 tierlist I don't really remember which others used this), can an argument be made that if a unit comes with a valuable item/weapon that we attribute their inventory towards part of their benefit to the team? This seems potentially really messy, and maybe it's been gone over before, I don't know.

afaik, inventory traditionally is not counted because you get that stuff even if the character doesn’t impact the chapter directly (moving, killing stuff, baiting AI, etc). The character isn’t “used” and there isn’t really player decision-making involved. However, these things can get into even trickier discussions such as, if a unit is dependent on using some resource like bexp, weapon, or a statbooster to perform well, does the other unit(s) that worked to obtain those resources get full, shared, or no downstream credit? Usually not, we don’t reward early joiners for simply getting us to the later chapters and also the later joiner’s contributions. But is this rational, for all the cases like thieving and such?

I'm on board with this, but given RD's awkward deployment situation, you're essentially forced to give people like Ed, Neph and Geoffrey big boosts to their tiering since they are basically mandatory for clearing some maps efficiently. I'm comfortable with this, but I'm not sure everyone else is.

Ah, I like to call this the “Edward clause”. >_>. I don’t see what it specifically has to do with resource allocation or deployment though. It’s a problem when considering more objective metrics for efficiency in general. It turns out using Edward is extremely efficient in every context. ;P

imo it’s actually an interesting problem. Why do we generally accept units like Sothe or the Royals can be (sometimes very) high tiered, who are good for a short part of the game and essentially nonexistent for a large part, yet not Edward who is really really good for an shorter amount of the game? Because not all turns saved are equal? Because prologue is “trivial”? Because single chapter outliers just feel extreme? Because Edward sucks later? (but is that truly worse than nonexistence because of the bench or just sitting in a corner?). Where’s the cutoff for similar units like FE6 Marcus, or Frederick in FE13 Lunatic?

Ultimately, the reason the issue comes up is we tend to think of characters in an in-play or not-in-play manner (kinda like on-off analytics in sports). I actually argue this is the proper way, in general, to think about them. I talk a lot about this, when in context, as the proper way to treat opportunity cost and that the true impact of a unit is observing the team’s performance in their absence - with all the resources saved then rationally re-allocated, and so on.

Normally, in a character’s absence, you re-allocate the resources and deploy the next best unit and determine how fast and reliable the clear now becomes. This is how a unit is “good”, because you’d rather use them to facilitate quick and reliable clears, relative to other members of the cast that could also potentially use that slot. However, in cases of forced deployment, this can break down. If we ask, “would we rather have Edward or Titania if we wanted to beat FE10 efficiently/quickly/reliably”, the answer might be Edward if he uniquely saves like 30 turns total early on and Titania is mostly replicable by Haar/Ike/Oscar/etc.

Does this completely dismantle the worth of such metrics? If we examine things more carefully, we find what happens is we’re essentially making trivial statements like “Edward is better than ‘nothing’”. Well duh, obviously Edward is better than ‘nothing’, especially if your only other units are base Micaiah and Leo (incidentally, this is nearly the stereotypical Jeigan argument, made extreme). Obviously Nephenee is better than ‘nothing’ if she’s one of two units. Their apparent contribution in these chapters is massive because instead of the opportunity cost of another unit’s contribution (that a blank deployment slot provides), there’s ‘nothing’. Instead of being compared to other units, like elsewhere when tiering, it’s a pure “because they exist” thing. That’s not very intuitively impressive, compared to units like the Royals who are used because they’re actually better than most others competing for that particular deployment slot. Not just simply better than ‘nothing’ (which of course they also are because it’s trivial).

Now clearly, there are conceivably pedantic cases where ‘nothing’ might be better such as lower tiers on more optimal teams (note: this is very very rare because of shoves/rescue). Still, these lower tiers are likely not worse than nothing on less optimal teams, and besides, to some extent we’re waiving the opportunity cost of deployment anyway, when indulging these suboptimal units.

To be the most self-consistent (if all turns are weighed the same), we should credit Edward for his significant contribution to prologue. I don’t really have a problem with that, despite how counterintuitive it may seem. However, in cases of near community consensus, we can vote or otherwise implicitly agree to discount, or weigh less, contributions along the lines of “because they exist”. We (most of us?) already do this intuitively to some extent, so nothing changes too much.

