The Great LTC Debate Thread (Yay? Nay? Burn in Hell?)

[dondon151]

There should be no "default" position. Many situations exist.

I don't agree with this at all. The consequence of this stance is that anyone can argue anything that they wish simply by tweaking the weighting of all possible positions. In multiplayer games, tier lists always take on a "default" position of the highest skill levels (at least, the highest skill levels known to players at the time). Why should that not be the same for a single player tier list?

[Othin]

You know what else multiplayer tier lists tell you? Which characters are good to use and which are not good to use.

Multiplayer tier lists also give you information in preparation for the uncertain circumstances of your opponent's abilities and choices. There is no point in preparing for the fixed circumstances of a pre-planned strategy that will work perfectly. There is merit in planning that strategy, but once planned, it becomes worthless, with no reason to revisit it.

[dondon151]

I am not saying the player is an idiot or doesn't know how to play the game, but rather things like resource distribution and strategies should not be set in stone, because once you do you start veering towards "how to play manual", "LTC", etc., which is again the whole point of this topic.

So, what, instead of making the tier list into a "how to play manual" for experts (which it isn't, by the way, because a tier list provides no information about how to play the game), you want to make it into a "how to play manual" for crappy players? Or an amalgam of both, which turns into a "how to play manual" for no one?

Multiplayer tier lists also give you information in preparation for the uncertain circumstances of your opponent's abilities and choices. There is no point in preparing for the fixed circumstances of a pre-planned strategy that will work perfectly. There is merit in planning that strategy, but once planned, it becomes worthless, with no reason to revisit it.

This is not true at all. Multiplayer tier lists yield absolutely no information about character matchups. You'd need a matchup guide or an in-depth character guide for that.

Furthermore, even a relatively rigid single player tier list (like some of the ones we have now) isn't limited to just one strategy. Yes, it is true that the strategy space is limited to a fairly small subset, but there are many cases where the player has options to do things differently that yield a similar outcome. We can still make informative generalizations about how well or poorly characters will perform.

Edited March 30, 2012 by dondon151

[Othin]

I didn't say anything about character matchups.

Let's look at the SSBB tier list. Ness is ranked fairly low, and I think it's a safe assumption that it's because he's generally ineffective to use. Meanwhile, in FE7 tier lists here, Athos is consistently ranked rather low. Is Athos ever ineffective to use? Of course not. His contributions are limited, but you'd still be a fool not to use him when you have him.

If the other options yield a similar outcome, then they're hardly different options at all; they're just equivalent options mirroring that one option.

[dondon151]

I didn't say anything about character matchups.

No, you didn't, but it is an important factor that you didn't consider. SSBB is not a good example to illustrate the distinction between character matchups and tier lists because the power levels between characters differ greatly and it's difficult for characters to compensate for innate weaknesses due to the game mechanics. But say that you have a mid tier character who has good or excellent matchups against 1-2 high tier characters but poor matchups against the rest of the high tier characters (I think SSBM Pikachu and Donkey Kong are adequate examples of this - or even Young Link). He would still be a mid tier character, but you cannot make the assumption that he's generally ineffective to use, even if you are inclined to do so based on his position in the tier list, because he does have a handful of good relevant matchups.

Actually, SSBM Young Link is a fantastic example of this. Armada defaults to Young Link against Jigglypuff players because the matchup is in Young Link's favor, but Young Link has problems against other top/high tier characters.

If the other options yield a similar outcome, then they're hardly different options at all; they're just equivalent options mirroring that one option.

If a certain character (say, Marcus) is ranked highly because of high movement, good offense, and good durability over a wide variety of maps, it does not make sense to not consistently apply that assumption. He cannot both be as good as he is and sandbagged simultaneously.

Edited March 30, 2012 by dondon151

[Toothache]

Perhaps the problem is that we're rating characters by a single metric. What if a tier list was rated on a combination of factors, rather than just how they fare in an LTC situation. You can consider combat potential, utility usage, and perhaps many other factors. Only then can you truly rate a unit in a tier list. The first issue with this approach is what categories of rating we want to choose, and then rating each unit in those seperate categories to get a more complete picture of their use as a unit within a tier list.

[dondon151]

You can consider combat potential, utility usage, and perhaps many other factors.

But what is the end goal of all of these considerations? Why are any of these factors good traits for a unit to have? The current tier lists still considers these factors, just with the goal of LTC in mind. The goal provides a perspective of which factors are more important that others.

[Toothache]

The problem is that the current LTC tiering criteria is somewhat vague. In breaking down the unit's worth against other units in different areas, there should be a clearer understanding about their tier position, rather than just however well that unit is argued up and down for that particular month.

[Progenitus]

But then this goes all the way back to the question of why you think turns is the only resource worth considering.

I don't agree with this at all. The consequence of this stance is that anyone can argue anything that they wish simply by tweaking the weighting of all possible positions. In multiplayer games, tier lists always take on a "default" position of the highest skill levels (at least, the highest skill levels known to players at the time). Why should that not be the same for a single player tier list?

