Warcraft 2; incorrect information?

[X3M]

When searching for information on damage calculation in Warcraft 2. Several different sites tell you something .....different.

 

While I consider a different story to be incorrect, when comparing with the official explanation on the blizzard wiki. Even then I find the game acting different than expected.

 

Here are some examples of explanations:

 

Whereas Armor, Piercing Damage and Basic Damage are the 3 variables that are used in the game.

And negative values occurring in a step are considered to be 0 instead.

Halves are rounded upwards.

 

Explanation 1:

Basic Damage minus Armor. A negative number becomes 0.

Now Piercing Damage is added.

Result is divided by 2, where a halve is rounded upwards. This is called the minimum damage.

The maximum damage is 2 times the minimum damage. This means that a result of 5, gives 2.5, which is 3, which gives 6 as the true maximum damage.

 

Explanation 2:

The same as 1. But now, the only 2 choices are the minimum or maximum damage. There is no continuous line. Whereas 3-5 would give 3, 4 or 5. In the second explanation it is either 3 or 5.

 

Explanation 3 (official wiki):

Basic Damage is divided by 2. This number is rounded upwards when needed and is the minimum Basic Damage. We only have 2 choices in this again like in the second.

Then both the minimum Basic Damage and the Basic Damage minus Armor.

Now we have 2 remaining numbers as choices.

Piercing Damage is divided by 2. This number is rounded upwards when needed and is the minimum Piercing Damage. We only have 2 choices; this half number or the whole.

Thus the first choice and second choice are added up now.

 

Explanation 4:

Same as 3, but Armor also serves with half or full.

 

Explanation 5 and 6:

Same as 3 and 4, but everything is continuous again. Sigh.

 

Who has the correct answer? So I decided testing!

 

---

 

Additional information, got this from a site that has a major flaw in one of the calculations and kept that flaw.

Unit Cease-fire cycles peon, grunt, ogre 25 skeleton 35 mage 40 archer 65 submarine 100 destroyer 120 cannon & guard tower 150 dragons 190 catapult 200 juggernaut 230

 

Thus even though this info is nice to have. I wonder if it is correct. Maybe it was created by tests. Then it could be influenced by the game speed just like in Dune2.

For this, having the source code would be nice.

However, a simple test shows me that the guard tower and cannon tower are completely different. The guard tower shoots about 3 times faster. Shame on that person who posted on that site in the first place.

 

---

 

Testing in Warcraft 2:

 

Test 1:

1 Piercing Damage. And attacked 0 Armor.

Actually, out of the 20 attacks, 6 happened to be 2 damage while the rest was 1. Thus a total of 28 damage while expecting only 20. How weird!

Another particular thing that I noticed is that how other units all have #-## as damage indicator, this unit had only a 1. Thus the game actually knows that there should be no other option. I would have expected 1-1.

 

Test 2:

Same as test 1, but instead of Piercing, I used 1 Basic Damage.

Once again, 2 damage in one strike occurs on a regular basis.

 

Test 3:

That same 1 Basic damage against 20 armor. The result was constantly 0.

Even 1 armor gives constantly 0.

 

Test 4:

30 basic damage against 20 armor.

The results in damage returned are: 6, 8, 5, 7, 7, 8, 9, 8, 7,10, 5, 7, 9, 9, 7, 5, 5, 5, 7, 9, 10, 8, 10, 10

Thus a yes or no choice in 50% or 100% damage is false. Explanations 2, 3, 4 and 6 are definitely scrapped now. Since all damages are under 10. This means armor itself is not random.

 

Test 5a:

1 piercing against 1 armor.

50 shots place 75 damage.

Test 5b:

2 piercing against 1 armor.

50 shots place, again 75 damage.

Ok, that is really odd. Somehow 1 and 2 piercing damage become the same value.

Test 5c:

2 basic against 1 armor.

50 shots placed 75 damage once again.

 

Could it be that there is an error in rounding the numbers upwards?

 

I think the only choice to find out what it really has to be is to look into the source code. But how to do that? Who has this power amongst you?

 

---

 

Meanwhile, I have come up with a little theory myself on how it could work. Regarding having just 1 basic or 1 piercing damage against 0 armor. However, if you continue to look at other tests. It is clearly false.

 

The game takes 50% of this 1. And randomly chooses anything between 0,5 and 1.

Once decided. The game then multiplies this with 2. There for having a number between 1 and 2.

The rounding only happens after the multiplication? Thus it then becomes either 1 or 2 damage. In the long run, we have 1,5 on average.

 

The reason why this is false is because we are supposed to get the same treatment with 2 damage or more.

By random, we have anything between 1 and 2.

Once decided. The game rounds this to 1 or 2.

There is no multiplication.

 

Perhaps the statement in the third explanation is correct. Before the numbers are chosen. This 0,5 is rounded upwards to 1. And the new maximum damage is 2 times this 1. Thus 0-1 effectively becomes 1-2. While 2 gives the exact same result.

