Unit Overall Rating

[Earthworm]

I have not yet completed my first game, but am really enjoying it. In trying to determine what units should compose my army, I have been looking at various threads related to Unit power. Such as Peasants vs Dragons, etc. None of them have satisfied my questions though, as there are quite a few good points.

A stack of 500 Peasants has a higher damage than a comparable Single Black Dragon, so on the surface the Peasants look to be the better deal, but when you factor in Attack and Defense Modifiers, the Dragon could do upwards of 300% damage to the Peasants.

So I decided to use the Unit Listing located at Gamefaq (So any wrong data in the unit info I got from there, which I think my stats for the standard Snake are wrong). And build a weighted Unit rating. I made it weighted because in my opinion Damage per Leadership is more important than Speed etc.

So this is what I based the ratings on. (#) = Weight

Damage/Ldrs = Average Damage ((LowDmg + HghDmg)/2) divided by Leadership cost per unit. i.e. a Peasant does .30 Average Damage per Point of Leadership, a Black Dragon does .05 Damage Per Leadership.

Hit Points/Ldrs = The Number of Hit points per unit divided by the Leadership cost per unit. i.e. Peasants have a 1 to 1 as they are 5 Leadership for 5 Hitpoints. Black Dragons are .32 to 1 as they only have 800 Hitpoints for 2500 Leadership cost

Attack = Straight 0 to 10 rating where Black Dragons = 10 as they have the highest Attack rating of 56, and Peasants are a 0 since they have an attack of 1.

Defense = Straight 0 to 10 rating where Black Dragons = 10 as they have the highest Defense rating of 56, and Peasants are a 0 since they have an attack of 1.

Speed = Straight 0 to 10 rating where Archedemons = 10 as they have the highest speed of 9, and Ancient Ents are a 0 since they have a speed of 1.

Initiative = Straight 0 to 10 rating where Archedemons = 10 as they have the highest Initiative of 7, and Ancient Ents are a 0 since they have an Initiative of 1.

Each of these Categories are rated from 0 to 10 based on comparison to all the other units, then Multiplied by the weight.

The Total is added up, and an overall rating is scored in comparison to the rest.

Here are My Results.

Non-Archer
Race Dmg(5) Atk(4) Dfs(2) Spd(1) Htpt(3) Init(2) Overall
Archdemon 1 9 9 10 1 10 10.00
Black Dragon 1 10 10 9 0 10 9.68
Red Dragon 1 9 9 8 1 8 8.97
Sprite 8 1 0 4 7 7 8.26
Emld Grn Dragon 1 8 9 6 1 10 8.23
Peasant 10 0 0 1 7 3 7.87
Thorn Warrior 8 1 1 3 7 5 7.58
Veteran Orc 4 4 4 3 5 8 7.49
Horseman 3 5 4 5 4 8 7.45
Black Unicorn 3 4 4 4 7 7 7.42
Bone Dragon 1 8 8 8 2 8 7.31
Lake Fairy 7 1 1 5 6 8 7.13
Demon 3 6 6 3 5 5 7.13
Ancient Ent 3 7 9 0 6 0 7.07
Griffin 2 3 3 5 9 7 6.98
Royal Snake 4 3 3 3 6 8 6.95
Giant 1 8 9 0 3 7 6.87
Furious Goblin 4 3 3 3 7 8 6.82
Orc 4 3 3 1 8 5 6.76
Cerberus 3 3 3 4 7 7 6.65
Berserker 4 4 1 3 6 8 6.61
Fire Spider 4 2 2 3 6 8 6.59
Unicorn 3 4 4 5 6 5 6.38
Dwarf 4 4 3 1 7 5 6.24
Black Knight 2 4 4 1 8 3 6.19
Sea Dog 4 3 2 1 6 7 6.16
Lake Dragonfly 7 1 0 4 4 8 6.06
Shaman 2 4 6 3 5 7 6.05
Snake 4 3 1 1 7 7 6.02
Necromancer 1 5 5 1 4 10 6.01
Ogre 1 7 7 1 3 3 5.96
Undead Spider 6 1 0 3 7 5 5.80
Marauder 4 2 1 1 7 7 5.79
Ent 3 5 6 1 5 2 5.69
Dryad 4 1 2 3 10 5 5.66
Polar Bear 3 4 5 1 5 5 5.65
Guardsman 4 3 3 1 6 5 5.64
Devilfish 5 1 1 3 6 8 5.58
Cave Spider 7 1 1 3 7 2 5.49
Knight 2 5 5 1 6 3 5.48
Venomous Spider 7 1 0 3 6 5 5.47
Skeleton 7 1 0 1 7 3 5.42
Miner 5 1 1 1 7 3 5.35
Fire Dragonfly 7 1 0 3 4 7 5.32
Werewolf Elf 4 3 3 1 5 7 5.26
Ancient Bear 4 3 4 1 6 3 5.23
Wolf 4 2 1 3 7 3 5.03
Ancient Vampire 2 4 4 1 3 8 4.97
Swamp Snake 4 2 1 1 6 5 4.83
Scoffer Imp 2 3 3 3 5 8 4.75
Demoness 2 5 4 1 2 8 4.74
Vampire 3 4 4 1 3 7 4.68
Anct Vampr(Bat) 1 4 4 5 2 10 4.58
Werewolf 3 3 2 4 5 7 4.57
Hyena 5 1 1 3 4 5 4.52
Swordsman 4 2 3 3 6 3 4.41
Pirate 5 1 1 3 5 5 4.37
Barbarian 4 2 1 3 6 3 4.07
Imp 3 3 2 4 3 7 4.03
Zombie 4 2 2 1 7 2 3.97
Robber 4 2 1 1 5 5 3.95
Vampire(Bat) 2 4 3 4 2 8 3.75
Bear 3 3 3 1 6 2 3.62
Decaying Zombie 4 2 2 1 5 2 3.16
Cursed Ghost 1 4 3 4 2 8 3.07
Cyclops 0 7 8 1 1 0 2.88
Ghost 2 3 2 4 2 5 2.14
Inquisitor 1 3 3 1 2 7 1.76
Archmage 0 4 4 1 1 7 1.53
Priest 1 2 2 1 2 5 0.29
Druid 1 3 4 1 1 2 0.00