For starters, the RNG can alter teams/strategies pretty significantly. For example, sometimes Nolan is so bad that you are better off dropping him after a certain point, and other times he becomes so good he can solo half the map. If you don't want to use Nolan as an example, then choose any unit whose performance can greatly differ depending on some RNG screwage/blessing (FE10 Jill or something).

Another point is that how do you know exactly what playthrough/team is the absolute best? There are many variables to determine it, so how would you knokw, going to Anouleth's example, whether or not comparing Artur vs Lute at 20/10 matters more than comparing them at 15/5 at some random chapter? And if your answer is "turns", then don't even bother to respond.

So, what, instead of making the tier list into a "how to play manual" for experts (which it isn't, by the way, because a tier list provides no information about how to play the game), you want to make it into a "how to play manual" for crappy players? Or an amalgam of both, which turns into a "how to play manual" for no one?

Strawman

Edited March 30, 2012 by IMPrime

[Lord Raven]

The problem is that the current LTC tiering criteria is somewhat vague. In breaking down the unit's worth against other units in different areas, there should be a clearer understanding about their tier position, rather than just however well that unit is argued up and down for that particular month.

How do you suggest weighing these factors against each other?

[Toothache]

I don't. My idea would be that each category would have equal weight, and therefore it would be a combination of factors that decide a unit's tier position. The main issue with this is deciding what categories (and how many) there should be, however.

[Seven Deadly Sins]

Another point is that how do you know exactly what playthrough/team is the absolute best? There are many variables to determine it, so how would you knokw, going to Anouleth's example, whether or not comparing Artur vs Lute at 20/10 matters more than comparing them at 15/5 at some random chapter? And if your answer is "turns", then don't even bother to respond.

You don't. That's why characters are compared in a vacuum, insofar as you only compare the effectiveness of Character A to Character B at any/all given points in time. As for the level argument, it's kinda silly. For the most part, assumed levels exist due to the effectiveness of a character (how much that unit levels on average assuming no significant favoritism, since favoritism skews tier lists) and the number of enemies present- we always know exactly how much EXP exists on most every map, it's just a matter of who gets it. A character is expected to hit a given level at a given point because of their performance in comparison to the rest of their team as well as their level and the map composition. For example: is it likely that Artur and Lute will be 20/10 at the same time? Is it likely they'll hit that level at the same time at all? Why is 20/10 the desired comparison point, rather than the level that the characters would be expected to be at given no significant favoritism.

Also, I'll submit the following for "how things really work" as opposed to the fabricated imagination of how those who attack efficiency play portray things.

http://serenesforest.net/forums/index.php?showtopic=25681&st=0&p=1531422entry1531422

Do you ever see turncount mentioned? Anything about specific strategies? Levels are given based on performance, available EXP, and how likely getting kills actually is. I even assume sub-optimal item usage for the sake of the comparison- Lute definitely has some access to items given her issues, but not even anywhere near optimal use.

[Snowy_One]

Taken from the first post of the New Mystery tier list.

Ground rules:

- Game is to be played on H3 (Lunatic) difficulty.

- Players have completed H1, and as such have access to Mixed Male Classes.

- Efficiency is defined as "Turncount per Chapter", not flat out lowest turncount.

There were more things, but I only needed to go up to three to find the counter to your claim. Turncounts are, indeed, not only mentioned but considered as part of the tiering standard. Since the only way to utilize turncounts is to follow specific strategies there is the second part (because otherwise you would get wildly divergent turncounts making it impossible to use as a standard. Additionally, you had to make it a rule that the goal was not a flat-out LTC, which implies that this has been a problem in the past when making lists. Not to mention that the character with the lower TC's are the ones who will be ranked higher, so even though it is not a 'flat-out lowest', LTC is still the goal of the tier list.

[Lord Raven]

They mean basically the same thing, what's your point? It's more or less a "not all turns are equal" concept. We still want a lower turncount per chapter ~_~

Edited March 31, 2012 by Mercenary Raven

[Othin]

Wait wait wait. THAT is what people were trying to bullshit as claiming there was some difference between LTC and Hivemind's twisted definition of "efficiency?

Hah! Ahahahahahahaha!

Never change, Hivemind. Never change.

[Aethereal]

There were more things, but I only needed to go up to three to find the counter to your claim. Turncounts are, indeed, not only mentioned but considered as part of the tiering standard. Since the only way to utilize turncounts is to follow specific strategies there is the second part (because otherwise you would get wildly divergent turncounts making it impossible to use as a standard. Additionally, you had to make it a rule that the goal was not a flat-out LTC, which implies that this has been a problem in the past when making lists. Not to mention that the character with the lower TC's are the ones who will be ranked higher, so even though it is not a 'flat-out lowest', LTC is still the goal of the tier list.