 

If this is true, then all the other odd numbers used in the calculations are influenced by this.

The worst one is the 3 piercing for the footman/grunt compared to the 4 piercing for the knight/ogre.

Both get a minimum of 2 and a maximum of 4? This calls out for a test.

 

Well, that test escalated quickly. They indeed give the same, exactly the same destructive power. 600 vs 600.

The question of course is, does this count for basic damage too?

Well, yes it does! Again 600 vs 600.

 

However, I took the liberty of combining 3 piercing with 3 basic. And 4 piercing with 4 basic. The outcome? While the first one does 450 damage, the other one reaches 600.

What happened there?

If I think about it, the only explanation would be that piercing is halved. And basic is halved. Then they are added up. And only after that, we have the rounding upwards.

Thus 50% of 3 is 1,5. We have 2 x 1,5 for piercing and basic. Thus 3 in total for the lower limit. 6 is the upper limit.

50% of 4 is 2, we have 2 x 2 = 4 for the lower limit. 8 is the upper limit.

The question is, is it possible to make 7 as an upper limit?

 

By having 3 piercing and 4 basic. We get 1,5 + 2 = 3,5. This is 4. So with the knowledge that we have. The upper limit would become 8?

With 4 piercing and 3 basic. We get 2 + 1,5 = 3,5. Same story, right?

Let's test it.

And thus after testing, we got twice 600 vs 600 damage in total. So the story is true indeed.

All I can say now is that using odd numbers is a bad thing to do in that game. You can't really track the true balance.

 

---

 

But wait, there is more. What about upgrades?

For the archers/axethrowers, we have +1 with each upgrade. And even +3 for the archer upgrade.

Not only that. Ships have +5 for each upgrade as well. These are odd numbers too!

 

With 6 piercing and 3 basic damage. It is obviously clear that the true damage against 0 armor would be 3 + 1,5 = 4,5

=5 and 5-10.

After 1 upgrade, we have 7 piercing and 3 basic. 3,5 + 1,5 = 5. Again 5-10.

After another upgrade, we have 8 piercing and 3 basic. 4 + 1,5 = 5,5 = 6. This time 6-12.

The +3 for the archers gives 11 piercing and 3 basic. 5,5 + 1,5 = 7. Now we have 7-14.

 

According to this theory, the very first upgrade is useless unless you do the second one as well.

Or, we do this upgrade and go for the +3 immediately afterwards. Ignoring the second upgrade.

Then we get 10 piercing and 3 basic, 5 + 1,5 = 6,5 = 7. Which also gives us 7-14.

Ignoring the second upgrade saves us 900 gold and 500 wood.

 

Of course, the story goes differently when facing structures. Well, I think that all basic damage would completely be nulified.

No upgrades gives 6 piercing, 3-6

+1 gives 7 piercing, 4-8. This time, the upgrade does have use!

+2 gives 8 piercing, 4-8. This time, the second upgrade has no use at all.

+5 gives 11 piercing, 6-12.

+4 gives 10 piercing, 5-10.

 

On a side note, other melee units will be having more armor with upgrades. This means that you only are dealing with your piercing damage. Thus it is important to use all upgrades by the humans. Yet the Orc's can still choose to ignore the second upgrade since it has no use against any armored unit. Unless, you use axe throwers to deal with 0 armored units like peasants, mages or balista's. But against those, you might as well use ogre's and grunts.

 

True formula of Warcraft 2:

Basic Damage - Armor = Remaining Basic Damage.

Piercing Damage + Remaining Basic Damage = Total Damage

0,5 x Total Damage = 50% Damage. This is rounded upwards.

50% Damage x 2 = 100% Damage.

A random value is chosen between 50% Damage and 100% Damage. This includes the 50% and 100%.

 

Ok, the editing ends here. Comments now please.

[Nyerguds]

With all that research, it might be easier to just disassemble the exe xD

[X3M]

Well, with the research I figured it out :D. As usual.

 

ROF remains. But that will be easy. Simply counting the shots by giving every unit 1 D and knowing it is 1,5 on average. ;)

When dealing with 18000 health (12000 shots). I can use a timer.

 

When I am done, I can post them here. Think I will have the most accurate list then.

 

---

 

I have my suspicion that the ROF of the archer compared to a footman is indeed correct according to his/her test.

Because I have the same answer at first as well. A 2,6 factor.

However, I did the test over and over. And eventually I got an average of 2,667 as factor. This is clearly a noticeable difference.

I also am having doubts in the other tests. I only went for 900 damage for a footman and noted the damage of the other units as reference. Thus perhaps I should change it into 18000 any way. And simply being very patient about the test.

 

Unless someone of you knows how to look at the ROF in the source.

