ITT I rank the characters

[Florete]

I think Smash is trying to say that Tormod needs to be compared to Leharan to show who is better during that time rather than comparing Leharan to a non-existing unit.

Lehran has a single map. Why should other characters care about his existence? Why should other characters be forced to go through suck just to be on the same map as him?

I have a feeling smash's reply won't be a pleasant one.

[Rodykitty]

Lehran has a single map.

Oh, so he's like a Gato/Athos/whatever unit?

Then yeah, Tormod should definitely be ranked higher than him due to Gato-like units being the absolute neutrality when it comes to liability/benefit balance scale. While it may be true that a utility unit may suck later on, you're not feeding a utility unit EXP or kills in order for them to have whatever utility they may have, while for a lot of units feeding EXP is a requirement for them to be useful, so in order for them to be good they have to be good throughout the game to make taking EXP worth it.

[Interceptor]

Think about it, if Tormod had a storyline death after part 1, would you still consider Lehran > him?

And this is basically the crux of the problem, in a one-sentence soundbite.

These are smash's rankings, subject to nobody's approval except his own. But, he posted them in a thread on a public forum, meaning that people are going to comment about them. I think that it's been correctly pointed out that assuming deployment for a ranked unit is completely at odds with the stated goals for ranking: beating HM as quickly and efficiently as possible. It works for excellent units, and it works for units that go from bad to good (aka Rolf), but it fails in situations where a unit starts excellent but then becomes a liability.

This is just another piece of data in a long series (last episode: let's ignore Shinon's manifest Enemy Phase problems) that increasingly point towards the rankings list being basically farcical, except that it's not really all that funny, either.

[Progenitus]

By getting less kills, they are generating less utility than the team, which means that overall they have negative utility since I could have deployed someone else that could actually cross the fucking water.

Except getting kills in 3-7 doesn't do anything towards helping us beat the chapter. Did you even read my part about 3-7?

Then again, I'm talking to the guy who lost to prog in a debate while defending a unit three tiers higher than his, and also has sasuke has his avatar. I shouldn't have expected anything better.

And of course he's taking exp away from the team. Everyone is doing this. E.g. Ike got 15 exp. Oops, he's taking exp away from the team. Haar gets 10 exp. Oops, he;s taking exp away from the team. Why don't we not kill anyone so that no exp is taken away from anyone/ That's a great idea, ja?

Well, you see, the difference is that only Boyd is the one taking any kills/exp from the team in 3-7, since Oscar isn't because he barely fights anything. Generally, in a comparison between two units where both exist in the same map, they both have the same disadvantage of taking exp away from the team when they get a kill, and so we cancel them out and not care about it. However, when one unit exists and the other doesn't, that disadvantage DOES appear because the other unit doesn't have it, and we DO care.

For example, if we had Lyre vs Stefan, we definitely care if Lyre is taking exp/has a crappy performance in part 3. 3-7 is the same case for Oscar, except we replace Lyre with Boyd (although he doesn't suck as much) and Stefan with Oscar. Boyd getting kills in 3-7 regardless of his actual performance doesn't give him a free ride towards garnering positive utility, even if we assumed getting kills actually helped us beat the chapter.

If your logic was actually correct, we would get ridiculous conclusions such as Lyre > Stefan because Lyre getting kills in part 3, regardless of her performance or the fact that the exp could've gone to other units, would make her accumulate positive utility.

Why does he have to "overcome" that to be considered better than Lehran? It doesn't make sense. Why should he acquire negative utility when he's already acquired enough positive to be considered better than Lehran and we have the choice to field him or not field him?

Because we're trying to measure a certain unit's performance, which requires that we actually field the unit.

Think about it, if Tormod had a storyline death after part 1, would you still consider Lehran > him?

If Tormod, or any unit for that matter, had a storyline death at any point in the game (unless it was after the final chapter), I probably wouldn't even rate him in the first place. I didn't rate the BK, who has a storyline death. Fuck, the BK isn't even seriously tiered in YOUR tier list topic. What a nice theoretical question you brought up. You'd go any lengths to try to disprove me, even if it includes using a double standard.

Oh, and why the fuck would we give Tormod a 4-E slot and not even use/field him? That's just simply retarded. We don't assume the player is retarded, smash.

Except we also assume that the player will occasionally play low tier units, even though that's generally seen as "retarded". So do we suddenly assume the player never uses them?

[dondon151]

Because we're trying to measure a certain unit's performance, which requires that we actually field the unit.

No we don't.

