fe10 balancing ideas

[nflchamp]

So, since I neglected to read this up until now, I'm just going to clarify. The goal is that every character will have equal usefulness in every chapter they're in. If that's right, this turns the tier list into an availability fest, even if smash and paperblade don't believe it. Read this and tell me that it isn't true.

Edit: Make a new thread to talk about the linked post, if you really want to.

Nah, I don't think I need to make a separate topic for it. I can finish this in a few posts. SOmeone else can make a new topic if they want to.

Negative utility doesn't spawn because you're performing worse than the average unit. Rather, negative utility is when the performance output does not exceed the resource input, in most cases the resources being kills/EXP (though in most cases you are weighed against the average unit anyway).

In your example, Edward would be doing things, or wahtever, and do it really badly because he's lolgarbage, and then eats up resources in the process. His bad performance does not justify all the kills poured into him. That's how he ends up with negative utility.

Ike can't fight in 1-4 or any DB chapter, but he's not taking any resources in the process. That's how he's neutral.

You can, in fact, use Edward less, to the point where his performance output equals his resource input. For example, instead of having Edward be an active tank and attacker, he does a couple of potshots and once in awhile tanks if absolutely necessary (like an enemy would otherwise reach Laura). In the process he takes a small amount of kills. He does a small amount of work, and took only a small amount of resources. So he ends up with about neutral utility.

But, how exactly do you measure Edward's usefulness? How many kills could you give to Edward before you say "nope, his performance of occasional potshots and face tanking does not justify the kills" and he goes from neutral utility to negative? There's enough bickering as is on just even kill distribution; we don't need to make it more complicated.

Plus, pretty much anyone can just do small work. Even crap like Lyre can shove or something and make themselves useful. That doesn't really tell us anything about their performance.

Thus, generally Edward/*insert crap unit* gets the same kills as the other units you're using, which means his performance has to be weighed against them. In other words, that's how we usually end up with Edward being weighed against the average unit.

You're only defining 'X', not negating the point. If everyone in a chapter is performing at a current "high/upper mid tier level," they're all getting positive utility in that chapter in comparison to someone who isn't in it. Sure, they're all net 0 when compared against someone else with in the chapter, but they're beating anyone who can't perform in the chapter. Now, if you could make everyone neutral utility in every chapter they're in, you'd be able to put everyone equal to each other despite they're availability differences. Just giving everyone the same level of performance, though, won't work.

[Progenitus]

You're only defining 'X', not negating the point. If everyone in a chapter is performing at a current "high/upper mid tier level," they're all getting positive utility in that chapter in comparison to someone who isn't in it. Sure, they're all net 0 when compared against someone else with in the chapter, but they're beating anyone who can't perform in the chapter.

What I mean by when everyone is performing at a "high/upper mid tier level", I mean that, relative to the enemies, our units are going to be about as good as someone like Oscar or so. That is, they won't be capable of soloing the map (lolike), but they aren't terrible at tackling the enemies (liek Soren). Basically, I don't want the game to turn into "what would happen if FEDS H5 didn't have Sedgar or Wolf" or "what would happen if everyone was as good as FE9 Titania".

Now, if you could make everyone neutral utility in every chapter they're in, you'd be able to put everyone equal to each other despite they're availability differences. Just giving everyone the same level of performance, though, won't work.

Yeah, I figure I would have to make chapters where you're given very few units into very easy chapters, since otherwise every unit in the map would gain positive utility despite all the PCs in the map being equal. I mean if Micaiah/Edward/Leo were all equal in 1-P, they'd all probably gain positive utility in that map since you're better off using all of them to clear it rather than neglect one.

...then again, maps with very few PCs available to you should probably carry less weight, since you're forced to use those units whether they're good or not. Kinda like Ike in FE9 Prologue where he's the only unit available, or FE10 1-9 where Micaiah lolsucks and BK basically soloes the map. These kinds of chapters tend to be easy and short anyway though, so meh.

[dondon151]

Negative utility doesn't spawn because you're performing worse than the average unit. Rather, negative utility is when the performance output does not exceed the resource input, in most cases the resources being kills/EXP (though in most cases you are weighed against the average unit anyway).

http://en.wikipedia.org/wiki/Economic_profit

http://en.wikipedia.org/wiki/Opportunity_costs

Your wrong. Bye.

