Character "Overall Power" Calculation Factors.

[uzy5o]

The problem with finding the value of stats purely theoretically is that you're assuming situations that won't actually happen. It's kinda ridiculous to consider scenarios where the characters do 1 damage, or have 0% change to hit, etc. You're also using a 5-30 stat spread, but at any given point in the game, your characters and enemies will have a much narrower stat spread.

Also, you're basing the value of a stat on average damage, even though damage may be overkill. For example, going from doing 9x2 damage to 10x2 to a 20 HP enemy is a huge improvement, whereas going from 10x2 to 19x2 on that same enemy is basically worthless. But if you only care about average damage, then 9x2 to 10x2 is a small improvement, whereas 10x2 to 19x2 is a huge improvement. That's probably why in your calculations, strength ended up being so valuable: more damage is always better.

[Narga_Rocks]

My problem with this whole thing is your conclusion doesn't match up with the value of various units in game. By your conclusions, a character with above average strength and defence but mediocre speed would be better than a unit with mediocre strength and defence but above average speed. This is almost never true. The ability to double stuff has always been the determining factor in which units are better (assuming you are comparing members of the same class with similar availability). You have healers to account for units with mediocre durability (dodge only about half the time and get 3HKOd, for example). You have good weapons to allow units to 2HKO many things regardless of actual strength (as long as it isn't completely terrible). Some games even give you forges and doubling allows you extra chances to crit if you don't have the strength or weapon to 2HKO and must rely on procs (or in the case of fe4/9/10 to proc skills or crits. Even fe7 it gives you extra chances to OHKO with Assassins). A speedwing is almost always a more valuable resource than an energy drop, and always more valuable than a dracoshield. The potential improvement that it gives to a unit is greater than nearly anything you'll get from energy drops/rings.

And like others have said, when you play a game you will make a lot more attacks than the enemy. Beyond how you are more likely to double than the enemy (particularly with slow fe7 enemies), you are also able to be smart about your attacks. Go after archers at 1 range. Go after 1 range guys at two range. And when you 3HKO and double, generally you attack with unit A (2 for you, 1 for the enemy) and then attack with unit B (1 for you, enemy dead). That's triple the number of attacks. On average throughout a playthrough, you'll almost certainly make three times the number of attacks as you will be attacked. Having strength = defence in importance is a travesty when you consider the reality of a well-thought FE playthrough.

When your conclusions utterly fail to line up with the reality of the game, that indicates that something along the way was wrong. Now, if your calculations were not the problem (and given the effort you've put in here I'll assume they aren't) then there must be something wrong with the assumptions. You say you are trying to calculate a character's "Overall Power", but power to do what? It certainly isn't beating on enemies better.

And yes, speed without strength fails. But strength/defence without speed fails too. Maybe not as much, and not if you actually get strong enough to OHKO or have enough defence to force tinks. But they do fail. Assuming a character's overall power = a₁x₁ + a₂x₂ + a₃x₃ + ... is probably part of the problem. Have you considered attempting to use a less linear function? A more appropriate function is probably more along the lines of

overall power = x₁^a₁x₂^a₂x₃^a₃...

I admit that there are problems with this function because if any stats (the x_is) are 0 then the overall power is 0, which isn't true. However, it might do more to display the interdepedency of the various stats. This function at least allows speed to be really important without ignoring how important strength is. Values of 2 strength and 30 speed would not be nearly as good as, say, 14 strength and 18 speed. Which is appropriate. And 18 strength + 14 speed could then be worse than 14 str + 18 spd without making lopsidedness appear to be better (2 str + 30 spd). There is still the issue that massive strength and defence overrides the need for speed and the new formula wouldn't account for that, however I think this new function at least allows for a more accurate conclusion in the middle ranges (between, say, 8 and 24, since that's where most stats end up). Oh, and in case you didn't know, the a_is are all 0 < a_i < 1. To make it more like Cobb-Douglas functions in economics (since that's basically what I'm going for here) you could make a₁ + a₂ + ... + a₈ = 1, but I don't think that is 100% necessary to make it work.

(In the second equation, I couldn't make the subscripts for the exponents look like actual subscripts. Sorry)

Edited August 11, 2010 by Narga_Rocks

[dondon151]

When your conclusions utterly fail to line up with the reality of the game, that indicates that something along the way was wrong. Now, if your calculations were not the problem (and given the effort you've put in here I'll assume they aren't) then there must be something wrong with the assumptions. You say you are trying to calculate a character's "Overall Power", but power to do what? It certainly isn't beating on enemies better.

I think he doesn't care about the conditions of the enemies in game. He basically wants stat configurations that win 50% of matchups and lose 50% of matchups, despite the fact that a vast majority of the matchups are never represented in-game.

[eclipse]

I don't think that I've ever before gotten sick of seeing the word "average" used so many times.

I'd have to agree.

Since this is FE7, how is CON going to be figured in? The heavier characters won't care, but it affects the lighter, speedier characters who may have to rely on heavier weapons for damage output.

[Narga_Rocks]

I'd have to agree.

Since this is FE7, how is CON going to be figured in? The heavier characters won't care, but it affects the lighter, speedier characters who may have to rely on heavier weapons for damage output.

CON only matters until you can start getting multiple silvers and killers. Looks like you have to wait a bit more in fe7 to get silvers than in fe6 (if you go sacae), but the points stands. Most lance users (like, the ones that aren't on pegasi) can even use Javelins after promotion for no AS loss. Granted Raven is a far cry from Dieck in terms of CON and thus cares about it for Hand Axes, but the Fighters have the CON for it. Even Vaida has the CON for it. CON may have more affect on units in fe7 than fe6 (beyond chapter 13) but it's not exactly a huge factor when the strongest weapons (killers and silvers) tend to only have 1 or 2 more wt than iron stuff (silver sword being the only exception with a +3). Oh, and wt has a much much larger impact on magic users than physical weapon users, of course, because it is so crazy heavy. In fe6, though, most of the good stuff didn't weigh anyone down or only did so by 2 or 3.

[eclipse]

I was thinking more of early-mid game, and a lot of those early units aren't exactly heavy. Also, I think that's where the Hammer/Poleax/other stuff in that vein get the best use (IIRC, most of the scarier late-game units are mages, and other things that don't care about such weapons).

It looks like Speed and Strength are the main points that are being brought up. If there's a unit that's extremely fast, but has the wrong combination of CON and Strength, that unit will be put at a greater early-mid game disadvantage. This will probably affect whether the unit is even considered for late game (Rebecca is a prime example of this).

[Anouleth]

CON only matters until you can start getting multiple silvers and killers. Looks like you have to wait a bit more in fe7 to get silvers than in fe6 (if you go sacae), but the points stands. Most lance users (like, the ones that aren't on pegasi) can even use Javelins after promotion for no AS loss. Granted Raven is a far cry from Dieck in terms of CON and thus cares about it for Hand Axes, but the Fighters have the CON for it. Even Vaida has the CON for it. CON may have more affect on units in fe7 than fe6 (beyond chapter 13) but it's not exactly a huge factor when the strongest weapons (killers and silvers) tend to only have 1 or 2 more wt than iron stuff (silver sword being the only exception with a +3). Oh, and wt has a much much larger impact on magic users than physical weapon users, of course, because it is so crazy heavy. In fe6, though, most of the good stuff didn't weigh anyone down or only did so by 2 or 3.

Well, Raven generally has enough speed and CON anyway that he doesn't mind being weighed down. It's a bigger issue for Isadora and the Pegasus Knights and Karla. Even KEs weigh down Isadora by 1.

Edited August 11, 2010 by Slowking

[Alondite]

The problem with finding the value of stats purely theoretically is that you're assuming situations that won't actually happen. It's kinda ridiculous to consider scenarios where the characters do 1 damage, or have 0% change to hit, etc. You're also using a 5-30 stat spread, but at any given point in the game, your characters and enemies will have a much narrower stat spread.

Also, you're basing the value of a stat on average damage, even though damage may be overkill. For example, going from doing 9x2 damage to 10x2 to a 20 HP enemy is a huge improvement, whereas going from 10x2 to 19x2 on that same enemy is basically worthless. But if you only care about average damage, then 9x2 to 10x2 is a small improvement, whereas 10x2 to 19x2 is a huge improvement. That's probably why in your calculations, strength ended up being so valuable: more damage is always better.

Yay another person who has no clue what I'm actually doing. Ok. I'm not battling characters with stats of 5 against characters with stats of 30. I'm measuring how much damage is added when going from 5-6 Strength (or speed, skill, luck...), then from 6-7, 7-8...all the way to 29-30, and then averaging the damage added for each point to find how much actual damage 1 point of strength adds.

Here is a realistic combat situation. We have character A attacking character B. What I will do is multiply the damage dealt by each of my weighted values that attributes to damage done which will show us how much damage each stat is directly attributing. Then I will add all of those resulting values, and if they match the actual damage, my values are correct. I'm guessing that the weighted value will be slightly higher than the actual damage, since my speed and luck values also include how much they count for defense (as avoid), and since avoid means nothing to the attacking unit, it will artificially inflate the value a little bit. Probably only like 5% or so.

Ok, character A (Attacking Character)

Str-20 Skill-16 Speed-18 Luck-14 Defense-18 Resistance-12

Here are the stats for character B (defending character, in the same order)

15 16 17 12 14 12

Let's say the weapon used has 7 Mt and 80 hit. The weapon doesn't matter, as the end result will be the same no matter what weapon is used (I've tried it).

Also note this will be a physical attack. My resistance value is based on the ratio of physical attackers to magic attackers for which I used player characters, since I'm applying these values to enemies (who have to face player characters), and I wasn't about to find the ratio of all physical:magic using enemies in the game.

Character A does 9.49 damage on average. Now multiply that by my weighted percent values for each stat that effects offense, and we get the following:

Strength accounts for 4.98 damage

Skill for 2.04

Speed for 2.27

Luck for 0.67 damage.

Total? 9.96. Very close. About 5% higher in fact. Again, that's because the defensive value of speed and luck does nothing offensively, but it's included in my value anyway since Speed and Luck are both one stat each.

My problem with this whole thing is your conclusion doesn't match up with the value of various units in game. By your conclusions, a character with above average strength and defence but mediocre speed would be better than a unit with mediocre strength and defence but above average speed. This is almost never true. The ability to double stuff has always been the determining factor in which units are better (assuming you are comparing members of the same class with similar availability). You have healers to account for units with mediocre durability (dodge only about half the time and get 3HKOd, for example). You have good weapons to allow units to 2HKO many things regardless of actual strength (as long as it isn't completely terrible). Some games even give you forges and doubling allows you extra chances to crit if you don't have the strength or weapon to 2HKO and must rely on procs (or in the case of fe4/9/10 to proc skills or crits. Even fe7 it gives you extra chances to OHKO with Assassins). A speedwing is almost always a more valuable resource than an energy drop, and always more valuable than a dracoshield. The potential improvement that it gives to a unit is greater than nearly anything you'll get from energy drops/rings.

Holy crap you are missing the point. Ok, here are two characters with the same base stat total, just to prove you wrong.

Character A

str-20 ski-16 spd-10 luck-14 defense-18 resistance-12

Character B (same stat order)

15 16 21 12 14 12

According to my values, character A has a weighted power value of 28.03, and B a weighted power of 25.79. Notice how this directly contradicts your statement about characters with above average Str and Def, vs above average Spd.

Now let's see who wins in combat (same 7 mt 80 hit weapon as in the other trial)

A does 8.45 damage on average per round.

B does 6.27 per round (that includes the double attack).

Hm...looks like you were wrong and I was right. Oops.

And like others have said, when you play a game you will make a lot more attacks than the enemy. Beyond how you are more likely to double than the enemy (particularly with slow fe7 enemies), you are also able to be smart about your attacks. Go after archers at 1 range. Go after 1 range guys at two range. And when you 3HKO and double, generally you attack with unit A (2 for you, 1 for the enemy) and then attack with unit B (1 for you, enemy dead). That's triple the number of attacks. On average throughout a playthrough, you'll almost certainly make three times the number of attacks as you will be attacked. Having strength = defence in importance is a travesty when you consider the reality of a well-thought FE playthrough.

Ahem. WRONG. 1 point of defence on average reduces the same amount of damage as one point of strength gives. Proof? Increase the Strength of character A by one and the Defence of character B by one and character A still does the same amount of damage. I'm not talking whole game here, I'm talking about a round of combat, and since the entire game is the sum of many rounds of combat...well....I'm sure you get the point.

