+/- Utility

[Reikken]

I fail to see how that has nothing to do with negative utility. I thought I explained it already, but I'll give it another shot.

A unit who performs worse than the worst unit's slot it would be taking in order to be fielded has negative utility, as your overall performance becomes weaker by adding that unit to your team. Pretty straightforward, right? Now, where the "likelyhood of a particular team" comes in is this. If a better unit would not have been fielded anyway, the unit whose utility is in question can not be taking the place of that unit. Thus this unit's opportunity cost will vary depending on the particular team that otherwise would have been used. To account for all situations and form an overall or average opportunity cost, take the cost for each team weighted by the "likelyhood of [that] particular team" and average them all together.

And as for the "arbitrary" bit: It's not arbitrary at all; it's just impossible/impractical to calculate, so I do the best I can practically do and estimate it. The value I come up with may be arbitrary, depending on your conception of that term, but the value that it's based on is certainly not.

Truly, my definition of negative utility is simply "A unit who performs worse than the worst unit's slot it would be taking in order to be fielded has negative utility" (That's not really a definition, though. I guess it's still "utility - opp cost".) And that's it. If we were doing a perfect playthrough tier list or whatever you call it, every unit used outside of the perfect team would have negative utility. The first paragraph of this quoted post does still apply even there. The "likelyhood of [whatever] particular team" in that case is 100% for the perfect team. Thus I must amend part of the second paragraph.

It's not arbitrary at all; it's just not always possible/practical to calculate, so if it's not, I do the best I can practically do and estimate it.

Edited August 19, 2009 by Reikken

[Reikken]

See, here's how the discussion will boil down.

1: debaters agree to discuss inefficient play.

2: one debater will point out that playing a character a certain way will decrease negative efficiency, and the other debater will attempt to counter it.

3: debate switches to discussing efficient play.

No. If they're agreed to discussed inefficient play, there will be no such switch. The point in step 2 may or may not be valid depending on the specifics of the agreement in step 1.

[Interceptor]

I fail to see how that has nothing to do with negative utility.

It has nothing to do with the definition of negative utility.

It does have something to do with negative utility as it applies to how Fire Emblem tier lists are structured, but that's something else entirely, isn't it? Your definition of negative utility depends on a tier list also representing playthroughs that are not the most efficient.

I understand all of your reasoning, so there's no need to keep rephrasing it. None of it is really germane to the point that I am trying to make (that you ignored).

As for the arbitrary bit, I am not accusing you of making things up, I am just pointing out your choice for what it is. You draw a line in the sand, you have your reasoning for it, but it's not the only logical way to construct a tier list and it's not necessarily the best way, either.

[CATS]

Whether we acknowledge or ignore the opportunity cost of deployment affects how non-ideal units are tiered. If the best way to play non-ideal units is to not play them at all, then every one of those units is virtually identical and cannot be tiered.

Not quite. Bartre shouldn't be played in the long term, and neither should Nino, but Bartre still contributes more towards an efficient run because he can help some during early chapters, whereas Nino brings absolutely miniscule contributions to the table in her join chapter. They're not the same. Likewise, the case from which this entire debaet stemmed, Karla, is undeniably worse than all (or at least most) of the other units who shouldn't be played. Karla forces an inferior unit into play for some length of time. That is, as far I can see, entirely unique to her.

Regardless, when you have two units who never contribute anything (or if neither is contributing anything significantly better than the other) without using up a deployment slot, you can then see what happens when they are indeed played to see who goes above who. This has always been done, and negative utility has never been officially ignored, and often acknowledged when it becomes relevant (such as in the case of Karla). There's no problem. If you want to say that Rath and Heath are technically equal and just put them in the same spot with a slash between them, go ahead. Or you can debaet them and put them right next to each other with one placed over the other, in the same way that ranks are ignored and such simply for the purposes of discussion. It doesn't really matter either way, and certainly doesn't mean that negative utility should be blatantly ignored in cases where it is strongly relevant.

Then does Bartre get credit for allowing a superior unit to take his slot if he's not played?