Sure, it is possible to discount or regard separately cases of forced deployment. You can then imagine all the fuzziness and subjectiveness that can arise. And this still results in issues like: should Frederick get any credit for being so much better than other scrubs, even though he primarily saved those 50 turns in prologue “because he exists”? How much credit is that, if we can’t compare that to turncount/reliability elsewhere? Aren’t we trying to be somewhat objective? (should Edward get some credit for being better than Micaiah/Leo in prologue, even if not top-tier credit? Should FE6 Marcus in his game? How much? Plus, Lords.)

So really I don’t see a clean way to do it (again, the cleanest is arguably counterintuitive, just put him in top). If it matters, some arbitrary clause like “discount the chapter where a character is expected to improve efficiency the most, from the set of chapters where a character is force deployed” could objectively deal with this, if somewhat inelegant. Alternatively, some people have proposed to weigh chapters by their triviality/complexity/enemy count/map size/color scheme/number of letters in the chapter title/whatever. Or we can all carry on with some fun hand-waving and fuzziness (I’ve sorta come around, it’s tbh not soo bad).

I will save my “are all turns equal” and “objective/subjectiveness of criteria” rants for another day. >_>

tl;dr: Edward for top tier 2015

Yea. But here, instead of debating the value of a certain character compared to another, we're debating just HOW the characters should even be compared in the first place and what is and is not fair. That seems like... well... a tier-list philosophy debate than one of actual character placement.

My point is nearly all controversial tiering discussion (that is, what actually resulted in hundreds/thousands of posts), was philosophy based at heart (“no u are comparing them wrong”), even if it wasn’t called that in the past in a self-aware manner. Well that and flaming and hypocrisy. If everyone’s values aligned perfectly, there would be no debate. We could certainly rank characters solely by how many turns are gained/lost in their absence in the absolute fastest known run, but there’s no discussion there. There’s an Answer, that really only 1 person is qualified to make. (sales pitch: reliability and varying team composition introduce a lot of discussion space and accessibility, but maybe that’s just me >_>)

There are different levels of philosophical discussion, admittedly. I already said I don’t like the meta “higher purpose of tiering” stuff. I don’t mind “how to allocate resources or properly measure utility/viability”, every once in a while.

Edited March 19, 2015 by XeKr

[-Cynthia-]

Soren seems to be missing by the way.

Regarding Aran, I guess it's because if we're not using Jill or Nolan for whatever reason, he's the next best choice we can dump resources on for him to be good for 3-6. He's not really good in 3-12, 3-13, or Part 4 though (too slow to double), so...it's really kind of a waste. He doesn't have much use when he joins either.

[Baldrick]

Lol..... it follows trivially from logic that, in the end, Ilyana is penalized for not being as good as Soren, since Ilyana's position in the tier list is determined based on both situations. Not the best counterargument when you actually unwittingly defend it.

Are you not aware that you could make a tier list using only the former situation, or only the latter situation?

[Irysa]

afaik, inventory traditionally is not counted because you get that stuff even if the character doesn’t impact the chapter directly (moving, killing stuff, baiting AI, etc). The character isn’t “used” and there isn’t really player decision-making involved. However, these things can get into even trickier discussions such as, if a unit is dependent on using some resource like bexp, weapon, or a statbooster to perform well, does the other unit(s) that worked to obtain those resources get full, shared, or no downstream credit? Usually not, we don’t reward early joiners for simply getting us to the later chapters and also the later joiner’s contributions. But is this rational, for all the cases like thieving and such?

How about this argument then - If a unit can't be credited just for having a Brave Lance in their starting inventory, then what about the recruitment scenario I raised before? Should Marcia be significantly credited with recruiting Haar, as what the player can hypothetically do now as a result of having Haar as an option/resouce to use is rather similar to the way a Thief gets us particular stolen items. Given that we make a big deal out of thieves getting us resources like Gold, a Silver Card or a Delphi Shield or an extra Silver Lance or what not (and prior to the knowledge of the Desert Item trick, we attributed Desert Items to them too), and those thieves don't actually exactly make use of the items themselves, they simply have an exclusive ability to let us able to use them, doesn't the logic follow that a unit who recruits a really good unit has to get credited? If the tier list assumes full recruitment then we can kinda waive that, but I'm not sure the current one does assume full recruitment.

Ah, I like to call this the “Edward clause”. >_> (I don’t see what it specifically has to do with resource allocation or deployment though. It’s a problem when considering more objective metrics for efficiency in general)

Well, I was trying to get at this point - If the tierlist accomodates suboptimal team setups, then we can actually start assuming that we're not even using Edward in Prologue, which means his actual contribution has to be given more weight instead of being assumed. I think there's an assumption that forced deployed units are actually used to clear maps though (since the deployment is "free"), but I'm not entirely sure about that given that we rank FE6 Roy as being pretty bad because he has negligble contributions even when trained to efficient clears, wheras FE12 Marth has significant contributions in earlygame despite swordlock and can use effective weaponry to a decent degree throughout the game.

