Heroes of Might and Magic Community
visiting hero! Register | Today's Posts | Games | Search! | FAQ/Rules | AvatarList | MemberList | Profile


Age of Heroes Headlines:  
5 Oct 2016: Heroes VII development comes to an end.. - read more
6 Aug 2016: Troubled Heroes VII Expansion Release - read more
26 Apr 2016: Heroes VII XPack - Trial by Fire - Coming out in June! - read more
17 Apr 2016: Global Alternative Creatures MOD for H7 after 1.8 Patch! - read more
7 Mar 2016: Romero launches a Piano Sonata Album Kickstarter! - read more
19 Feb 2016: Heroes 5.5 RC6, Heroes VII patch 1.7 are out! - read more
13 Jan 2016: Horn of the Abyss 1.4 Available for Download! - read more
17 Dec 2015: Heroes 5.5 update, 1.6 out for H7 - read more
23 Nov 2015: H7 1.4 & 1.5 patches Released - read more
31 Oct 2015: First H7 patches are out, End of DoC development - read more
5 Oct 2016: Heroes VII development comes to an end.. - read more
[X] Remove Ads
LOGIN:     Username:     Password:         [ Register ]
HOMM1: info forum | HOMM2: info forum | HOMM3: info mods forum | HOMM4: info CTG forum | HOMM5: info mods forum | MMH6: wiki forum | MMH7: wiki forum
Heroes Community > Heroes 4 - Lands of Axeoth > Thread: Creature Analysis: a good way to objectively compare units!
Thread: Creature Analysis: a good way to objectively compare units! This thread is 2 pages long: 1 2 · NEXT»
Wub
Wub


Responsible
Famous Hero
posted September 14, 2002 01:14 AM bonus applied.
Edited By: Hexa on 17 Oct 2002

Creature Analysis: a good way to objectively compare units!

This thread is about comparing creatures. I will start explaining extensively how to do it and after that give some examples. If you really don't like long posts shame on you ) you can still read the summary at the end of the post. Let's get started!

Introduction
I have been here at heroescommunity for a while now and I have read many threads about creature comparison. I read posts of some people saying that it is impossible to compare creatures, because it just depends on the situation in which they are used. Of course, such factors are very important but sometimes one unit is just better than the other. Nobody will deny that a bone dragon on the battlefield is in all ways better than a skeleton. What I want to say is: independent of situational factors, it is often possible to make valid pronunciations of which unit is better.

Some people want to find out which creature is the best by letting them fight on the battlefield. That is fine and it can give you some information, but you shouldn't put too many faith in this way of comparing creatures. Why not? Because all creature statistics are blent in to a random degree. For example: when you let a black dragon fight two faerie dragons, the magic resistance of the black dragon will weigh much heavier than it usually does. And you don't take the magic mirror ability of the faerie dragons into consideration.

How to do it
Then how should you compare creatures? It depends on what you need the comparison for. In this post, I will show a way that lets you compare which unit performs better on the battlefield. Stats like adventure map movement, creature cost and building prerequisites will therefore not be taken into account. Also, the taxpayer ability of peasants and the stealth of bandits are no factors on the battlefield. But I will use creature growth in the comparison. The idea is to systematically compare all of the stats that differ between units.

The first step to take is to summarize the 'strength' of a creature into one number. The strength of a unit is determined by its base damage, attack and defense skill and hitpoints. To calculate that number I will just multiply all four values with each other. If you really want to know why this is a good method to determine creature strenght, you can visit the newspage on h3trio.com; go to the news of a few weeks ago and click on the articles section. I wrote an article about attack and defense skills and creature comparison there.

All right, next step is to include creature growth in that number. It is extremely complicated and cannot be done very precisely. That is because it depends upon the defense/attack ratio of the units you are fighting against and the number of troops you have. For example: when 1 black dragon fights against 2 behemothes, it will have another outcome than when 100 dragons fight 200 behemothes. Therefore I can only give an indication how much creature growth weighs by giving a minimum limit and a maximum limit. For example: unit A is 10-50% 'stronger' than unit B. If you are really interested I can eloborate on the exact way to do it, but right now that may be a bit boring. But let's just say it is a very valid method.

The final step is to list all creature stats in which units differ. Then you will have all pro's and contra's of a unit on a row. It may seem hard so I give an example. We are going to compare bandits and pirates at land.

