Major Imbalances!!!

[Karawasa]

Jolin012 has brought to light some major imbalances within the map. I will first layout the problem, and then propose some changes.

Problem: Super Triples are too strong relative to Super Duals. Dual level 3 seems too weak.

[*]Foundry Tower ~ 100% increased damage[/*:m:19z3d5hu]

[*]Waterfall Tower ~ 100% increased damage[/*:m:19z3d5hu]

[*]Hydrochloric Tower ~ 100% increased damage (15 armor is equal to about 50% damage reduction, removing that is basically doubling damage)[/*:m:19z3d5hu]

[*]Monsoon (etc.) Tower ~ 80% increased damage (40% slow means 80% more time spent getting attacked)[/*:m:19z3d5hu]

[*]Witchcraft Tower (for comparison) ~ 40% increased damage[/*:m:19z3d5hu]

As you can see, armor reducers and slows are MUCH too powerful. It also appears that level 3 dual towers are too weak for their cost and limitation.

Solution: Nerf armor reducers and slows, buff duals.

[*]Double armor on all difficulties (6/12/18/24/30), but modify armor tables so damage reduction on creeps remains the same[/*:m:19z3d5hu]

[*]Reduce slow on 4 AoE Super Triples from 20/40% to 15/30%[/*:m:19z3d5hu]

[*]Reduce cost of duals, from 500/2000/8000 (total cost) to 500/1625/5281[/*:m:19z3d5hu]

[*]Make dual damage gain from upgrade go from 5/5/7 to 5/5/5 (compensation for reduced cost)[/*:m:19z3d5hu]

Thoughts?

[0rb3r]

Double armor on all difficulties (6/12/18/24/30), but modify armor tables so damage reduction on creeps remains the same

Reduce slow on 4 AoE Super Triples from 20/40% to 15/30%

Reduce cost of duals, from 500/2000/8000 (total cost) to 500/1625/5281

Make dual damage gain from upgrade go from 5/5/7 to 5/5/5 (compensation for reduced cost)

Good point, i mostly dont use lvl3 duals, hmm...

I think the main problem actually with the lvl3 duals is their exorbitant price, reducing it would be a good balancing change, they are just too expensive for the fact that they dont have an aoe.

Corrosion/Hydrochloric should have -3/-9 armor reduction and Enchantment/Incantation should have -5/-15 armor. I think that the difference between lvl1 and lvl2 of these triples is just too high. Doubling armor like you purposed would make the the lvl1 triples pretty worthless (the difference on a creep having 30 armor and 27 armor after being hit by a corrosion tower would be how much? 5% less damage reduction?), i think it's more correct this way (quoted: -9 Armor means about a 30% less armor reduction, receiving 60% more damage).

Nerfing a bit triples slow and the damage of lvl3 duals sounds good to me, the actual windstorm and supernova are a bit imba atm.

And Voodoo/Witchcraft needs a buff, maybe 30%/60% extra damage should be equal to the new slow value of triples (quoted: 30% slow means 60% more time spent getting attacked) and a small buff on aoe (2 voodoos should be enough to maledict all creeps).

This way all should fit in all corners.

[Guest Sancdar]

Your calculations don't work out at all.

Monsoon (etc.) Tower ~ 80% increased damage (40% slow means 80% more time spent getting attacked)

Suppose you have to go 100 miles, and you travel at 100 mph. It takes you 60 minutes to travel those 100 miles. If you were traveling at 60 mph (40% slower), it would take you 100 minutes to go 100 miles, which is a 2/3 (67%) increase in time, not an 80% increase. Slows increase damage time by (1 - 1/[1-slow]), not by (2[slow]); 99.9999% slow sure isn't just doubling the time you're attacked. The main draw of AoE slows for damage are that they tend to increase clumping substantially. That's why any AoE slow build except for one that ONLY uses Monsoon has an incredibly hard time with Mech levels, because you lose the clumps. The average amount of slowing isn't too much worse, but your damage ability drops a ton.

Hydrochloric Tower ~ 100% increased damage (15 armor is equal to about 50% damage reduction, removing that is basically doubling damage)

On VH this is close to true. Going from 47% to 0% reduction is only an 89% increase in damage, but it's still up there. However, on VE you're going from 15% reduction to 52% increase, only a 79% increase in damage. Increased amounts of negative armor have diminishing returns, while decreased amounts of positive armor always helps the attacker the same amount. It's hard to word that in an understandable way, but basically armor reducers only work at full effectiveness when the total armor of the target remains non-negative.