If we field a unit that's terrible enough to be a negative, the ultimate goal is to raise that unit to the point where he's a positive. For a unit like Tormod, where it is literally impossible to make him a nonnegligible positive after 4-4, there is no point in fielding him.

[Vykan12]

Consider this: If 2 units have negative utility their entire existence (Eg/ Kyza and Pelleas) then we optimize their use by not fielding them. That sure makes a great comparison: an empty slot vs an empty slot.

Similarly, we are walking a fine line when sub-par units with high availability only get examined in the period where they offer positive utility. Instead of Edward vs Ilyana being 18 chapters vs 21, it gets reduced to like 5 chapters vs 3. For the raw sake of argument, that's not an interesting premise to debate under as it oversimplifies too many comparisons.

[Interceptor]

Consider this: If 2 units have negative utility their entire existence (Eg/ Kyza and Pelleas) then we optimize their use by not fielding them. That sure makes a great comparison: an empty slot vs an empty slot.

Your premise is wrong before you even started your point, since those units don't have "negative utility" for their entire existence.

Aside from that, the issue is not that Pellas vs. Kyza comparisons where both units are deployed in their full suckiness can't be done, it's that forcing crappy units on the field runs directly counter to the stated goals of this ranking thread.

[dondon151]

Similarly, we are walking a fine line when sub-par units with high availability only get examined in the period where they offer positive utility. Instead of Edward vs Ilyana being 18 chapters vs 21, it gets reduced to like 5 chapters vs 3. For the raw sake of argument, that's not an interesting premise to debate under as it oversimplifies too many comparisons.

And we also examine them in the period where they suck, and determine who sucks more, so it's still 18 chapters against 21.

[Progenitus]

If we field a unit that's terrible enough to be a negative, the ultimate goal is to raise that unit to the point where he's a positive. For a unit like Tormod, where it is literally impossible to make him a nonnegligible positive after 4-4, there is no point in fielding him.

If Tormod is so bad after he rejoins in part 4 that we literally never get anything good out of him, that says something about his actual performance.

And we also examine them in the period where they suck, and determine who sucks more, so it's still 18 chapters against 21.

So why should Tormod get special treatment and we go "let's ignore his period of suck and just drop him"?

[dondon151]

If Tormod is so bad after he rejoins in part 4 that we literally never get anything good out of him, that says something about his actual performance.

So why should Tormod get special treatment and we go "let's ignore his period of suck and just drop him"?

You quoted my two posts that exactly answer your question.

[Vykan12]

Your premise is wrong before you even started your point, since those units don't have "negative utility" for their entire existence.

If you want to walk that road, then no unit really has negative utility when used as sparingly as possible. That still doesn’t change the point though: Low tiers are more a competition of who gets forced than who is actually better in the long run.

I don’t mind if Smash’s topic adheres to such logic, my complaint is mainly directed towards tier lists that have more intricate criteria for the sake of having more interesting arguments.

[Progenitus]

You quoted my two posts that exactly answer your question.

So you're telling me you used a double standard for Tormod? Okay.

I don’t mind if Smash’s topic adheres to such logic, my complaint is mainly directed towards tier lists that have more intricate criteria for the sake of having more interesting arguments.

Well, this topic is basically a different form of a tier list. That was what I had in mind when I made it.

[Narga_Rocks]

If Tormod, or any unit for that matter, had a storyline death at any point in the game (unless it was after the final chapter), I probably wouldn't even rate him in the first place. I didn't rate the BK, who has a storyline death. Fuck, the BK isn't even seriously tiered in YOUR tier list topic. What a nice theoretical question you brought up. You'd go any lengths to try to disprove me, even if it includes using a double standard.

You know, that's something I don't quite get, anyway. BK is available in 1-9, 1-E, 3-6, and sorta 4-3. He has way more availability than Lehran, Gareth, and Nasir. Especially since 4-E-4 and 4-E-5 can be very short. He even arguably has more availability than Stefan and Renning and Caineghis and Giffca and Ena and Kurth, given how short all of the 4-E chapters are. Why don't we tier the Black Knight? Who cares that we kill him later on and he isn't available for the final chapter.

Except we also assume that the player will occasionally play low tier units, even though that's generally seen as "retarded". So do we suddenly assume the player never uses them?

Um, no we don't. At least, I wouldn't when you are looking at rating Gatrie's or Mia's performance, or really anyone in high or top tier. I use units like Ilyana and others who aren't so good because I like them, but I'm not going to say that a single high tier unit is entitled to a resource because he/she uses it better than a bunch of low tiers and we might not use the other high tiers. We compare Gatrie's crown usage to all the other units, but mainly the good ones. The only reason we assume the player is using low tier units is so that we can place them. Really there are lots of units that in a normal playthrough are effectively equal, because nobody is going to use them. But in order to determine where those units place, we have to assume they are in play and see which one isn't quite as bad an idea as some other unit.