It's obvious that you just have no idea about what my numbers actually MEAN. They don't mean a character with higher strength and defenese is always better than a faster one. They are weighted numbers used to calculate how much each stat contributes to combat damage.

When your conclusions utterly fail to line up with the reality of the game, that indicates that something along the way was wrong. Now, if your calculations were not the problem (and given the effort you've put in here I'll assume they aren't) then there must be something wrong with the assumptions. You say you are trying to calculate a character's "Overall Power", but power to do what? It certainly isn't beating on enemies better.

You have yet to prove that my "assumptions" are wrong at all, or that they fail to line up with the reality of the game. I have used characters cased on these values in game, and I can assure you that they DO line up with the reality of the game, since they are calculated from the game mechanics! while I have used several examples to show that they are not. Character A in the above example would perform better against all enemies than character B on average. Character B might perform better against particular enemies, but overall character A would have a higher net damage (given-taken).

And yes, speed without strength fails. But strength/defence without speed fails too. Maybe not as much, and not if you actually get strong enough to OHKO or have enough defence to force tinks. But they do fail. Assuming a character's overall power = a₁x₁ + a₂x₂ + a₃x₃ + ... is probably part of the problem. Have you considered attempting to use a less linear function? A more appropriate function is probably more along the lines of

overall power = x₁^a₁x₂^a₂x₃^a₃...

I admit that there are problems with this function because if any stats (the x_is) are 0 then the overall power is 0, which isn't true. However, it might do more to display the interdepedency of the various stats. This function at least allows speed to be really important without ignoring how important strength is. Values of 2 strength and 30 speed would not be nearly as good as, say, 14 strength and 18 speed. Which is appropriate. And 18 strength + 14 speed could then be worse than 14 str + 18 spd without making lopsidedness appear to be better (2 str + 30 spd). There is still the issue that massive strength and defence overrides the need for speed and the new formula wouldn't account for that, however I think this new function at least allows for a more accurate conclusion in the middle ranges (between, say, 8 and 24, since that's where most stats end up). Oh, and in case you didn't know, the a_is are all 0 < a_i < 1. To make it more like Cobb-Douglas functions in economics (since that's basically what I'm going for here) you could make a₁ + a₂ + ... + a₈ = 1, but I don't think that is 100% necessary to make it work.

(In the second equation, I couldn't make the subscripts for the exponents look like actual subscripts. Sorry)

So please tell me then how stats should be weighted. Give me your values. And then we will make weighted equal characters based on those values, and if they really are equal, than they should fare equally well against each other.

Here is an example of the growths I used for Sain and Kent to be weighted equals. I made Sain the fast one since the green knight is supposed to be the fast one >:o

Name HP S/M Skl Spd Lck Def Res

Sain 80 42 46 46 40 35 20

Kent 85 45 35 35 29 42 20

Which one would you take? I bet they perform VERY similarly on average.

[dondon151]

Ahem. WRONG. 1 point of defence on average reduces the same amount of damage as one point of strength gives. Proof? Increase the Strength of character A by one and the Defence of character B by one and character A still does the same amount of damage. I'm not talking whole game here, I'm talking about a round of combat, and since the entire game is the sum of many rounds of combat...well....I'm sure you get the point.

I honestly think you don't understand anything that we are telling you. I don't know if it's because you don't understand it, or because you refuse to accept any criticism contrary to your assumptions.

In any case, your models fail to factor in the most important element of FE gameplay: a human player plays against an AI; 2 AIs don't play against each other. Now, your averages nonsense would be all fine and dandy if you pit 2 equally thoughtless AI against each other and watched them duke it out. But for the player to be playing against the AI, well, that's already heavily skewed in the player's favor. The AI can't plan ahead, can't take advantage of baits and lures, and can't take advantage of positioning other than their starting positions. Not only that, but the player can accurately predict each of the AI's moves with enough knowledge, place baits and lures, and use positioning to their advantage. The simple fact that a conscious player can work with tiles on a map and can choose weaponry against the enemy greatly skews the weighting of the stats.

As for the example that I quoted, it is false logic to equate rounds of combat with the number of attacks made and taken. If your character is doubled by more than half of the enemies in the game and doubles less than half of the enemies in the game(i.e. has below average AS), and we ignore all other factors, then defense ends up saving more damage than strength gives. If your character has poor avo and poor hit, then that works in defense's favor as well. After that, we can factor in everything else that isn't strictly determined by combat parameters, but determined by player choice instead. If your character ever needs to finish off enemies, then he won't take a counter, so that favors strength. If your character is an archer, and taking attacks from most enemies (basically, not locked to 2 range) is therefore inefficient, the worth of defense is greatly diminished. If your character is a staff user with sufficient staff rank for, say, Recover, then magic is fairly worthless and defense is fairly worthless as well.

And finally, not all marginal points of damage taken or saved are created equal. I don't think I need an example to illustrate this, because it's fairly obvious that, for example, doing 16x2 instead of 15x2 to a 30 HP enemy accomplishes nothing new. Maybe this system of averages would be true of we conglomerated all player units into 1 unit and all enemy units into 1 unit, with combined HP, and then traded hits until one or the other fell, but that's already far, far removed from the game.

Name HP S/M Skl Spd Lck Def Res

Sain 80 42 46 46 40 35 20

Kent 85 45 35 35 29 42 20

Which one would you take? I bet they perform VERY similarly on average.

Easy. I would take Sain over Kent, any day. Assuming that their bases are the same, Sain pretty much always has +2 str over Kent, and outstrips him in spd at roughly 10/0, which is shortly after they join on HHM. Kent, with his 1 AS advantage, is not doubling anything significant, but lategame, Sain will double enemies faster and will do 4 more damage when both double. This is a blowout in Sain's favor. The slightly worse defensive parameters are trivial because neither will be experiencing survival issues.

Edited August 11, 2010 by dondon151

[Anouleth]

Yay another person who has no clue what I'm actually doing. Ok. I'm not battling characters with stats of 5 against characters with stats of 30. I'm measuring how much damage is added when going from 5-6 Strength (or speed, skill, luck...), then from 6-7, 7-8...all the way to 29-30, and then averaging the damage added for each point to find how much actual damage 1 point of strength adds.

Say we have an enemy with 30HP. Character A does 20 damage and Character B does 25 damage. Now, the two are functionally equal, since both need two rounds of combat to 1-round. Once you can 2HKO the enemy, additional strength does not help your case until you can reach OHKO strength.

While your simulation is fine to accurately judge how useful stats are for raw damage, the game is about much, much more than 'who has the highest number'.

So please tell me then how stats should be weighted. Give me your values. And then we will make weighted equal characters based on those values, and if they really are equal, than they should fare equally well against each other.

Uh, why? The idea of this is presumably not to create a version of FE7¨such that the Link Arena is perfectly balanced. It's to create a version of FE7 such that each character is useful.

It's possible to have three characters A B and C such that A can beat B can beat C can beat A. Does that make one equal or does it make one the best?

Name HP S/M Skl Spd Lck Def Res

Sain 80 42 46 46 40 35 20

Kent 85 45 35 35 29 42 20

Which one would you take? I bet they perform VERY similarly on average.

Well, no shit sherlock, the biggest gap in their growths is 11%, of course they're going to be similiar. But still, I think I'd probably pick Sain.

[Narga_Rocks]

And here I was trying to be nice.

Your idea is worthless as far as the real fe7 is concerned. It in no way measures the relative powers of the characters involved. You couldn't even recognize that Sain wins handily with your new growths. Even if you make two characters start with the same bases, the Sain growths still win. After 20 levels, that's a 2 point speed win and in return only a 0.6 str loss. hp and def aren't nearly as important since neither character has survival issues. There is more than enough variance in enemy AS for Sain to double a significant number of enemies that Kent does not in each chapter. Outside of that, the small difference in strength will very frequently not even be 1. Even when it is, 1 damage is rarely enough to change the #HKO values of the units involved. A couple of instances in which Kent 2HKOs and Sain 3HKOs when neither/both double is not nearly enough to overcome the many times in which Sain would double and Kent would not. At least, when the difference in strength is never more than 1.

Now, if you are setting up some kind of VS rating system in which the only important thing is who comes out on top in a 1 on 1 with no interference from any other units and no beating of chapters involved, your idea has some merit. Large enough differences in str and def are enough to overpower a spd advantage in a 1 on 1. But that doesn't help you calculate "Overall Power" as far as the actual game is concerned.

Yay another person who has no clue what I'm actually doing. Ok. I'm not battling characters with stats of 5 against characters with stats of 30. I'm measuring how much damage is added when going from 5-6 Strength (or speed, skill, luck...), then from 6-7, 7-8...all the way to 29-30, and then averaging the damage added for each point to find how much actual damage 1 point of strength adds.

Here is a realistic combat situation. We have character A attacking character B. What I will do is multiply the damage dealt by each of my weighted values that attributes to damage done which will show us how much damage each stat is directly attributing. Then I will add all of those resulting values, and if they match the actual damage, my values are correct. I'm guessing that the weighted value will be slightly higher than the actual damage, since my speed and luck values also include how much they count for defense (as avoid), and since avoid means nothing to the attacking unit, it will artificially inflate the value a little bit. Probably only like 5% or so.

Ok, character A (Attacking Character)

Str-20 Skill-16 Speed-18 Luck-14 Defense-18 Resistance-12

Here are the stats for character B (defending character, in the same order)

15 16 17 12 14 12

Let's say the weapon used has 7 Mt and 80 hit. The weapon doesn't matter, as the end result will be the same no matter what weapon is used (I've tried it).

Also note this will be a physical attack. My resistance value is based on the ratio of physical attackers to magic attackers for which I used player characters, since I'm applying these values to enemies (who have to face player characters), and I wasn't about to find the ratio of all physical:magic using enemies in the game.

Character A does 9.49 damage on average. Now multiply that by my weighted percent values for each stat that effects offense, and we get the following:

Strength accounts for 4.98 damage

Skill for 2.04

Speed for 2.27

Luck for 0.67 damage.

Total? 9.96. Very close. About 5% higher in fact. Again, that's because the defensive value of speed and luck does nothing offensively, but it's included in my value anyway since Speed and Luck are both one stat each.

Way to ignore one of his better points. Strength and Defence are only important insofar as they help you to kill things. Why did you not respond to his more important point?

Character A does 9x2 damage to an enemy with 20 hp.

Character B does 10x2 damage to an enemy with 20 hp.

If I increase Character A's strength by 1, he now does 10x2 and kills the thing. If I increase Character B's strength by 9, he now does 19x2 and kills the thing.

Character B had 0 improvement from the +9 strength I just gave him. Your calculations give this a rather large importance despite the fact it does nothing. Character A went from 2RKOing to ORKOing. A rather large improvement. Your calculations see this as a much much smaller improvement than what Character B got. This is a pretty common occurrence over the course of playing fe7. You will have these situations pop up lots of times. Probably hundreds, where one character gets a large improvement from +1 or +2 str and another gets no relevant improvement from +8 or +9 str. And yet your set-up does not acknowledge this scenario's existence. And you don't even see how that's a problem.

Let's go back to one of your earlier "counters" that I left alone. The bold was you, the rest was me.

How much are these worth:

40 hp, 28 str, 14 skl, 16 spd, 15 luck, 30 def, 15 res.

40 hp, 30 str, 14 skl, 12 spd, 15 luck, 30 def, 15 res.

Which is worth more by your system? Which one do you think will win? With any weapon that has at least 4 mt, the first is almost guaranteed to win.

The second is like a tenth of a point stronger overall, but the first wins because the average damage increase from 4 points of speed over your opponent is higher than 2 points of strength. But stats don't cap at 4 obviously. What about a speed difference of 8 and a strength difference of 4? Same difference proportionally, but I bet the outcome may be different. If speed really is more important, than the advantage should be fairly constant, but it's not. What about each of them against an enemy with these stats?

40 hp, 29 str, 14 ski, 14 spd, 15 luck, 30 def, 15 res.

I bet the stronger character fares better. That's because speed ONLY gives you an offensive bonus when the difference between attack speeds is 4 (or higher, but that gives you the same bonus, not an additional one, so it increases average damage by 0). Assuming the characters remain "equal" for every variation, and speed difference below or above 4 presents NO offensive bonus. Therefor, on average, one point of strength is worth more than one point of speed.

Who fares better is 100% dependent on the weapon used. Let's make both characters 3HKO. 16 mt weapon allows this. As does 17 and 18 and 19 mt. Now who does better against your random enemy? Mine. Why? Mine goes first. Yours goes second. Mine only gets hit 2 times, yours gets hit 3 times and dies. If you let the enemy go first, mine is more likely to dodge and not die. What if either yours or mine misses the enemy if we both go first? Mine is more likely to the extra attack that the enemy gets.