[Progenitus]

So you don't understand what negative utility is.

Edited November 10, 2009 by 8========================D

[Narga_Rocks]

So you don't understand what negative utility is.

We can't help it if you use an illogical definition. dondon is just stating what the logical definition of negative utility would be: when economic profit is negative. Seems to me that you don't understand what a logical definition of negative utility is.

Unless you mean to tell us that you think using an illogical definition is superior to using a logical definition. Go ahead: I'd like to see that argument.

Edited November 10, 2009 by Narga_Rocks

[Zkirsche]

What? The line I quoted didn't mention other units at all.

So? You just picked that line without reading on. If you actually put it into context and quoted more than just a random line from his post, then you'll see this bit:

Now, however, any unit can take Vantage + Cancel and not need adept and significantly reduce the chance of getting hit. More than Mia, in fact, since an unaltered Mia still isn't 2HKOing anything aside from swordmasters and sages/bishops and dragonmasters.

I was talking generally about how forges are broken, and he assumed I was talking specifically about Mia. =/

Well, then change what you quote, putting this:

Yeah, since Mia is the only unit who can use forges.

below this:

Besides, she can still grab a steel blade and 3HKO almost as much, and in a map or two exactly as much.

makes it look like you're speaking out of your ass.

[Progenitus]

Let me clarify for dondon and nflchamp....

Both "performance output - resource input" or "profit" (or rather, performance - opportunity cost) have the same idea. The difference is that you people, if I'm not mistaken, take opportunity cost as the value of another unit you could've put in that slot (of course there are other forms of opportunity cost like stat boosters and skills, but that's not really relevant right now, as we're not arguing for such things on units). For example, if in a given GM chapter, you field Boyd instead of Oscar, then the opportunity cost would be Oscar, and Boyd's net utility would be Boyd's performance - Oscar's performance. If I'm mistaken, then clarify for me, because I can't read minds.

Resource input, on the other hand, is that PLUS kills/CEXP. Kills/CEXP ARE resources, and therefore they must be considered as part of opportunity cost. If, for example, I gave a kill to Edward in 1-4, that's a kill that could've gone to another unit, like Nolan. Sure, Ike can't kill anything in 1-4 due to not existing, but he's not taking kills from anyone either.

Seems to me that both ideas are similar, just with different names, and the fact that I'm also considering kills/CEXP as a resource.

Edited November 10, 2009 by 8========================D

[dondon151]

Seems to me that both ideas are the same thing, just with different names.

It's the same general thought with some key differences. Opportunity costs are tied in to the very basic decision of even using a unit. That allows us to quantify a unit's performance by simply doing unit's performance - performance of unit being replaced (which you understand). But your idea of quantifying utility is performance output - resource input, and you implied that there was a certain point where performance output can exceed resource input for non-optimal units. But there isn't one if you're sub-optimal, because opportunity cost dictates that not going with the first best alternative yields negative economic profit.

I'm also not sure how you got "negative utility doesn't spawn because you're performing worse than the average unit" and then concluded your post with "that's how we usually end up with Edward being weighed against the average unit." Because he shouldn't be weighed against the average unit, period.

[nflchamp]

You're only defining 'X', not negating the point. If everyone in a chapter is performing at a current "high/upper mid tier level," they're all getting positive utility in that chapter in comparison to someone who isn't in it. Sure, they're all net 0 when compared against someone else with in the chapter, but they're beating anyone who can't perform in the chapter.

What I mean by when everyone is performing at a "high/upper mid tier level", I mean that, relative to the enemies, our units are going to be about as good as someone like Oscar or so. That is, they won't be capable of soloing the map (lolike), but they aren't terrible at tackling the enemies (liek Soren). Basically, I don't want the game to turn into "what would happen if FEDS H5 didn't have Sedgar or Wolf" or "what would happen if everyone was as good as FE9 Titania".

This is nonsensical in terms of making units equal without giving everyone equal availability. They're all beating the map better than someone who isn't there, so they're all getting positive utility in comparison to someone who isn't there. This makes it into an availability fest.