Maybe if recruiting him or keeping him alive or doing something with him somehow created a new deployment slot for that superior unit. But obviously, this is not the case. You can kill him off in Ch 12 and the number of available slots later stays the same. His existence has no effect on the slots and who goes in them; he could be removed from the game entirely and you'd still have the same units going in the same slots.

"They're related. Optimal character choice is a part of efficient play."

"The second part of your statement, after the word "yet," I don't see anywhere in the tier list FAQ, nor do I remember seeing it anywhere else."

And the statement that you were referring to: "The tier list is created under the premise of high efficiency play yet it entails selecting negatively efficient units for comparison."

So where's the part where I say that using bad characters doesn't matter?

I don't think anyone would disagree with me that the basic principle of a tier list is efficiency, and I don't think I had to cite it outright, though anyone with the ability of inference can deduce that much.

The point is that contributions towards an efficient run is clearly labelled as the defining criteria of the list. Nowhere does it say that negatively efficient units must be judged as if they were deployed in each and every chapter in which they are available. I can use Nino only in BBD where she joins, while never recruiting Karla, and Nino contributes more towards an efficient run. Whatever other complications might come up in other cases are irrelevant to this, and do not change the fact that it is the reality of the situation.

You can argue that negative utility should be put aside in other cases for whatever reason, but it should be done on a case-by-case basis if it's done at all. You can't just universally ignore it in all instances.

Are you even reading my posts now? This will be the fourth time that I said that the only opportunity cost that will be ignored is the cost of deployment. Stuff like the cost of weapons, support slots, stat boosters, and promotion items are not ignored.

I was referring to this statement:

Yes, if we didn't consider Wil's existence to be negative before he can comfortably double most enemies, then he would be above Vaida. His "good" period begins long before Vaida joins

You'll also notice that I only said Wil is consuming a large amount of resources, and that deployment slots are indeed a resource. Arguably your most important one.

See, here's how the discussion will boil down.

1: debaters agree to discuss inefficient play.

2: one debater will point out that playing a character a certain way will decrease negative efficiency, and the other debater will attempt to counter it.

3: debate switches to discussing efficient play.

Unless the scenario of efficient play has already been considered and was inconclusive, as I said earlier. Or if people just feel like discussing things a different way for no particular reason, ala Conditional Debates in the debaet challenges forum. I mean, if discussion of efficient play is still the deciding factor in a comparison, then ofcourse people will discuss it in those terms. I don't see anything wrong with it.

Edited August 19, 2009 by CATS

[Reikken]

I fail to see how that has nothing to do with negative utility.

It has nothing to do with the definition of negative utility.

It does have something to do with negative utility as it applies to how Fire Emblem tier lists are structured, but that's something else entirely, isn't it? Your definition of negative utility depends on a tier list also representing playthroughs that are not the most efficient.

I understand all of your reasoning, so there's no need to keep rephrasing it. None of it is really germane to the point that I am trying to make (that you ignored).

As for the arbitrary bit, I am not accusing you of making things up, I am just pointing out your choice for what it is. You draw a line in the sand, you have your reasoning for it, but it's not the only logical way to construct a tier list and it's not necessarily the best way, either.

So you agree with me except on tier list goals? If not, you'll have to do a better job of explaining where the discrepancy is.

[Rodykitty]

So you're saying that I am wrong because you don't think that I'm right even though I have a logical argument and you do not. (There is no argument in your post, just a voicing of disagreement.)

btw, units are lower on the tier list because they are worse, not the other way around. Their position on the tier list doesn't directly have anything to do with "negative utility".

It's not like "This unit is bad. Oh, and they also have negative utility which pulls them down even more." Instead, "negative utility" is a reflection of them being bad, or simply another way of looking at how bad they are. It's not 'in addition to' anything.

Units can either be more of a positive than negative, or more of a negative than positive depending on how much they give or take away from a team. This reasoning of yours is implying to me that most units are more of a negative than positive because they're using up a unit slot someone else can take, but really if they can still contribute positively to a team, then they are not a negative.