At some mt values you'll win more often. Like 11 mt. You are able to 4HKO, the enemy is able to 4HKO, but mine 5HKOs.

Also, don't forget that there are actually a large number of enemy speed values that give me a massive advantage. You only really get a minor advantage on the rest because you only win when your #HKO value is better than mine and the enemy shares your HKO value rather than mine.

With anything from 9 to 12 speed, I double and you don't. With anything from 16 to 19 speed, you get doubled and I don't. As for the rest, we are both pretty much screwed at 20+. Chances are, the enemy will slaughter us. For some weapon mt values, you'll do better than me since you may only need to get lucky and dodge once (and you need to hit the enemy each time) whereas I may need to dodge twice (and hit the enemy each time). With other mt values, I have the advantage since if we both only need to dodge once I am more likely to do so. With anything that has 8 or less speed, we both pretty much obliterate anyway so who cares? Unless you want to give us weapons that only have 3 or less mt, we are both pretty much going to win easily. That means I have a significant advantage for 8 speed values (9 to 12 and 16 to 19), you have a minor advantage at best (after averaging over the various weapon mt possibilities) for 3 speed values (13 to 15), and you have a minor advantage at best for the other 11 (20 to 30), though we are both pretty much screwed in that instance. I'd say when you consider the various speed values of the enemy (31 possibilities if you include 0) while keeping the rest of the stats constant, I win. Also, don't forget weapon wt. My unit can switch weapons to sacrifice a little speed if the enemy has 13 to 15 speed. If you sacrifice a little speed, you may end up getting doubled.

And let's not forget how other enemies exist that don't have 30 def. As you decrease enemy def, our characters are more and more likely to KO the enemy in the same number of rounds. Which tends to either make us equivalent (since neither has any chance of death and will generally KO in the same number of rounds) or make mine win since mine may actually dodge something. Particularly since our units may miss the enemy allowing it to have enough swipes that it has the potential to kill one of us. If we 2HKO and it 3HKOs, well, we may only need to miss once for a COD > 0%. Mine will be less.

Holy crap you are missing the point. Ok, here are two characters with the same base stat total, just to prove you wrong.

Character A

str-20 ski-16 spd-10 luck-14 defense-18 resistance-12

Character B (same stat order)

15 16 21 12 14 12

According to my values, character A has a weighted power value of 28.03, and B a weighted power of 25.79. Notice how this directly contradicts your statement about characters with above average Str and Def, vs above average Spd.

How does that contradict anything? I said you'd value A more than B. You do.

Now let's see who wins in combat (same 7 mt 80 hit weapon as in the other trial)

Back to the vs. tournament, eh? What about the enemies? I specifically said "your conclusion doesn't match up with the value of various units in game."

Why are you countering a statement about how the game works with a vs. example? What about the game? First off, you won't be stuck with a 7mt weapon for the entire game. You'll have one near the beginning, but at that point you won't have those stats. Divide everything by 2 if you want to use weaker weapons. That shouldn't change anything since you made it all linear. Unfortunately you gave me two odd numbers, so I'll round str up and spd down.

Character A

str-10 ski-8 spd-5 luck-7 defense-9 resistance-6

Character B (same stat order)

8 8 10 6 7 6

Okay, so per hit you do 10 damage. I do 6 damage per hit. Oh look, I'm already winning before we account for misses since I double. 12 > 10. HP matters, though, since timing could still give you the victory. (MYMYMMMYM if I go first, and you do 20 damage before I can get past 12. However, I do 24 and 30 before you can get past 20. If you go first in the first round it just shifts around the hp values at which I have the advantage.) Now, hit matters as well, and if we go with fe7 then hit = skill*2 + luck/2. We both have 19 base hit, but I have 26 avo and you have 17 avo. I have +2 hit and you have -7 hit. It's not huge, but it can make a difference in my favour for the unfavourable hp values. It doesn't do a thing for you when hp is unfavourable to you. All that does is increase my win%.

(If I go first and nobody ever misses, MYMYMMMYMYMMMYMYMM, I win when the hp for both of us start at: 1-6, 11-12, 21-36, 41-n as n->infinity. At infinity, obviously, nobody wins since nobody can ever get hit to 0. You can only win for 7 to 10, 13 to 20, and 37 to 40. A lot less.)

Use something with more umph and increase the stats (say, x1.5) and it's the same. Multiply by 1.5 and use 10 or 11 mt weapons and see what happens. Again, if you are simply looking at vs. tournaments and forcing weak weapons (so, I guess, arena only?) then your stuff works out. Think closer to the game and not so much.

Let's go back to your numbers there. 14 mt weapon, now. You do 20 damage per hit. Go you. I do 11 per hit and double. Not quite the same as before, but I still outdamage you by 2 each round. You do a bit better in the comparison over the various hp values, but I'm guessing I still win. And when you start accounting for misses, I do even better. We are fighting each other, and since I have a 20 avo advantage you are more likely to miss than I am. I have to take two hits, so on the one hand I may have less chance to do full damage, but at the same time I have more chance to do some damage. 14 mt is a silver lance (one example of 14 mt, anyway). 75 hit.

I have 38 base hit. You have 39 base hit. I have 54 avo. You have 34 avo. I get +4, you get -15.

79

60.

Let's ignore true hit since you said you never accounted for that. Of course, the game uses true hit which might change your average damage with those pathetic iron lances a little, but who cares, right? (It's possible you did use true hit for that, though I don't feel like recalculating your calcs to find out)

My chance of hitting twice? 62.41%. Gee, this is going to be sad for you, I can already tell. My chance of not hitting? 4.41%. I'm left with 33.18% chance of hitting once.

I do 11 per hit, so: 17.38 damage per round.

Your chance is 60% and you don't double and you do 20 damage. 20 * .6 = 12.

I outdamage you by 5.38 per round. That's pretty impressive, eh? I think I could even win with 10 mt weapons. You do 16 when you hit, I do only 7 per, but that's still 14 total. I don't know if there are any 75 hit 10 mt weapons without crit rates, but I don't care at the moment. I hope you don't, either.

7 * 2 * .6241 + 7 * .3318 = 11.06 (you can simplify the first part to 1.58 * 7)

16 * .6 = 9.6

Wow. I still win. You have to look at pretty weak weapons before you get to start winning, dude. Which unit in the game is expected to be mainly using iron weapons when they have stats like 20 str, 18 def, 21 spd? (I mean, sure, my 20/2 Dieck in fe6 will sometimes run around ORKOing with iron swords, but only on the weak pirates in 12x and places like that. He still uses silver and killers for the ones that aren't pathetic)

If you like, I'll calculate what happens with 9 mt, 75 hit, 30 crit killer weapons. (wow, in a game where nearly every weapon got +10 hit, killers got -5 or +0 hit. Granted, they were pretty psychotic in fe6) You know, it's a good thing that this also has 75 hit, otherwise I wouldn't do this next part.

6 for me, 15 for you. Who do you think will win when we account for crit rates?

Without crit, 1.58 * 6 = 9.48 and .6 * 16 = 9. I win (barely) without accounting for crit. What do you think happens with crit?

A does 8.45 damage on average per round.

B does 6.27 per round (that includes the double attack).

Hm...looks like you were wrong and I was right. Oops.

How do you figure?

My problem with this whole thing is your conclusion doesn't match up with the value of various units in game. By your conclusions, a character with above average strength and defence but mediocre speed would be better than a unit with mediocre strength and defence but above average speed. This is almost never true. The ability to double stuff has always been the determining factor in which units are better (assuming you are comparing members of the same class with similar availability). You have healers to account for units with mediocre durability (dodge only about half the time and get 3HKOd, for example). You have good weapons to allow units to 2HKO many things regardless of actual strength (as long as it isn't completely terrible). Some games even give you forges and doubling allows you extra chances to crit if you don't have the strength or weapon to 2HKO and must rely on procs (or in the case of fe4/9/10 to proc skills or crits. Even fe7 it gives you extra chances to OHKO with Assassins). A speedwing is almost always a more valuable resource than an energy drop, and always more valuable than a dracoshield. The potential improvement that it gives to a unit is greater than nearly anything you'll get from energy drops/rings.

Look at the bolded stuff. First off, you didn't look at the two characters you created as they would perform in various chapters. Second, I said there are some times where the str/def character would win. Maybe yours would do better in the game, maybe he wouldn't (I don't think he would), but even if you showed yours to win that wouldn't actually prove anything. The third thing is that I said that the game gives you good weapons. You used 7 mt pieces of crap. And I mentioned crit. You could at least give them 9mt killing edges and take a look at what happens. What even has 7 mt and 80 hit? An iron lance? Why would you ever be using an iron lance when your characters are clearly promoted (21 spd isn't possible otherwise). How do you figure that you are "right" when your example has little in common with my statements? We aren't even arguing the same thing at the moment so it's impossible for you to prove me wrong with this. And the best part is you are still wrong with your example units unless you force iron weapons onto them. That's awesome!

If yours isn't doubling enemies in the game, mine could probably grab a silver to 2HKO and thus ORKO while yours is not likely to pull off the OHKO. Mine could grab a killer to 4HKO or 3HKO if it can't 2HKO with silver and would still get over 50% to KO an enemy. If yours picks up killer and somehow has the str to 2HKO (which would mean mine could usually 2HKO with silver since the difference is only 5 points and that's only one more than the difference between killer and silver) then you still won't get much past 30% to ORKO. I say you need to 2HKO since if you are 3HKOing while I'm 4HKOing then I 2RKO if I fail to crit and you 3RKO if you fail to crit. That's obviously an advantage for me. In the game, unless enemies are insanely slow (never get past 6), I'll choose the unit you created for me in almost any game.

If you are worried about how your guy has 4 more def than mine, don't forget mine has 20 more avo. And that line in italics where I pointed out that the game has healers for a reason. Since mine is clearly stomping yours in offence, it is worth it to heal mine from time to time if it fails to dodge enough to overcome your 4 point defence lead.

Ahem. WRONG. 1 point of defence on average reduces the same amount of damage as one point of strength gives. Proof? Increase the Strength of character A by one and the Defence of character B by one and character A still does the same amount of damage.

Then we have another problem. You are including enemies into your calculations. What is troublesome in an enemy is different from what is important in a PC. Sure, if you add 1 str to a PC or subtract 1 def from an enemy it has the exact same effect. But then you make your "Overall Power" stat even more worthless because it reflects the relative power of PC units even less than before. What we are saying is that looking at unit A, +1 str gives it more damage about 3 times as often as +1 def makes it take 1 less damage. Therefore it is pretty clear that str plays a factor for PCs about 3 times as often as for the enemy.

Troublesome enemies in a rout map could be units with massive def and res with terrible str, skl, and spd. It wouldn't matter that it can't hurt you because it would prolong the map as you are causing only 1 or 2 damage to 40+ hp. Thanks to healers, strong enemies aren't actually that problematic. If they have crit then they can OHKO and that can't be fixed (unless you can avoid getting attacked completely), but anything short of that can be dealt with. Just look at dondon's 0% growths run of fe11. Or fe1DS if you like calling it that. If you are calculating a unit's power, shouldn't you focus on unit A alone? How much does +1 def help unit A? You are equating str and def because 1 str does the same for unit A as -1 def does for unit B. But can't you see the flaw in that? As far as unit A is concerned, how much does 1 extra damage help? How much does 1 less damage received help? How often does one situation occur compared to the other? You shouldn't be looking at both unit A and unit B when considering the importance of a stat for unit A.

Enemies are on the other side of the 3:1 ratio I told you about. They are getting swung at 3x as often as they are able to attack you. Including them in your "Overall Power" for PCs will heavily skew your results away from what is relevant for PCs.

I'm not talking whole game here, I'm talking about a round of combat, and since the entire game is the sum of many rounds of combat...well....I'm sure you get the point.

Each round of combat can be very different. Many rounds only see the PC attack. I've already told you why. In the very thing you quoted, in fact. Attack an archer at 1 range. Attack a myrm at 2 range. There you go. You've just made defence mean nothing for unit A. It is the sum of many different types of rounds of combat. You can't extend one type over the whole game.

It's obvious that you just have no idea about what my numbers actually MEAN. They don't mean a character with higher strength and defenese is always better than a faster one. They are weighted numbers used to calculate how much each stat contributes to combat damage.

Then you are still mixing between unit A and unit B. Unit B's def doesn't contribute to unit A at all. Unit A's strength helps him to cause more damage. Unit A's defence helps him to avoid more damage. The question is how often does unit A get attacked? Also, I highly doubt you even did what you are claiming to have done.