Witchcraft Tower (for comparison) ~ 40% increased damage

If Witchcraft's tooltip is accurate, it pings twice for damage. The first time it does 40% increase on 2.5 seconds of damage, the second time it does 40% increase on 5 seconds of damage, including the 40% increase the first time. If you have 100 damage done per second, that's 500 base + 250*.4 from ping1 on base + 500*.4 from ping2 on base + 250*.4*.4 from ping2 on ping1, for a total of 840 damage, a 68% increase.

Percent Damage Increase per Thousand Gold Spent:

Monsoon: 10.9865964% (area)
Hydrochloric: between 13.0191167% and 14.6671061% (area)
Enchantment: between 16.6737757% and somewhat higher (multiple target)
Witchcraft: 11.2063283% (area)
Well/Blacksmith: 12.5% (multiple target)

So, the numbers aren't anywhere near as bad as you thought, but the damage buffs are still not as much better efficiency for their cost as they should be compared to the triples. I'd suggest changing Acid to -4/-12 and Enchantment to -7/-21 for now. Right now the first levels of these towers aren't that great. Enchantment especially, since it is only a single-target reduction. I do think that Dual3s need a cost nerf though, but not back down to 5k. Maybe like 7k? It's just a huge investment right now that's hard to make if you aren't able to sell.

[Karawasa]

Thank you for pointing out the inaccuracies in calculation, I did this rather hastily (but I used ~ to represent about not actual). At any rate, the conclusion still stands.

It is clear that simply nerfing the values of armor redux is the easier way to go (as opposed to modifying armor value and armor tables). I like Orber's 3/9 for Corrosion, and 5/15 for Enchantment.

It is debatable as to whether slows need to be nerfed as well. How does the proposed 15/30% sound? It may be noteworthy to know that this change would allow two level 2 slows to work in tandem, as opposed to being wasted. Maximum slow in-game is 66% of base, compare 60% combined with 80% combined.

As for the cost redux of duals, everyone agrees it is needed. The proposed changes would slightly increase DPG, but more importantly, will make them more accessible. 7000k is too high for a tower that you have to dedicate 3 in 2 elements for, in my opinion.

[Twilice]

Tried a game of short mode, random and chaos yesterday (very hard ofcourse)

Mire and Sniper and Hydro towers seems balanced. but I didn't got level 3 dual as I wanted to try that. But I don't think it would have been worth it. Poison......... just suck now; seems bugged somehow.

Sniper should have a little bit faster attack,

hydro should have a faster upp and down animation but instead a little higher chance or better speed AND it still looks like a water tower, make it attack water!!!! (most important make it attack water!)

The build had very much earth so were pretty weak vs nature. But thanks to mire towers and mass focused darkness saved me. I died at double mechs at level 55 and 56. but the thing that made me a bit mad, I had mayor earth damage, but level 55 were nature and I leaked 8 or something. But the next were normal and i like leaked 15 or something like that. ( i had lost a bit of lives before with healing levels. specially nature healing)

[Guest Sancdar]

Thank you for pointing out the inaccuracies in calculation, I did this rather hastily (but I used ~ to represent about not actual). At any rate, the conclusion still stands.

Does it? I don't think that being slightly more powerful than other towers necessitates the nerf stick.

It is clear that simply nerfing the values of armor redux is the easier way to go (as opposed to modifying armor value and armor tables). I like Orber's 3/9 for Corrosion, and 5/15 for Enchantment.

The first level of Acid seems pretty much useless to me, as does Enchantment.

It is debatable as to whether slows need to be nerfed as well. How does the proposed 15/30% sound? It may be noteworthy to know that this change would allow two level 2 slows to work in tandem, as opposed to being wasted. Maximum slow in-game is 66% of base, compare 60% combined with 80% combined.

You already probably are going to build for 2 AoE slows if you want a slowing build, why force people to do it by making a single AoE slow weak?

[0rb3r]

Does it? I don't think that being slightly more powerful than other towers necessitates the nerf stick.

You are right at this point, but its a balancing try worth.

We have 4 main slowers, Roots, Sludge, Storm and Nova. On 5 element builds you can combine 3 slow-triple types to reach max slow, so the first tiers should add together 66%, max slow. On 4 element builds you can mostly build 2 lvl2 slow triples excepting the case is that the elements you miss are L+D or any of them, in which case you get Well and Blacksmith to compensate. 2 lvl2 Triples with slow should add 66% max slow. I exclude 3 element builds as they are rare and mostly not focused on support but more on single damage and duals.