Examples

Strength of a bandit: 2(average damage)x9(attack skill)x9(defense skill)x10(hitpoints)=1620. Including growth of 26: strength is 42.120-1.095.120
Strength of a pirate at land: 1485. Including growth of 24: 35.640-855.360

Comparison of bandit and pirate at land:

Bandits: 18-28% stronger, +2 speed, +50% movement.
Pirates at land: -

Conclusion: Bandits are better than pirates at land. No if's and but's (read: no situational factors). They are just better!


A comparison between bandits and pirates at sea:

Bandits: -
Pirates at sea: 74-89% stronger.

Conclusion: Pirates at sea are flat out better than bandits.

more examples
Sometimes, the comparison can be made very easily but it is just hard to interpret the results. For example:

Nomad: +33.3% movement
White tiger: +26% strength

I can't say which is more important. It is too dependent on the situation. But it does give a very good comparison between both creatures.

Sometimes, creature specialties can be processed in a relatively easy way. For example:

Objective part:
Gargoyle: 92-116% stronger, stone skin, +1 speed, elemental
Ghost: insubstantial, aging, undead

Now I'm going to include the stone skin and insubstantial specialty:

Semi-objective part:
Gargoyle: 20-35% stronger, elemental, +1 speed
Ghost: aging, undead.

And now the interpretation; the subjective part.
In my opinion, the elemental and undead specialty is roughly equal, so they can be left out of the comparison. We then see that ghosts have an aging attack, while gargoyles have +1 speed and are 20-35% stronger. I can't give a definite opinion: it's a too close to call!

I hope you get the idea, let's give one final creature comparison:

Objective part:
Bone dragon: 58.7% stronger, terror, skeletal, flying, +86% movement, undead.
Hydra      : +1 speed, multiple attack, no retaliation.

Subjective part:
I think the bone dragon is a lot better. Terror and no retaliation aren't too much different. Hydra has +1 speed, but overall the bone dragon is much more mobile. The multiple attack doesn't compensate for that mobility and the skeletal and undead ability. Conclusion: the bone dragon will in most cases be a better choice than the hydra
(subjective, but at least well stated).


Summary

It is possible to give a reliable creature comparison. You just summarize the strength of a creature in one number and compare all other statistics that differ as well. Very often you will still need to rely on your common sense, but at least you can see very clearly what are the pro's and the contra's. When comparing creatures this way, you can see that for example bone dragons are in most cases better than hydra's and that bandits are better than pirates at land.

If you want me to make a certain comparison between creatures, just say so in this thread and I will give it a try.

Edit by hexa: just the kind of threads we need now and here!@




____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Wub
Wub


Responsible
Famous Hero
posted September 17, 2002 05:41 PM
Edited By: Wub on 17 Sep 2002

Some examples of creature analysis

Now the first expansion is released and the stats of the new creatures are revealed, let's see if those new units are any good.

First, the Dark Champion.

Dark Champion - lvl.4 Death-aligned
Damage: 30-42
Hit Points: 200
Attack: 40
Defense: 40
Combat Movement: 15
Speed: 6
Shots: 0
Spell Points: 18
Special Abilities: Charge Bonus, Terror, Undead, and Regeneration
Weekly Growth: 2

Maybe it would be nice to compare this unit with (a weekly growth of) Champions (level 4 castle unit):

Champions: First strike.
Dark Champions: 2,4%-104,8% stronger, terror (3x), undead, regeneration, +50% movement.

I think it is obvious that the Dark Champion is very often a better unit than the Champion. They may even be overpowered.

Now let's look at another new creature, the Gargantuan.

Gargantuan - lvl.4 Nature aligned
Damage: 28-40
Hit Points: 300
Attack: 32
Defense: 32
Move: 7
Speed: 4
Shots: 16
Spell Points: 0
Special Abilities: Ranged Attack, Shoots twice, Ranged Area Attack, and No Melee Penalty
Weekly Growth: 2

I will compare that Gargantuan with a respected level 4 creature, the Titan.

Gargantuan: 15.6% stronger, shoots twice, area attack.
Titan: +2 speed, chaos ward (edit, I forgot chaos ward...how stupid ).

Again, it is obvious that the Gargantuan is the better choice. I mean, it shoots twice, no ranged retaliation, can hit multiple units...move over titan!

Those new units don't seem to balanced. What's next? A level 1 unit with 300 hitpoints, area attack, spellcaster ability and life draining ability??? I hope 'The gathering storm'  will not imbalance gameplay too much...


With thanx to 'lich_king' from whom I used a post for the creature stats.