[Reikken]

The portion of units that are negative depends on the particular game and tier list/judging criteria. It could indeed be that more are negative than positive.

[Rodykitty]

The portion of units that are negative depends on the particular game and tier list/judging criteria. It could indeed be that more are negative than positive.

Precisely. I've been saying that usually endgame units like Athos or Gotoh are pretty close to "neutral" (but not actually neutral), but they could be lower than mid tier depending on how many units are more of a positive than negative.

(The four endgame Bishops in FE3 Book 2 are the closest to true neutral utility, never being a negative to your team but only contributing maybe a heal at most due to almost nonexistent availability).

Edited August 19, 2009 by Rody

[Interceptor]

So you agree with me except on tier list goals? If not, you'll have to do a better job of explaining where the discrepancy is.

It's more accurate to say that you're in agreement with me, since I've been saying in this thread from the beginning that strict adherence to negative utility isn't useful for Fire Emblem tiering. This is not something that I think you'd disagree with.

My only point here is that a discussion of negative utility in a void is pointless, because we only care about negative utility as it applies to tiering. This is a problem that's begging to be solved at the individual tier list level, with specific goals laid out up-front. The fact that nobody has voted for the strictly correct answer, #3, means that people recognize this. Now we're waiting on step #2, for more people who are arguing to discover that they aren't arguing about negative utility, but rather they are arguing about which way is better for tiering.

As for tier list goals, we may or may not disagree on them (probably disagree), but I don't see how that matters since we're not making a tier list here.

[Narga_Rocks]

Every single time we start comparing two units that should never ever be deployed in an efficient playthrough we are doing a tier list entailing selecting negatively efficient units for comparison. Every time we put a really bad unit on the field to see how good it does we are already operating in a way mutually exclusive with "Characters are compared based on how they contribute towards an efficient playthrough." Because the way in which Lyre contributes towards an efficient playthrough is by entertaining Gatrie between chapters so he can be more relaxed during a map. She does not contribute towards an efficient playthrough by being fielded. By fielding her we make a mockery of the very idea of an efficient playthrough.

It's not like discussion of what happens when the units are actively used couldn't still happen. It's just that one would also have to acknowledge what they are able to contribute without taking away from anything else.

I'm not saying we can't discuss those units. I'm just saying, you disagreed with dondon saying:

"The tier list is created under the premise of high efficiency play yet it entails selecting negatively efficient units for comparison."

I'm saying every time we discuss those units we are making a tier list that entails selecting negatively efficient units. I don't see how your statement counters dondon's statement. Sure, we acknowledge that they are taking away from the better team choices, but that doesn't change the fact we deployed them in the first place which already makes our efficient tier list playthrough not so efficient.

By selecting "mid tier is roughly neutral utility" that is basically stating that even if we could know the perfect team choice then mid tier would still be roughly neutral utility.

No. That's obviously not what it says. It doesn't say "Assuming we know the perfect team choice, mid tier is still neutral utility."

It's a logic thing.

Mid tier is roughly neutral utility

Negative utility doesn't exist

Assuming there's a perfect team choice, anyone who doesn't belong to it has negative utility of some sort.

Other (please specify)

By choosing any one of these options, you are saying the other 3 are untrue. As such, if we say midtier is roughly neutral utility, we say the other 3 are untrue.

therefore:

negative utility does exist.

Assuming there is a perfect team choice, anyone who doesn't belong to it does not necessarily have negative utility.

That's the way it works. Choose one, make the others untrue.

Therefore, by saying mid tier is roughly neutral utility, you are in fact saying:

"Assuming we know the perfect team choice, mid tier is still neutral utility."

Because you are saying:

Assuming there is a perfect team choice, anyone who doesn't belong to it does not necessarily have negative utility.

and

Mid tier is roughly neutral utility

Combine those two statements, you get:

"Assuming we know the perfect team choice, mid tier is still neutral utility."

And that's how logic works. Everything is connected, blah blah blah.

