Jump to content

Tier List Ideas, and why they fail


Recommended Posts

Note: There are multiple attractive ways to rigorously define what efficiency is.

Exclusion

Idea: A unit is as strong as how much harder the game becomes without that unit.

Problem: Some strong units will be at the very bottom of the tier list because of an anti-synergy with another unit.

Participation

Idea: A unit is as strong as how much it's utilized in-battle with perfect play.

Problem: How to measure utilization is arbitrary, unit impact is not fully assessed, and units recruited late are penalized in a peculiar fashion.

Frequency

Idea: A unit is as strong as how often it is deployed with perfect play.

Problem: Frequency of use can be measured in different ways and might not correspond to how impactful the unit is.

Power

Idea: A unit is as strong as its statistics against other enemy units are good.

Problem: Power ignores tactical considerations.

Efficiency

Idea: A unit is as strong as how useful it is in an efficient playthrough.

Problem: This isn't well-defined.

Economical

Idea: A unit's strength is determined by how expensive a perfect opponent would make that unit, in the context of distributing costs over all units, to lessen a player's chances of victory if the player had only so much buying power.

Problem: How much buying power the player is allowed is arbitrary. It also can't address synergies and anti-synergies properly.

Meta

Idea: There must be some criteria which are well-defined and fully satisfying.

Problem: There are none because the nature of characters in the game don't lend themselves well to linear concepts of what makes a unit "good." Units have synergies and anti-synergies which keep them from being properly considered in isolation. Furthermore, there are many chapters and units change over time.

Conclusion

The point of this is to introduce the idea that it might be that there is no such thing as a well-defined and fully satisfying tier list criteria. Here the term "fully satisfying" is not given a definition but it means essentially that the criteria shouldn't have arbitrary aspects, overlook something important, or overestimate/underestimate factors. I've worked on all of these ideas so if anyone wants more explanation about an idea/problem then feel free to ask.

I think tier lists should exist as quick advice for non-expert players on how useful each unit is. They should not be studied to death to try to "carve into stone" the "correct tier list" because there is no such thing. Instead a more interesting use of time and energy would be to study team management strategies (strategies for short). Well-defined and fully satisfiable criteria exist for strategies, unlike for units by themselves. The only problem is that there are generally 10^(some large number) number of strategies, which is too many to organize into a list. Instead they can be grouped into categories and the categories can be ranked. The best strategies can also be sought after.

I know this probably won't happen but please be respectful with disagreements.

Edited by tacticsfan999
Link to comment
Share on other sites

I haven't been in any tier list discussions, but I've never liked them anyways. They can be a quick way to get a general grasp of how a character works, but there is plenty more information than some placement on a graphic. Then there is the context of the rankings. Apparently there is this character Athos that is super busted, but tier list give him a relatively low rank because his availability is bad. And I'm assuming Villager characters score low because you have to piggy back them a little, obviously this is bad for speed runs, but they become great in a normal play though.

I disliked them from fighting game perspective, because people like to think "Oh look this character is A tier! If I use them I'll be great at the game" when stuff like player actual skill is a larger factor.

Link to comment
Share on other sites

Just as there are multiple different ways to rate units, there are also multiple different ways to rate tier lists. I would say that a good tier list should ideally be well defined, rigorous, clear, and useful. However, there are often trade-offs between these different categories, which means that it's essentially impossible to create a perfect tier list. Here's what I mean by those different criteria.

First, "well-defined". By this, I mean that a tier list should state upfront what it is trying to measure. A tier list that has no accompanying text and just lists the units and their rankings is not well-defined at all. One that has accompanying text "tier list of strength of units" is slightly better, but not much. One that is labelled "tier list of units based on how useful they are on hard difficulty, assuming a player who is a series veteran but has never played this game before" is considerably better.

"Rigorous" refers to the determination of what unit goes in what tier, what rules are followed, how much room there is for argument, etc. A tier list based purely on the opinions of a single person is not at all rigorous. One arrived at purely by mathematical methods with no opnion at all is extremely rigorous.

"Clear" means how easy it is to understand the tier list's methodology and purpose and how concisely it is presented. In a lot of ways, this is why tier lists even exist in the first place. It's a lot easier to take in the information presented by a tier list than it is to read a 50,000 word essay about relative unit strengths and weaknesses. The essay probably conveys a lot more information, but the tier list is clear, quick and easy to understand.