*snip*

I think that a lot of these postulations come down to this; Is this list aiming to be a ranking of units for "Contributions towards efficiently beating the game" or is it ranking units for "Usefulness (throughout the game) for efficiently completing maps". They are close, but not quite the same thing. In the latter scenario, we can remove the turns Edward or Neph saved us earlier by how many turns they're going to cost us to keep deploying instead of someone else (raises a problem if the game has very generous deployment limits but most are superfluous but I think this problem does not generally occur since just about anyone can help with Rescuing or Shoving), but if it's the former then every single unit who contributes needs to have their turns weighted equally.

Edited March 19, 2015 by Irysa

[Chiki]

Are you not aware that you could make a tier list using only the former situation, or only the latter situation?

I'm so confused. If you're using a tier list using the former situation, then you're penalizing Ilyana for being not as good as Soren, which is what I'm arguing for. I don't know whether or not you're trying to make a point for and against my view.

[Baldrick]

But if you use a tier list with the latter situation, Ilyana is not being penalised.

For a concrete example, it's like Anouleth's Kyza argument Soul cited. It's giving him resources that better units could use for a better payoff. But he is still considering the scenario where Kyza gets those resources even though it's suboptimal, because he wants to analyse the optimal Kyza.

If you're arguing that the tier list should penalise units for not being in the optimal team, fair enough. What I am saying is you can choose not to, like Anouleth does in the Kyza argument.

[Chiki]

But if you use a tier list with the latter situation, Ilyana is not being penalised.

For a concrete example, it's like Anouleth's Kyza argument Soul cited. It's giving him resources that better units could use for a better payoff. But he is still considering the scenario where Kyza gets those resources even though it's suboptimal, because he wants to analyse the optimal Kyza.

If you're arguing that the tier list should penalise units for not being in the optimal team, fair enough. What I am saying is you can choose not to, like Anouleth does in the Kyza argument.

Tier lists consider every situation, not just the latter or former. It's why Jill is so high up in the FE9 tier list even though Marcia exists.

[Zkirsche]

I think Anoulethe explained it pretty well, here:

Ok, but what about Aran? He gets doubled by pretty much everything so his durability is bad and doesn't ORKO because he doesn't double. He's locked in at 6-7 mov, is competing with units far superior to him the whole game and brings nothing to the table in return. At least Taur/Muarim are good for one chapter. Tanith is much better than him as already shown in the thread, even without LTCing the fact that she's very good at helping things go super fast is a testament to how not bad she is. Lions should probably move up for endgame shenanigans.

What is Kieran doing so low compared to these two? He's much easier to fix than Kyza is and actually will kill and has about the same availability (and is more useful in part 2 and 3-9 than Kyza is in 3-4), not to mention better movement, 1-2 range and no weed addiction.

Mid needs a major reshuffle. Geoff could probably stand to move there to.

Stefan arrives pretty much when the chapter ends and Volke has one chapter that he might as well sit out of. I put Stefan above him (only switched their spots around) because he has Alondite access, which allows for flexibility in positioning, thanks to 2-range. There's also the fact he can bless a Wyrmslayer, if you wish to count flower petals.

Volke is a much more reliable killer in 4-E-1 though and that's the only map that matters because lolkillboss. Both 2RKO almost every non-mage so rely on skills and crit, and Volke has a better proc chance of both with baselard/forged silver knife vs Vague Katti/tempest blade.

Eventual Rend + 9 Mov is a bit hard to get to when you're really far from doing any bit of substantial damage...she's really awful, and it Cat gauge doesn't help. I'd really like to raise her, but the best she's got is probably Shoving (and even other NON-Laguz are better for that than her, unfortunately). And I thought Oliver could do something with Rescue, as redundant as it may seem (yeah yeah, we have Micaiah, but that doesn't take away his capability of doing and adding flexibility to the team)

Lyre with like paragon just taking pot shots and getting some bexp will get level 30 and Rend and then contribute to 4-4 or something.

Rescue!Oliver is basically pointless with Micaiah/Rhys/Mist/Elincia/Bastian(?) and herons.

I've never even unlocked Lehran, honestly. Does he contribute in killing Auras?