Archer
Race Dmg(4) Atk(3) Dfs(1) Spd(1) Htpt(1) Init(3) Overall
Hunter 2 9 5 5 3 10 10.00
Cannoner 0 10 6 5 3 10 9.53
Alchemist 2 8 10 5 4 8 9.42
Evil Beholder 2 7 8 10 2 8 8.35
Skeleton Archer 9 0 0 5 0 5 3.84
Elf 1 7 4 5 2 8 4.00
Catapult 1 7 4 5 3 5 2.76
Goblin 3 5 3 5 1 5 1.33
Beholder 2 6 7 10 2 5 4.08
Thorn-Hunter 10 0 0 10 0 0 1.47
Royal Thorn 2 10 9 0 10 0 3.25
Bowman 2 5 3 5 1 5 0.00


Keep in mind that this does not compare specials, like The Archer Ability, so really only compare like type units. It also does not take into consideration Damage Type, or Resistances, so use it as a loose guideline. ;)

I hope it helps.

[Earthworm]

My current stats on the Snake are as Follows, if they are incorrect can you let me know?

Leadership 5
Damage 1 -2
Attack Rating 14
Defense Rating 8
Speed 2
Hitpoints 8

It seems odd that the Snake would score so high even in comparison to the peasants.

[Ryastar]

Well, for one, your stats on the snakes are wrong. The correct info is:
leadership 30
Dmg 3-6
hp 28

Seeing as that information is wrong, that puts all the other information you have into doubt. I'm not saying that this is your fault, but you really should have done the work to double check all the facts. It took me less than five minutes to check the stats on the snakes and about half of that was loading time - either of the game or of a save. The rest was simply in finding some of them, in this case in a battle.

Apart from the doubts about the accuracy of your information, your process itself has several serious flaws that detract from its usability as a guideline.

1) Your spectrum system that gives you such nice numbers from 0 to 10 for each stat is seriously flawed, so flawed that it is one of the biggest problems with the whole idea. Just considering attack, if 1 attack gives a rating of zero, and 56 a rating of ten, and seeing that you only use whole numbers, this means that each rating MUST have about 4-6 different attack values, since there are only 11 ratings to account for 56 different attack values. This means that two creatures with the same rating might have a difference of say 5 attack. 5 difference in attack is, in the perfect fight against a zero defence enemy, approx. 17% more damage. And with speed, you are using eleven ratings for only 9 different values, meaning that no unit ended up with a rating of 2 or 7, this skews your results since this gives an extra bias towards the top of the rating spectrum. All of this is a HUGE problem with your rating system that proceeds to affect every single number you have created since you rated EVERYTHING on that scale.