____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Djive
Djive


Honorable
Supreme Hero
Zapper of Toads
posted September 17, 2002 06:29 PM

Well, I'd be inclined to say that any creatures that are not included in a town should be at least slightly better than the ones that are. usually, you will be able to build for the units in the town a lot sooner than you can get to the external dwellings and that means you will get fewer creatures from the external dwellings. To compensate these creatures will have to have a somewhat better worth for value, otherwise you would not consider buying them.

You forgot the chaos ward for the Titan. It can come in handy at times. (Most direct damage spells are chaos aligned for one thing, so they do half damage vs. the Titans)


____________
"A brilliant light can either illuminate or blind. How will you know which until you open your eyes?"

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Wub
Wub


Responsible
Famous Hero
posted September 17, 2002 07:16 PM

Quote:

You forgot the chaos ward for the Titan.



Yep, that was a bit dumb of me . Thanx, I have edited it.

Quote:

Well, I'd be inclined to say that any creatures that are not included in a town should be at least slightly better than the ones that are.



That sure makes sense, but for some reason other creatures that are only recruitable from external dwellings aren't more powerful. A gargoyle isn't really superior, nor is a mummy, a troglodyte,  an ice demon, a peasant or a troll. But you are probably right that it isn't too imbalancing.

I made some other creature comparisons to show that an ogre magi really is an inferior unit. And yes, I did include growth! They fall behind on almost every level 3 walker, while we already knew they should not be chosen over a cyclops.

Ogre mage: bloodlust
Unicorn: 0%-20% stronger, +2 speed, +66% movement, blinding

I'd prefer unicorns.

Ogre mage: bloodlust
Naga: 5% weaker-42% stronger, +3 speed, no retaliation

Again, I wouldn't go for ogres.

Ogre mage: bloodlust
Nightmare: 16%-75% stronger, +2 speed, +50% movement, terror

Nightmare is better.

Ogre mage: bloodlust
Ice demon: 7%-60% stronger, immune to cold spells, freezing attack, +1 speed.

And finally, I would also prefer ice demons over ogre magi.

Now maybe I underestimate the bloodlust spcialty, but I think that ogre magi are inferior units. Good thing that Might can also build the cyclops.
____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Lich_King
Lich_King


Honorable
Supreme Hero
posted September 17, 2002 07:30 PM

Quote:
I think it is obvious that the Dark Champion is very often a better unit than the Champion. They may even be overpowered.


Could be, but Champion growth rate is 4 a week, while Dark champion's only 2 (But Undead with regeneration - I had never seen such ability's together, it will be tougher than I thought. Don't forget the terror )

____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Wub
Wub


Responsible
Famous Hero
posted September 18, 2002 01:52 AM
Edited By: Wub on 17 Sep 2002

Quote:

quote:
--------------------------------------------------------------------------------
I think it is obvious that the Dark Champion is very often a better unit than the Champion. They may even be overpowered.
--------------------------------------------------------------------------------



Could be, but Champion growth rate is 4 a week, while Dark champion's only 2 (But Undead with regeneration - I had never seen such ability's together, it will be tougher than I thought. Don't forget the terror )



Yes, but in the comparison the growth already is included. So, the Champions have double growth, but the overall picture of hitpoints, damage, attack and defense skill and growth of the Dark Champion is still better. And on top of that they have great specialties. You are right: Death is great

It would be most useful to compare creatures that cannot be built simultaneously. For example, ballistae and pikemen or minotaurs and medusas. Unfortunately, this is very hard in most cases because you will almost always have to compare a flyer with a spellcaster or a grunt with a shooter etc. But I will try to make some comparisons as well as possible. Remember that all stats are included except adventure map movement, cost prize and -very important- building prerequisites. Let's begin with the griffin and the unicorn:

Griffin: +30% movement, flying, unlimited retaliation
Unicorn: 1%-27% stronger, blinding

When you look at this, it is hard to say which unit is better in my opinion. It heavily depends on the situation. But people who think that the griffin is better and that therefore there is the high building prerequisite of a citadel, don't seem to be entirely correct.

The efreet and the nightmare.

Efreet: Immune to fire spells, fire shield, flying, +44% movement.
Nightmare: 39% stronger, terror.

I myself have a preference here, but it is hard to say from the comparison which is better.

The hydra and the black dragon:

Hydra: Multiple attack, no retaliation.
Black Dragon: 54%-207% stronger, immune to magic, breath attack, flying, +1 speed, +114% movement.

Look how much stronger the black dragon is! And how much more mobile! Sure, there are situation in which I would prefer a hydra -when I have an extremely powerful magic hero for example- but the Black Dragon seems preferable to me in most cases.