And finally, "useful", which is, essentially, a measure of whether the tier list is giving information that anyone actually cares about. If someone looks at a tier list, it then informs which units they pick in a game, and after they have finished playing, they think that the tier lsit was helpful, then it has been useful. If they read a tier list and then ignore it, or if they follow the tier list but later regret doing so, then it has not been useful.

Of these, well-defined stands somewhat apart from the other three, since I think that it's something that all tier lists can and should aim for. The other three, however, often stand in opposition, and you often end up with a "pick any two" situation. For instance, "tier list of units by sum of all stat caps" is clear and rigorous but isn't useful. "Tier list of drafting priority in LTC drafts based on the opinions of an expert on the format" is clear and useful but not rigorous. "Tier list based on the criteria laid out in my accompanying post-doctoral mathematics thesis" could be rigorous and useful but it wouldn't be clear.

The important part in the construction of any tier list is to know from the beginning exactly what it is that you're trying to accomplish and who the target audience is, and then to focus on creating and communicating that specific tier list. If you successfully communicate what you were trying to do, then that implicitly acknowledges what you weren't trying to do and so (in theory) reduces the possibility of criticism for not being something it was never intended to be.

Link to comment
Share on other sites

I've always thought of efficiency as basically how to reduce the time of the playthrough most. That means going quickly, but above all avoiding resets.

Link to comment
Share on other sites

3 hours ago, LoneStar said:

I haven't been in any tier list discussions, but I've never liked them anyways. They can be a quick way to get a general grasp of how a character works, but there is plenty more information than some placement on a graphic. Then there is the context of the rankings. Apparently there is this character Athos that is super busted, but tier list give him a relatively low rank because his availability is bad. And I'm assuming Villager characters score low because you have to piggy back them a little, obviously this is bad for speed runs, but they become great in a normal play though.

I disliked them from fighting game perspective, because people like to think "Oh look this character is A tier! If I use them I'll be great at the game" when stuff like player actual skill is a larger factor.

Fighting games are an interesting example for tier lists. I think, in any gaming community, the tendency is usually to go straight for tier lists. However imagine instead a "tier matrix" where each fighter is ranked against each other fighter. If you had this plus character usage stats then you might even be able to project the tier matrix into a tier list. I say project to suggest that information would be lost in the process, like projecting a cube onto its shadow. Yeah availability is a whole other topic but I don't think it's a good fundamental concept.

 

12 minutes ago, lenticular said:

Just as there are multiple different ways to rate units, there are also multiple different ways to rate tier lists. I would say that a good tier list should ideally be well defined, rigorous, clear, and useful. However, there are often trade-offs between these different categories, which means that it's essentially impossible to create a perfect tier list. Here's what I mean by those different criteria.

First, "well-defined". By this, I mean that a tier list should state upfront what it is trying to measure. A tier list that has no accompanying text and just lists the units and their rankings is not well-defined at all. One that has accompanying text "tier list of strength of units" is slightly better, but not much. One that is labelled "tier list of units based on how useful they are on hard difficulty, assuming a player who is a series veteran but has never played this game before" is considerably better.

"Rigorous" refers to the determination of what unit goes in what tier, what rules are followed, how much room there is for argument, etc. A tier list based purely on the opinions of a single person is not at all rigorous. One arrived at purely by mathematical methods with no opnion at all is extremely rigorous.

"Clear" means how easy it is to understand the tier list's methodology and purpose and how concisely it is presented. In a lot of ways, this is why tier lists even exist in the first place. It's a lot easier to take in the information presented by a tier list than it is to read a 50,000 word essay about relative unit strengths and weaknesses. The essay probably conveys a lot more information, but the tier list is clear, quick and easy to understand.

And finally, "useful", which is, essentially, a measure of whether the tier list is giving information that anyone actually cares about. If someone looks at a tier list, it then informs which units they pick in a game, and after they have finished playing, they think that the tier lsit was helpful, then it has been useful. If they read a tier list and then ignore it, or if they follow the tier list but later regret doing so, then it has not been useful.