Comparing Kyza vs. Pelleas, like Vykan suggested, isn't the same as comparing Tormod and Lehran. Assuming vykan's premise is correct, whether it is or not, then they are negative their whole time, but since we can't leave two units equal in a tier list you have to have some kind of tie breaker, so they are fielded until they achieve their maximum utility, then you drop them. Then compare the two maximum utilities. If they never achieve a maximum because they aren't that great until 4-E, then they are fielded the entire time in the comparison. If Kyza somehow became awesome in 3-E and 4-2 but started sucking again in 4-5, then the comparison of Kyza vs. Pelleas would then go like this:

The suck of Kyza before 3-E and 4-2, plus how good he is in 3-E and 4-2, minus the cost of the loss of experience in training Kyza when we know we are going to drop him later, versus Pelleas' career. Pelleas' career could play out similarly to Kyza, raising then dropping, or it could be raising till the end, depending on what his situation is.

Because Kyza doesn't become awesome for two chapters, we don't compare them that way, and we have to assume usage throughout. But the idea is that we are still attempting to achieve Kyza's maximum utility, and Pelleas' as well.

Of course, you could argue that they reached their maximum utility before joining because it is downhill from there, but you have to compare them anyway so that's what you are left with. Basically, it goes down before it goes up, yes, but it eventually goes up because you can train them and so they'll get better and start being useful. Once that curve stops going up and instead starts going down again, you cut it off at that point, even if the local maximum is not higher than the 0 at the beginning. Unfortunately it's the only way to compare two not so great units.

If my point isn't clear, then for any two units, say they both max out at 0, so it's a tie. Now you have to break the tie. So you look at the next best local maximum on their utility curve. If that's a tie, look at the next one, etc. If unit A's next best spike is better than unit B's, the win goes to unit A. (Oh, and the "utility curve" is of course cumulative, not how good a unit is in individual chapters.)

A unit like Tormod, on the other hand, is good when he joins, then sucks later. As such, he already achieved his maximum utility at the end of 1-E. So why keep using him afterwards? All you are doing at that point is driving down his utility when he already spiked. And he didn't use experience for 5 chapters to become good, unlike what my imaginary Kyza would have done. So Tormod just went up from the beginning and then he'll go back down in 4-4, but he doesn't have to exist anymore.

So Tormod vs. Lehran is basically comparing how good he is in 1-7, 1-8, 1-E minus the experience that went to him rather than Sothe/Volug/whatever versus Lehran's use in 4-E-5. If Tormod wasn't good enough in those chapters, then fine. If he was, then fine.

Edited July 28, 2009 by Narga_Rocks

[Interceptor]

If you want to walk that road, then no unit really has negative utility when used as sparingly as possible. That still doesn’t change the point though: Low tiers are more a competition of who gets forced than who is actually better in the long run.

Like I said in the tier thread, I don't subscribe to smash's theory of negative utility, I am just pointing out where he's using double-standards being inconsistent.

I don’t mind if Smash’s topic adheres to such logic, my complaint is mainly directed towards tier lists that have more intricate criteria for the sake of having more interesting arguments.

I'm all for tier lists that encourage interesting arguments, since that's all that tier lists are good for. Arguments, that is. But this is not a tier list. This is "smash ranks the characters", and he can't even stay focused on his mission statement. The whole idea of throwing a good-unit-turned-bad into Endgame and souring their entire contribution to efficiency by sucking at life is failing to measure the unit's actual contribution to ganme completion accurately.

[Narga_Rocks]

Like I said in the tier thread, I don't subscribe to smash's theory of negative utility, I am just pointing out where he's using double-standards being inconsistent.

I don't know. I'm kinda okay with the idea of negative utility. I don't think I agree with smash's version of negative utility, of course, but when one unit is so much worse than another that it takes 1 or 2 more turns to finish a map and I miss the max bexp cutoff because of it, I don't like that. On the other hand, we could just add up everything good a unit does and go from there. It would probably work just fine. But I'm not sure how to go about doing that without rewarding availability more than maybe we should.

[dondon151]

So you're telling me you used a double standard for Tormod? Okay.

So you're telling me that your level of reading comprehension isn't sufficient for FE debates? Okay.