Well he has the best stats in the game period so yeah he can do that. Have micaiah rescue him to the right spot then trade with her for a tome. Being good for one chapter > being mediocre in like 5. Push him to top of low imo.

Haar doesn't save any turns in chapters other than maybe 2-E

Maybe not flat out but in 3-11 he must make it significantly more reliable unless I'm misremembering? I can't imagine ferrying with Tanith is great and killing the boss would be harder.

[XeKr]

How about this argument then - If a unit can't be credited just for having a Brave Lance in their starting inventory, then what about the recruitment scenario I raised before? Should Marcia be significantly credited with recruiting Haar, as what the player can hypothetically do now as a result of having Haar as an option/resouce to use is rather similar to the way a Thief gets us particular stolen items. Given that we make a big deal out of thieves getting us resources like Gold, a Silver Card or a Delphi Shield or an extra Silver Lance or what not (and prior to the knowledge of the Desert Item trick, we attributed Desert Items to them too), and those thieves don't actually exactly make use of the items themselves, they simply have an exclusive ability to let us able to use them, doesn't the logic follow that a unit who recruits a really good unit has to get credited? If the tier list assumes full recruitment then we can kinda waive that, but I'm not sure the current one does assume full recruitment.

Well yes, that’s pretty much the trickiness I was referring to.

Considering our intuition, afaict, is to not count Haar’s contributions as Marcia’s, I think that’s also something we just implicitly decide on (we actually do this for a lot of things like why Seizing doesn’t count, and much more). Recruitment costs are often traditionally waived for the recruited being analyzed, and we can also waive recruitment “benefits” for the recruiter (note this doesn’t necessarily have to mean full recruitment).

I suppose, as in other similar situations, an argument can also be made regarding the limited information that method of ranking would convey; it says very little about Marcia as a unit combat/utility wise besides the trivial fact she recruits Haar. Compare to seizing, etc. but I find those discussions boring

Well, I was trying to get at this point - If the tierlist accomodates suboptimal team setups, then we can actually start assuming that we're not even using Edward in Prologue, which means his actual contribution has to be given more weight instead of being assumed. I think there's an assumption that forced deployed units are actually used to clear maps though (since the deployment is "free"), but I'm not entirely sure about that given that we rank FE6 Roy as being pretty bad because he has negligble contributions even when trained to efficient clears, wheras FE12 Marth has significant contributions in earlygame despite swordlock and can use effective weaponry to a decent degree throughout the game.

(I’m not entirely sure I get your point here)

Isn’t it that even in the optimal deployment case, you should consider his theoretical absence to measure his relative impact? Or else, no one “saves” any turns, because they’re either deployed or not all the time. That’s the whole “only 2 tiers exist, used vs. not-used” criticism.

I think that a lot of these postulations come down to this; Is this list aiming to be a ranking of units for "Contributions towards efficiently beating the game" or is it ranking units for "Usefulness (throughout the game) for efficiently completing maps". They are close, but not quite the same thing. In the latter scenario, we can remove the turns Edward or Neph saved us earlier by how many turns they're going to cost us to keep deploying instead of someone else (raises a problem if the game has very generous deployment limits but most are superfluous but I think this problem does not generally occur since just about anyone can help with Rescuing or Shoving), but if it's the former then every single unit who contributes needs to have their turns weighted equally.

I honestly don’t think it really matters even if we’re forced to deploy Edward/Neph in every chapter, because it’s not like we’re losing the best unit (potentially Jill/Haar) from our nonspecific team, we’re losing the 8-12th best unit which is probably a shove/rescue guy anyway. Maybe it matters a bit, but not compared to what they save in their 1 special chapter which is just abnormally super-inflated. The negatives are not high enough, because as possible, we should not be deliberately making inefficient tactical and strategic decisions (outside the act of deployment itself, ofc, as discussed earlier in detail).

So we just barely use them, even if we have to deploy them. And subsequently, they barely impact things (well I suppose there are some fringe contexts where Neph/Edward should be trained and actually significantly help later but those aren’t negative ones, just minor positive ones).

edit: And even if the negatives later were high enough, is it still fair that Edward gets so much credit for earlygame (because he's being compared to 'nothing' as I ranted about)?

Also, can you clarify what you mean by “but if it's the former then every single unit who contributes needs to have their turns weighted equally”? Should Edward/Micaiah/Leo (or Frederick/Chrom/Robin/Lissa) get equal credit for the reliable 5-6 turn of Prologue? (they shouldn't, no?)