[Nyerguds]

Took me about a year to get familiar with TD's internals... would also take a while to get to know the WC2 ones :-\

[X3M]

Well, I think my ROF will be estimates. I really need a pro on this one :D. So get working on it :D.

 

I have spend 1 hour now on only testing the archer compared with a melee unit. I have seen the ratio going between 2,5 and 2,7.

Eventually I did put a 6000 health building to death (0 armor, and each unit 1 basic damage). When the footman finishes. The archer did 2349 damage.

This is about 2,55 to 2,56 as ROF ratio.

My mistake might be using health as a guide. Of course I activated my units and await the result. But the damage is random.

 

I really have to start counting. Which is something that I rather don't spend time on.

And for accuracy, I need to count to,... 100 shots?

[MrFlibble]

The official Blizzard website from 1997 had an explanation of what they called "The Combat Equation":

http://www.blizzard.com/tips/tip2.htm

I have no idea if this is how it works in-game, but I suppose that at least it's how the developers intended it to work.

The Combat Equation is also used to explain how the Bloodlust spell works:

http://www.blizzard.com/tips/tip6.htm

About your tests, what if the damage values that you gave the tested units somehow affected the outcome? It is my understanding that in tests 4a and 4b you used 0 basic and 0 piercing damage respectively, right?

[X3M]

Regarding the workings explained by Blizzard. Most other sites say, that Blizzard has it wrong. And testing did show that the game works differently.

The damage displayed in the unit box is in most times different than what it actually is.

Examples;

- Footman has 6 basic, 3 piercing. The box shows 2-9. But it has to be 2-10.

- Knight has 8 basic, 4 piercing. The box shows 2-12. Which is correct.

- Archer has 3 basic, 6 piercing. The box shows 3-9. But it has to be 3-10.

 

Thus the maximum damage is displayed wrong in most cases. When this one is an odd number, add 1 for the correct number.

Not only that, but they take 15 armor into account when the basic damage goes above 15 armor. It would have been better to use 20 instead. Or even 25 since the Battleships can reach this armor.

Ballista has 80 B and 0 P. The box shows 25-80. Which insinuates that the basic damage is decided by random first. Then armor is subtracted before piercing is even added. If this was not insinuated but the armor is subtracted before a random number is chosen like it should be. Then we would get 33-80.

If the box would show the result of the real mechanic and using the 20 armor of the structures, then we would have 30-80.

 

---

 

Yes, when I put in 30 piercing damage, the basic was indeed 0 damage. And vice versa.

But I see I have posted something wrong there by mistake?

4b was a correct test and is the same as what to be expected with the knowledge that I have now.

But 4a surprises me and I will do that one again to be sure. Because armor does not influence piercing damage. And returns like 11, 12, 13, 14 and 15 should be expected as well.

But then again, 4a happened after 4b, that I remember. Perhaps I didn't save the alteration. Silly me. But then those numbers of 4a belong to 4b as well.

 

30 basic versus 20 armor.

If armor is not random, to expect: 5 to 10.

If armor is random, to expect: 5 to 20. Thus values above 10.

 

30 piercing versus 20 armor.

Well, that test has notching to do with the randomness of armor. Since Piercing is not influenced by Armor.

No need for doing those tests again.

 

Thank you for pointing that out though. I will correct 4a and 4b into 4.

And reconfirming resulted in reconfirmed. :)

 

If you still feel like the numbers are influenced in a way. Please tell me in what way you think they are influenced. Because then I can conduct another test for that.

[MrFlibble]

If you still feel like the numbers are influenced in a way. Please tell me in what way you think they are influenced. Because then I can conduct another test for that.

Well, my only thought is that maybe the game's damage calculating algorithms are designed in such way that they work for the existing ranges of values that are actually used by game units, but not for values like basic damage = 0 or 1 (IIRC all units have damage values greater than one).

Perhaps another possible cause for the values not being what the official equation predicts could be the way this is coded? Remember how in Dune II v1.07 building repair costs get much lower than expected because of the very approximately rounded numbers? Maybe something like that happens here as well? I mean, the coders did take the Combat Equation as their starting point but the actual in-game effects are different due to how the calculation algorithm is implemented?

 

The damage displayed in the unit box is in most times different than what it actually is.

Examples;

- Footman has 6 basic, 3 piercing. The box shows 2-9. But it has to be 2-10.

- Knight has 8 basic, 4 piercing. The box shows 2-12. Which is correct.

- Archer has 3 basic, 6 piercing. The box shows 3-9. But it has to be 3-10.

Thus the maximum damage is displayed wrong in most cases. When this one is an odd number, add 1 for the correct number.

Hmm, the values you quote are in accord with the Combat Equation - at least, in what concerns maximum damage versus 0 armour: for Footman it would be (6 - 0) + 3 = 9, and for the Elven Archer (3 - 0) + 6 = 9. But I assume both of them can actually inflict a maximum of 10 damage in the game, right?