Let's show it this way.

U (Utility) = P (Performance) - C (Cost)

Utility is simply what we're finding, Performance is the measured performance a character has within a chapter, and Cost is the cost of using resources (whatever they may be).

A unit who doesn't exist simply has a utility that looks like this:

U = 0 - 0 = 0

AKA no performance and no cost, so no utility.

A unit who does exist, of course, has a utility that looks like our original equation:

U = P - C

If P = C, then U = 0. However, this means that I find no difference in using the resource and not using the resource. This of course cannot be true. So long as using the resource promotes winning, I'd always prefer to pay the cost to get the performance than not pay the cost and not get the performance.

Of course, this can be shifted to be neutral by having an expected performance, or rather, by expecting a unit to help beat the game. This changes the equation though.

U (Utility) = P (Performance) - C (Cost) - X (Expected Performance)

So now if P - C = X, U = 0. That is, if the difference in gains between a unit's performance and cost is equal to what we expect of a unit in the chapter, then they have no utility. Note that X is a constant; I'm now expecting every unit to be able to outperform their cost by a certain amount, including those units that aren't in the chapter. As such, the utility of have a unit not in the chapter is:

U = 0 - 0 - X = -X

The unit still has no performance and no cost, but now they aren't living up to my expectations.

Edit: Removing the stupid end of the post. Forgot to get rid of it when writing.

Edited November 11, 2009 by nflchamp

[dondon151]

Actually, if all units are equal, then every unit's economic profits would be 0 (because the next best unit is effectively the exact same thing). And 0 times anything is still 0.

Of course, if we only looked at normal profit, then it turns into an availability fest.

[Narga_Rocks]

Actually, if all units are equal, then every unit's economic profits would be 0 (because the next best unit is effectively the exact same thing). And 0 times anything is still 0.

Of course, if we only looked at normal profit, then it turns into an availability fest.

Considering that we already know there are issues with a tier list that strictly adheres to all aspects of economic profit (specifically the issue is the effect of including the opportunity cost of deployment), it turns into an availability fest since we don't want to bother with the opportunity cost of deployment. So we focus on your second paragraph.

[Progenitus]

nflchamp, if I'm not mistaken, isn't C and X the same thing in your example? C is the cost of using the unit, which is supposed to include the opportunity cost of using a unit. And X is the... expected performance? I dunno wtf that is, but I'm assuming it's some sort of opportunity cost where the unit is being weighed against the average unit, but the opportunity cost is supposed to already be weighed against the average unit (since you can't assume that a given unit will always replace a specific unit), and this was already considered with the variable C.

Now, instead of using variables that come out of nowhere, let's instead use a different example.

Suppose we have a game with 2 chapters in it. In the first chapter there's unit A and B. Unit C joins in the second chapter.

Let's say that we've done thorough calculations and we've determined that unit A and B are exactly equal to each other in both chapters. Unit C's performance is such that it will be equal to the unit who was used in chapter 1 and got exp (it will beat the unit that wasn't used).

Now lets assume that both chapters can only field one unit each. That is, in chapter 1, we can only field either unit A or B. And in chapter 2, we can likewise only field one of the three units.

Unit C is not worse than unit A or B. Suppose we did a comparison between unit A and unit C. Unit A does not win because I could simply use unit B in chapter 1 with the same result as if I used unit A in chapter 1. Then for chapter 2, I drop unit B and use unit C to get the same result.

If you still disagree with this, then I suppose we'll just have to agree to disagree. I'm sure that by now, continuing this argument will get nowhere, and I'd like to get back to our scheduled television program.

[nflchamp]

nflchamp, if I'm not mistaken, isn't C and X the same thing in your example? C is the cost of using the unit, which is supposed to include the opportunity cost of using a unit. And X is the... expected performance? I dunno wtf that is, but I'm assuming it's some sort of opportunity cost where the unit is being weighed against the average unit, but the opportunity cost is supposed to already be weighed against the average unit (since you can't assume that a given unit will always replace a specific unit), and this was already considered with the variable C.