Even if we accept your assumptions as reasonable:

If you have 60hp max and 30 max for the other stats (combining str and mag into one like in fe7), you have 60 * 31⁶ different possibilities (unit A's) for a character. That character must then fight against 60*31⁶ different stat set-ups for the enemy. Then you must consider how much raising each statistic helps those unit A's as they go from whatever their actual value is up to the maximum in that stat. Against each of the other enemies. This is a fair amount beyond 3600 * 31¹² (~10²¹) different calculations. Excel is not capable of that. But that's what you must do if you want to see how much each stat contributes to damage. Take unit A against unit B. Raise each stat from unit A to the maximum (one point at a time). See what the improvement in damage is as you do that (so for 10 str, you go from 10 to 30. That's 20*60*31⁶ just to deal with the strength of one unit A). And you have to do that for every single weapon in the game (multiply again by, well, how many weapons are there?). Now do that for the other stats that aren't strength. Then you have to do that for each of the other 60*31⁶ possible unit As. Or around that, anyway. Pretty sure that would take your computer a pretty long time to compute. Also I doubt that excel is capable of that. How's your C?

Oh, and since you'd have to be you not to, I'd recommend including some test to see if the improvement actually improves unit A's chance of winning in X rounds. In other words, if damage goes from 7 to 8 on an enemy with 41 or 42 hp, they kill it in the same number of hits. 6 hits. So the improvement of 1 str does zilch. However, improvements of skl or spd or lck will help because it helps you get your 6 hits in quicker or it prevents the enemy from getting however many hits they need to kill you.

You might, perhaps, be able to get away with only giving each unit A +1 one at a time in each stat and comparing each result to all the other possible enemies. That's still going to be 7*3600*31¹², which is ~1.98 * 10²². The observable universe only contains around 3 to 7 * 10²² stars, by the way, according to wikipedia. Oh, right, I forgot weapons. Well, there are at least 10 of those so you are already looking at more calculations than stars in the sky. Literally. I never knew excel was so powerful.

You have yet to prove that my "assumptions" are wrong at all, or that they fail to line up with the reality of the game. I have used characters cased on these values in game, and I can assure you that they DO line up with the reality of the game, since they are calculated from the game mechanics! while I have used several examples to show that they are not. Character A in the above example would perform better against all enemies than character B on average. Character B might perform better against particular enemies, but overall character A would have a higher net damage (given-taken).

What? Which one is A and B? The ones where I showed that any weapon with 9 mt or more gives the advantage to character B against A? And which enemies does A win on? Show me. Look around whatever chapter you think most units are going to be promoting, since the stats you are giving are pretty much late tier 1 or early tier 2. As far as durability is concerned, character B has 20 more avo, character A has 4 more def. Character B has plenty enough defence to not have survival issues unless you don't believe in healing. What the hell is character A doing more damage on? B doubles, A doesn't. In order for A to do more damage per round, you'd have to pretend that A is doing more than double the damage that B is per hit. With 15 str, I find it hard to believe that B will fail so utterly on enemies that it won't be more likely to kill sooner than A. There's only a 5 damage gap, which means that you must be assuming unit B will only be doing 4 damage or less per enemy. Which enemies are so durable that, say, a killing edge (24 mt) would have that happen? How many enemies even have 20 defence? 15 def enemies take 18 from B and take 14 from A. If B gets just one crit, it causes 36 hp damage. How much is that not killing? B is more likely to get one crit than A is since B doubles. 30% crit means 51% in two swipes. 14 * 3 = 42 damage. You'd basically need the enemy to have between 37 and 42 to get a win for A, since otherwise 51% > 30% for chance of killing. In two rounds, A does 28 without a crit and B does 36 without a crit. B has 4 chances at a crit to get up to 54 damage. That kills a lot. A only has 2 chances at a crit. Granted it would reach 56 damage in that case, so A wins on 55 to 56 hp enemies, but that still means B wins on anything from 14 to 36 and 43 to 54. A only wins on anything with 37 to 42 and 55 or 56. Above 56 it gets more complicated. "What is a win?" becomes the question. Technically A can kill in two rounds. It just needs to double crit. That is incredibly unlikely. B is more likely to win in 3 rounds. One crit out of 6 can mean 72 damage total. No crits for unit A means 42 damage in 3 rounds, and a single crit only does 70 damage total and B is far more likely to get at least 1 than unit A is to get at least 1.

Also, um, given-taken? We've spent many posts trying to explain to you why that reasoning is flawed. They can attack stuff that doesn't counter. They can attack stuff that is weakened by other units that failed to kill. There are multiple common situations where a good player will make a kill without even getting attacked. And let's compare, say, 30% hit to 50% hit. If an enemy is going to cause 10 damage to unit A and 14 damage to unit B, unit A takes 5 damage on average while unit B takes 4.2 damage on average. So unit B not only causes more damage to the enemy, it also takes less in return. 50% vs. 70% makes them equivalent (.7 * 10 = 7 and .5 * 14 = 7). You have to get some pretty accurate enemies before unit A starts taking less damage in return, and unit B is still causing significantly more due to doubling. Hence, you are wrong about what you said and what you said didn't reflect reality very well anyway. You are doubly wrong.

So please tell me then how stats should be weighted. Give me your values. And then we will make weighted equal characters based on those values, and if they really are equal, than they should fare equally well against each other.

You are the one doing the calculating. I see this as a pointless endeavour since you can't account for the extremes without invalidating the formula at the more common values. I've also told you that how they do against each other is not an indication of how they do over the game. Why is "Overall Power" about the vs. that never happens in the game? Why isn't it about beating on the enemies?

Plus, my values would be for the superior formula that is not linear. I think it's harder to come up with values for the exponents than for the coefficients of your formula. Plus I don't actually see any utility in completing the calculations. Plus I know that my formula still isn't perfect. There needs to be some linear portion to at least strength and defence. If there isn't then I can't account for how epic strength/defence can override the need for speed. 40 str, 40 skill, 0 spd, 40 def, 40 res and 15 mt weapons if no enemies have more mt than 40 and no enemies have a higher hp + def than 55 will destroy all without ever dying. Replacing any of those 40s with "speed" and making the other stat 0 would likely result in an inferior character. The first will probably never miss, doesn't care about getting doubled, and kills everything in 1 hit. 0 strength means the 40 speed one will never kill anything. 0 skill makes it much harder to hit anything. 0 defence makes it far more reliant on dodging (though 80 avo might make up for that, it is still inferior if anything has enough hit to get hit%>0), and 0 res is the same as 0 defence but is a problem less frequently. That's the problem with trying to account for the extremes, though. Your linearity makes spd look bad no matter what the other values are.

I'm guessing you've never actually seen a Cobb-Douglas function. Or done any Economics courses that deal with utility or production.

Let's go with a simple one. P = L^0.25K^0.75

The idea behind it is that 40 labour with 0 capital won't let you build anything. 0 labour with 40 capital also sucks. If L + K = 40, your best bet is L = 10 and K = 30.

P ~= 22.8 at that point. 20 labour and 20 capital results in P = 20, which is inferior. 5 labour and 35 capital results in P ~= 21.5, which is also inferior.

Now, with a linear function, let's set something up so that 10 and 30 give you 22.8. 10(.3) + 30(.66) = 22.8. Obviously, at this point I can make L=0 and K=40 and I'll get something better than what I've got. 26.4. A linear function means you can take anything to the extreme and you'll have a better overall value with the same total points in the 7 stats. If you give me a unit that has points in each stat and 100 total points, I can make a unit with 1 hp and 99 strength and your stupid linearity will tell me that my 99 strength 1hp 0 everything else unit is superior to the unit you created that is actually somewhat balanced. But I'm fairly certain that whatever you made will kill my unit far more often. If it goes first, all it needs to do is hit and it will kill (regardless of the weapon). If yours goes second, it needs to dodge. In 50% of the scenarios all you have to do is hit, and decent skill with a decent hit weapon should make that 100%. In the other 50%, you dodge you win. You'll have more than a 50% chance to win. But the sucky unit that I create with 1 hp and 99 str will probably nearly double the total "value" that your unit has. That's the trouble with linearity and that's why I recommend something that accounts for needing a variety of stats to be good. Something that has exponents between 0 and 1. That's the only way.

Here is an example of the growths I used for Sain and Kent to be weighted equals. I made Sain the fast one since the green knight is supposed to be the fast one >:o

Name HP S/M Skl Spd Lck Def Res

Sain 80 42 46 46 40 35 20

Kent 85 45 35 35 29 42 20

Which one would you take? I bet they perform VERY similarly on average.

Um, dondon responded to this pretty well. Sain has more strength forever (even at the end, .3 * 38 levels isn't even going to cover the 2 point difference). It only takes 9 levels for him to tie-game in speed, and as the game goes on he'll reach the doubling speed for each enemy sooner than Kent will. The durability difference is pretty irrelevant given how neither are having issues. Sain and Kent don't have trouble staying alive in the game as is, and you gave them even more defence than before. Clearly not an issue for Sain. Oh, and Sain continues to have better luck for annoying crit-wielding enemies.

[Alondite]

I honestly think you don't understand anything that we are telling you. I don't know if it's because you don't understand it, or because you refuse to accept any criticism contrary to your assumptions.

OR because you just don't get what I'm doing it. Test it yourself. Take two characters and pit them in battle against each other. Plug their stats into the damage formula, and then start messing with stats to see which one effects damage the most. I guarantee you 1 point of strength or defensive increases/decreases damage by more than any other stat in virtually every situation. Go ahead and do it, then tell me I'm wrong. The reason I'm refusing to yield is because I did hundreds upon hundreds of calculations, maybe even thousands. EVERY SINGLE POSSIBLE battle scenario, and these are the figures I got. There is absolutely NO situation that is going to come up in game that I didn't include in my averages. Do you honestly think I'm that stupid? I've been playing Fire Emblem for like 15 years. I know how it works.

In any case, your models fail to factor in the most important element of FE gameplay: a human player plays against an AI; 2 AIs don't play against each other.

And your criticisms fail to be relevant. Game mechanics aren't different for human and AI players. Two units are going to have the same result in battle whether they are both human, AI, or one of each.

Now, your averages nonsense would be all fine and dandy if you pit 2 equally thoughtless AI against each other and watched them duke it out. But for the player to be playing against the AI, well, that's already heavily skewed in the player's favor. The AI can't plan ahead, can't take advantage of baits and lures, and can't take advantage of positioning other than their starting positions.

Since when does any of that happen IN BATTLE? My numbers are used to find a character's battle capacity.

Not only that, but the player can accurately predict each of the AI's moves with enough knowledge, place baits and lures, and use positioning to their advantage. The simple fact that a conscious player can work with tiles on a map and can choose weaponry against the enemy greatly skews the weighting of the stats.

Again, not relevant in battle. The player can't manipulate anything once battle has been initiated.

As for the example that I quoted, it is false logic to equate rounds of combat with the number of attacks made and taken. If your character is doubled by more than half of the enemies in the game and doubles less than half of the enemies in the game(i.e. has below average AS), and we ignore all other factors, then defense ends up saving more damage than strength gives.

"If." "If" means nothing. In the average situation (yes average, meaning that it includes every possible scenario) Strength and Defense are equal. I did the math. All of it. Every possible character matchup. Do you understand what that means. Every possible relevant stat combination against every other combination. I have an Excel sheet for each stat, and my tables are massive.

If your character has poor avo and poor hit, then that works in defense's favor as well.

No, if your character has poor hit it effects strength...equally to how your avoid effects the importance of defense. That's why Strength and Defense don't have weighted values of 100%. They only do when you (or your enemy) have 100% hit, but that's certainly not an average is it?

After that, we can factor in everything else that isn't strictly determined by combat parameters, but determined by player choice instead. If your character ever needs to finish off enemies, then he won't take a counter, so that favors strength. If your character is an archer, and taking attacks from most enemies (basically, not locked to 2 range) is therefore inefficient, the worth of defense is greatly diminished. If your character is a staff user with sufficient staff rank for, say, Recover, then magic is fairly worthless and defense is fairly worthless as well.

The player has no choices in battle. The choice is made before battle. Do you understand that my numbers do not mean that one character is absolutely better than another 100% of the time? You could argue on the opposite point that since an archer can't counter melee units, that offensive is greatly diminished in that situation. You can't just pick and choose what arguments you want to make and leave out the ones that don't work in your favor, because I'm not an idiot.