Another question is the minimal damage. Blizzard's old site says that minimal damage is exactly 50% of maximum damage. In fact, this means that a Footman's minimal damage against 0 armour would be 4.5, presumably rounded up to 5 (maybe this explains the max being 10 rather than 9?).

But where does the min damage of 2 come from? Is that for the cases when the target's armour is greater than the Footman's basic damage, thus leaving him with piercing armour only? Then an unupgraded Footman's max damage against a building would be 3 and min damage would be 1.5 rounded up to 2?

I guess this kind of makes sense, right?

[X3M]

Well, my only thought is that maybe the game's damage calculating algorithms are designed in such way that they work for the existing ranges of values that are actually used by game units, but not for values like basic damage = 0 or 1 (IIRC all units have damage values greater than one).

Perhaps another possible cause for the values not being what the official equation predicts could be the way this is coded? Remember how in Dune II v1.07 building repair costs get much lower than expected because of the very approximately rounded numbers? Maybe something like that happens here as well? I mean, the coders did take the Combat Equation as their starting point but the actual in-game effects are different due to how the calculation algorithm is implemented?

In Warcraft 2; 0 Damage = 0 Damage ;) I also conducted that test.

In Starcraft; 0 Damage = 0-1 Damage. 0,5 on average. Always following the sequence 0101010101 but starting somewhere randomly.

Both events only happen when 0 Damage is the end result.

 

To bad that my SC CD exploded in my PC. Or else I could perform a test on the hydra, seeing if it randomly does 7 or 8 damage on a vulture. (75% of 10 = 7,5)

 

Hmmm, saying that the event only happens when I apply 1 damage is a good remark then.

I will conduct some tests with 3 or 4 damage, either piercing or basic. Results will be posted soon.

 

 

Hmm, the values you quote are in accord with the Combat Equation - at least, in what concerns maximum damage versus 0 armour: for Footman it would be (6 - 0) + 3 = 9, and for the Elven Archer (3 - 0) + 6 = 9. But I assume both of them can actually inflict a maximum of 10 damage in the game, right?

Yes.

 

 

Another question is the minimal damage. Blizzard's old site says that minimal damage is exactly 50% of maximum damage. In fact, this means that a Footman's minimal damage against 0 armour would be 4.5, presumably rounded up to 5 (maybe this explains the max being 10 rather than 9?).

And that is indeed the reason.

 

 

But where does the min damage of 2 come from? Is that for the cases when the target's armour is greater than the Footman's basic damage, thus leaving him with piercing armour only? Then an unupgraded Footman's max damage against a building would be 3 and min damage would be 1.5 rounded up to 2?

The display box takes a value of 15 armor for calculating the minimum damage.

For units with basic damage of 15 and less. Basic damage is ignored and you get the damage only caused by piercing damage. In case of 3 piercing damage, your minimum is indeed 2. However, in the game your actual damage is 2-4, while the box displays 2-3.

For units with basic damage above 15. Basic damage divided by 2. Subtract 15 and round upwards.

Cannon Tower: 50 basic, 50%=25, but 15 is subtracted. Thus the cannon tower has 10-50 in the display box. The true damage against 0 armor is 25-50 and against structures (20 armor) is 15-30. An complete upgraded battleship (25 armor), the damage is 13-26.

Thus as you can see, 10 damage will never, ever happen with the cannon tower, even though it is displayed.

 

---

 

I conducted a balance test, where a number of footman take on a number of knights.

With this test I have a certain equal ROF. Testing against other units like the archer is still a no-no.

I tested with math on paper, but also in game to reconfirm.

I did assume that each unit is targeted by all others.

 

Here is the result by math:

Number of Knights get defeated when facing Number of Footman:

1 2

2 4

3 6

4 7

5 9

6 10

 

Whereas 3 against 5 gives a tie.

 

Of course, the Knights are better in game. There are several reasons:

 

- This due to different costs (600 G versus 800 G-100 W). However, the game does hint that 100 W equals 300 G. Thus the real costs of a Knight is 1100 G according to the game designers. (PS. Oil equals Gold).

Well, even though I have mentioned this difference of 600 versus 800. Keep in mind, it does take 3 times longer to gather 100 wood. Until, well, a lumber mill increases it to 125. While gold can only be increased to 120.

Thus after upgrading every wood is actually a factor 0,96 compared to gold.

Right? No wrong, wood is harder and harder to get. Since the distance increases to the drop off point. So where the balance is, I don't know. In general, wood still goes faster than gold. And gold is used more frequently when you reach the combat fase (no base buildings needed any more). It really depends on the player how expensive or cheap the knight really is.

 

I have seen players on youtube who actually win by not using knights at all. And somehow I do get the feeling that they do a better job too on the mission.

 

- Knights have more durability and damage, this is well balanced in open field combat (in other words, singularities against each other). But when 2 armies face each other while the forces are packed. Then Knights have an advantage. This is due to it's size. The size is 1, just like the footman. This always causes imbalance when using huge, packed armies.