Of these, well-defined stands somewhat apart from the other three, since I think that it's something that all tier lists can and should aim for. The other three, however, often stand in opposition, and you often end up with a "pick any two" situation. For instance, "tier list of units by sum of all stat caps" is clear and rigorous but isn't useful. "Tier list of drafting priority in LTC drafts based on the opinions of an expert on the format" is clear and useful but not rigorous. "Tier list based on the criteria laid out in my accompanying post-doctoral mathematics thesis" could be rigorous and useful but it wouldn't be clear.

The important part in the construction of any tier list is to know from the beginning exactly what it is that you're trying to accomplish and who the target audience is, and then to focus on creating and communicating that specific tier list. If you successfully communicate what you were trying to do, then that implicitly acknowledges what you weren't trying to do and so (in theory) reduces the possibility of criticism for not being something it was never intended to be.

I appreciate the flexibility with which you view tier lists. I agree with what you're saying. I actually put the "stat point sum" tier list on my blog as an example of something clearly defined but not terribly interesting. I think your "rigorous" might be what I'm referring to as well-defined. To me something is well-defined if it is defined unambiguously. For example, pi is twice the first positive zero of the cosine function. So when I speak of well-defined tier lists, I mean tier lists defined with equivalent precision. 

 

7 minutes ago, Parrhesia said:

I've always thought of efficiency as basically how to reduce the time of the playthrough most. That means going quickly, but above all avoiding resets.

That's what I had in mind too. I tried to capture that idea in terms of a reliability-speed tradeoff. Like you said, you don't want to die too much (reliability) but you also want to go fast (speed). So I think there are many "efficiencies" because there are many different tradeoffs between reliability and speed that could be considered.

Link to comment
Share on other sites

13 hours ago, tacticsfan999 said:

Idea: A unit is as strong as its statistics against other enemy units are good.

Problem: Power ignores tactical considerations.

Hard disagree, mainly because being the amongst the best in accuracy, defense, and offensive power means that unit can reliably solo a certain section of the map. Or it relegates them into a certain role, like having Ignatz kill off the high evasion units, and using Lysisthia to kill an Pegasus Knight with Luna. Plus, the A.I. is kind of predictable in 3H.

Edited by Armchair General
Link to comment
Share on other sites

11 minutes ago, Armchair General said:

Hard disagree, mainly because being the amongst the best in accuracy, defense, and offensive power means that unit can reliably solo a certain section of the map. Or it relegates them into a certain role, like having Ignatz kill off the high evasion units, and using Lysisthia to kill an Pegasus Knight with Luna. Plus, the A.I. is kind of predictable in 3H.

There are factors this misses, though. An FE8 Swordmaster will be worse than a same-stats FE8 Berserker, for instance, because only the latter has broad access to 1-2 range. Meanwhile, an FE9 Warrior will be worse than a same-stats FE9 Axe-Bow Paladin, because the latter has better mobility. Heightened attack range and mobility means more chances to kill enemies.

14 hours ago, tacticsfan999 said:

The point of this is to introduce the idea that it might be that there is no such thing as a well-defined and fully satisfying tier list criteria. Here the term "fully satisfying" is not given a definition but it means essentially that the criteria shouldn't have arbitrary aspects, overlook something important, or overestimate/underestimate factors. I've worked on all of these ideas so if anyone wants more explanation about an idea/problem then feel free to ask.

Was expecting this to be a general hate-jerk about tier lists, but honestly I agree with your general thesis. Namely, that tier lists have utility, but their validity is dependent on a certain standard-of-judgement (the nature of which is not widely agreed upon). Good post!

Link to comment
Share on other sites

1 hour ago, Shanty Pete's 1st Mate said:

There are factors this misses, though. An FE8 Swordmaster will be worse than a same-stats FE8 Berserker, for instance, because only the latter has broad access to 1-2 range. Meanwhile, an FE9 Warrior will be worse than a same-stats FE9 Axe-Bow Paladin, because the latter has better mobility. Heightened attack range and mobility means more chances to kill enemies

My first FE game was Awakening and I never really had the chance to buy the older games. Although, what you said does carries over to Awakening.

 

Now that I've thought about it, Fates is  probably the only one with throwable swords that isn't infused by the power of magic.