Here's what I said: a unit will only be used when he is a negative if the long term goal is to make him a positive. Tormod rejoins in 4-4 with almost no chance at being a positive. The Edward and Ilyana examples have units gaining utility as they progress through the game; therefore there is incentive to use them when they have negative utility because it will balance out or go in the green in the long run. There is no incentive to use Tormod when he rejoins in 4-4 because it will not balance out or go in the green in the long run (ignoring his part 1 performance, of course).

[Vykan12]

Here's what I said: a unit will only be used when he is a negative if the long term goal is to make him a positive.

The problem is whether that unit will be a net positive or not. If it's the latter case, there's no point in using the unit to begin with, so the best position they can manage is mid tier by not being deployed ever.

In regards to the specific examples used, many would argue that Edward's part 1+3 struggle easily outweigh whatever good combat he's providing in part 4, hence no green light for him.

I don't subscribe to smash's theory of negative utility

Can you elaborate on this please? As in, your view on +/- utility, not how you interpret "smash's theory".

Edited July 28, 2009 by Vykan12

[dondon151]

The problem is whether that unit will be a net positive or not. If it's the latter case, there's no point in using the unit to begin with, so the best position they can manage is mid tier by not being deployed ever.

What makes you think that 0 utility is mid tier? Under this system, 0 utility is bottom tier.

Anyway, I don't really subscribe to the theory of negative utility either because all of those circumstances can be easily avoided by just... avoiding them. Plastering negative utility on characters seems to be consistent with plastering idiocy on the assumed player. Additionally, the amount of negative utility that a character accumulates is largely variable based on playstyle.

[Narga_Rocks]

What makes you think that 0 utility is mid tier? Under this system, 0 utility is bottom tier.

Anyway, I don't really subscribe to the theory of negative utility either because all of those circumstances can be easily avoided by just... avoiding them. Plastering negative utility on characters seems to be consistent with plastering idiocy on the assumed player. Additionally, the amount of negative utility that a character accumulates is largely variable based on playstyle.

I'm kind of wondering really if calling something negative utility is just semantics. At least, among units with equal availability. If you stick your 0 line at the bottom to make everything positive, it isn't much different than if you stick it half way up (the middle unit's goodness, or whatever) and call everything underneath negative.

The trouble is this makes it very hard for units that are merely okay for 20 chapters to be better than a unit that is great for 1 chapter, when those units that existed for 20 chapters were clearly more useful to us throughout the game. So the system of negative utility punishes availability for units that aren't so great, or at least it punishes them for being 5th best out of 8 or something similar.

Without the system of negative utility, however, a unit that really sucks for 20 chapters might then be considered better than a unit that is great for 1 chapter, simply because adding up tiny points along the way could result in making a unit that isn't better than another unit better. Or whatever.

Neither system is flawless. Done well either system could probably get it right, at least with a couple of loopholes along the way to avoid silly things like a unit being good in one chapter becoming better than a unit that is great in two chapters because the second unit ends up not so great even though we don't have to use the second unit beyond the point it reached maximum utility. Also we don't have to apply the same tactics to a unit that is never good as we do to a unit that starts great and ends bad. Or to a unit that starts bad and ends great. They are all completely different scenarios and units, so why should they be treated the exact same?

[Florete]

Because we're trying to measure a certain unit's performance, which requires that we actually field the unit.

But Tormod was already fielded. He's already gained positive utility. Why should he be forced to go into the negative when the choice to not do that is there?

If Tormod, or any unit for that matter, had a storyline death at any point in the game (unless it was after the final chapter), I probably wouldn't even rate him in the first place. I didn't rate the BK, who has a storyline death. Fuck, the BK isn't even seriously tiered in YOUR tier list topic. What a nice theoretical question you brought up. You'd go any lengths to try to disprove me, even if it includes using a double standard.

I've brought up possibly tiering the Black Knight before, but it probably got lost. I even rated him in my ranking topic before it got eaten in the time warp. There are no double standards here, smash.

So, let me ask the question again, with a bit of an add-on: If Tormod had a storyline death, and you still decided to rank him, would Lehran still be > him?

Except we also assume that the player will occasionally play low tier units, even though that's generally seen as "retarded". So do we suddenly assume the player never uses them?

We only assume the player will use them so we can rank them. If we never used them, the tier list would end at Upper Mid. Characters like Tormod, Sothe, and Geoffrey have periods where we use them because they're good, and so they get ranked well for it. Characters like Lyre and Fiona suck forever (or at least for quite a long time), and they have to go through that suck to be rated, so they get a low rating as a result.

It's retarded to field a unit for no reason, or to field them for negative utility when they've already gone through good times. You yourself said, and I quote:

This is why we have to at least dedicate a unit slot for Tormod in every chapter he's available in. I never said we should necessarily USE Tormod, or even field him.