During my research on other games, I have learned that the true worth of an unit is its initial cost (1100) divided by the basic unit cost (I take the footman for now, 600). = 1,8333. This factor gets rooted. = 1,354 And then we are left with a factor to multiply with the initial cost of 1100. = 1489,4 Let's say, 1500 Gold. They are weaker in a chaos attack. But still stronger in a planned attack.

 

However, Knights and Ogres eventually get to use magic. Thus you would expect them to be more expensive in general.

On the other hand, other units sort of ignore the armor of Knights and Ogres in a mid and late game (Archers and Axethrowers) Thus that effect needs to be taken into account as well.

Then again, food usage is the same for every unit as well. In that regard, a cheap unit actually needs a relatively more expensive preparation.

[X3M]

I conducted some tests again. The results reconfirm the formula.

 

Against 0 armor.

3 Basic gives damages of 2, 3 and 4

4 Basic gives damages of 2, 3 and 4

3 Piercing gives damages of 2, 3 and 4

4 Piercing gives damages of 2, 3 and 4

3 B and 3 P gives damage of 3, 4, 5 and 6

3 B and 4 P gives damage of 4, 5, 6, 7 and 8

4 B and 3 P gives damage of 4, 5, 6, 7 and 8

4 B and 4 P gives damage of 4, 5, 6, 7 and 8

 

Against 2 armor.

3 Basic gives damages of 1 and 2

4 Basic gives damages of 1 and 2

3 Piercing gives damages of 2, 3 and 4

4 Piercing gives damages of 2, 3 and 4

3 B and 3 P gives damage of 2, 3 and 4

3 B and 4 P gives damage of 3, 4, 5 and 6

4 B and 3 P gives damage of 3, 4, 5 and 6

4 B and 4 P gives damage of 3, 4, 5 and 6

 

These where short tests. I wonder if the lowest and highest damage value will occur less in certain situations.

Can a programmer tell me if it is possible to take a random number between 3 and 6 and letting every number occur equally?

Thus with 1000 tests, we get around:

250 times 3,

250 times 4,

250 times 5 and

250 times 6?

When rounding occurs, the outcome is different.

 

One of the sites says 1..2 x lowest damage (which was already rounded upwards). Thus I conclude that there is another rounding afterwards.

In that case, I may expect:

166 times 3,

333 times 4,

333 times 5 and

168 times 6?

 

Is that correct?

By testing, which will take a long time this time. I will know for sure. :)

However, life is taking over, thus this experiment is saved for another time.

[X3M]

With 4 basic on 0 armor. We would expect values of 2 to 4.

With that same statement as in the previous post. We would expect the following returns:

 

OR

333 times 2 (1)

333 times 3 (1)

333 times 4 (1)

 

OR

250 times 2 (1)

500 times 3 (2)

250 times 4 (1)

 

So I tested that, 200 shots.

Know what I got?

58 times 2 (1,1)

90 times 3 (1,7)

52 times 4 (1,0)

 

It is clearly pointing towards my rounding theory.

Now for the average roll, it doesn't really matter.

Thus the long a fight occurs, the less it matters.

However, for a veteran in calculating possibilities, this will be showing us a difference in game play. When trying to design a Sniper game.

 

If you give your opponents 4 Health. Then there is only 25% chance that they die after 1 shot, instead of 33%. The game will be 33% harder then previously expected.

If you give your opponents 3 Health. Then there is only 25% chance that they live after 1 shot, instead of 33%. The game will be 33% easier then previously expected.

Thus in a Sniper game, having units with 3 AND 4 Health will still be balanced. But the difference in those 2 units has become bigger. I am leaving that calculation to you guys :).

 

Remaining test: The ROF. I really am to lazy for that one :D.

[MrFlibble]

Not only that, but they take 15 armor into account when the basic damage goes above 15 armor. It would have been better to use 20 instead. Or even 25 since the Battleships can reach this armor.

Ballista has 80 B and 0 P. The box shows 25-80. Which insinuates that the basic damage is decided by random first. Then armor is subtracted before piercing is even added. If this was not insinuated but the armor is subtracted before a random number is chosen like it should be. Then we would get 33-80.

If the box would show the result of the real mechanic and using the 20 armor of the structures, then we would have 30-80.

The display box takes a value of 15 armor for calculating the minimum damage.

For units with basic damage of 15 and less. Basic damage is ignored and you get the damage only caused by piercing damage. In case of 3 piercing damage, your minimum is indeed 2. However, in the game your actual damage is 2-4, while the box displays 2-3.

For units with basic damage above 15. Basic damage divided by 2. Subtract 15 and round upwards.