And finally, not all marginal points of damage taken or saved are created equal. I don't think I need an example to illustrate this, because it's fairly obvious that, for example, doing 16x2 instead of 15x2 to a 30 HP enemy accomplishes nothing new. Maybe this system of averages would be true of we conglomerated all player units into 1 unit and all enemy units into 1 unit, with combined HP, and then traded hits until one or the other fell, but that's already far, far removed from the game.

Ok...but what about an enemy with 60 HP? 45? 16? Not all enemies have 30 HP, so using one single example is pointless. I used the average of all of them. And your last bit is so far off the wagon that I don't even know how to respond. My numbers are about any unit vs. any other unit, not all players against all enemies.

Easy. I would take Sain over Kent, any day. Assuming that their bases are the same, Sain pretty much always has +2 str over Kent, and outstrips him in spd at roughly 10/0, which is shortly after they join on HHM. Kent, with his 1 AS advantage, is not doubling anything significant, but lategame, Sain will double enemies faster and will do 4 more damage when both double. This is a blowout in Sain's favor. The slightly worse defensive parameters are trivial because neither will be experiencing survival issues.

Their bases are weighted equal too. What about an enemy that Sain can't double? Or one with high defense? Or one that Kent can double too? On top of that, those are their growths for MY game, you don't know what my enemies are like. Their overall combat proficiency is 50/50. I tested it. My chart looks like this:

Stats are in the same order as in-game, and HP is a separate calculation so it's not included.

0 0 0 0 0 0

1 0 0 0 0 0

2 0 0 0 0 0

. . . . . . etc...

30 0 0 0 0 0

0 1 0 0 0 0

1 1 0 0 0 0

2 1 0 0 0 0

ect... All the way to

30 30 30 30 30 30

Trust me, EVERY possible enemy was accounted for, and they perform 50/50. It may seem like a ton of work, but excel formulas make it nothing more than click and drag and built-in formulas.

I would not have even bothered posting this unless I was confident in my numbers, because I know how nitpickey and objective the FE community is. Argue all you want, but unless you have some cold hard proof I'm sticking to my hard work.

[Narga_Rocks]

OR because you just don't get what I'm doing it. Test it yourself. Take two characters and pit them in battle against each other. Plug their stats into the damage formula, and then start messing with stats to see which one effects damage the most. I guarantee you 1 point of strength or defensive increases/decreases damage by more than any other stat in virtually every situation. Go ahead and do it, then tell me I'm wrong. The reason I'm refusing to yield is because I did hundreds upon hundreds of calculations, maybe even thousands. EVERY SINGLE POSSIBLE battle scenario, and these are the figures I got. There is absolutely NO situation that is going to come up in game that I didn't include in my averages. Do you honestly think I'm that stupid? I've been playing Fire Emblem for like 15 years. I know how it works.

I sincerely doubt that you accounted for 10²¹ different possible scenarios. And that's just with one weapon for all those scenarios. You are looking at over 10²² when you want to consider multiple weapons.

And your criticisms fail to be relevant. Game mechanics aren't different for human and AI players. Two units are going to have the same result in battle whether they are both human, AI, or one of each.

Since when does any of that happen IN BATTLE? My numbers are used to find a character's battle capacity.

Again, not relevant in battle. The player can't manipulate anything once battle has been initiated.

"If." "If" means nothing. In the average situation (yes average, meaning that it includes every possible scenario) Strength and Defense are equal. I did the math. All of it. Every possible character matchup. Do you understand what that means. Every possible relevant stat combination against every other combination. I have an Excel sheet for each stat, and my tables are massive.

No, if your character has poor hit it effects strength...equally to how your avoid effects the importance of defense. That's why Strength and Defense don't have weighted values of 100%. They only do when you (or your enemy) have 100% hit, but that's certainly not an average is it?

The player has no choices in battle. The choice is made before battle. Do you understand that my numbers do not mean that one character is absolutely better than another 100% of the time? You could argue on the opposite point that since an archer can't counter melee units, that offensive is greatly diminished in that situation. You can't just pick and choose what arguments you want to make and leave out the ones that don't work in your favor, because I'm not an idiot.

The player chooses which battles to commence, idiot. If you have two units with 20 spd and 15 spd, and two enemies with 15 and 10, which one do you choose to fight which unit? I choose the set up that gives me the best chance to kill both enemies. What do you do? Set up a dart board with the locations of enemies and name each of your darts for your units, put on a blindfold and toss it at the board to determine who fights which enemy? The player sets it up so that you minimize counters, maximize damage (on the enmy), and take advantage of the weapon triangle to further minimize the chance that those counters actually hurt you. The AI does almost none of this because it lacks the opportunity. Some of it is programmed, but a good player can take advantage of how the game is programmed. Seriously, do yourself a favour and check out some of dondon's 0% growth playthroughs and see if you can't figure out why we are saying that you are being an idiot when you are valuing these things the way you do.

Ok...but what about an enemy with 60 HP? 45? 16? Not all enemies have 30 HP, so using one single example is pointless. I used the average of all of them. And your last bit is so far off the wagon that I don't even know how to respond. My numbers are about any unit vs. any other unit, not all players against all enemies.

Okay? And the enemies that have 60 follow the same principle. Some strength improvements mean squat because they don't change the #HKO. If you seriously have managed to find a way to account for 10²² different battle scenarios, then I recommend that you consider including a way to check out the #HKO rather than the damage.

Seriously, if you are really going for this moronic "vs" thing, then find out if unit A 3HKOs unit B or 4HKOs. Set up your calculations so that if you add 1 str and it doesn't change the #HKO value then it doesn't give a bigger improvement to Overall Power. If you are really looking at each scenario, you should be looking at how changes affect the ability to win, not to do more damage. The examples that we are showing with 30 hp or with 20 hp are meant to explain to you an idea that you seem incapable of grasping. It isn't about 30 hp specifically, it is about the fact that there are multiple occasions in which +1 str does nothing and you aren't accounting for that fact.

Their bases are weighted equal too. What about an enemy that Sain can't double? Or one with high defense? Or one that Kent can double too?

Did you ignore how even if you start them with the same base that Kent will never win by more than 1? How often do you honestly think that 1 str is changing the #HKO value? I can guarantee you that the different enemy AS values at which Sain doubles and Kain doesn't or Kain gets doubled and Sain doesn't will vastly outweigh the rare instances in which 1 str makes the slightest bit of difference.

On top of that, those are their growths for MY game, you don't know what my enemies are like. Their overall combat proficiency is 50/50. I tested it. My chart looks like this:

Stats are in the same order as in-game, and HP is a separate calculation so it's not included.

0 0 0 0 0 0

1 0 0 0 0 0

2 0 0 0 0 0

. . . . . . etc...

30 0 0 0 0 0

0 1 0 0 0 0

1 1 0 0 0 0

2 1 0 0 0 0

ect... All the way to

30 30 30 30 30 30

Trust me, EVERY possible enemy was accounted for, and they perform 50/50. It may seem like a ton of work, but excel formulas make it nothing more than click and drag and built-in formulas.

do you really have 31⁶ = 887 503 681 rows in your excel spreadsheet? And do you really compare each row against all 887 503 681 rows?

What was your mt value? Or did you even bother with that? Is damage = str - def? That will make a significant impact on the validity of your work. I admit that a difference of 1 str can probably make a bigger difference on some enemies when you are doing damage that way. It still takes over 30 levels before Kent will win str by 1 on average. There are still tons of enemy def values at which a difference of 1 str pails in comparison to a difference of 3 spd. Like I said, if they have the same #HKO, the str doesn't matter.

More importantly, there isn't actually any point of comparing them to all 887 503 681 enemies (ignoring hp). There are quite clearly not 887 million enemies in the game.

I would not have even bothered posting this unless I was confident in my numbers, because I know how nitpickey and objective the FE community is. Argue all you want, but unless you have some cold hard proof I'm sticking to my hard work.

I ask you again: What are you actually measuring? What is the point about comparing units to enemies that don't exist? What is the point of balancing for the link arena (as someone else pointed out). What about the game itself? Why wouldn't you try to make a way to translate units' bases and growths and turn it into a sort of "power level" for when they go through the game? I just don't see how there is any point in doing it this way.

Oh, and str =/= def anyway since you are combining str and magic. The str value needs to account for both physical and magical damage, whereas the def value only counters the str portion of your str column. It should be easy to see, by your stupid "1 str = 1 def in damage" idea, that it should be "1 str = 1 def + 1 res" or something similar. In other words, if you have str = 1.23 and res = 0.23, then def can't be any more than 1.00 since it appears that magic is supposed to account for 18.699% of the attacks. Unless you split str and mag (which makes no sense from GBA FE perspective) you can't have str = def otherwise res is getting its worth from nowhere. I love how you got even this much wrong.

[Panthera]

You seem to be ignoring that the argument that dondon is making against you is that your "overall power" idea is inherently flawed because it takes into account situations that are unlikely to happen, and also more importantly because it over-emphasizes direct combat, when the actual issue of balancing quality of characters against one another involves a lot more than just how each character fares in direct combat against one another (or against any hypothetical set of enemies) due to the fact that tactical flexibility comes into play. You say that two units will have the same outcome regardless of whether they're human or AI controlled; while this is true, it ignores the fact that there are more than just those two units. I've beaten some enemies in FE who are hugely superior statistically by your calculations (or by my own approximations; I sure wouldn't object to some of those fully capped final bosses joining me!) because I have access to all kinds of stuff that isn't just "A vs B". You're building a theoretical basis for one on one matches (although you're having some flaws pointed out in that area too), but a rather important point is that 1v1 combat is extremely pointless in Fire Emblem, a series where 1v5 (one powerful player character taking on armies of enemies at once) is way more common, and 5v1 (when players need healing, support characters and multiple attackers to handle some bosses efficiently) is probably more common than even numbers. Direct 1v1 comparisons just don't really matter much in Fire Emblem.

[Anouleth]

OR because you just don't get what I'm doing it. Test it yourself. Take two characters and pit them in battle against each other. Plug their stats into the damage formula, and then start messing with stats to see which one effects damage the most. I guarantee you 1 point of strength or defensive increases/decreases damage by more than any other stat in virtually every situation. Go ahead and do it, then tell me I'm wrong. The reason I'm refusing to yield is because I did hundreds upon hundreds of calculations, maybe even thousands. EVERY SINGLE POSSIBLE battle scenario, and these are the figures I got. There is absolutely NO situation that is going to come up in game that I didn't include in my averages. Do you honestly think I'm that stupid? I've been playing Fire Emblem for like 15 years. I know how it works.

Well, congratulations, but it doesn't mean squat. Take a gander at the FE9 tier list. Guess who's not at the top? Tauroneo, Gatrie, Brom and Haar who have the best STR/DEF in the game. Guess who's above them all? Mia, Zihark and Stefan, who have relatively awful STR/DEF. Even though they deal more damage, it's generally overkill since we have a pile of resources that can fix Mia/Zihark/Stefan's strength issues but nothing can fix the General's flaws.

And your criticisms fail to be relevant. Game mechanics aren't different for human and AI players. Two units are going to have the same result in battle whether they are both human, AI, or one of each.

Sure, except that human characters are going to be able to do a fuckton of things outside of battle that affects what happens in battle, like whip out a horseslayer and OHKO, or sit next to a support partner, or position themselves on a forest, or etc etc etc. Unless you are playing FE9 and switch the AI to Roam and just watch them bump into each other, there is always a marked difference between the player characters and the enemies.

Since when does any of that happen IN BATTLE? My numbers are used to find a character's battle capacity.

But it still has an effect on what happens in battle.

Again, not relevant in battle. The player can't manipulate anything once battle has been initiated.

How is equipping a Horseslayer and rendering Strength redundant irrelevant to what happens in battle?

"If." "If" means nothing. In the average situation (yes average, meaning that it includes every possible scenario) Strength and Defense are equal. I did the math. All of it. Every possible character matchup. Do you understand what that means. Every possible relevant stat combination against every other combination. I have an Excel sheet for each stat, and my tables are massive.

No, if your character has poor hit it effects strength...equally to how your avoid effects the importance of defense. That's why Strength and Defense don't have weighted values of 100%. They only do when you (or your enemy) have 100% hit, but that's certainly not an average is it?

I know a guy on this board who eats tables for breakfast.

The player has no choices in battle. The choice is made before battle. Do you understand that my numbers do not mean that one character is absolutely better than another 100% of the time? You could argue on the opposite point that since an archer can't counter melee units, that offensive is greatly diminished in that situation. You can't just pick and choose what arguments you want to make and leave out the ones that don't work in your favor, because I'm not an idiot.