No, the idea isn't that everyone becomes good eventually. The idea is that in, say, chapter 8, when you have 10 slots and 13 units, and 10 of the units are clearly better than the other 3, those three units do not automatically have negative utility by being fielded. This is because on most maps all 10 of those "better units" are not always operating at peak output. There are only so many enemies, only so many places through which enemies attack, and only so many of your units will be in position to do something on enemy phase. As such, it is entirely possible that whatever your tenth unit is doing in a normal playthough could very easily be done by one of the other 3. If this is the case, then those 3 units do not have negative utility for being fielded. They aren't hurting us in any way whatsoever.

I don't quite understand your argument here. Are you saying that here, negative utility doesn't exist because a lesser unit can get the job done anyway?

Effectively. If a lesser unit is doing what a better unit is doing anyway, it isn't negative utility. Thus, negative utility is often completely cancelled, and even when it isn't completely cancelled it is still mitigated significantly.

See, if that not so good unit is able to do the exact same thing as the better unit would have done for 8 turns out of 10 turns, it only really has 2 turns of "negative utility", because on the other turns it's basically neutral. As such, after 10 chapters if it gets good it doesn't have nearly as much "negative utility" to make up for as it would have if you simply look at deployment slots.

Anyway, I'm only saying this if we accept the existence of negative utility. If we don't, then for those 8 turns it accumulates just as much positive as the other guy could have, and on the other 2 turns it accumulates less. Well, that's a simplification, but whatever.

But if A is bad for 18 chapters and B is bad for 6 chapters, shouldn't B have to be more than three times as bad as A is in order to be ranked lower?

If you deploy both units in every chapter that they're available, then sure, something like that. There's no reason to assume that's what happens, though.

It's kind of annoying that everyone has different ideas of what a tier list requires. A lot of people want the unit in question to be deployed all the time, others are okay with playing the unit smarter to maximize what it is doing.

You mean the logic against sticking Mr. Useless above those 10 guys that are 5RKOd and 3HKO all the enemies? I just find it strange that a unit that does nothing ever gets a pass simply because it doesn't kick out other units.

Oh, indeed. Them's the breaks, though. In this case, the fact that he isn't consuming a limited and important resource (i.e. your deployment slots) is a pretty big advantage for him, even if he sucks at the rest of everything, just like in FEDS prepromos get points because they don't consume a limited and valuable resource (Master Seals).

Well, that's one way of looking at it. It is the proper conclusion of negative utility, as I see it.

Of course unit B would be better if it showed up at the same time as unit A. It has everything to do with availability.

Well now I'm just confused, because you're agreeing with my points while arguing against them. You're trying to show that greater availability can be a disadvantage when you acknowledge negative utility, correct? I'm telling you otherwise. Adding more availability to a unit will only ever cause them to be ranked higher on the tier list. Assume that Nino joins in Ch 11 instead of BBD, for example.

Your own example does not compare two equal units joining at different times; you added more stats and levels to the later joining unit. That's an advantage for that unit; he has higher stats without requiring any resource investment. The two are not equal. If you truly keep everything else the same and only adjust availability, the earlier joining unit is always better. Greater availability is never a disadvantage in and of itself, negative utility or not.

Well, what's the point in comparing two units that are equal to you? The answer is obvious. Same stats/growths/level, earlier join, better. Obviously. I have to add stats and levels to make a point in comparing them. To me, as of chapter 15 when the other guy joins, they are equal from that point on. This makes it interesting because it means the earlier unit had to get kills and use weapons and stuff to reach the other guy's base level. If this guy was the best unit for those extra chapters, we'd likely say it accumulated more utility for being around longer and is better. But if he is below average, he still killed things during his time and led us towards the chapter goals. Why shouldn't he get credit for those chapters? Why must he be penalized for not being top of the line? We already rate him lower than those units he was competing against in those chapters, why do we have to hurt him twice for the same crime?

My problem with the scenario is that a lot of people say "negative utility" this and "negative utility" that, but then say Kyza > Fiona. For the sake of logical consistency, I'd much rather see Fiona > Kyza.

Well phooey on those people then.

Well, at least it seems that unlike some others you are more consistent. I can't complain about that.