It's like every response in the last few pages just went over your head. Obviously the goal in a list aimed at going faster is to have lower turncounts. This does not imply that tier list debaters assume any specific strategies. You're being an ass, this isn't even a debate any more, it's just you venting your misguided frustrations.

[Lord Raven]

Wait wait wait. THAT is what people were trying to bullshit as claiming there was some difference between LTC and Hivemind's twisted definition of "efficiency?

Hah! Ahahahahahahaha!

Never change, Hivemind. Never change.

Lack of reliability will always be something counted against a unit. Always. LTC in its purest form allows for less reliability, and Efficiency allows for more; it just so happens that LTC as it's being used by SDS refers to what everyone else calls Efficiency. Read between the lines.

Why can't you people who dislike LTC be like Paperblade and actually put up a decent argument with regards to it without being condescending assholes? I'm serious, by the way. I get that you dislike LTC, but you don't have to come off as an asshole, because people respond to people being assholes by being assholes!

[Othin]

Hard to read between lines that don't exist. And for a community so obsessed with turn count as a solid metric, it's hard to have a solid metric for reliability when it's just sort of "assumed" without any mention of a basis to use to relate it to turns.

Being a condescending asshole can be fun at times. Also, you answered your own question:

because people respond to people being assholes by being assholes!

Edited April 1, 2012 by Othin

[OrangeCrush980]

It's like every response in the last few pages just went over your head. Obviously the goal in a list aimed at going faster is to have lower turncounts. This does not imply that tier list debaters assume any specific strategies. You're being an ass, this isn't even a debate any more, it's just you venting your misguided frustrations.

I'm playing FE6 and am on chapter 6. I'm required by the game to deploy Roy, Saul, and Dorothy. There are zillions of different strategies that can be used. But if I'm playing by SF efficiency standards, then I have to also deploy Rutger, Marcus, and some taxis and send them straight to the boss. There's still many different ways to do this, but the number of strategies is limited and they all share many similarities.

[Lord Raven]

And for a community so obsessed with turn count as a solid metric, it's hard to have a solid metric for reliability when it's just sort of "assumed" without any mention of a basis to use to relate it to turns.

I've suggested such a basis in the past and it mathematically worked out - Weighted TC = Turncount/Probability (N/P) is what a probability distribution/summation eventually becomes (considering it's what you get when you are summing a relevant geometric series). And it works damn well, because it allows "weighted" turncount outdo "shown" turncount. A lower reliability strategy creates a smaller P value with an N value, meaning the weighted turncount is higher.

Don't believe me? Look through the Jill v Marcia debate in the FE9 tier list, and look for a thread named "Defining Efficiency" in General. The implications are that a 50% reliable strategy means you multiply your turncount by 2, and a 66% reliable strategy means you multiply your turncount by 1.5, and a 1% reliable strategy means you have 100 times the weighted turncount. Simple method with a reliable basis.

Edited April 1, 2012 by Mercenary Raven

[Naglfar]

Okay, this thread is retarded, so let me lay out some reasons why efficiency tier lists are generally accepted.

-easily understood (turns, effort and reliability, what's not to get?)

-all-encompassing (every single significant factor is evaluated)

-goal-oriented (said factors are measured against the threefold values of efficiency, nothing justifies itself)

-practical (units are analysed in context, not in a vacuum)

If you have a problem with LTC tier lists, you're in luck, because I don't recall any LTC tier lists existing.

If you have a problem with efficiency tier lists, please provide a standard superior to the above for us to use instead, otherwise shut the fuck up, do your own thing and let the rest of us do ours. I am sick and fucking tired of your shitty persecution complexes.

[General Banzai]

Okay, this thread is retarded, so let me lay out some reasons why efficiency tier lists are generally accepted.

-easily understood (turns, effort and reliability, what's not to get?)

-all-encompassing (every single significant factor is evaluated)

-goal-oriented (said factors are measured against the threefold values of efficiency, nothing justifies itself)

-practical (units are analysed in context, not in a vacuum)

If you have a problem with LTC tier lists, you're in luck, because I don't recall any LTC tier lists existing.

If you have a problem with efficiency tier lists, please provide a standard superior to the above for us to use instead, otherwise shut the fuck up, do your own thing and let the rest of us do ours. I am sick and fucking tired of your shitty persecution complexes.

umad

Tier lists should be guides for new players, that's what I think.

[-Cynthia-]

If tier lists are supposed to be guides for new players, then maybe they should be on more of a chapter by chapter basis? For example I would suggest using Athos whenever he is available for an easier clear of Fire Emblem 7, but low availability makes him a Low tier unit.

[dondon151]

Tier lists should be guides for new players, that's what I think.

Please inform us how an ordered list of units can adequately serve as a guide for new players, much less anyone.

[General Banzai]

Please inform us how an ordered list of units can adequately serve as a guide for new players, much less anyone.

The same way a number from 1 to 10 can.