Well, if your numbers aren't tell us whether characters are better than each other, then what the hell was the point of calculating them?

Ok...but what about an enemy with 60 HP? 45? 16? Not all enemies have 30 HP, so using one single example is pointless. I used the average of all of them. And your last bit is so far off the wagon that I don't even know how to respond. My numbers are about any unit vs. any other unit, not all players against all enemies.

There will always come a point where additional damage is worthless. And if the enemy has 60 or 45 HP, it's still the same since it's a 2RKO. Extra damage is not always productive when it's overkill.

Their bases are weighted equal too. What about an enemy that Sain can't double? Or one with high defense? Or one that Kent can double too? On top of that, those are their growths for MY game, you don't know what my enemies are like. Their overall combat proficiency is 50/50. I tested it. My chart looks like this:

Stats are in the same order as in-game, and HP is a separate calculation so it's not included.

0 0 0 0 0 0

1 0 0 0 0 0

2 0 0 0 0 0

. . . . . . etc...

30 0 0 0 0 0

0 1 0 0 0 0

1 1 0 0 0 0

2 1 0 0 0 0

ect... All the way to

30 30 30 30 30 30

Trust me, EVERY possible enemy was accounted for, and they perform 50/50. It may seem like a ton of work, but excel formulas make it nothing more than click and drag and built-in formulas.

And like I said before, unless your game actually contains such a range of enemies, uniformly spread across every possible permutation of stats, who cares? Unless the 'O stats in every area' and '30 stats in every area' guy are actually in the game as enemies, then does it matter how Sain and Kent perform against them?

I would not have even bothered posting this unless I was confident in my numbers, because I know how nitpickey and objective the FE community is. Argue all you want, but unless you have some cold hard proof I'm sticking to my hard work.

Somehow, I don't think people are taking issue with your numbers...

[Janissary]

This formula ignores the fact that Magic is much more desirable than Strength and is therefore flawed.

[uzy5o]

Can you post your Excel file? I would like to see what you actually did, and I believe others would like to as well.

[SuperTroll Maxim Lapierre]

It would have been better if you simply left your calculations as something that models or approximates combat ability rather than some end-all be-all to determining combat ability. What you've created is the FE version of the Body Mass Index.

[Charpig]

This formula ignores the fact that Magic is much more desirable than Strength and is therefore flawed.

How is Magic more desirable than Strength? Doesn't it depend on whether you use weapons or magic? And in the GBA games characters had either Strength OR Magic, not both.

Edited August 12, 2010 by Charpig

[Parrhesia]

How is Magic more desirable than Strength? Doesn't it depend on whether you use weapons or magic? And in the GBA games characters had either Strength OR Magic, not both.

Yeah, but by the same token, characters have both Def and Res, and usually their Res is rubbish.

[Charpig]

Yeah, but by the same token, characters have both Def and Res, and usually their Res is rubbish.

I see what you're trying to say here. But what if the opponent has higher Res than Def (which is often the case with magic users)? Then you would want to target Def, so Str is better in that case. I don't think that Magic is always much more desirable than Strength. If I have a physical fighter who cannot use magic weapons, I wouldn't want Magic for him, he can't do anything with it. No, I would want him to have as much Strength as possible. And most of the time enemy units don't have good Def either unless they are Armor Knights/Generals. The topic creator is working on making the hack for Blazing Sword, in which case a unit has either Str OR Mag. So which one is better depends on the unit. On average, do you have more physical fighters or magic users? I would say physical fighters. I think Alondite was essentially considering Str and Mag the same, because for magic users in Blazing Sword, Magic IS Strength. You just apply Res to the damage done rather than Def.

[Parrhesia]

I see what you're trying to say here. But what if the opponent has higher Res than Def (which is often the case with magic users)? Then you would want to target Def, so Str is better in that case.

Well, yeah, but the majority of enemies will be physical units. And the difference between killing a Mage with magic and an armour knight with a weapon is that magi have far fewer HP. On average, you'll mostly be facing dudes with pathetic Res and at least mediocre Def.

I assume Janissary was saying that in the case of a mage with 8 magic and a mercenary with 8 strength the mage will do more damage in most cases (if neither/both double, but that's a different stat entirely)

Edited August 12, 2010 by Agent Dale Cooper

[Alondite]

And here I was trying to be nice.

Your idea is worthless as far as the real fe7 is concerned. It in no way measures the relative powers of the characters involved. You couldn't even recognize that Sain wins handily with your new growths. Even if you make two characters start with the same bases, the Sain growths still win. After 20 levels, that's a 2 point speed win and in return only a 0.6 str loss. hp and def aren't nearly as important since neither character has survival issues. There is more than enough variance in enemy AS for Sain to double a significant number of enemies that Kent does not in each chapter. Outside of that, the small difference in strength will very frequently not even be 1. Even when it is, 1 damage is rarely enough to change the #HKO values of the units involved. A couple of instances in which Kent 2HKOs and Sain 3HKOs when neither/both double is not nearly enough to overcome the many times in which Sain would double and Kent would not. At least, when the difference in strength is never more than 1.

Prove it. Prove that Sain averages more net damage in any combat situation.

Character B had 0 improvement from the +9 strength I just gave him. Your calculations give this a rather large importance despite the fact it does nothing. Character A went from 2RKOing to ORKOing. A rather large improvement. Your calculations see this as a much much smaller improvement than what Character B got. This is a pretty common occurrence over the course of playing fe7. You will have these situations pop up lots of times. Probably hundreds, where one character gets a large improvement from +1 or +2 str and another gets no relevant improvement from +8 or +9 str. And yet your set-up does not acknowledge this scenario's existence. And you don't even see how that's a problem.

Bullshit my setup doesn't acknowledge this. If it didn't then strength would have a 100% weighted value (where +1 strength always means +1 damage). When +1 speed doesn't allow you to double, what is it's value? Nothing offensively. Defensively? +2 avoid, which amounts to a 2% damage reduction. +1 Strength adds 1*hit% damage. The scenarios where +1 strength is more valuable than +1 speed is better than 2:1. You keep picking isolated situations where speed wins out, but that's not the only situation you're going to run into. Logical fallacy used to hide the weaknesses of your argument.

Let's go back to one of your earlier "counters" that I left alone. The bold was you, the rest was me.

Who fares better is 100% dependent on the weapon used. Let's make both characters 3HKO. 16 mt weapon allows this. As does 17 and 18 and 19 mt. Now who does better against your random enemy? Mine. Why? Mine goes first. Yours goes second. Mine only gets hit 2 times, yours gets hit 3 times and dies. If you let the enemy go first, mine is more likely to dodge and not die. What if either yours or mine misses the enemy if we both go first? Mine is more likely to the extra attack that the enemy gets.

Weapons are independent of character stats, and don't change the ratio of of how each stat attributes to damage. I was doing my calculations with weapons until I realized that it wasn't effecting my relative results.On top of that, show me a weapon with 16 mt and I'll show you 30 that don't 16 mt is by no means an average weapon. See how far you have to stretch to prove me wrong? That actually weakens your argument too. Take a logic course.

At some mt values you'll win more often. Like 11 mt. You are able to 4HKO, the enemy is able to 4HKO, but mine 5HKOs.

11 my is far closer to average than 16. Thanks for lending credence to my argument.

Also, don't forget that there are actually a large number of enemy speed values that give me a massive advantage. You only really get a minor advantage on the rest because you only win when your #HKO value is better than mine and the enemy shares your HKO value rather than mine.

With anything from 9 to 12 speed, I double and you don't. With anything from 16 to 19 speed, you get doubled and I don't. As for the rest, we are both pretty much screwed at 20+. Chances are, the enemy will slaughter us. For some weapon mt values, you'll do better than me since you may only need to get lucky and dodge once (and you need to hit the enemy each time) whereas I may need to dodge twice (and hit the enemy each time). With other mt values, I have the advantage since if we both only need to dodge once I am more likely to do so. With anything that has 8 or less speed, we both pretty much obliterate anyway so who cares? Unless you want to give us weapons that only have 3 or less mt, we are both pretty much going to win easily. That means I have a significant advantage for 8 speed values (9 to 12 and 16 to 19), you have a minor advantage at best (after averaging over the various weapon mt possibilities) for 3 speed values (13 to 15), and you have a minor advantage at best for the other 11 (20 to 30), though we are both pretty much screwed in that instance. I'd say when you consider the various speed values of the enemy (31 possibilities if you include 0) while keeping the rest of the stats constant, I win. Also, don't forget weapon wt. My unit can switch weapons to sacrifice a little speed if the enemy has 13 to 15 speed. If you sacrifice a little speed, you may end up getting doubled.

You win 8 scenarios, I win (or get beat less badly) in the other 22. You're throwing out stuff like "more likely" a lot. How about you throw out a variety of combat scenarios and show me some hard proof.

And let's not forget how other enemies exist that don't have 30 def. As you decrease enemy def, our characters are more and more likely to KO the enemy in the same number of rounds. Which tends to either make us equivalent (since neither has any chance of death and will generally KO in the same number of rounds) or make mine win since mine may actually dodge something. Particularly since our units may miss the enemy allowing it to have enough swipes that it has the potential to kill one of us. If we 2HKO and it 3HKOs, well, we may only need to miss once for a COD > 0%. Mine will be less.

The fundamental flaw in your argument is you are pointing out the very narrow range where speed is more valuable than strength. At the +4 to like +6 or +7 range. Less than a +4 speed spread has virtually no value. 6 avoid at most. Yay. 1 point of defense is more valuable than that, unless your enemy's hit is below 6%. Of you have a +0 speed advantage, than +3 strength is was more valuable overall (again, on average, not against one particular enemy as you keep pointing out). And then if you already have a +4 speed advantage, than what does more speed do? +2 avoid per point is all. Again..yay. More strength is only useless in 2 conditions: when you have 0 hit, or when it doesn't allow you to kill your enemy faster. It does, however, do MORE damage in the second case, and makes it easier for another character to finish off said enemy.

Character A

str-10 ski-8 spd-5 luck-7 defense-9 resistance-6

Character B (same stat order)

8 8 10 6 7 6

Ok, since you insist on using just enough speed gap to give speed an advantage, I'll use the average value of 5 points of speed instead of 30. If I do that The weighted values in the same order as your stats (minus resistance because we aren't using it) become:

50.00% 29.00% 100.00% 5.00% 50.00%

Which makes each characters weighted power:

A: 17.17

B: 20.12

Now of COURSE B is going to win, he has a higher weighted value. The rest of your argument falls apart at that fact. I used the average of all possible gaps from 0-30 for each stat and averaged them to find the absolute value of a stat. Apparently that's not "realistic." My goal here wasn't to give be-all end-all values, it was to hopefully contribute something useful to the community. My numbers are correct for the average gap of 0,1,2,3...all the way to 30, but if I should use a different gap then I will redo it all using that gap. I just don't know what the average stat gap you're likely to encounter is in-game, so I just used the absolute average.

So tell me, what then should I do? Note that weapons do not affect the percentage of which one stat increases damage. Whether you do 1 damage or 30, a double attack still adds 100% damage. I will use a universal hit% though. 50% is the average of 0=100%, all of which you're likely to see in game. Just give me the word and I'll do the math so it's right by you guys. Just know that my values are not wrong for an absolute average.

Edited August 12, 2010 by Alondite

[Narga_Rocks]

Prove it. Prove that Sain averages more net damage in any combat situation.

I would if I had enemy stats on this site. I could grab statistics from games like fe9 if you are too dense to see how 3 speed is a better win than maybe having 1 strength.

Bullshit my setup doesn't acknowledge this. If it didn't then strength would have a 100% weighted value (where +1 strength always means +1 damage). When +1 speed doesn't allow you to double, what is it's value? Nothing offensively. Defensively? +2 avoid, which amounts to a 2% damage reduction. +1 Strength adds 1*hit% damage. The scenarios where +1 strength is more valuable than +1 speed is better than 2:1. You keep picking isolated situations where speed wins out, but that's not the only situation you're going to run into. Logical fallacy used to hide the weaknesses of your argument.

You are considering an improvement of 1 strength to make an improvement of 1 * hit%. How is that possibly accounting for when extra strength doesn't help at all? You are focusing on damage, moron. If it doesn't change the #HKO value, it is worthless. Even moreso than when speed doesn't change not doubling to doubling or getting doubled to not getting doubled. At least speed still gives +2 to your evade. It's more than strength gives.

Let's look at an enemy with 30 hp and 10 def. This is the #HKO values for the various mt values you can have:

1: 40+

2: 25 to 39

3: 20 to 24

4: 18 and 19

5: 16 and 17

6: 15

7: doesn't happen

8: 14

9: doesn't happen

10: 13

etc.