Cannon Tower: 50 basic, 50%=25, but 15 is subtracted. Thus the cannon tower has 10-50 in the display box. The true damage against 0 armor is 25-50 and against structures (20 armor) is 15-30. An complete upgraded battleship (25 armor), the damage is 13-26.

Thus as you can see, 10 damage will never, ever happen with the cannon tower, even though it is displayed.

Actually if the Combat Equation as described at Blizzard's website is indeed true, then we have to assume that the game takes 30 as maximum armour when calculating minimum damage.

According to Blizzard, minimum damage would be calculated as follows:

minD = (B - maxA + P) / 2

For the Ballista example above, the equation would be

(80 - maxA + 0) / 2 = 25

Initially I assumed that maxA would be 20, the armour of buildings, but this equation suggests that apparently it is 30, not 20.

The same works for the Cannon Tower:

(50 - maxA + 0) / 2 = 10

Again the answer is 30.

The question though is whether there are any units or buildings in the game that have an armour rating of 30. And if there is and you change it to something else in the editor, is it going to affect stats display for units on that map? Or is 30 some hypothetical maximum armour that isn't actually used in the game?

[X3M]

Subtracting 30 then dividing with 2.

Or dividing with 2, then subtracting with 15.

Same thing.

 

Although, 30 does sound more logical since the maximum possible armor is 25.

 

Perhaps the 30 is a subtraction on splash damage?

With that I mean;

damage - 30 = splash damage.

 

I don't know for sure, but this could make sense since all the units with basic damage above 15 are splash damage units.

Except the submarine and giant turtle. Well, they are splash, however, it never can be used since they can only attack units on water.

[X3M]

Bloodlust effects.

Piercing damage is doubled, Basic damage is doubled, Upgrade damage is doubled.

Then they are used in the combat calculation.

 

Grunt.

6 Basic (+0), 3 Piercing. After Bloodlust, 12 Basic, 6 Piercing.

Against 0 Armor: 5-10 goes to 9-18 (+80%)

Against 2 Armor: 4-8 goes to 8-16 (+100%)

Against 4 Armor: 3-6 goes to 7-14 (+133,3%)

Against 6 Armor: 2-4 goes to 6-12 (+200%)

Against 8 Armor: 2-4 goes to 5-10 (+150%)

Against 20 Armor: 2-4 goes to 3-6 (+50%)

6 Basic (+4), 3 Piercing. After Bloodlust, 20 Basic, 6 Piercing.

Against 0 Armor: 7-14 goes to 13-26 (+85,7%)

Against 2 Armor: 6-12 goes to 12-24 (+100%)

Against 4 Armor: 5-10 goes to 11-22 (+120%)

Against 6 Armor: 4-8 goes to 10-20 (+150%)

Against 8 Armor: 3-6 goes to 9-18 (+200%)

Against 20 Armor: 2-4 goes to 3-6 (+50%)

 

Ogre

8 Basic (+0), 4 Piercing. After Bloodlust, 16 Basic, 8 Piercing.

Against 0 Armor: 6-12 goes to 12-24 (+100%)

Against 2 Armor: 5-10 goes to 11-22 (+120%)

Against 4 Armor: 4-8 goes to 10-20 (+150%)

Against 6 Armor: 3-6 goes to 9-18 (+200%)

Against 8 Armor: 2-4 goes to 8-16 (+300%)

Against 20 Armor: 2-4 goes to 4-8 (+100%)

8 Basic (+4), 4 Piercing. After Bloodlust, 24 Basic, 8 Piercing.

Against 0 Armor: 8-16 goes to 16-32 (+100%)

Against 2 Armor: 7-14 goes to 15-30 (+114,3%)

Against 4 Armor: 6-12 goes to 14-28 (+133,3%)

Against 6 Armor: 5-10 goes to 13-26 (+160%)

Against 8 Armor: 4-8 goes to 12-24 (+200%)

Against 20 Armor: 2-4 goes to 6-12 (+200%)

 

Now you know, why it is considered to be OP.

Really, the 50 magic consumption is way to low. It should have been at least 150.

 

Next time, the dragon + bloodlust + haste (ROF x 2)

[MrFlibble]

Perhaps the 30 is a subtraction on splash damage?

With that I mean;

damage - 30 = splash damage.

Heh, this is exactly what I thought :)

BTW, which unit/structure has 25 armour?

[X3M]

The Battleship/Juggernaught, after 2 armor upgrades, has 25 armor.

 

---

 

Dragon

 

The Dragon does twice it's damage. This due to the splash damage of the second explosion right behind the first explosion. There is no subtraction on this splash damage!

0 Basic, 16 Piercing. After Bloodlust, 0 Basic, 32 Piercing.

Against any Armor: 2x 8-16 goes to 2x 16-32 (+100%)

 

Haste effectively doubles the attack speed of only a few units. The Dragon is one of them.

This means that a Dragon that has Bloodlust and Haste is indeed 4 times stronger in it's attack.