Edited March 20, 2015 by XeKr

[Irysa]

I suppose, as in other similar situations, an argument can also be made regarding the limited information that method of ranking would convey; it says very little about Marcia as a unit combat/utility wise besides the trivial fact she recruits Haar. Compare to seizing, etc. but I find those discussions boring

I'm honestly not too bothered about it since it was just a random thought in the first place. I think it's probably best to simply make it clear that the list assumes full recruitment so someone can't get points for recruiting someone else, that's just a mandatory requirement, similar to keeping everyone alive, even though we can throw a lot of units to the wolves and not give a shit in this game.

(I’m not entirely sure I get your point here)

Isn’t it that even in the optimal deployment case, you should consider his theoretical absence to measure his relative impact? Or else, no one “saves” any turns, because they’re either deployed or not all the time. That’s the whole “only 2 tiers exist, used vs. not-used” criticism.

Unless I wholly misunderstood your post a page back or so (quite possible I suppose), when you said the tierlist should/does assume "optimal resource allocation with varying team composition", I see that as confirming that we're allowed to point out that someone may actually not even attempt to use Edward whatsoever in prologue. He's force deployed, but a player's team composition could mean we they won't use Edward even once for combat. This is why I asked if we just automatically assume force deployed units are going to be utilised, since there's an argument that since we assume a high level of play, it's completely inefficient not to actually use them.

I want to ask whether or not the premise of "not making use of Edward in Prologue" (in terms of allowing for varying team composition) is something we're actually even allowed to consider. If it isn't, then Edward doesn't have to be credited that much since we basically assume that you're going to use every single resource you have at your disposal on the map to beat it as efficiently as possible.

I honestly don’t think it really matters even if we’re forced to deploy Edward/Neph in every chapter, because it’s not like we’re losing the best unit (potentially Jill/Haar) from our nonspecific team, we’re losing the 8-12th best unit which is probably a shove/rescue guy anyway. Maybe it matters a bit, but not compared to what they save in their 1 special chapter which is just abnormally super-inflated. The negatives are not high enough, because as possible, we should not be deliberately making inefficient tactical and strategic decisions (outside the act of deployment itself, ofc, as discussed earlier in detail).

So we just barely use them, even if we have to deploy them. And subsequently, they barely impact things (well I suppose there are some fringe contexts where Neph/Edward should be trained and actually significantly help later but those aren’t negative ones, just minor positive ones).

That's fair, there are quite a few maps in this game where half the units barely do anything. At worst, we've lost a filler rescuebot/canto unit, which isn't a huge deal so doesn't really mean we can penalise them for hurting us later by continously using them. Perhaps there needs to be a clause stating that continuing to "use" a unit actually means they have to actually do something on the map? Then we can look at if theres any scenario where they can actually do anything (via investment) and compare how much they need to help compared to other units.

edit: And even if the negatives later were high enough, is it still fair that Edward gets so much credit for earlygame (because he's being compared to 'nothing' as I ranted about)?

I think it's a bit...unadvisable to go too far down that line of thinking because we can start saying that since our last couple of deployment filler slots are relatively uncontested for many maps, that all the units in that position are only being compared to "nothing" when they get any points at all for maybe shoving someone on turn 1.

Also, can you clarify what you mean by “but if it's the former then every single unit who contributes needs to have their turns weighted equally”? Do Edward/Micaiah/Leo (or Frederick/Chrom/Robin/Lissa) get equal credit for the reliable 5-6 turn of Prologue? (they shouldn't, no?)

Well, that can be explained by pointing out that removing Ed/Fred has the most significant impact on turns, wheras we don't hurt the strategy as much by removing any other single unit. But again this comes back around to what I said before in just assuming that if it's forced deployment and they significantly contribute, that you're going to use them?

[Chiki]

One way to make Edward and Nephenee's turns count for less is to consider something we already do. When a lord seizes a throne, it enables us to complete the game. It follows just from reasoning that it also enables us to complete the game in the fewest number of turns possible. If it weren't for Ike being able to seize the throne, LTC playthroughs wouldn't even be possible at all. So should Ike go into top tier because the game forces us to use him? No.

Similarly, we could just say that the turns forced units save count for less simply because they're forced in a context where you don't have much of a choice. The fewer forced units there are in a given chapter, the less the turns matter. The more forced units there are, the more the turns matter.

Edited March 20, 2015 by Chiki

[XeKr]

I kinda addressed this specific point when I asked about whether Edward/Nephenee/Frederick/Marcus should get any credit for their “forced” contributions and how much compared to others in the rest of the game.