So you sent Tormod to 4-E and didn't even assume he was fielded. Please explain how that isn't completely retarded.

[Vykan12]

Anyway, I don't really subscribe to the theory of negative utility either because all of those circumstances can be easily avoided by just... avoiding them. Plastering negative utility on characters seems to be consistent with plastering idiocy on the assumed player.

Are you saying negative utility doesn't exist so long as the player has common sense?

If you're in the bottom tier, then basically anyone else in the game can be deployed in your place while giving higher returns. Moreover, the combat returns you get from using someone like Lyre probably doesn't even outweigh the basic resources she gets (olivi grass, concoctions), not to mention how badly she'll damage your enemy phase if she isn't protected.

You might wonder "by that logic, wouldn't anyone besides the top 10 units in a given army have negative utility? After all, there's always a better unit to take their place." Try looking at it from a probabilistic standpoint. If you're upper-mid tier, it's more likely that you're taking up an inferior unit's slot than you are a superior one's. If you're mid tier, the likelihood is equal both ways.

The other interpretation I have is comparing a unit to the benefits of an empty slot (ie more concentrated resources for your team in play). Obviously you get more benefit from Ike and Haar's exp gains going up by 10% than you do from adding Lyre to the roster.

Edited July 28, 2009 by Vykan12

[-Cynthia-]

I still don't understand how Heather is so high under this method, since her negative utility in chapters where she isn't stealing/finding items in enormous (hello 4-E).

[Vykan12]

Assuming we use Heather in 4-E, our best course of action would be to hide her in a corner. Then the loss in efficiency is only the loss of 1 combat unit. Even assuming each combat unit contributes equally, that's a drop of efficiency of 10%, and that's if we disregard forced units (eg/ a heron, kurth, ena). In practice, Heather would more likely be replacing a bad unit than a good one since subbing Tibarn for Heather is moronic, so she's hurting efficiency even less than that.

Though just for kicks, you could use Heather to steal some brave weapons or maybe an extra tempest blade if you're lucky with disarm. That's certainly helpful for certain army configurations, mainly units who don't mind sacrificing accuracy for raw damage.

[dondon151]

If you're in the bottom tier, then basically anyone else in the game can be deployed in your place while giving higher returns. Moreover, the combat returns you get from using someone like Lyre probably doesn't even outweigh the basic resources she gets (olivi grass, concoctions), not to mention how badly she'll damage your enemy phase if she isn't protected.

If you're not in the top or high tiers, then other people can be deployed in your place while giving higher returns. Including opportunity cost of deployment (among other resources) gets the debater to the conclusion that fielding anyone outside of the top x units is an inefficient and therefore impossible choice. After excluding opportunity cost, every character in the game has a positive contribution.

You might wonder "by that logic, wouldn't anyone besides the top 10 units in a given army have negative utility? After all, there's always a better unit to take their place." Try looking at it from a probabilistic standpoint. If you're upper-mid tier, it's more likely that you're taking up an inferior unit's slot than you are a superior one's. If you're mid tier, the likelihood is equal both ways.

I don't buy this. Probability doesn't nullify the fact that unwise decisions are unwise (in fact, probability doesn't even matter in this scenario; if you're taking up the slot that a superior unit can use, then you're taking up that slot).

[-Cynthia-]

Assuming we use Heather in 4-E, our best course of action would be to hide her in a corner. Then the loss in efficiency is only the loss of 1 combat unit. Even assuming each combat unit contributes equally, that's a drop of efficiency of 10%, and that's if we disregard forced units (eg/ a heron, kurth, ena). In practice, Heather would more likely be replacing a bad unit than a good one since subbing Tibarn for Heather is moronic, so she's hurting efficiency even less than that.

Though just for kicks, you could use Heather to steal some brave weapons or maybe an extra tempest blade if you're lucky with disarm. That's certainly helpful for certain army configurations, mainly units who don't mind sacrificing accuracy for raw damage.

Can't we also just hide Tormod/Geoffrey etc. in a corner, and therefore only drop efficiency by 10% as well? I don't really understand the second part of your post, Heather probably isn't replacing Tibarn, but she's probably replacing a unit much better than she is, I doubt she would be replacing a "bad" unit persay, maybe someone like Mia, which is a pretty big difference.

Eh, if Heather starts attempting to steal after the very unpredictable Disarm, she's going to get in a lot of trouble (since she might get attacked then). Obviously this stealing is useless on 4-E(3) and beyond as well.