I'm getting to the point; I think, even if it may sounds a bit weird, that the values for tier 1/2 of slowing triples should be 22% slow and 33% slow. First to encourage building many and secondly because it gets closer to game practice and less to pure mathematics. The actual 40% slow is too much for a lvl2 triple, just think how valuable windstorm is for its price.

Quoting your calculations:

(1 - 1/[1-0,22]) = 28% more damage on 22% slow --> cost 1517 gold

(1 - 1/[1-0,33]) = 51% more damage on 33% slow --> cost 6068 gold

(1 - 1/[1-0,44]) = 79% more damage on 44% slow (2 lvl1 triples) --> cost 3034 gold

(1 - 1/[1-0,66]) = 200% more damage on 66% slow (3 lvl1 triples, 2lvl2 triples) maximum. --> cost 4551, 12136 gold

Now, bearing in mind that the towers do damage as well as they have slow, and that the damage is about 5x higher on lvl2 triples (1x tier1 triple= 1 damage value, 1x tier2 triple= 5 damage value):

1 lvl1 triple: (cost 1517 gold/(1,28*1 damage value))= 1185 inverted efficiency per gold (the lower the better)

1 lvl2 triple: (cost 6068 gold/(1,51*5 damage value))= 803 inverted efficiency per gold

2 lvl1 triples: (cost 3034 gold/(1,79*2 damage value))= 847 inverted efficiency per gold

3 lvl1 triples: (cost 4551 gold/(3*3 damage value))= 505 inverted efficiency per gold

2 lvl2 triples: (cost 12136 gold/(3*10 damage value))= 404 inverted efficiency per gold

This value is just indicative for mentioned towers soloing, so, extra towers make more slow builds greatly more efficient then above stated ratio comparing i.ex. 1 lvl1 triple and 2 lvl2 triples, but it shows that the values i purpose for slow% are pretty fair comparing it to invested money.

The first level of Acid seems pretty much useless to me, as does Enchantment.

Here a real table of the armor strenth token on beta 7:

• Armor --> Damage%

0 --> 0%

3 --> -15%

6 --> -26%

9 --> -35%

12 --> -42%

15 --> -47%

0 --> 0%

-3 --> 15%

-6 --> 26%

-9 --> 35%

-12 --> 42%

-15 --> 47%

The problem is that the basic Wc3 formula makes the % of reduction increase, lower the higher the armor points are, so that a 100% reduction can only be reached on infinite armor points. If you watch closely on the list, you'll realize that -3 armor benefits hugely VE players and does not benefit much VH Players.

If i remember correctly this formula can be changed in the editor, and it should be changed to be something like this:

• Armor --> Damage%

0 --> 0%

3 --> -10%

6 --> -20%

9 --> -30%

12 --> -40%

15 --> -50%

30 --> -100%

The armor reduction would be this way not be affected by difficulty and thus provide always a same damage increase.

So, comparing it to the values of extra damage of slow for tier1/tier2 (28% extra damage on 22% slow, 51% extra damage on 33% slow), if this change is applied, corrosion tower should be -9/-15 armor for having the same extra damage which is 30%/50% (and -15/-25 armor for enchantment which is 66% stronger then corrosion).

You already probably are going to build for 2 AoE slows if you want a slowing build, why force people to do it by making a single AoE slow weak?

I think the fact that a tier 1 triple combined with a tier 2 triple is already enough to reach the top is a bad thing, specially knowing that the difference on extra damage is very high over 46% slow (85% extra damage) up to 66% (200% extra damage). Your calculations gave me the best reason to insist that it needs to be nerved, specially to reach top slow. Making it necessary to have 2 tier2 tripples to reach the top makes the game much better under the light of the facts.

[Guest Sancdar]

Here a real table of the armor strength token on beta 7:

• Armor --> Damage%

0 --> 0%

3 --> -15%

6 --> -26%

9 --> -35%

12 --> -42%

15 --> -47%

0 --> 0%

-3 --> 15%

-6 --> 26%

-9 --> 35%

-12 --> 42%

-15 --> 47%

The problem is that the basic Wc3 formula makes the % of reduction increase, lower the higher the armor points are, so that a 100% reduction can only be reached on infinite armor points. If you watch closely on the list, you'll realize that -3 armor benefits hugely VE players and does not benefit much VH Players.

Your values aren't correct, at least on the negative armor (-12 is -52% reduction, I don't know others), but I can still show the calculations for Acid/Hydrochloric on VH and VE. It's true that -3 armor benefits VE more than VH, but -15 doesn't.