C and X are in no way, shape, or form related. Especially since one is a variable and the other is constant. To put it simply, X is exactly how much I expect every unit to further my goal of beating the game within the chapter. This only makes sense, of course, after admitting that P > C so long as increasing C promotes winning. You seemed to skipped that entire progression.

But to really make my point, let's use your example to show the exact same thing.

Now, instead of using variables that come out of nowhere, let's instead use a different example.

Suppose we have a game with 2 chapters in it. In the first chapter there's unit A and B. Unit C joins in the second chapter.

Let's say that we've done thorough calculations and we've determined that unit A and B are exactly equal to each other in both chapters. Unit C's performance is such that it will be equal to the unit who was used in chapter 1 and got exp (it will beat the unit that wasn't used).

Now lets assume that both chapters can only field one unit each. That is, in chapter 1, we can only field either unit A or B. And in chapter 2, we can likewise only field one of the three units.

Alright, I can live with that.

Unit C is not worse than unit A or B. Suppose we did a comparison between unit A and unit C. Unit A does not win because I could simply use unit B in chapter 1 with the same result as if I used unit A in chapter 1. Then for chapter 2, I drop unit B and use unit C to get the same result.

It's cool and all that you get the same result, but that doesn't mean they've helped the same. So to see who is helping me the most, let's add up all the times they could be helping me and subtract all the times another unit could be helping me, by the respective amounts of help I get of course. Let's look at all the scenarios first, as it is easy enough to do:

2A

1A + 1C

1B + 1C

2B

A could be used in 3 possible chapters, B in 3 possible chapters, and C in 2 possible chapters. As A = B = C whenever they are being used, let's just set them all to equal to X. So let's look at how much they help me out.

U_A (Utility of A) = 3A - 3B - 2C = 3X - 2X - 2X = -2X

U_B = 3B - 3A - 2C = 3X - 3X - 2X = -2X

U_C = 2C - 3A - 3B = 2X - 3X - 3X = -4X

As you can see A = B > C. Now you could say that this is only true so long as X >= 0, which is true. However, it is impossible of X=0 (that was part of the U = P - C; go back and read it) and if X was negative it would imply that using the characters hinders my winning, which is silly.

If you still disagree with this, then I suppose we'll just have to agree to disagree. I'm sure that by now, continuing this argument will get nowhere, and I'd like to get back to our scheduled television program.

Well, I'll let you go back to creating a game in which I can rank the characters due to availability then.

[Progenitus]

Since you still failed to account for consumption of exp/kills/resources, I'll have to simply leave it at that.

[nflchamp]

Since you still failed to account for consumption of exp/kills/resources, I'll have to simply leave it at that.

How did I fail to account for that stuff? You're the one who said we already calculated all of that to be equal.

[Progenitus]

Fiona

Problems

lolz. Well, considering she's part of the unholy trinity, she has nowhere to go but up. I mean it's pretty hard to make Fiona worse than she already is.

Changes

- +4 to all stats.

- DOuble mag growth (for imbue lulz),

- Available to fight in 1-8.

Results

With this we'll get Fiona into what IS tried to make her into; a scary tank. She'll still have low str as well as some issues doubling initially, but that shouldn't be a big problem with her growths and lolearth.

[NinjaMonkey]

Changes

- Available to fight in 1-8.

You do know that 1-8 is a swamp level, right? This means that her movement is going to be severely limited (I doubt that she could get to the piece of land where the boss is, for starters).

[Ether]

Still better than,loluiznotallowedtogoplz

[NinjaMonkey]

Still better than,loluiznotallowedtogoplz

Given that there are only, like, 5 enemies on the middle island, is it really worth her coming to this level? You'd get more EXP Boss abusing the armoured guy in 1-7.

[Narga_Rocks]

Still better than,loluiznotallowedtogoplz

Given that there are only, like, 5 enemies on the middle island, is it really worth her coming to this level? You'd get more EXP Boss abusing the armoured guy in 1-7.

I believe as far as being "balanced" is concerned, needing BA when other units do not would most certainly not qualify.

Actually, I'd suggest available to be recruited in 1-6-2. Having her earlier than 1-6-2 would cause changes to the game, so that doesn't work, but being able to make her blue in 1-6-2 (edit, oops on the 1) would be nice. And the moment you make her blue her stupid friends should either disappear or become yellow. (oh, and make available in 1-8 like he said, of course.)