If we assume that units that are force deployed are always used, so their contributions are free (or discounted, weighed less, etc), then you run into issues in this game where often your whole team is force deployed (or at least free to deploy). And this can also apply to other cases like early FE6 chapters with Marcus but SF dislikes lowering Jeigans.

More later, busy today.

Edited March 20, 2015 by XeKr

[Chiki]

If we assume that units that are force deployed are always used, so their contributions are free (or discounted, weighed less, etc), then you run into issues in this game where often your whole team is force deployed (or at least free to deploy). And this can also apply to other cases like early FE6 chapters with Marcus but SF dislikes lowering Jeigans.

There's more forced units in those chapters, so Marcus's contributions are worth more. Nephenee's contributions on the other hand aren't worth anywhere near as much because only her and Brom are forced.

To consider an even more extreme example of this, Ike's contributions in the Prologue of FE9 are worth absolutely nothing since he's the only forced unit.

I think this is the best, most simple way to say "all turns are not equal" without needing to add in stuff like complexity and etc. All I'm doing is just slightly updating criteria that we already have (that is, Ike doesn't get any credit for being able to seize thrones, etc.).

Edited March 20, 2015 by Chiki

[XeKr]

That's fair, there are quite a few maps in this game where half the units barely do anything. At worst, we've lost a filler rescuebot/canto unit, which isn't a huge deal so doesn't really mean we can penalise them for hurting us later by continously using them. Perhaps there needs to be a clause stating that continuing to "use" a unit actually means they have to actually do something on the map? Then we can look at if theres any scenario where they can actually do anything (via investment) and compare how much they need to help compared to other units.

The "best use" (most efficient/fast/reliable/etc) of a unit, in a given team, might just be to do nothing. Obviously depending on the unit and the team. I consider that similar to the bench and not being available. And that’s fine to me.

Yeah it should be recognized if a unit can do something remarkable with investment. I think the way to look at that is to think about how well the (potential) contribution stacks up relative to other options. Who else would you rather have to invest, who else wouldn’t you rather have. Where is said unit in this list? For example, Edward is probably middling on the list of units to invest/use for later, below Jill/Nolan/Volug/maybe Sothe, even-ish with Aran, while above Micaiah, Leo, Meg, Fiona. We can make this determination in the method you suggest, see if there are scenarios they can do stuff and how much they need to do so.

What are the chances that all of Jill/Nolan/Volug/(Sothe/Aran) are all not in play? Intuitively probably not very likely because of how suboptimal that seems. I would then say that Edward’s invested contributions are therefore probably not very significant (an improbable, so small, fraction in the ensemble of various contexts).

(fwiw, there’s more to be said about how exactly the team composition is varied, which might be worth discussing.)

I think it's a bit...unadvisable to go too far down that line of thinking because we can start saying that since our last couple of deployment filler slots are relatively uncontested for many maps, that all the units in that position are only being compared to "nothing" when they get any points at all for maybe shoving someone on turn 1.

Well I call low tiers “fringe contexts” for a reason. Low tiers don’t matter much (turncount/reliability impact is typically low if used), and don’t matter very often (only make sizable contributions in very suboptimal contexts).

Many units are compared very nearly to ‘nothing’ in the optimal case, true. However, this doesn’t hold if we allow a varied team composition. If they’re good if given a chance, they get to flex that. The whole reason Edward and other forced characters are special is because “varying” the team in those cases is far trickier.

For a more standard example. FE9!Jill does little/nothing when Marcia is in play. When she is not (a slightly suboptimal context, i.e. “what are the chances Marcia is not in play?), Jill now has massive contributions across the game. We give her a decent amount of credit for this, exactly how much is dependent on how you prefer to weigh different team compositions (indeed, “what are the chances Marcia is not in play? Again, lots to potentially say there)

If I’m misunderstanding you, please give an example of a unit that would be compared to ‘nothing’, or to later filler deployment slots that are essentially ‘nothing’, but is otherwise considered high-ish tiered. I’m pretty sure varying teams deals with this and those who still don’t get much credit are (or should be) just lower tiered.

There's more forced units in those chapters, so Marcus's contributions are worth more. Nephenee's contributions on the other hand aren't worth anywhere near as much because only her and Brom are forced.

To consider an even more extreme example of this, Ike's contributions in the Prologue of FE9 are worth absolutely nothing since he's the only forced unit.

I think this is the best, most simple way to say "all turns are not equal" without needing to add in stuff like complexity and etc. All I'm doing is just slightly updating criteria that we already have (that is, Ike doesn't get any credit for being able to seize thrones, etc.).