2) Your weight system is totally arbitrary, and you need to give more reason for the weights. The one that surprised me the most is the weight of zero that you gave to initiative, by no means an unimportant stat, since it determines who gets to deal damage first. Initiative is the biggest decider in a battle between two stacks of equal strength since the first one to deal damage will take less because there are fewer units left in the other stack. Your other weights are similarily fairly arbitrary, I do not agree, for example that attack and hp per leadership are of equal value or that speed is of so little value, considering speed is why peasants SUCK (their speed is actually 2, despite what your info may have said, and any non-ranged unit of speed less than or equal to 2 that doesn't have tons of hp (note that I don't say hp per leadership) is of suspect usefulness). Also, attack is not only 60% as important as damage per leadership, they are actually far closer in importance (see point 4 below).

3) This is a minor one, but you cannot compare ranged units and melee units with statistics, ranged units will tend to do less damage and be more fragile, but are attacked less and suffer no retaliation.

4) Any stat that is "per leadership" lies slightly since it is balanced by the increased survivability of the higher hitpoint units and the higher attack of the more damage units. Consider your black dragon vs 500 peasants analogy. If 500 damage is applied to each stack, you still have a stack of one black dragon, but you now only have 400 peasants. Also, that stack of one black dragon will do far more damage that the 500 peasants.

Note that all of this is not to say that your system is worthless, just that it has some major bugs that need to be worked out.

[Earthworm]

Ryastar,

Thank you for the feedback. I can address your points here.

I made this table while only having access to the statistics that are on the web, as I am currently at work, and cannot verify the numbers. Also as I said, I have not completed my first game yet, so I do not have access to all of the possible units in the game to check for accuracy, so I based them off the numbers I could find for the units.

As for your individual points, I will respond to them, without re-quoting what you wrote to help save on vertical space.

1. My Overall Rating is not based on the rounded numbers, I just did not feel it necessary to post all the Individual ratings out to n decimals, but I assure you in doing the math, no rounding was used. So while true a rating of 5 for attack maybe for a lot of units, the exact rating differs based on exact numbers, and in the list in the forum you only see the rounded number. I needed a standard scale for being able to average the ratings properly, and for ease of understanding.

2. I agree the weight system is arbitrary, and may need some refining, which I can do. From the posts I have read here, most people seem to feel that Damage per Leadership is a more important stat than say Attack rating, esp since Attack Rating is easily adjusted by the Hero's stats, and regalia. If you have some suggestions on the weight, I am definitely open to them. Also You are right, I should add a Category for Initiative as it is an important statistic of the Unit.

3. I agree with you, and actually debated breaking the ranged units into a separate list, since obviously the weights of the categories would have dmg much more important than speed and defense. Mostly I wanted to get the list up, and get feedback on it before spending a bunch more time.

4. While I agree with you that the per leadership stat for Hit Points is a little misleading, it is a very important statistic for measuring your army. If you are going to take 1000 points of damage to a stack of 500 Peasants, you will lose 1000 Leadership worth of peasants and 40% of the stack, where you will lose 2500 Leadership worth of Black Dragons, and 100% of the stack. I wrestled with how to rate Hit points, and felt as though the per leadership was a good way, but I am not set 100% in that thinking. Again Like I said, i have not completed a full game yet.

[Elwin]

Well its just theory and judging units just by stats is senseless, according to this support units would never be good. I would never choose you units accroding to these table ;) I dont find archdemon so powerfull ... hes goood against red/black dragons but at other fights there are way more better units than him. And also there are items and other features which change units power for example snakes with frog feanora and snake ring became very powerfull units

[Ryastar]

One quick response, which is all the time I have for now, about point 4, one thing I did not stress is that short of rage abilities, if you apply 500 points of damage to both stacks, the dragons won't take nearly as much damage. Let's assume that the attacker had 61 attack so the 500 is the full expression of damage ie. it is 300% of the base damage of the attacker so that means that base damage was 500/3=166.666etc. The damage to the black dragons will be only 194.44 (166.666*(3.33*5)). That is why the hp per leadership lies somewhat because the creatures with a lower hp/ldr tend to be the one with high def. So you cannot just apply x damage to both since the one wouldn't have received nearly as much damage. The same thing applies for dam/ldr and attack.

Oh, and since you aren't actually using only whole numbers for your rating system, then I concede that my point #1 isn't valid.

[Stepsongrapes]

First off, thanks for work. Each column of info is separately very intresting. I think overall rating causes some headaches, though.

A quick recommendation on weighting:

Damage and attack should be related to generate a shared weight rating. The seperate system you have now gives way too much weight to high attack, which is only useful to the extent it can modify the base damage. 3X crap damage is usually still crap.

Similarly, HP and defense should be related. What use is damage reduction without the HP to stand behind it?