Edited November 12, 2009 by Narga_Rocks

[NinjaMonkey]

I believe as far as being "balanced" is concerned, needing BA when other units do not would most certainly not qualify.

If smash had upped her base level to 11, I'd agree with you. But he didn't, which means she still needs a lot of experience if she's gonna hit 20/1 (at the very least) by the end of 1-E (even on easy).

[Progenitus]

Actually, before posting more changes, here's the first draft to the new caps.

Note: 1st tier physical units have 40 HP, 10 mag, and 20 in every other stat (including lck). 1st tier magical units are, obv, str and mag switched.

[Florete]

I believe as far as being "balanced" is concerned, needing BA when other units do not would most certainly not qualify.

If smash had upped her base level to 11, I'd agree with you. But he didn't, which means she still needs a lot of experience if she's gonna hit 20/1 (at the very least) by the end of 1-E (even on easy).

With +4 to all her bases, she might not need to hit 20. Perhaps she'll be fine at 16/1.

[Narga_Rocks]

I believe as far as being "balanced" is concerned, needing BA when other units do not would most certainly not qualify.

If smash had upped her base level to 11, I'd agree with you. But he didn't, which means she still needs a lot of experience if she's gonna hit 20/1 (at the very least) by the end of 1-E (even on easy).

Still, she'd pretty cool and you can just seal her in 3-6. You can have 4 without paying for them.

29 hp, 12 str, 10 mag, 12 skl, 14 spd, 11 luck, 12 def, 10 res.

45% hp, 40% str, 30% mag, 40% skl, 60% spd, 55% lck, 55% def, 50% res.

Probably doesn't double anything, but so what? She's less durable than Jill (assuming levels in 1-6), but it still takes at least 27 mt to 2RKO her, which almost nothing has in 1-7. She starts okay, but not too good.

level 20 stats:

33.95 hp (/40), 16.4 str (/20), 10 mag (/10), 16.4 skl (/20), 20 spd (/20), 17.05 lck (/20), 18.05 def (/20), 10 res (/20).

I'd worry about what her 20/1 looks like, but she still won't likely reach level 20 by 3-6 in HM, even if you include my suggestion about 1-6-2. So she's not about to start 3-6 at 20/1 with 36 hp, 18 str, 21 spd, 20 def. I'd worry about balance if she did, but she won't, so its fine. I mean, I'd seriously worry if I thought it was possible to hit that level. Anyway, starting at level 9 means she gains faster exp before sealing. Hopefully she can hit level 15 or more, but I don't know.

She might be balanced, or she might be underpowered, but she's way better than she was and is most certainly a viable character on NM. As for HM, it depends on exp gain. I'd check her level 16/1 stats, but I don't feel like it. Her 20/1 stats scare me, so being unable to hit level 20 is a good thing.

Maybe she could use a boost to her hp growth, but 3.75 for her growths is probably high enough. To get a better hp growth, the other growths would need to be altered.

(enemy warriors in part 4 scare me, since they should gain the benefits of tier 3 boosts to crit and tend to have really high mt values. Fortune is my new favourite skill. Warrior in 4-E-2, assuming skill ~ spd, since I don't remember: 49 crit. Egad. Good thing it is only 2x now.)

Also, while I'm sad to see Micaiah's 25 luck at level 20 and 40 luck at level 20/20 go away, it is probably for the best now that he upped her spd so much. That actually mostly wipes away my concerns about it since now she doesn't stay ahead of Zihark in avo. And she won't be immune to the crits from everything not using special weapons. I mean, in part 4.

Edited November 12, 2009 by Narga_Rocks

[NinjaMonkey]

I believe as far as being "balanced" is concerned, needing BA when other units do not would most certainly not qualify.

If smash had upped her base level to 11, I'd agree with you. But he didn't, which means she still needs a lot of experience if she's gonna hit 20/1 (at the very least) by the end of 1-E (even on easy).

With +4 to all her bases, she might not need to hit 20. Perhaps she'll be fine at 16/1.

That's still 11 levels of EXP needed though. Besides, I don't like the idea of missing out on 4 levels of stat increases.