Fair enough, this is precisely what I meant by weighing those contributions less, the weight you propose is a function of the number of forced/free characters for that chapter, something that’s quite reasonable and relatively intuitive to do in a broad sense. As noted, we usually did this for cases truly mandatory to beat the game, such as seizing and if only 1 character was available. These are easier to resolve because they’re binary – it’s actually forced to happen instead of merely being slower and/or less reliable. If yes, then discount fully.

Do you have a proposal for elegantly, self-consistently, and numerically determining more precise weights? almost certainly maybe I can’t think right now, but all I have is multiply by the number of forced characters (N-1 maybe if we wish to simultaneously account for the 1 unit case giving no credit) and normalize somehow? But the normalization has be contextual and compare to the rest of the game somehow and it seems to get messy. >_>

But I do feel this framework for thinking helps the fuzziness issue wrt forced characters be a bit more solid.

Edited March 21, 2015 by XeKr

[Junkhead]

hey guyz nealuchi for king of heroes

https://www.youtube.com/watch?v=k6Xkir8xqjc

[Irysa]

The "best use" (most efficient/fast/reliable/etc) of a unit, in a given team, might just be to do nothing. Obviously depending on the unit and the team. I consider that similar to the bench and not being available. And that’s fine to me.

I'm sorry, can you expand this? Are you agreeing that, if a unit essentially does nothing on the map that it may as well count as not deploying/not being used?

Yeah it should be recognized if a unit can do something remarkable with investment. I think the way to look at that is to think about how well the (potential) contribution stacks up relative to other options. Who else would you rather have to invest, who else wouldn’t you rather have. Where is said unit in this list? For example, Edward is probably middling on the list of units to invest/use for later, below Jill/Nolan/Volug/maybe Sothe, even-ish with Aran, while above Micaiah, Leo, Meg, Fiona. We can make this determination in the method you suggest, see if there are scenarios they can do stuff and how much they need to do so.

What are the chances that all of Jill/Nolan/Volug/(Sothe/Aran) are all not in play? Intuitively probably not very likely because of how suboptimal that seems. I would then say that Edward’s invested contributions are therefore probably not very significant (an improbable, so small, fraction in the ensemble of various contexts).

Hmm. I mostly agree with this. It gets significantly fuzzier with the less strict segments of the game though, such as most of Part 3. There, many more characters are capable of making notable contributions if they are being utilised as our main combat units. It's not nearly as suboptimal to assume that some of the best GMs aren't being used for combat basically, since many of them come relatively combat capable and aren't as much of a pain in the ass to use as the majority of the Dawn Brigade or CRKs.

(fwiw, there’s more to be said about how exactly the team composition is varied, which might be worth discussing.)

I don't feel that there's a very effective way to determine how many different combinations of units should be considered valid. Having given it thought, all I can think of us arbitrating some kind of baseline for the Maximum TC and the Minimum Reliability of each individual map, but that seems like pretty steep amount of effort. Perhaps the individual chapter by chapter tierlist with a total points ranking to determine the final positions would be easiest after all.

Well I call low tiers “fringe contexts” for a reason. Low tiers don’t matter much (turncount/reliability impact is typically low if used), and don’t matter very often (only make sizable contributions in very suboptimal contexts).

Many units are compared very nearly to ‘nothing’ in the optimal case, true. However, this doesn’t hold if we allow a varied team composition. If they’re good if given a chance, they get to flex that. The whole reason Edward and other forced characters are special is because “varying” the team in those cases is far trickier.

For a more standard example. FE9!Jill does little/nothing when Marcia is in play. When she is not (a slightly suboptimal context, i.e. “what are the chances Marcia is not in play?), Jill now has massive contributions across the game. We give her a decent amount of credit for this, exactly how much is dependent on how you prefer to weigh different team compositions (indeed, “what are the chances Marcia is not in play? Again, lots to potentially say there)

I was going to type out a more detailed reponse to this segment but a lot of it has to do with a subjective interpretation of units like Neph and isn't really terribly meaningful. Ultimately I think Chiki's point about "less units force deployed = less weight on your turns" solves the problem mostly, but to respond to your earlier FE6 Marcus criticism, Marcus is practically essential for fast and reliable clears even past the forced deployment maps simply due to the fact he has promoted mount movement, Silver Lance access, and acceptable durability. He is helpful on basically every map up till 16x (Desert and 14x aside), and can somewhat contribute even in Ilia with a Brave Lance.

I will however, address this.