[Earthworm]

Elwin, I agree that Units should not be picked solely on this table, as there are Tactics that make almost any unit here much better than raw numbers, which is one of the things that will give this game significant replayability. And as I said, the list does not take into account any unit specials, like Snake's Lunge ability or Dragon's resistance to fire. I was providing it more as a unit comparison, so if you were thinking of ditching one unit for the other, this gives a good break out of the two units, not for selecting the top 5 units from this list and always winning the game.

Ryastar, I also agree that the Dmg/Ldship and Hitpoints/leadership can be misleading which is why I did not rate the units based solely on either of those stats. As you pointed out, the ones with lower Dmg/Ldrs tend to have higher defense or other stats, so defense is included in the rating, as not all units with poor hit points per leadership totally make up for it in defense rating or attack rating.

Maybe I should remove the Overall rating, which would eliminate the arbitrary Weight issue. I guess I was using the overall to get an idea of the better units, then compare what category is lowering its rating below the rest, and determine if it is important to me. That is the main reason I showed the individual category ratings themselves, as I find them more important for comparison, but the Overall gives me a good starting point when deciding between Ancient Bears, and Polar Bears for instance.

[Ryastar]

I like the new table and weights, it's not perfect, but perfection here is impossible since a) it is too much of a headache to include damage type, resistances, critical chance, specials, and abilities, especially since the last two cannot be quantified (other than how many of each a unit has) and b) different people would weight all those things differently (I more or less agree with the current weights). The only things that are still to be done is 1) a confirmation of the data used, a time-consuming task, but I can do it, if desired. I should have enough saves to be able to find every unit fairly quickly. 2) maybe just a column that says whether a unit is ranged or not, because that really is an important data point, or, as suggested, a separate table for the ranged units.

P.S. Yay for Sprites being the best non-level five unit, and only surpassed by archdemons, black dragons, and red dragons.

[Leind]

I like the new table and weights, it's not perfect, but perfection here is impossible since a) it is too much of a headache to include damage type, resistances, critical chance, specials, and abilities, especially since the last two cannot be quantified (other than how many of each a unit has) and b) different people would weight all those things differently (I more or less agree with the current weights). The only things that are still to be done is 1) a confirmation of the data used, a time-consuming task, but I can do it, if desired. I should have enough saves to be able to find every unit fairly quickly. 2) maybe just a column that says whether a unit is ranged or not, because that really is an important data point, or, as suggested, a separate table for the ranged units.

P.S. Yay for Sprites being the best non-level five unit, and only surpassed by archdemons, black dragons, and red dragons.

Even if the classifying method is not perfect, it show correctly which units are the best/the worst. ie Sprite among the first, archer among the last.

To improve it you should put damage, attack & leadership in a single column : Average damage for 1pt of leadership against an enemy with, say, a defense of 20. This allow to take into account all basic damage parameters (number of units in troop, attack, damage.)
You can even add the critical chance as a multiplier, with for example an arbitrary 130% critical dmg.


The same can be done with defense & HP : how much effective HP for 1pt of Leadership, against an ennemy with an attack of 20.

Thus for Ancient Bear :
9.8 average dmg against an enemy with 20 defense for each bear.
With the 30% critical taken into account, that increase to 10.682 average dmg.

each pt of leardership give you 0.136 dmg and 0.875 pt of HP (supposing an ennemy with 20/20 att/def).

If you already have some excel chart with all units it should be easy to generate that.

As for the snake, I saw that some website mentioned a Lvl1 Snake. I guess its was transformed to a Lvl2 in later versions of the games.

[Earthworm]

Made an edit to the original Table and split out the Archers from non-archers, and rebalanced Weights for the Archers as Initiative and Dmg/Attack are way more important for Archers where defense is not so important.

Weights are now listed in the Table itself in the headings since each section has different weights.

I was unsure if I should have put the Dwarven Alchemist as an archer since its range is only 3 spaces, so it is much more likely to be attacked than the others, but I put it in there anyway for now, let me know what you think.

Also what to you think of the new weights for Archer class? Are they good or should they be tweaked.

I will post my raw Unit stats I based these rating off of a little later today so that anyone can verify the correctness of them.

[Ryastar]

Necromancers, Cyclopses, Archmagi, Inquisitors, druids, priests are all ranged with their basic attack, so you need to modify the table slightly.

[Earthworm]

Necromancers, Cyclopses, Archmagi, Inquisitors, druids, priests are all ranged with their basic attack, so you need to modify the table slightly.

Yah, I was not sure how to qualify the Magic Caster Archers, I was thinking about a third table for the Necro, ArchMagi, Inqui, Druids, and Priests, as really most people don't take them for their Archer ability so much as their specials. Not sure how I missed the Cyclopses in the list.