Link to comment
Share on other sites

3 hours ago, Armchair General said:

My first FE game was Awakening and I never really had the chance to buy the older games. Although, what you said does carries over to Awakening.

 

Now that I've thought about it, Fates is  probably the only one with throwable swords that isn't infused by the power of magic.

In fact, Radiant Dawn also has widely available 1-2 range swords. The Wind Edge, namely - but also the Storm Sword and Tempest Blade. Aesthetically, they somewhat resemble magical swords from prior games, attacking with a blade of wind. But like Fates' Kodachi, they deal physical damage.

Link to comment
Share on other sites

11 hours ago, Armchair General said:

Hard disagree, mainly because being the amongst the best in accuracy, defense, and offensive power means that unit can reliably solo a certain section of the map. Or it relegates them into a certain role, like having Ignatz kill off the high evasion units, and using Lysisthia to kill an Pegasus Knight with Luna. Plus, the A.I. is kind of predictable in 3H.

I should have maybe said that power doesn't take into account all tactical considerations. Power would be a formula based on unit statistics, but how do you factor in things like movement, range, flying, and staves? There are more subtle things to consider too. For example, maybe a unit is really good against several units on a map, but maybe those units aren't bunched up into a difficult mob. An even more subtle consideration would be effective weapons. Maybe a unit can use an effective weapon to OHKO one enemy unit, and a silver weapon to 2HKO another enemy unit, but it can't necessarily do both at the same time (say if it was baiting both enemies). I wouldn't argue that power isn't really important but I would argue that if you were trying single out a uniquely special mathematical formula for it, with the hopes of explaining unit strength in a satisfying and generalized sense, then you wouldn't be able to. Let me know if you still disagree or if you have any thoughts.

 

11 hours ago, Shanty Pete's 1st Mate said:

Was expecting this to be a general hate-jerk about tier lists, but honestly I agree with your general thesis. Namely, that tier lists have utility, but their validity is dependent on a certain standard-of-judgement (the nature of which is not widely agreed upon). Good post!

Thanks! I am quite critical here but not of tier lists in general. Personally I've used a tier list for every Fire Emblem game I've played. I'm more critical of Fire Emblem character tier list discourse when it attempts to elevate itself to some kind of level of debating "the truth," often as if every single unit placement has a right answer. My theory is that the reason this happens is because people don't put more effort into defining what their tier lists are trying to do in the first place. Everyone has a different perspective on the same words and concepts. I go a little further and say that if people did try to really define what it is that they're trying to do with tier lists, then their definitions would never be wholly satisfactory in terms of what players care about. There would always either be some seemingly arbitrary choice involved or some aspect of gameplay that's not really addressed in a satisfying way, even if it still leads to a fairly meaningful tier list. So your "standard-of-judgment" I think is closely related to what I mean by "well-defined criteria."

Edited by tacticsfan999
Link to comment
Share on other sites

10 hours ago, tacticsfan999 said:

I should have maybe said that power doesn't take into account all tactical considerations. Power would be a formula based on unit statistics, but how do you factor in things like movement, range, flying, and staves? There are more subtle things to consider too. For example, maybe a unit is really good against several units on a map, but maybe those units aren't bunched up into a difficult mob. An even more subtle consideration would be effective weapons. Maybe a unit can use an effective weapon to OHKO one enemy unit, and a silver weapon to 2HKO another enemy unit, but it can't necessarily do both at the same time (say if it was baiting both enemies). I wouldn't argue that power isn't really important but I would argue that if you were trying single out a uniquely special mathematical formula for it, with the hopes of explaining unit strength in a satisfying and generalized sense, then you wouldn't be able to. Let me know if you still disagree or if you have any thoughts.

 

 

It kind of depends on exactly what were talking about, here. As in, there's a lot of roles to fulfill in FE, that where one guy doesn't really fill all of them unless they can have high mobility or they can use magic and conventional weaponry. For instance, you can keep a Dark Knight in 3H as an Swiss Army knife and pair him with another one of my Bow Knights or a Great Knight. You could also have the Bow Knight with an anti-cav spear and now the duo can reliably fend off half of the Master Classes. Mines usually hold their areas, but I also have the luxury of doing a hit & run most of the time. 