[CATS]

I'm not saying we can't discuss those units. I'm just saying, you disagreed with dondon saying:

"The tier list is created under the premise of high efficiency play yet it entails selecting negatively efficient units for comparison."

I'm saying every time we discuss those units we are making a tier list that entails selecting negatively efficient units. I don't see how your statement counters dondon's statement. Sure, we acknowledge that they are taking away from the better team choices, but that doesn't change the fact we deployed them in the first place which already makes our efficient tier list playthrough not so efficient.

The implication is that subpar units must be judged as if they are constantly deployed (or if it isn't, then you're agreeing with me, so w/e). I'm saying this isn't true or at all reasonable.

Effectively. If a lesser unit is doing what a better unit is doing anyway, it isn't negative utility. Thus, negative utility is often completely cancelled, and even when it isn't completely cancelled it is still mitigated significantly.

Worse is worse. Maybe that other guy might be able to perform similarly in some situations, but he can't in all of them, that's why he's worse. So negative utility is still there. And like I said earlier, the magnitude of it is irrelevant to the fact that it exists, which is what this discussion is about, so w/e.

It's kind of annoying that everyone has different ideas of what a tier list requires. A lot of people want the unit in question to be deployed all the time, others are okay with playing the unit smarter to maximize what it is doing.

Well again, phooey on those people. It doesn't particularly matter--the criteria of the list is quite specific. "Characters are compared based on how they contribute towards an efficient playthrough." If they can contribute more by only being fielded once or twice, then obviously that's what happens.

Well, what's the point in comparing two units that are equal to you? The answer is obvious. Same stats/growths/level, earlier join, better. Obviously. I have to add stats and levels to make a point in comparing them.

And in doing so you defeat the point of your example. #2 is only better after you add those additional stats and levels to him, to make up for the fact that he joins later. If he joined earlier then he wouldn't need more levels and stats to stay competitive. If he is better than #1, then it's clearly because of those extra stats and levels you added to him, not because #1 joins earlier. Again, compare them if they joined at the same level, or if #2 joins on a higher level with more stats. We kept jointimes the same; the only thing different is that #2 has higher stats and a higher level in the second case, so clearly that's what's making the difference, not the availability.

[dondon151]

And in doing so you defeat the point of your example. #2 is only better after you add those additional stats and levels to him, to make up for the fact that he joins later. If he joined earlier then he wouldn't need more levels and stats to stay competitive. If he is better than #1, then it's clearly because of those extra stats and levels you added to him, not because #1 joins earlier. Again, compare them if they joined at the same level, or if #2 joins on a higher level with more stats. We kept jointimes the same; the only thing different is that #2 has higher stats and a higher level in the second case, so clearly that's what's making the difference, not the availability.

I don't think you're understanding his example. If unit 1 is generally subpar, then he's worse than unit 2 for being around longer. If unit 1 is generally good, then he's better than unit 2 for being around longer. The core of the example is that when unit 2 joins he's equal to what unit 1 would be at that point in the game; this assumption couldn't possibly defeat the point of the example.

[Narga_Rocks]

I'm not saying we can't discuss those units. I'm just saying, you disagreed with dondon saying:

"The tier list is created under the premise of high efficiency play yet it entails selecting negatively efficient units for comparison."

I'm saying every time we discuss those units we are making a tier list that entails selecting negatively efficient units. I don't see how your statement counters dondon's statement. Sure, we acknowledge that they are taking away from the better team choices, but that doesn't change the fact we deployed them in the first place which already makes our efficient tier list playthrough not so efficient.

The implication is that subpar units must be judged as if they are constantly deployed (or if it isn't, then you're agreeing with me, so w/e). I'm saying this isn't true or at all reasonable.

Well, I think we do agree on some things. I'd rather they get to choose when they are deployed, and the only reason to force them to be deployed more often is if they tie with another unit when they choose not to be deployed too much.

Effectively. If a lesser unit is doing what a better unit is doing anyway, it isn't negative utility. Thus, negative utility is often completely cancelled, and even when it isn't completely cancelled it is still mitigated significantly.