The Champion and Angel:

Champion: charge, first strike.
Angel: 9%-108% stronger, flying, +4 speed, +50% movement, resurrection.

Hard to make the final conclusion, but I would be inclined to say that the Angel should often be preferred. But...judge for yourself.

The Cerberus and Ghost:

I will include the insubstantial specialty in the strength rating of the ghosts.

Cerberus: 8% weaker-18% stronger, 3-headed attack, no retaliation.
Ghost: aging, flying, undead, +86% movement.

I don't know. Seems a too close to call. At least to me, that is.

The devil and the bone dragon:

Devil: +2 speed, teleport, summon ice demon, life ward.
Bone Dragon: 42% stronger, terror, skeletal, undead.

Again a too close to call. That means it depends very heavily on the situation.

The harpy and the nomad:

Harpy: Flying, strike and return, no retaliation.
Nomad: 52%-103% stronger, +3 speed, first strike.

I'll let you do the subjective part for yourself, beacouse I find it hard to say.

Well, I tried to compare some creatures that cannot be built simultaneously and that is hard. As I said, the units are so different that the comparison often depends very much on the situation. But you may have an idea now of which units you personally prefer. Personally I like more to compare units that are quite similar, however.




____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
dragonsister
dragonsister

Hero of Order
MapHaven administrator
posted September 18, 2002 09:26 AM

Quote:

Well, I tried to compare some creatures that cannot be built simultaneously and that is hard. As I said, the units are so different that the comparison often depends very much on the situation. But you may have an idea now of which units you personally prefer. Personally I like more to compare units that are quite similar, however.



In other words: your method of comparing creatures *isn't* the be-all and end-all of creature comparison ... there *is* some scissors-paper-stone stuff going on.

Although it *is* better than anything we had for HOMM3, at least partly because of the simple relationship between stats and damage in HOMM4.  And I would like to see some explanation of how you allowed for creature growth!  :-)

Details for Evil Sorceress now out - it looks like they do some massive spell-casting; move over genies!  We've now seen three different 'new L4 creatures' - I'd guess there are another three we'll hear about between now and the game's release.  One thing I find very interesting - the new creatures seem to fit town themes rather than rounding out alignments; the Sorceress has much in common with the Titan and Genie (and Mage; she's fragile!  Granted order *can* do hefty melee lineups); the Dark Champion is a regenerating melee unit, a bit like the Vampire, where Death could really benefit from another shooter.  The Gargantuan joins elementals, waspworts, and elves in the Nature lineup; I could also count Faerie dragons as ranged units.  (Elementals and Waspworts count for something because of the Creature Portal.)

So what will we see for Might, Life, and Chaos?

Might unit probably *won't* be ranged, given how little ranged power there is in the might alignment.  It will probably be a fast tough walker.

Life - may well have a ranged unit; if so, there's a good chance it won't suffer ranged penalties!  It may have Death Ward.

Chaos: is a pretty varied town, making predicting hard.  I bet the creature's artwork will be muddy or dark colours, though, and it'll probably have high hitpoints.

Them's me guesses.  Soon we'll see how wrong I am :-)

DragonSister
____________
MapHaven

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Wub
Wub


Responsible
Famous Hero
posted September 19, 2002 12:08 AM

How creature growth influences creature performance.

Quote:

In other words: your method of comparing creatures *isn't* the be-all and end-all of creature comparison ... there *is* some scissors-paper-stone stuff going on.



Of course I can (in most cases) not definitely say which unit is better than the other, because it is often too dependent upon the situation. And to be honest, I wouldn't want it to be else, because then the game would grow boring pretty soon. But I do claim that this way of comparing creatures gives a good idea of the strengths and weaknesses of units. Only the interpretation is subjective, the comparison itself is purely factual.

Quote:

And I would like to see some explanation of how you allowed for creature growth! :-)



Well, you asked for it ! I will try to explain it, but it will become a long story. In fact, it will become a long, theoretical and mathematical story. That's why I refrained from explaining it before. I will do my best to make it understandable and not too theoretical (which is hard).

First step to take in any creature comparison is calculating the strength of one single unit by multiplying attack and defense skill, hitpoints and base damage. Thus you get a certain number, for example 1620 for a bandit. Now it is the trick find a specific multiplier for that number in such a way, that the growth of 26 is included. What number is that multiplier?