 

Also got a War Master set up to crit someone when he's punching them for when someone doesn't entirely go down in one round, but it's not really worth it to send someone else to back him up.

 

The rest are mostly mages who I only use to help clear a stalemate that I set up and for healing.

 

Overall, it ultimately depends on how you play and how's the level is set up. Since 3H gives you 10 units to work with, which is barely enough to fight in 3 directions. Other games give you enough deployment slots to have four teams of soldiers that are essentially a pair guys going at it and someone replacing them when I have to heal them. Plus, you'll have room for your thief so you can raid the occasional chest.

 

However, there isn't a stat to rule them all, because 4 of them are tied to two different damage formulas. You might argue that Speed is the more important one, since it reduces the likelihood of getting hit (twice); but Skill/Dexterity are also important for the opposite reason, along with triggering certain skills.

Edited by Armchair General
Link to comment
Share on other sites

I should've done this sooner, but I wasn't here.

On 5/6/2021 at 6:35 PM, tacticsfan999 said:

Exclusion

Idea: A unit is as strong as how much harder the game becomes without that unit.

Problem: Some strong units will be at the very bottom of the tier list because of an anti-synergy with another unit.

What are specific examples of this?  Preferably across multiple games.

On 5/6/2021 at 6:35 PM, tacticsfan999 said:

Participation

Idea: A unit is as strong as how much it's utilized in-battle with perfect play.

Problem: How to measure utilization is arbitrary, unit impact is not fully assessed, and units recruited late are penalized in a peculiar fashion.

You'll have to define "utilized", since staffbots tend to be rated pretty darn high, yet they see no combat.

On 5/6/2021 at 6:35 PM, tacticsfan999 said:

Frequency

Idea: A unit is as strong as how often it is deployed with perfect play.

Problem: Frequency of use can be measured in different ways and might not correspond to how impactful the unit is.

Again, what are examples of this?  And why is it important to a tier list?

On 5/6/2021 at 6:35 PM, tacticsfan999 said:

Power

Idea: A unit is as strong as its statistics against other enemy units are good.

Problem: Power ignores tactical considerations.

This covers some of the unit considerations, not all.  I don't think this is a particularly honest comment, if you're going to point out tactical considerations as an issue.

On 5/6/2021 at 6:35 PM, tacticsfan999 said:

Efficiency

Idea: A unit is as strong as how useful it is in an efficient playthrough.

Problem: This isn't well-defined.

That's up to the person making the tier list to define.  Which is on the maker of the tier list more than anything.

On 5/6/2021 at 6:35 PM, tacticsfan999 said:

Economical

Idea: A unit's strength is determined by how expensive a perfect opponent would make that unit, in the context of distributing costs over all units, to lessen a player's chances of victory if the player had only so much buying power.

Problem: How much buying power the player is allowed is arbitrary. It also can't address synergies and anti-synergies properly.

Are you talking about item allocation?

On 5/6/2021 at 6:35 PM, tacticsfan999 said:

Meta

Idea: There must be some criteria which are well-defined and fully satisfying.

Problem: There are none because the nature of characters in the game don't lend themselves well to linear concepts of what makes a unit "good." Units have synergies and anti-synergies which keep them from being properly considered in isolation. Furthermore, there are many chapters and units change over time.

Er, that usually falls under the purpose of the tier list (efficiency, LTC, full recruitment, etc.).

---

It's well and good to criticize tier lists, but don't fall into the same traps as the tier lists you're criticizing.  Explain yourself, using concrete examples as well as tier lists.

Link to comment
Share on other sites

I don't think your "this comment isn't honest" is an honest comment, especially since I already explained my point of view on it in this thread, which wasn't addressed. I also said that I could elaborate on anything so all anyone needs to do is ask and they'll get plenty of details. And it's extremely awkward to have to deal with rudeness from a moderator. I'm locking myself out bye all. 

Link to comment
Share on other sites

. . .well, that's an even better way to convince me that I was right about this entire thing NOT being honest.  Thanks for saving me a lot of time~!

Link to comment
Share on other sites

Guest
This topic is now closed to further replies.
 Share

  • Recently Browsing   0 members

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