VH:
   15 base armor = 47% reduction, or 53% damage taken
   12 acid armor =  42% reduction, or 58% damage taken
   0 hydrochloric armor = 0% reduction, or 100% damage taken
   acid damage / base damage = .58/.53 = 1.09433962 = ~9.5% increased damage
   hydrochloric / base = 1/.53 = 1.88679245

VE:
   3 base armor = 15% reduction, or 85% damage taken
   0 acid armor = 0% reduction, or 100% damage taken
   -12 hydro armor = -52% reduction, or 152% damage taken
   acid damage / base damage = 1/.85 = 1.17647059
   hydrochloric / base = 1.52/.85 = 1.78823529

The reason that Acid is better on VE than VH is that effective HP from armor is gained linearly, so the closer to zero positive armor the creeps are, the more reducing 1 point gains you proportionately. The reason that Hydrochloric is better on VH than VE is because effective HP isn't reduced linearly on armor values below zero, so the farther from zero negative armor you are, the less each further point of reduction matters. I'm not sure what the most efficient difficulty level is for the -15 armor on Hydrochloric, since I don't have time at the moment to find all the negative armor reduction values.

If i remember correctly this formula can be changed in the editor, and it should be changed to be something like this:

• Armor --> Damage%

0 --> 0%

3 --> -10%

6 --> -20%

9 --> -30%

12 --> -40%

15 --> -50%

30 --> -100%

The armor reduction would be this way not be affected by difficulty and thus provide always a same damage increase.

I don't think that's possible and it wouldn't work the way you think anyway. Going from 90% to 100% damage is less of an increase than going from 50% to 60%. The first is an 11% increase, the second a 20% increase. That would just make armor reduction more powerful on VH. The way armor works in WC3 is that it increases reduction on a logarithmic scale, which increases effective HP linearly. If you make armor reduction follow a linear scale, effective HP increases exponentially instead.

I think the fact that a tier 1 triple combined with a tier 2 triple is already enough to reach the top is a bad thing, specially knowing that the difference on extra damage is very high over 46% slow (85% extra damage) up to 66% (200% extra damage). Your calculations gave me the best reason to insist that it needs to be nerved, specially to reach top slow. Making it necessary to have 2 tier2 tripples to reach the top makes the game much better under the light of the facts.

Your calculations on how much gold you need to reach max slow currently makes me doubt your solution even more. I agree that needing 2 tier2 triples to max out is good, but maxing with 3 tier 1 triples is a bad thing, I think. My current suggestion is raising the max slow cap to 70% (a small increase), and changing AoE slow % to 20/35. Two maxed slows would cap out, but level 1 slows would still be effective to stack and have a noticeable effect by themselves. I think that 20% is really the lower bound for an AoE slow.

[0rb3r]

but maxing with 3 tier 1 triples is a bad thing, I think.

It is not, think about how often you get 5 elements on random but don't get enough of them up to level 2 to get good tier2 triples. You depend absolutely on supports like slow and armor reducers.

Btw, changing AoE slow % to 20/35 sounds pretty decent to me, I'm not convinced that max slow cap really needs to be increased to 70% but it sounds alltogether like a good purposal.

[Guest Sancdar]

Maxing slow with 2 tier2 triples is 8 element picks and 12k gold. Maxing slow with 3 tier1 triples is 5 element picks and 4.5k gold. How is that at all balanced?

Edit: random SHOULD be a challenge. If it's easy to get max slow with a widespread random pick, you'll just see the same couple strategies every game. I'd rather see a random that forces players to learn how to use uncommon towers effectively.

[Karawasa]

Increasing the slow percentage to 70% from 66% may not be a large change, but it will nevertheless impact the game and end up buffing slows. We all agree that level 1 slows should not be able to be utilized in the way that Sancdar said above. The crucial question, is whether 15% slow is too weak on its own. If it proves sufficient, then I believe that 15/30% would probably be the best solution. It will make it harder to achieve high slowing with only level 1 (4 vs. 3 towers needed), and it would also make it worth getting two level 2 slows.

[Guest Sancdar]

Increasing to 70% would set tier 2 slows farther apart from stacked tier 1 slows without making tier 1 slows suck (like Nova). You'd need 2 tier 2s to max out, but all 4 tier 1s. With 15/30 you'd need two tier 2s and a tier 1 and have some overslow, or 1 tier 2 and 3 tier 1s.