Unfortunately, that multiplier is not a fixed number. It depends on things such as the defense/attack ratio of a unit and the number of weeks growth. For example: 1 black dragon vs 2 behemothes will have an other outcome than 50 black dragons fighting 100 behemothes or 300 fighting 600. Yet the proportion between both creatures resembles the actual growth of both units; 2:1. But even though I can't give a fixed multiplier for those reasons, I can still indicate the lower limit and the upper limit to which growth influences creature performance. Thus, I can say thing like: creature X is at least 20% stronger and at most 60%. Now how are those lower and upper limits defined?

Almost every heroes fan will have experienced that once you have a very big army, you will hardly take any losses when engaging smaller armies. Bigger armies suffer relatively few losses against smaller armies. Now let's make this concrete with an example.

A stack of 10 bandits fights a stack of 5 bandits. How many bandits are left after this fight?

It may be tempting to say that there will be 5 bandits left, but then, don't forget that bigger armies take relatively few losses! In fact, the stack of 10 bandits will win with (on average) 8 bandits remaining. Now let's continue fighting with those 8 bandits against stacks of 5 bandits. It appears that you can take out 2 extra stacks of 5 bandits. Conclusion: while the large stack of bandits was only 2 times as big as the small stack, it could finish 3 of them. Or: in this case the growth is 2 times as high, but it makes the creatures roughly 3 times more effective.

This is not coincidently. When creatures are X times as numerous, they will be more than X times as effective. Seems we have found our lower limit here.

And now for the upper limit. Imagine the following situation. A stack of 10 unicorns fights 1 black dragon. Now when you look at the stats of both creatures, it appears that 10 unicorns do exactly the same amount of damage as 1 black dragon and they can take exactly the same amount of damage. Now who will win?

Again, it may be tempting to say that the result will be a draw. However, the dragon appears to win with ease. Now how is this possible? At the beginning of the combat, the fight will go as planned: the black dragon destroys 20% of the hitpoints of the unicorns and the unicorns damage the black dragon for 20%. But after the first blow, two unicorns have perished while the dragon can still do full damage. And at the last round only 2 unicorns are left, doing 16.5 damage while the black dragon still does 165 damage.

Summary: the reason that the dragon wins, is that it is more 'compact', it will not suffer a loss of damage during the fight. The unicorns will suffer a loss of damage during the battle, because they are a stack and therefore constructed out of separate units.  
Conclusion: If you 'compress' the stats of a stack of units to those of 1 unit, you are overestimating the creature performance of the stack. Or: if you say (like I did in my argument)

Growth x attack x growth x defense = total strength unit

you are overestimating creature performance. So: if you want to include growth, you should never multiply with more than the square of the creature growth.

Or in simple words:

If creature growth is 10, the multiplier for creature growth will never be more than 10²=100.
If creature growth is 5, the multiplier for creature growth will never be more than 5²=25. Etc.

Ok, now I have defined the lower and upper limit to which growth influences creature performance. The conclusion is:

If creature growth is 4, the multiplier for creature growth will be at least 4 and at most 16.
If creature growth is x, the multiplier for creature growth will be at least x and at most x².

All right, if you read until here, your head will probably be swimming. Let's just say I also tested this theory extensively and it seems to be right. I agree that the multiplier for creature growth isn't always too accurate, but it always contains the extent to which creature growth influences creature performance.

Let's give a final example of a creature comparison. I will compare orcs with centaurs, two roughly similar creatures that mainly differ in growth.

Strength of 1 centaur: 4 x 20 x 10 x 10 = 8,000.
Including lower limit of creature growth: 9 x 8,000 = 72,000.
Including upper limit: 9² x 8,000 = 648,000.
So, the strength of the weekly creature growth of centaurs is between 72,000 and 648,000.

Strength of 1 orc: 2 x 12 x 11 x 9 = 2,376.
Including lower limit of creature growth: 16 x 2,376 = 38,016
Including upper limit: 16² x 2,376 = 608,256.

Comparing lower limits of both creatures:
72,000/38,016 x100%= 89.3%

Comparing upper limits of both creatures:
648,000/608,256 x 100%= 6.5%

Conclusion: Centaurs are 7-89% stronger. And then the total comparison:

Orc: +2 shots
Centaur: + 66% movement, 7%-89% stronger.

Interpretation: Centaurs should almost always be preferred.


____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
afm
afm


Hired Hero
posted September 19, 2002 08:12 AM
Edited By: afm on 19 Sep 2002

Unfortunately (or is it fortunately?) the comparison of two units is not that easy, even if you have the weekly growth in your formula.