So, for a large number of strength increases, strength changes nothing about the combat whatsoever. How are you accounting for that? And before you give your retarded "that's only one situation", this type of thing happens with any hp and def value that you can come up with. ANY. This happens everywhere for all of your calculations, yet based on what you have told us, you will still give full credit to a +1 strength that changes mt from 22 to 23 when it does NOTHING. Literally nothing. Not even giving a slightly higher chance to dodge. And you still don't understand what we are getting at. That makes you an idiot.

Weapons are independent of character stats, and don't change the ratio of of how each stat attributes to damage. I was doing my calculations with weapons until I realized that it wasn't effecting my relative results.On top of that, show me a weapon with 16 mt and I'll show you 30 that don't 16 mt is by no means an average weapon. See how far you have to stretch to prove me wrong? That actually weakens your argument too. Take a logic course.

11 my is far closer to average than 16. Thanks for lending credence to my argument.

You are using stats in the 30s. Why the hell would you be using weaker weapons when you are at the end of the game? And which enemies even have 30 defence? If you are going after an enemy with 30 defence while using weak weapons, you are the dumbest player ever. If you encounter enemies that are that strong, use the S rank weapons. Problem solved.

Oh, and I forgot you were so stupid otherwise I would have listed more of the mt values that give my unit an advantage. Sorry about giving you enough credit to think that you could figure it out for yourself. I'll work harder to remember that you are too incompetent to figure out anything not spoonfed.

40 hp, 28 str, 14 skl, 16 spd, 15 luck, 30 def, 15 res.

40 hp, 30 str, 14 skl, 12 spd, 15 luck, 30 def, 15 res.

40 hp, 29 str, 14 ski, 14 spd, 15 luck, 30 def, 15 res.

Take a look at 12 mt weapons. I 4HKO. You 4HKO. Which one of us gets hit more? How about 13 mt weapons? Oh look, we still both 4HKO.

And then there is 10 mt. Sure, you 4HKO and I don't. But the enemy 5HKOs as well which means I can win anyway. And since we should allow the fastest unit to go first in the first round (that's what arenas do, right? You are basically making your stats to go off of arena setups) then I get to whack 5 times before it goes 5 times. As for you? You whack 4 times before it goes 5 times. We are basically even since we are both in trouble if we don't dodge. Except wait...I have more avo.

You win 11 mt outright. I win 10, 12, and 13. I win 3 out of 4. You win 1 out of 4. Go you. But hey, I mean, you win 14 and 15 mt, right? Wait, no. The same thing happens at 14 as it did at 10. You may 3HKO, but the enemy 4HKOs just like me. I win (If you set up MEMEMEME instead of MEEMMEEM. The second one lets the enemy get 4 before I get 4 so that could be troublesome. With 10 mt this isn't a big problem though since either set up lets me get 5 first).

You win 11 and 15. I win 10, 12, 13, 14, 16, 17, 18, 19. And you think yours is better? Okay, you could argue we kinda tie-game for 10 and 14, but that's about it.

And if you want to use anything with less than 10 mt then you are being ridiculous. Who the hell picks up an 8 mt weapon to go up against something like that?

You win 8 scenarios, I win (or get beat less badly) in the other 22. You're throwing out stuff like "more likely" a lot. How about you throw out a variety of combat scenarios and show me some hard proof.

Um, I godstomp you in 8 scenarios, and in 9 scenarios (0 to 8) we tie game since we both godstomp the enemy. In the scenario where the enemy has 13, 14, or 15 speed, see above. I win with reasonable mt values on our weapons. And what are you complaining about "more likely"? If we both need to dodge x number of attacks, and I have more avo, I am clearly "more likely" to win out. What more do you want from me? Do you need me to run the numbers to tell you precisely how much more likely I am to win? I should think that is unnecessary. A win is a win.

Now, for the rest of the scenarios (20 to 30 speed) we are both screwed. Do you want me to run those numbers, too? I'll give you one beneficial to you weapon. You can have anything from 10, 11, 14, 15. I don't care which (Unlike when neither of us double and 10 and 14 were arguably in my favour, when we are both doubled 10 and 14 are probably even better for you than 11 and 15 since 10 and 14 give you a #HKO advantage over the enemy). Then pick one that is good for me. 12, 13, 16, 17, 18, 19. I don't really care. Pick two hit values for the weapons. Low and high, so around 55 to 65 and then 75 to 90. We can then take a look and see if either of us can even get past 20% to win considering how we are both doubled. Pick a speed for the enemy, as well, since that affects how often we actually hit. You can pick two if you want.

Now, considering from 9 to 19, the more common speed values, I win 8 out of 11 and the other 3 are arguable for either of us, I think that makes mine look better. Consider further how very screwed we both are for 20 speed enemies and above. Does it really matter if you have a 10% chance of winning and I have a 4% chance of winning, or whatever? We both suck, so why do those speed values mean anything for you? You fail less? Oh boy. And 8 and under we will annihilate anyway.

Overall, how can you think your character does better in this set up?

The fundamental flaw in your argument is you are pointing out the very narrow range where speed is more valuable than strength. At the +4 to like +6 or +7 range. Less than a +4 speed spread has virtually no value. 6 avoid at most. Yay. 1 point of defense is more valuable than that, unless your enemy's hit is below 6%. Of you have a +0 speed advantage, than +3 strength is was more valuable overall (again, on average, not against one particular enemy as you keep pointing out). And then if you already have a +4 speed advantage, than what does more speed do? +2 avoid per point is all. Again..yay. More strength is only useless in 2 conditions: when you have 0 hit, or when it doesn't allow you to kill your enemy faster. It does, however, do MORE damage in the second case, and makes it easier for another character to finish off said enemy.

Oh, so now you care about the rest of the team. Yes, doing more damage means certain other weaker friendlies are able to kill something instead. Joy? How about the fact that my 4 point speed advantage lets a unit like mine double more enemies?

I don't have stats for fe7, but let's look at fe9.

Enemy speed goes up faster in fe9 than fe7, from what I've heard about people complaining of slow enemies in fe7. I'll take chapter 24, then (out of 29 + P).

2x Fighter lv 20 (steel axe)

40 hp, 27 atk, 12 AS, 92 hit, 29 avo, 11 def, 6 res, 5 crit, 5 cev

1x Warrior lv 5 (steel axe)

44 hp, 28 atk, 12 AS, 91 hit, 28 avo, 11 def, 7 res, 5 crit, 4 cev

1x Warrior lv 6 (silver axe)

47 hp, 33 atk, 13 AS, 99 hit, 31 avo, 12 def, 7 res, 6 crit, 5 cev

5x Sword Knight lv 15-17 (steel sword)

31 hp, 20 atk, 12 AS, 96 hit, 29 avo, 13 def, 6 res, 4 crit, 5 cev

2x Sword Knight lv 19 (steel sword)

32 hp, 20 atk, 15 AS, 98 hit, 35 avo, 14 def, 8 res, 4 crit, 5 cev

2x Lance Knight lv 17 (steel lance)

31 hp, 22 atk, 11 AS, 90 hit, 26 avo, 13 def, 6 res, 4 crit, 4 cev

1x Lance Knight lv 19 (steel lance, javelin)

32 hp, 23 atk, 14 AS, 93 hit, 33 avo, 14 def, 6 res, 4 crit, 5 cev

3x Bow Knight lv 19-20 (steel bow)

33 hp, 22 atk, 14 AS, 91 hit, 33 avo, 14 def, 7 res, 4 crit, 5 cev

1x Paladin lv 7 (silver sword)

37 hp, 28 atk, 17 AS, 107 hit, 39 avo, 16 def, 11 res, 5 crit, 5 cev

2x Archer lv 17 (steel bow)

27 hp, 18 atk, 12 AS, 106 hit, 28 avo, 11 def, 7 res, 8 crit, 4 cev

1x Archer lv 18 (steel bow, Iron B)

28 hp, 19 atk, 11 AS, 109 hit, 27 avo, 11 def, 7 res, 8 crit, 5 cev

1x Archer lv 20 (steel bow, Iron B)

30 hp, 19 atk, 14 AS, 111 hit, 33 avo, 12 def, 7 res, 9 crit, 5 cev

1x Sniper lv 4 (longbow, Iron B)

31 hp, 18 atk, 14 AS, 106 hit, 33 avo, 12 def, 9 res, 24 crit, 5 cev

1x Sniper lv 5 (steel bow)

31 hp, 22 atk, 14 AS, 109 hit, 33 avo, 12 def, 9 res, 23 crit, 5 cev

1x Sage lv 7 (bolganoe, mend)

30 hp, 26 atk, 12 AS, 114 hit, 27 avo, 10 def, 17 res, 6 crit, 3 cev

1x Myrmidon lv 17 (steel sword)

28 hp, 19 atk, 16 AS, 110 hit, 37 avo, 8 def, 4 res, 7 crit, 5 cev

4x Myrmidon lv 20 (steel sword)

30 hp, 21 atk, 17 AS, 112 hit, 39 avo, 9 def, 4 res, 8 crit, 5 cev

1x Swordmaster lv 8 (iron blade)

35 hp, 23 atk, 18 AS, 114 hit, 42 avo, 11 def, 7 res, 24 crit, 6 cev

1x Swordmaster lv 8 (longsword, vulnerary)

33 hp, 21 atk (27 eff), 20 AS, 127 hit, 46 avo, 11 def, 7 res, 24 crit, 6 cev

2x Wyvern Rider lv 20 (steel lance, 1 javelin)

35 hp, 27 atk, 10 AS, 102 hit, 24 avo, 18 def, 7 res, 7 crit, 4 cev

1x Wyvern Rider lv 20 (killer lance)

36 hp, 27 atk, 10 AS, 103 hit, 25 avo, 19 def, 7 res, 37 crit, 5 cev

1x Wyvern Lord lv 5 (silver lance)

37 hp, 32 atk, 11 AS, 107 hit, 26 avo, 18 def, 8 res, 7 crit, 4 cev

5x Soldier lv 16 (steel lance)

32 hp, 20 atk, 8 AS, 97 hit, 19 avo, 10 def, 5 res, 6 crit, 3 cev

1x Halberdier lv 2 (steel lance, vulnerary)

35 hp, 22 atk, 12 AS, 96 hit, 26 avo, 12 def, 7 res, 6 crit, 2 cev

Rikard lv 15 (silver blade, short axe, elixir)

48 hp, 34 atk, 19 AS, 107 hit, 45 avo, 20 def, 16 res, 10 crit, 7 cev

Turn 3 Reinforcements:

2x Lance Knight lv 19-20 (steel lance)

32 hp, 23 atk, 14 AS, 93 hit, 33 avo, 14 def, 6 res, 4 crit, 5 cev

3x Wyvern Rider lv 19 (steel lance, 1 javelin)

35 hp, 25 atk, 11 AS, 101 hit, 27 avo, 18 def, 8 res, 6 crit, 5 cev

1x Wyvern Rider lv 19 (short spear)

34 hp, 26 atk, 10 AS, 99 hit, 25 avo, 19 def, 7 res, 6 crit, 5 cev

1x Wyvern Lord lv 4 (silver lance, short spear, vulnerary)

38 hp, 31 atk, 11 AS, 106 hit, 25 avo, 17 def, 7 res, 7 crit, 3 cev

1x Bandit lv 16 (steel axe)

40 hp, 25 atk, 11 AS, 82 hit, 25 avo, 8 def, 3 res, 3 crit, 3 cev

1x Bandit lv 18 (hand axe)

40 hp, 23 atk, 13 AS, 73 hit, 30 avo, 8 def, 3 res, 3 crit, 4 cev

1x Berserker lv 7 (killer axe)

46 hp, 31 atk, 16 AS, 90 hit, 37 avo, 13 def, 3 res, 50 crit, 5 cev

There's, what, 56 enemies or something? Mid 50s.

Let's count things that have either 9 to 12 or 16 to 19.

3 out of 4 warrior/fighters.

8 out of 14 mounted knights/paladins.

3 out of 6 archer/snipers.

1 out of 1 sage.

6 out of 7 myrm/swordmasters.

4 out of 4 wyverns.

1 out of 6 soldier/halbs.

1 out of 1 bosses.

6 out of 10 reinforcements.