If I’m misunderstanding you, please give an example of a unit that would be compared to ‘nothing’, or to later filler deployment slots that are essentially ‘nothing’, but is otherwise considered high-ish tiered. I’m pretty sure varying teams deals with this and those who still don’t get much credit are (or should be) just lower tiered.

I think the classical example of this would be games or maps that facilitate warp/rescueskipping. If all I need to beat the map in 1 or 2 turns is the Warper and the Bosskiller (and maybe the force deployed Lord and a Dancer if it's seize), every slot but the ones occupied by those units is now being compared to "nothing". Even if good units would be occupying those slots, they just simply don't contribute. This gets taken to extremes in games like FE11 where the latter half of the game becomes Warpskip city and the majority of the cast's contributions stop existing.

Edited March 21, 2015 by Irysa

[Chiki]

Do you have a proposal for elegantly, self-consistently, and numerically determining more precise weights?

No, that's impossible to do objectively, I wouldn't bother.

Edited March 21, 2015 by Chiki

[XeKr]

hey guyz nealuchi for king of heroes

https://www.youtube.com/watch?v=k6Xkir8xqjc

Personally, I’m looking forward to what other wackiness you’ll end up doing. ^_^

I'm sorry, can you expand this? Are you agreeing that, if a unit essentially does nothing on the map that it may as well count as not deploying/not being used?

I agree with that statement in general. However, one (small) advantage of being deployable instead of not being around, is that when the better characters are not in play, that unit can make some sizable contributions. It’s small, but rewards availability and potential accordingly.

Hmm. I mostly agree with this. It gets significantly fuzzier with the less strict segments of the game though, such as most of Part 3. There, many more characters are capable of making notable contributions if they are being utilised as our main combat units. It's not nearly as suboptimal to assume that some of the best GMs aren't being used for combat basically, since many of them come relatively combat capable and aren't as much of a pain in the ass to use as the majority of the Dawn Brigade or CRKs.

This is all reasonably resolvable by the expected turncount framework, afaict.

I don't feel that there's a very effective way to determine how many different combinations of units should be considered valid. Having given it thought, all I can think of us arbitrating some kind of baseline for the Maximum TC and the Minimum Reliability of each individual map, but that seems like pretty steep amount of effort. Perhaps the individual chapter by chapter tierlist with a total points ranking to determine the final positions would be easiest after all.

The simplest and cleanest way is to say the varied team is uniformly random. Which is okay, but I feel people intuitively like to think higher tiers are in-play more often (this has lots of precedence in various discussions about assumed supports and such). There are the obvious dangers here with slippery slopes and circular arguments, so it’s a tricky topic.

A minor issue with chapter-by-chapter tierlists is continuity. That which contributes to the low turncount/high reliability in a single given chapter, might make sacrifices in later chapters (that are not counted?).

...but to respond to your earlier FE6 Marcus criticism, Marcus is practically essential for fast and reliable clears even past the forced deployment maps simply due to the fact he has promoted mount movement, Silver Lance access, and acceptable durability. He is helpful on basically every map up till 16x (Desert and 14x aside), and can somewhat contribute even in Ilia with a Brave Lance.

I actually agree, we tend to think Marcus as still very good for a decent amount of time, just not overwhelmingly good solely because of the very earlygame. I mentioned that we intuitively think in this manner already, which is visible by the fact Marcus does not (iirc) top the list. I primarily brought him up to ask, how much credit, if any, does he get for those very early chapters? It definitely feels like he should get some. It’s acceptable if it’s a decent amount, because he is so much better than the scrubs and your forced team is relatively big, but not super-inflated (like top of top tier worth).

I think the classical example of this would be games or maps that facilitate warp/rescueskipping. If all I need to beat the map in 1 or 2 turns is the Warper and the Bosskiller (and maybe the force deployed Lord and a Dancer if it's seize), every slot but the ones occupied by those units is now being compared to "nothing". Even if good units would be occupying those slots, they just simply don't contribute. This gets taken to extremes in games like FE11 where the latter half of the game becomes Warpskip city and the majority of the cast's contributions stop existing.

That’s the nature of the game. I would then consider how “good” characters are in large part how good they are as bosskillers and warpers.

The majority of cast can still contribute in suboptimal teams. The degree of which, is then how they’re roughly ranked.

No, that's impossible to do objectively, I wouldn't bother.

Hence, fuzziness.

edit: can't grammar

Edited March 21, 2015 by XeKr

[Radiant head]

Wait, why didn't Ludveck attack at two range?