1:
In my first online game, I had lots of wood and ore, but only a small number of the other resources. There was no doubt to go for champions instead of angels, although I think that angels are usually the better choice.

2:
You also have to bear in mind: The costs for the building and for the necessary prebuildings may differ a lot. Maybe you can get a weaker creature days before you can get your favorite - and it makes a difference in online play if you have lvl4 creatures and your opponent has not.

3:
Usually you don't have money to waste so the cost of a creature is an important factor, too.

Just my 2 cents
afm

____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Wub
Wub


Responsible
Famous Hero
posted September 19, 2002 05:55 PM

You are completely right of course, afm. I would even say, as you suggested, that it is fortunate that creature comparison isn't so easy. It would make the game boring. Also, I wouldn't dare to deny that things such as building prerequisites are important factors. But if you bear those things in mind, it is still useful to know where certain units excel.

If I compare griffins and unicorns and I think they are equally strong, it doesn't make sense to build griffins since they require an expensive citadel. And if I have a lot of money, cost price of units do not make a difference. So the bottom line is that even when not all factors are taken into account, the comparison may still give you useful information.

By the way, this doesn't mean that I don't fully agree with your post!


____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Holger
Holger


Adventuring Hero
King of Silence
posted October 20, 2002 03:43 AM

Quote:
First step to take in any creature comparison is calculating the strength of one single unit by multiplying attack and defense skill, hitpoints and base damage. Thus you get a certain number, for example 1620 for a bandit. Now it is the trick find a specific multiplier for that number in such a way, that the growth of 26 is included. What number is that multiplier?

Unfortunately, that multiplier is not a fixed number. It depends on things such as the defense/attack ratio of a unit and the number of weeks growth. For example: 1 black dragon vs 2 behemothes will have an other outcome than 50 black dragons fighting 100 behemothes or 300 fighting 600. Yet the proportion between both creatures resembles the actual growth of both units; 2:1. But even though I can't give a fixed multiplier for those reasons, I can still indicate the lower limit and the upper limit to which growth influences creature performance. Thus, I can say thing like: creature X is at least 20% stronger and at most 60%.


There could be a problem, though.

let's compare centaurs with skeletons:

<b>centaur</b>
Minimum: 72,000
Maximum: 648,000

<b>skeleton</b>
Minimum: 43,200
Maximum: 1,080,000

Difference:
Minimum: 43,200-72,000/72,000*100%=-40%
Maximum: 1,080,000-648,000/648,000*100%=67%

Which concludes that the skeleton is 40% weaker - 67% stronger than the centaur in raw damage fight...?!

If you try that out in real life you will learn, that the skeletons beat the snow out of the centaurs - without taking speed and shooting into consideration... and even if you do include speed and shooting the centaurs are in trouble - yet the statistics say that the outcome should be difficult to guess...:-)

Statistics are wonderful they proof what you want them to proof.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Wub
Wub


Responsible
Famous Hero
posted October 21, 2002 12:15 AM
Edited By: Wub on 20 Oct 2002

Quote:

Which concludes that the skeleton is 40% weaker - 67% stronger than the centaur in raw damage fight...?!

If you try that out in real life you will learn, that the skeletons beat the **** out of the centaurs - without taking speed and shooting into consideration... and even if you do include speed and shooting the centaurs are in trouble - yet the statistics say that the outcome should be difficult to guess...:-)



Thank you for criticizing the analysis. Now I realize that I didn't explain some things too well. At first, your calculation is flawless...the result indeed says that skeletons are 40% weaker -67% stronger than centaurs. Now what does this exactly say?

First, keep in mind that such a creature analysis does the following. First, it estimates how strong centaurs are in comparison with all other creatures. Then it estimates how strong skeletons are in comparison with all other creatures. Finally, it compares the results of both comparisons. That's one reason why you can't pinpoint exactly how much stronger one creature is as the other...because you don't calculate the outcome of a stack of centaurs fighting skeletons but both of them fight a variety of creature stacks. And that's the only way to go, otherwise the comparison would be meaningless.

Second, you cannot say exactly how different growth rates influence creature performance. And when you compare creatures with very different growths (such as skeletons and centaurs) the range in which creature strengths vary can be huge, such as -40 to +67%. But the actual value can be -20% as well as 0% or +65%. And when one creature is 65% stronger than another, you can hardly call it a close battle.