Worse is worse. Maybe that other guy might be able to perform similarly in some situations, but he can't in all of them, that's why he's worse. So negative utility is still there. And like I said earlier, the magnitude of it is irrelevant to the fact that it exists, which is what this discussion is about, so w/e.

Ah, but the issue is very much the magnitude. If we accept negative utility, then we have to calculate how much negative they have to make up for. If there are only a small number of situations in which they aren't quite as good, then there isn't as much to make up for as if we just say the unit is bad so every chapter in which it is deployed it is worse for the entire chapter. Note that this argument I'm making is not to say there is no negative utility for deployment. This is in the case of the existence of negative utility, it can be mitigated in some cases, if not totally eliminated.

It's kind of annoying that everyone has different ideas of what a tier list requires. A lot of people want the unit in question to be deployed all the time, others are okay with playing the unit smarter to maximize what it is doing.

Well again, phooey on those people. It doesn't particularly matter--the criteria of the list is quite specific. "Characters are compared based on how they contribute towards an efficient playthrough." If they can contribute more by only being fielded once or twice, then obviously that's what happens.

Which is why I think we agree on some points.

Well, what's the point in comparing two units that are equal to you? The answer is obvious. Same stats/growths/level, earlier join, better. Obviously. I have to add stats and levels to make a point in comparing them.

And in doing so you defeat the point of your example. #2 is only better after you add those additional stats and levels to him, to make up for the fact that he joins later. If he joined earlier then he wouldn't need more levels and stats to stay competitive. If he is better than #1, then it's clearly because of those extra stats and levels you added to him, not because #1 joins earlier. Again, compare them if they joined at the same level, or if #2 joins on a higher level with more stats. We kept jointimes the same; the only thing different is that #2 has higher stats and a higher level in the second case, so clearly that's what's making the difference, not the availability.

I don't think you're understanding his example. If unit 1 is generally subpar, then he's worse than unit 2 for being around longer. If unit 1 is generally good, then he's better than unit 2 for being around longer. The core of the example is that when unit 2 joins he's equal to what unit 1 would be at that point in the game; this assumption couldn't possibly defeat the point of the example.

Yeah, there is no point to the example without the assumption. I don't see how the very assumption that makes the example worthwhile could possibly defeat the point. This example is about the affect of negative utility. Not about which fictional fictional character is better. (yes, there is two fictionals because I made up a made up character.) The entire point of the example is to discuss how bad unit 1 needs to be in the chapters it has without unit 2 around in order to be worse than unit 2 overall. If they didn't have the same stats when unit 2 joins, then the example is asking for something completely different. If unit 2 doesn't have the same stats as what unit 1 would have when unit 2 joins, I am no longer asking about the affect of negative utility on unit 1 in the 10 chapters unit 2 isn't around. I'm asking about something completely different at that point. Simply put, does he have to just be below average but not taking a slot to be worse overall? Or does he have to be bad enough that he takes a slot away from someone else in order to be worse? If there are like 30 units before unit 2 appears and 7 slots, and unit 1 is above average during this time but not quite top 10, is he then worse than unit 2 overall? The only way these questions mean anything is if unit 2 gets the stats unit 1 would have when unit 2 joins. Otherwise I'm asking not asking about the affects of positive/negative utility, I'm asking about the affect of stats. And we know the affect of stats already. Higher is better.

[Progenitus]

Breathing some new life into this topic.

And in doing so you defeat the point of your example. #2 is only better after you add those additional stats and levels to him, to make up for the fact that he joins later. If he joined earlier then he wouldn't need more levels and stats to stay competitive. If he is better than #1, then it's clearly because of those extra stats and levels you added to him, not because #1 joins earlier. Again, compare them if they joined at the same level, or if #2 joins on a higher level with more stats. We kept jointimes the same; the only thing different is that #2 has higher stats and a higher level in the second case, so clearly that's what's making the difference, not the availability.

I don't think you're understanding his example. If unit 1 is generally subpar, then he's worse than unit 2 for being around longer. If unit 1 is generally good, then he's better than unit 2 for being around longer. The core of the example is that when unit 2 joins he's equal to what unit 1 would be at that point in the game; this assumption couldn't possibly defeat the point of the example.