 

---

 

Another hint in the game besides of the Gold versus Wood costs.

Is the health/damage ratio that the designers have chosen.

With the intention of bloodlust only doubling the average damage. And Healing of "doubling" a units health.

We can compare the magic cost of the 2. Where 50 is compared with 6 magic.

This means that there is supposed to be 50 health for every average damage of 6.

The ratio is 8,333.

 

My friend previously suggested me to look at the basic unit, namely the Footman/Grunt for this health/damage ratio. Which would be 10,000 in that case. But the Footman/Grunt are clearly meant as slightly meaty.

If both fighters are fully upgraded, the ratio is still 10,000. Upgrading your melee units is literally an arms race.

 

I always love to see units suicide. This gives already good numbers to work with. But those ROF are needed for accurate numbers, the proof is seen with the Ballista and Battleship. With the ROF there is a little simple trick to make the following numbers more accurate. (But I guess, I am a bit too lazy still :) )

Archer: 5,333 (4,762 after ALL upgrades)

Axethrower: 5,333 (5,556 after ALL upgrades)

Knight: 15,000

Demo: 8,889 (normal fight)(5,333 after ALL upgrades)

Mage: 8,000 (normal fight)

Ballista: 1,833

Peasant: 6,667

Gryphon: 4,167 (Damage is 12 average times 2)

Elven Destroyer: 5,333

Battleship: 1,538

Submarine: 1,600

 

If all the units really would try to suicide. The Submarine is the fastest. Although, the Submarine always needs at least 2 shots. When magic is researched: A mage can polymorph himself.

 

Now then, actually suicide is possible with certain units, even in a wasteland :)

Mage can actually blizzard himself.

In addition, the Death Knight can cast Whirlwind.

The Ogre can use Runes.

And the Demo squads simply explode.

 

---

 

Some more suggestions for adjusting blood lust (if they ever would)

Since Elven Archers can have a +3 upgrade, while the Axethrowers can not. How about making Bloodlust an upgrade as well, even though temporary. Of +1 or +2 to any unit. This can be piercing as well. I think this would be a fine balance, not?

It also makes sense since Bloodlust is not supposed to double the damage, but actually add 12%! instead. If you wonder how I got to that number ,let me know, I'll only explain then.

 

Another way to balance bloodlust is to combine this piercing bonus with a number of hits.

With 6 magic for healing 1 health.

With 50 magic for dealing 8,333 damage. With a +2 bonus, we could say, 5 hits worth for each bloodlust.

 

If only I could mod the game in this regard.

[X3M]

Just finished the human campaign beyond the dark portal. It was a long run. And often it left me by playing a map twice or three times before knowing what to do exactly.

A trick that I have discovered is that with the beginning of each mission. Your units basicly do a suicide run against the first line of opponents. This is not supposed to be a suicide run!

To counter this, simply put your game speed on the speed that it is supposed to have. Keep a close eye on any unit that does something stupid. Save them when needed. And rush in onto at least 1 opponents base after the suicide run. Killing that opponent makes each mission at least twice as easy as before. Don't go in if an opponent has cannon towers. And if there are Guard towers, simply use all your archers focus fire on that tower while your meat takes the hit. Keep your forces condensed.

 

I even managed to kill all opponents in the last mission by strategically retreating all units. You get overrun by about 30 grunts from teal and some axethrowers. And put the mage in front with blizzard on those massive grunts. As a result, 1 squad of footman and 1 squad of archers remained. Killing white in the south just seconds after teal has been annihilated. This time, also the hero's make their enter. Once you dealt with white, you got your third gold mine ready. But keep going!

Now go to the upper left corner where you dragon opponent is. He has some nice towers, but by now you have plenty of mages with blizzard. Make sure you kill him completely too. And you have your fourth gold mine.

Now you have a choice, if you are to weak now. You only have to deal with red. You can muster up your forces. If you have plenty of units ready like I did (keep pushing the barracks production) Then you have at least 2 extra squads by now fully upgraded. Just do a suicide run on red.

I bested this mission under 25 minutes. Only then I started using Khadar for destroying the portal.

 

---

 

Has anyone found a file that contains the cooldown of each weapon?

I have found out that guessing is not really an option.

[X3M]

I guess that guy wasn't really that off.

Cooldown/ROF of units:

25 peon/peasant, grunt/footman, ogre/knight/ogre, mage/paladin and goblin/dwarven demolition squad

35 skeleton

40 death knight/mage

60 guard towers

65 axethrower/archer

100 giant turtle/submarine

120 destroyers

150 cannon towers

190 gryphon rider/dragon

200 balista/catapult

230 battleship/juggernaut

 

Edit (8-4-2017): It seems that Guard towers are slightly faster than archers and axe throwers. Perhaps it is the animation of shooting/throwing, that makes the difference.

Edited April 8, 2017 by X3M

[X3M]

Well, the formula is almost complete for checking true balance.