But of course it is true: skeletons slaughter centaurs. When 25 skeletons fight 9 centaurs, they will win with roughly 12-13 left (yes, I also test what I say ). Therefore it may indeed be misleading to say that skeletons are -40% weaker to +67% stronger. But nevertheless it is true. You see, it may seem impressive to have half of your skeletons left after the battle, but in terms of actual strength of the stack, your skeletons are 'nearly destroyed'. In other words: you have 50% of your skeletons left, but your army strength may very well be down to something like 30%.

An example: if you let those 12-13 skeletons fight another stack of 9 centaurs, all skeletons will be slaughtered while only 2 centaurs or something perish. So, these 25 skeletons can't kill much more than 9 centaurs actually.  

Of course I can more or less pinpoint how much stronger a weekly growth of skeletons is as a weekly growth of skeletons (when fighting solely each other, that is). It would be something like 25%-50%, which is perfectly within the -40% to +67% range.

Conclusion

Saying that a unit is, say, 50-100% stronger than another is indeed misleading, because it doesn't directly say how many units will remain in a direct fight. But it isn't untrue.


Quote:

Statistics are wonderful they proof what you want them to proof.



Undoubtedly . But that wasn't my plan.

____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
csarmi
csarmi


Supreme Hero
gets back
posted October 21, 2002 03:57 PM

Centaurs vs Skeletons

Well, I tested the

9 Centaurs vs. 25 skeletons battle

and found that 4 skeletons were left (and even the skeletons were lucky - they got morale at precious time)
____________
Yes, I play the game only on the forums.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
DarkVenom
DarkVenom


Adventuring Hero
Grandmaster Daywalker
posted October 22, 2002 07:28 AM

Wow...I can't believe how Wub can write such a long letter (just with his mere hand of leaves and branches ;-)
____________
A dead has no weakness of a living.

 Send Instant Message | Send E-Mail | View Profile | PP | Quote Reply | Link
Wub
Wub


Responsible
Famous Hero
posted October 23, 2002 05:50 PM

Quote:

Well, I tested the 9 Centaurs vs. 25 skeletons battle and found that 4 skeletons were left



Hmmm....you probably fought to the best of your ability with the centaurs and/or the skeletons, using any creature specialties to the fullest. I let them fight as follows:

Centaurs make melee attack, skeletons make melee attack, centaurs make melee attack, etc. No tricks, no morale, no specialties, just to measure the strength of both stacks. In that test, +/- 13 skellies were left.

Quote:

Wow...I can't believe how Wub can write such a long letter (just with his mere hand of leaves and branches ;-)



Yeah, I can't believe how I manage to do it either . You have no idea how annoying it is to automatically bind any key of my keyboard that I touch to my fingers. Not to mention the painly and dreadfully boring process of unbinding those keys .
____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Djive
Djive


Honorable
Supreme Hero
Zapper of Toads
posted October 23, 2002 06:03 PM
Edited By: Djive on 23 Oct 2002

You didn't even shoot with the centaurs?

Also for a reasonable tactics Centaurs should wait before they fire because that way the skeletons may be forced to move closer and reduce the short range penalty.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
csarmi
csarmi


Supreme Hero
gets back
posted October 24, 2002 12:01 PM

Wub, you should decide on what you want to compare. Do you want to compare L1's with L3's for example? If so, the maths need some adjusting (and it IS possible). So, what's your aim? Maybe focusing on 1vs1 comparisons? If so, then I might have some ideas to make it better.

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Shadow_Elf
Shadow_Elf


Wandering Thief
posted October 24, 2002 01:48 PM

what if u did it with towns im only in 6th rade math so i dont really understand but if you found the supreme creature for each level the town with the most would win !
____________
Dont mess with the best cuz u aint got no contest.
Forget the rulz they were made to be broken

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
csarmi
csarmi


Supreme Hero
gets back
posted October 24, 2002 01:57 PM
Edited By: csarmi on 24 Oct 2002

This is far from being true - there are so much factors. Team work, for example.

It is not complicated anyways:

Wub states, that (including growth)

Gr*HP*Off*Def*Dam < UnitStr < Gr*Gr*HP*Off*Def*Dam

that's all.

As for the comparison itself: it is still hard to compare shooters and foot troops, L1's and L4's, spellcasters and tank troops, etc...

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Wub
Wub


Responsible
Famous Hero
posted October 24, 2002 04:02 PM
Edited By: Wub on 24 Oct 2002

I see that some things are not completely clear. Let me clarify, without using those maths. Let's say I compare two creatures. I am going to list all plusses of both creatures. So for example, creature 1 has a skeletal specialty and creature 2 has +4 speed. Now assume that all other creature statistics are the same. You can now make a (subjective) decision on which unit you prefer.