Now, granted, with the stats that I gave you a while back (30 def), neither of us are going to die so we could perhaps ignore the 16 to 19 guys. But without epic weapons (some enemies even require fe4 holy weapons) you aren't OHKOing much anyway. And since you don't seem to approve of using 16+ mt weapons... I win on 33 out of ~53 enemies. There will be chapters in which 12 and 16 are similar, but even in those cases there will be a different spread of speed where it is relevant. The thing about these games is that generally your characters' speed grows with the enemies. There will always be situations like what I just showed you where there is a significant number of enemies against whom a 4 speed difference in your own units is extremely relevant. There's a reason why we tend to stick doublers so high in tier lists. If a unit 3HKOs and doubles, that is nearly always better than a unit that 2HKOs and doesn't double. Why? Killer weapons exist and they tend to result in more opportunities to proc crit and kill the thing. 50% is > 30%. Obviously. Even 4HKOing and doubling will be better than 2HKOing and not doubling since criticals give 3x damage. Differences of 2 or 3 str generally pale in comparison to differences of 3 or 4 spd, yet given your weighting system 3 str is more important than even a difference of 6 spd. I think even more than 7. And yet the faster unit will always have better offence as long as killer weapons exist. Or things like adept and masteries in fe9 and fe10. Even ignoring those weapons, you have supports to give crit or you are simply doing 70% damage instead of 60%. (or like 80% over 60%) Like you said above, it is easier to KO an enemy with less hp.

Ok, since you insist on using just enough speed gap to give speed an advantage,

What? I took your numbers and divided by 2. What the hell? You insisted on using a 7 mt weapon with high stat characters. I thought that was beyond stupid. I divided by 2 because a 7 mt weapon is much more appropriate when your stats are lower. I took your numbers. What's your problem?

"Why are you countering a statement about how the game works with a vs. example? What about the game? First off, you won't be stuck with a 7mt weapon for the entire game. You'll have one near the beginning, but at that point you won't have those stats. Divide everything by 2 if you want to use weaker weapons. That shouldn't change anything since you made it all linear. Unfortunately you gave me two odd numbers, so I'll round str up and spd down. "

I can see that you enjoy ignoring statements that are damaging to your case, but come on.

I'll use the average value of 5 points of speed instead of 30. If I do that The weighted values in the same order as your stats (minus resistance because we aren't using it) become:

50.00% 29.00% 100.00% 5.00% 50.00%

Which makes each characters weighted strength:

A: 17.17

B: 20.12

Now of COURSE B is going to win, he has a higher weighted value. The rest of your argument falls apart at that fact. I used the average of all possible gaps from 0-30 for each stat and averaged them to find the absolute value of a stat. Apparently that's not "realistic." My goal here wasn't to give be-all end-all values, it was to hopefully contribute something useful to the community. My numbers are correct for the average gap of 0,1,2,3...all the way to 30, but if I should use a different gap then I will redo it all using that gap. I just don't know what the average stat gap you're likely to encounter is in-game, so I just used the absolute average.

What? So you just go and change your weighting system when it is shown that yours sucks?

"character A has a weighted power value of 28.03, and B a weighted power of 25.79"

I divided every stat by 2. Your system is linear. You take your weighting and multiply it by the stat and then add them up. Dividing each stat by 2 divides the total weight by 2.

character A has a weighted power value of 14.015. B has 12.895.

Sadly, I couldn't do this precisely because of the odd numbers on character B. However, I highly doubt that turning 7.5 into 8 and 10.5 into 10 is going to make 12.895 increase beyond 14. In fact, I know it won't.

Why are you changing your weightings? And if you have a weighting that values speed more highly, I say go for it. I approve of any weighting system that has spd > str > def. I'm even somewhat okay with str > spd > def. Only somewhat, though.

All I did was take your numbers and take a look at what happens when you have stats closer to the early-game. Like, when it is actually appropriate to be using iron lances. If you have a problem with that, I suggest you rethink this. Seriously, stop using sucky weapons when characters have good stats. It just doesn't work.

So tell me, what then should I do? Note that weapons do not affect the percentage of which one stat increases damage. Whether you do 1 damage or 30, a double attack still adds 100% damage. I will use a universal hit% though. 50% is the average of 0=100%, all of which you're likely to see in game. Just give me the word and I'll do the math so it's right by you guys. Just know that my values are not wrong for an absolute average.

But here's the problem: weapons do matter. They actually matter one hell of a lot. You may notice that even when I took your initial values and didn't divide them, my character had an advantage for anything 9 mt and above. Yours only had an advantage for 8 and below. In fact, it was a stomping for anything above 12 mt weapons.

Whatever you think you know that lead you to the conclusion that weapons don't matter, it was very clearly wrong.

Consider, 29 str and 27 def. That's a gap of 2. When you give +1 str, that's a 50% improvement in damage. Against an enemy with x hp, instead of dying in x/2 hits, it dies in x/3 hits. That's a pretty significant difference. At 30 hp, that's a difference of 5 hits. At 60 hp, that's a difference of 10 hits. Huge difference by adding 1 strength.

But what if there is a weapon? Defence isn't improved in this game by very much. Just Armads. However, weapons give you better damage and def can't counter it. Take a 10 mt weapon. Now you are doing 12 damage. Enemy dies in x/12 hits (round up). Let's add 1 strength. 13 damage. x/13 hits (round up). Often, this doesn't even change the #HKO value at all. Like, say, 30 hp. It takes 3 hits from either.

Remember how +1 str at 30 hp turned 15 hits into 10 hits? That was a huge improvement. It did 50% more damage and caused a 33% reduction in number of rounds needed to KO.

You have excel, use it. Spread hp from 1 to 60. Calculate the reduction in number of rounds needed to KO for each hp value when you increase hp by 1. Average that out. Do that for starting values of 29 mt and for 39 mt. Compare the average % reduction in number of rounds needed to KO.

This happens for all your strength values everywhere. Instead of both str and def going from 0 to 30, str goes from 10 to 40 while def still goes from 0 to 30. You are too focused on absolute increase in damage. Absolute increase in damage doesn't actually matter all that much. If you are trying to set up situations where two characters with equal "overall values" with have a 50/50 chance to kill each other, weapons matter as a result of the above statements. Weapons are like a free +10 str without having to pay for it. You can already see what happened to these characters:

Character A

str-20 ski-16 spd-10 luck-14 defense-18 resistance-12

Character B (same stat order)

15 16 21 12 14 12

when given +7 to the weapon.

Let's go back to your numbers there. 14 mt weapon, now. You do 20 damage per hit. Go you. I do 11 per hit and double. Not quite the same as before, but I still outdamage you by 2 each round. You do a bit better in the comparison over the various hp values, but I'm guessing I still win. And when you start accounting for misses, I do even better. We are fighting each other, and since I have a 20 avo advantage you are more likely to miss than I am. I have to take two hits, so on the one hand I may have less chance to do full damage, but at the same time I have more chance to do some damage. 14 mt is a silver lance (one example of 14 mt, anyway). 75 hit.

I have 38 base hit. You have 39 base hit. I have 54 avo. You have 34 avo. I get +4, you get -15.

79

60.

Let's ignore true hit since you said you never accounted for that. Of course, the game uses true hit which might change your average damage with those pathetic iron lances a little, but who cares, right? (It's possible you did use true hit for that, though I don't feel like recalculating your calcs to find out)

My chance of hitting twice? 62.41%. Gee, this is going to be sad for you, I can already tell. My chance of not hitting? 4.41%. I'm left with 33.18% chance of hitting once.

I do 11 per hit, so: 17.38 damage per round.

Your chance is 60% and you don't double and you do 20 damage. 20 * .6 = 12.

I outdamage you by 5.38 per round. That's pretty impressive, eh? I think I could even win with 10 mt weapons. You do 16 when you hit, I do only 7 per, but that's still 14 total. I don't know if there are any 75 hit 10 mt weapons without crit rates, but I don't care at the moment. I hope you don't, either.

7 * 2 * .6241 + 7 * .3318 = 11.06 (you can simplify the first part to 1.58 * 7)

16 * .6 = 9.6

Wow. I still win. You have to look at pretty weak weapons before you get to start winning, dude. Which unit in the game is expected to be mainly using iron weapons when they have stats like 20 str, 18 def, 21 spd? (I mean, sure, my 20/2 Dieck in fe6 will sometimes run around ORKOing with iron swords, but only on the weak pirates in 12x and places like that. He still uses silver and killers for the ones that aren't pathetic)

If you like, I'll calculate what happens with 9 mt, 75 hit, 30 crit killer weapons. (wow, in a game where nearly every weapon got +10 hit, killers got -5 or +0 hit. Granted, they were pretty psychotic in fe6) You know, it's a good thing that this also has 75 hit, otherwise I wouldn't do this next part.

6 for me, 15 for you. Who do you think will win when we account for crit rates?

Without crit, 1.58 * 6 = 9.48 and .6 * 16 = 9. I win (barely) without accounting for crit. What do you think happens with crit?

You used those two characters earlier to "show" me that your strength oriented character would win more often. You gave them 7 mt weapons and, yes, yours did more per round. 8.45 over 6.27. However, all I had to do was increase the weapon damage just a little bit and I blew you away. Just +7, creating a 14 mt weapon (like, say, silver lance, which are really easy to get in fe7 starting in 27E/29H). The victory shifted to my side with 17.38 damage over 12. An even greater victory than you had earlier. Even by %. 44.83% compared to ~35%. Weapons quite clearly change things tremendously. I even won with a 10 mt weapon. 11.06 over 9.6. Quite convincing win, there, actually. All it took was +3 mt on top of the weapon you initially gave, and I won. Now, yes, you could simply add 3 str to each unit and then the same thing happens. However that wouldn't actually make a difference for our relative "Power". The difference between our powers would still be 2.24. You'd still have more "power", but you'd obviously lose.

As near as I can tell, your values apply a lot stronger when str is close to def and damage is lower. And that's to be expected since you apparently compared every possible str to every possible def. I shouldn't be surprised that this is what you'd end up with. However, weapons exist and enemy units normally have less def than units have str, and even when they don't (like armors) you can still just pull out a silver weapon or even a killer weapon.

Oh, and one other thing that I want you to look at:

Oh, and str =/= def anyway since you are combining str and magic. The str value needs to account for both physical and magical damage, whereas the def value only counters the str portion of your str column. It should be easy to see, by your stupid "1 str = 1 def in damage" idea, that it should be "1 str = 1 def + 1 res" or something similar. In other words, if you have str = 1.23 and res = 0.23, then def can't be any more than 1.00 since it appears that magic is supposed to account for 18.699% of the attacks. Unless you split str and mag (which makes no sense from GBA FE perspective) you can't have str = def otherwise res is getting its worth from nowhere. I love how you got even this much wrong.

[Alondite]

What? So you just go and change your weighting system when it is shown that yours sucks?

No. You consistently picked out specific situations where speed is at the advantage, so I adjusted my values to reflect that EXACT situation. My original values were designed to account for the "average" situation, not to be true for every possible situation, because that's not feasible.

"character A has a weighted power value of 28.03, and B a weighted power of 25.79"

I divided every stat by 2. Your system is linear. You take your weighting and multiply it by the stat and then add them up. Dividing each stat by 2 divides the total weight by 2.

Right, but since you were using 1 singular situation, I adjusted my values to accommodate that one singular situation. Again, my original values are supposed to represent the "average" situation, which as an absolute amounts to 15 in each stat (30 HP). My values then measure the average value of one point in a stat in that "average" situation. It's a STARTING POINT as I said. I posted it here to get some constructive criticism, not "it's totally wrong and unrealistic and has no value because there's no way to calculate the average value of a stat because it is different in every situation." The numbers I got are correct for what I did, and without knowing what the "average" character actually is, it was the best solution I had: Just include them all, even if you'll never see them.

Why are you changing your weightings? And if you have a weighting that values speed more highly, I say go for it. I approve of any weighting system that has spd > str > def. I'm even somewhat okay with str > spd > def. Only somewhat, though.

Ok. I'm just going to make statement here. If you consider every possible character/enemy that could ever possibly be in the game, even if you will never see them, but if they are at least possible, even only through hacking, then my values are correct (because that's what I did). Now since I included unrealistic situations, my values won't be true for some in-game situations, but they WILL be true in others. Unless my math is wrong (and since I used excel, I'm sure it's not) this is the case. If you know what the average stat spread of your characters and your opponents are, let me know and I'll use those numbers instead. My goal here is to make something of value.

Want to give me values I should use? I'll use game average weapon mt, overall hit%, and stat spreads and recalculate. A few things I can assure you will not change: Strength and Defense are exact polar opposites, so no matter what I do they will come out as equal, and Resistance will not be as valued as Defense because there are fewer magic users.

I have the tables and the formulas already made, all I need are the numbers.