But what bothers me is the cost weight of armor.

 

The hint that the games gives us that 1 piercing damage is worth 2 basic damage.

Armor is substracted from basic damage.

 

It makes no sense to bother with balancing different armor classes, let alone having a continieus range of armor classes. It would really depend on what you decide on before hand. And the piercing damage is simply a 100% against the highest armor. That being said, knowing that one damage would equal 8,333 health makes no sense either any more. Using that number is simply a guideline, so there is no real need for that.

 

---

 

This reminds me of my C&C style maps on Starcraft BroodWar. There I had infantry with 0 armor, tanks with 24 armor, super heavy armor and even "supreme infinite" armor class.

Each armor class had a score to multiply with the health. 0,5 was a minimum damage, thus this was regarded the ultimate piercing damage.

Each class would have a factor to multiply with the health.

 

Piercing of 0,5 would always penetrate even the highest armor class, thus was worth 0,5 whatever score all classes contained.

 

Now I don't know any more what the exact numbers where, but to use this class system. You need a worksheet. Depending on the number of classes that you want, the worksheet is different every time. If any one is interested in how it works, please ask.

 

---

 

To get an instant balanced map, I have to decide on armor classes first, take into account the upgrades. And then decide on the true damage values. All other values are easy to deal with or written in stone. The damage/health ratio is on to decide on myself.

 

Of course I need to take into account, magic effects. But screw magic, gona ban them again :D. I banned most energy based weapons in my C&C style map too!

 

I think I better do sloppy adjustments and simply go with a feeling from play testing.

[X3M]

Done analysing.

When not adjusting the statistics of all units. Yet wanting to have a slightly better balance. Implement the following price adjustments:

 

Normal maps // A lot of choke points in the map

gold equals 1/3th wood equals oil, meaning 50 wood is worth 150 gold or 150 oil, adjust for personal taste if you like.

 

Peasant/Peon (gets cheaper)

350 gold // 300 gold

 

Knight/Ogre/Paladin/OgreMage

1000 gold, 100 wood // 1200 gold, 100 wood

 

Balista/Catapult

1250 gold, 400 wood // 1600 gold, 500 wood

 

Mage/Death Knight

1450 gold // 1700 gold

 

Battleship/Juggernaught

1100 gold, 550 wood, 1100 oil // 1200 gold, 600 wood, 1200 oil

 

Submarine/Sea Turtle (gets cheaper)

750 gold, 150 wood, 900 oil // 750 gold, 150 wood, 850 oil

 

Gryphon Rider/Dragon

2550 gold // air is always open, thus only medium adjustment

 

Flying Machine/Zeppelin

350 gold, 100 wood // air is always open, thus only medium adjustment

 

Guard Tower Upgrade (gets cheaper)

450 gold, 150 wood // 350 gold, 150 wood

 

Cannon Tower Upgrade

1050 gold, 350 wood // 1250 gold, 350 wood

[X3M]

Apparently, there is magic and other techniques that doesn't give you points. So if you are a score lover, don't use these techniques. But fight bravely.

 

No points are given when you kill with:

- Runes.

- Polymorph.

- Hovering over destroyers with air (mostly eye of killrogg), such that he sinks his own ships.

- Forcing the enemy to catapult his own troops. This one is dodgeable.

- Forcing the enemy to use his air on his own troops. This one is slightly dodgeable, but chances are low.

- Of course killing your own troops.

 

If there is other magic or  that has the same problem, I will expand this list, I suspect:

- Kaboom!

- Fireball

- blizzard/death and decay

- death coil

- Whirlwind

- flame shield

- under construction, the peasant/peon within might not be counted.

- filled transport, the units within might not be counted. But how to test this?

 

---

 

PS. Transport still sailing around? They don't have to die for the victory to occur.

[Nyerguds]

- Hovering over destroyers with air (mostly eye of killrogg), such that he sinks his own ships.

I never knew this worked! I assumed their arrows would only damage the air unit then xD

[X3M]

The destroyer in warcraft 2 uses a cannon.

 

I figured how to get the enemy to use a transport for only 1 unit.

Now to figure how to kill the transport while the 1 unit is in it. I think reducing the health to 1 would do :D.

[Nyerguds]

ohh, right. The Destroyer's own cannons. I assumed it worked for all weapons.

Honestly never used that as an actual tactic to take out their ships xD

[X3M]

Well, it is the only weapon that has this effect. Due to splash.

It also works around. If you position a ground unit under his air. And the destroyer wants to shoot your ground unit. The air unit of him gets hit as well.

 

Dragons and Gryphons also have this effect. But they do not target other air except for dragons and gryphons. Thus you cannot use an eye of killrogg. Ground units on the other hand still get targeted. I once did this in multiplayer: luring 2 dragons to his peons. While I only had a paladin stationed there. His own splash, killed the majority of his peons. :)