Another example, trolls and gold golems. This should look something like this:

Things in favor of the golems: they are somewhat 'stronger' (I'll explain in a minute), they have extra combat movement, they have 75% magic resistance, they are mechanical.

Things in favor of the trolls: they have regeneration.

And finally you weigh out all plusses and decide what creature is best or what creature is best in what situation.

As you can see, I said that golems were 'stronger' than trolls. I know it sounds a bit lame. Usually I say something such as 'absolute force' myself. So, 1 golem has a higher 'absolute force' than 1 troll. But I don't want to force anybody to use those words, so I kept saying strength everywhere. The most basic idea of strength/absolute force is this:

(All damage that a unit can do) X (all damage that a unit can take).

All damage that a unit can do is defined by its base damage and attack skill. All damage that a unit can take is defined by its hitpoints and its defense skill. If you compress all these 4 statistics in one number, you have the strength of one unit.

You may object that a medusas stone gaze increases the damage done and a minotaurs block increases the damage that can be taken. Well, agreed. And strictly speaking, these specialties should be included in the strength number. But it often is not that easy, so I leave those strength influencing specialties out and only add them at the subjective part of the comparison.

Alright, now you can calculate the strength of one creature. But often there is a difference in growth. For example, if you want to compare orcs and centaurs, you need to know how important growth is if you are to make a meaningful comparison. So I developed the formula that csarmi mentioned:

Gr*HP*Off*Def*Dam < UnitStr < Gr*Gr*HP*Off*Def*Dam

With this, you can calculate the strength of a one week growth of creatures. I thought it sounded quite theoretical and as always, you need to test such a thing. So I did. I let creatures fight in a way that they use only their strength. I didn't let them do any tricks etc. etc. After all, specialties such as flying and unlimited retaliation and also speed and movement are included in the comparison later. So to answer your question, Djive:

Quote:

You didn't even shoot with the centaurs?

Also for a reasonable tactics Centaurs should wait before they fire because that way the skeletons may be forced to move closer and reduce the short range penalty.



Believe me that I would have done that (along with some other tricks) when fighting to the best of my ability. But I just wanted to test if the comparison was right about the strength of both creatures. And the short range specialty is included only later.

Quote:

Wub, you should decide on what you want to compare. Do you want to compare L1's with L3's for example? If so, the maths need some adjusting (and it IS possible). So, what's your aim? Maybe focusing on 1vs1 comparisons? If so, then I might have some ideas to make it better.



Well, the basic idea was to be able to compare a weekly growth of unit X with a weekly growth of unit Y. But, as you say yourself, csarmi:

Quote:

As for the comparison itself: it is still hard to compare shooters and foot troops, L1's and L4's, spellcasters and tank troops, etc...



That is right. It is rather meaningless to compare a shooter with a spellcaster. Or a level 1 with a level 4. So generally, I restrict myself to using the comparison for similar creatures. For example: griffins and efreets (same level, both flyers) or orcs ans centaurs (same level, both short range). If I wouldn't do so and I would compare faerie dragons with phoenixes, you get this comparison:

Faerie dragon: Spellcaster, magic mirror
Phoenix: 192% stronger, rebirth, fire spell imunity, breath attack, +5 speed, +15% movement.

And still you know virtually nothing. Just because, as you said, there are too many factors left. But you said you may know some things to improve the comparison. I am very curious to hear those.

Quote:

Wub, you should decide on what you want to compare.



Well, I focused on the creature performance on the combat map. Therefore I didn't include adventure map movement, creature cost and building cost(s) in the comparison. You could also try to make a comparison to decide which creature is a better build.

Quote:

what if u did it with towns im only in 6th rade math so i dont really understand but if you found the supreme creature for each level the town with the most would win !



Indeed Shadow_Elf, that is what I was after, ultimately. Although to decide which town is best, you should also look at your heroes and the magic they can obtain etc. But you're right, I can say that in most cases 3 vampires are better than 4 griffins. So I can also say that either Death has one of the better level 3 units or Nature has one of the worse level 3 units. The ultimate goal is to say which creatures of which towns are strong and on which creature level a town is vulnrable.


____________

 Send Instant Message | Send E-Mail | View Profile | Quote Reply | Link
Jump To: « Prev Thread . . . Next Thread » This thread is 2 pages long: 1 2 · NEXT»
Post New Poll    Post New Topic    Post New Reply

Page compiled in 0.1396 seconds