So, say that unit A is level 1 in chapter 1, and if unit A was given the same number of kills as every other unit on the team, he would be level 10 in chapter 10. And unit B joins in chapter 10 at level 10 with the same stats/weapon levels/etc as unit A.

Who would be better would depend on the number of units you're using from chapters 1-10. If, say, you were using 5 units, and we had 10 units to choose from, and unit A is our 10th best unit, then it's extremely unlikely we'll field unit A to get those kills and thus reach level 10 at chapter 10. And unit B would be better. On the other hand, if unit A was the 3rd best unit, then unit A is probably better, as he's better at helping through those 9 chapters than the others below him, which means using him to kill enemies and clear chapters is more efficient, and getting him exp is just a bonus for him. If unit A was 6th or 7th best, then I'd wager he'd be about neutral, and the comparison would be roughly tied.

Of course, as a disclaimer, everything has exceptions. I'm pretty sure there's an exception or two for this that I'm missing.

[Zkirsche]

Why does it make him worse? If Unit A actually contributed in said 10 chapters by helping to finish things off/killing thigns outright, then what does being teh 10th string unit matter as he's still helping more towards game completion than Unit B is as he's helping from chapter 1 onwards, unlike unit B who's helping from chapter 10 onwards. This is why Unit A > Unit B when comparing combat utility.

[Tino]

What you're saying, kirsche, is actually what we're trying to achieve with this thread. Does negative utility exist, and if it does, what constitutes as negative utility. If negative utility does exist, then merely being able to help finish things off whereas others are 1-2RKOing things is negative utility, if you ask me. After all, an additional cow in a herd of 1,000 only consumes more grass and makes the pastures more crowded. So even though she might provide dairy products, she's still negative utility.

Edited September 14, 2009 by Tino

[Zkirsche]

The problem with your way of thinking is that availability can't hurt a unit. What happens if Unit A isn't even deployed again at all in the game. Surely then Unit B can onyl be equal to Unit A if not deployed and worse for being deployed for the reasons you mentioned. Basically, Unit B can only be equal to or worse than Unit A, making Unit A overall superier to Unit B.

[Tino]

In other words, buy the cow and pretend it doesn't exist? Sounds like some faulty logic to me.

[Zkirsche]

In other words, buy the cow and pretend it doesn't exist? Sounds like some faulty logic to me.

It would if it actually costs you something, but then you can say the same thing about Unit B. You shouldn't force a unti to exist if it's bad, as with your logic about negative utility, it's actually better to bench him/her, so why wouldn't you?

[Narga_Rocks]

The problem with your way of thinking is that availability can't hurt a unit. What happens if Unit A isn't even deployed again at all in the game. Surely then Unit B can onyl be equal to Unit A if not deployed and worse for being deployed for the reasons you mentioned. Basically, Unit B can only be equal to or worse than Unit A, making Unit A overall superier to Unit B.

In other words, buy the cow and pretend it doesn't exist? Sounds like some faulty logic to me.

It would if it actually costs you something, but then you can say the same thing about Unit B. You shouldn't force a unti to exist if it's bad, as with your logic about negative utility, it's actually better to bench him/her, so why wouldn't you?

Actually, the original intent of my example was for all cases, not just where the unit goes from bad -> bad and they are equal if neither needs to be deployed ever.

In the case where unit A goes from bad -> good and just happens to start hitting it's stride when unit B shows up, that's where things get interesting. If unit A isn't ever deployed, unit B is clearly better since unit B is top 5 or whatever from when it shows up until the game is finished and unit A (still at base level since it wasn't ever deployed) could have been top 5 (from that point on) if it just went through it's bad phase, for example.

So how much did taking the slot and getting the kills to reach that level cost the unit/team and is A > B despite the cost (because it was doing stuff too I guess) or is B > A?

From the standpoint where units get credit for helping out and finishing chapters, A > B. From the standpoint where if you aren't worthy of the slot you get a major penalty against your overall score, B > A.