The theta parameter doesn’t swing the damage a ton in practice. But it does produce some variation.
1 Like
Hello, i tried to make some comparisons myself in a raid and it appears that in the same raid (not war), the +20% defense boost for defense team is sometimes applied and sometimes not applied. Have someone encountered the same kind of results ?
I compare my Caedmon’s special skill multiple times with and without Boldtusk’s +48% attack buff against Dark Lord. But it only increase from around 450 to 550… 
I wouldnt expect much more to be honnest.
An incredibly good piece of work! Thank you
BT’s buff moves Caedmon’s special damage from 345% to 345%+48%=393%. That is a 13.9% increase in pre-exponentiation attack power. Putting that through the exponential gives a 19.2% damage increase.
550 is 122% of 450, so it sure looks about right to me.
3 Likes
Isn’t +48% is increase in attack stat, not special skill power?
All the multipliers that aren’t troops get added together before multiplication.
My Caedmon has 714 base attack.
The troop give 20% attack bonus.
Caedmon attack without buff = 714 × 120% = 856
Boldtusk buff +48% attack = 48% × 856 = 410… It’s 406 in reality due to rounding.
With 856 attack stat, Caedmon tile hit Kappa around 50 damage while the special hit for around 600 damage.
With 856 + 406 = 1262 attack stat, Caedmon tile hit Kappa around 85 damage while the special hit around 700 damage.
It seems you are correct when you said that the 48% were added to %damage. But that would make it as a bug as it is against the attack stat displayed.
2 Likes
A bug is a thing which is unintended behavior. I’m confident this is the intended behavior.
The display shows the state of the hero at present, not the state under activating a special. If this display were updated to show the state under special, it would momentarily show an attack stat that was much higher.
The problem, as I see it, is the way the information is being displayed, and the conclusions that display leads people to draw.
I think SG should simply publish the damage equation, since they’ve said before that it’s not a secret.
5 Likes
I think the problem is they would assume that the “468% damage” combined multiplicatively with that attack stat in some way.
But the whole way of talking about “attack” and “damage” on the cards is intrinsically misleading. The actual damage done is never increased by 468% over anything when Joon’s special fires.
The “Attack buff” does not combine in any intuitive way with that either.
Essentially, what happens is this:
(Attack stat)*(468% + Attack buff %) = (Attack stat)*468% + (Attack stat)*(Attack buff %)
We can also factor this as:
((Attack Stat)/(Defense Stat))^1.35 * (468% + Attack Buff %)^1.35
But again, it’s not a 468% increase in damage. Nor is the fact that the “attack buff” is a linear increase in the special damage % an obvious or intuitive fact.
2 Likes
For Titans, the primary damage driver is tile damage. Because the multiplier for a regular attack is 100%, attack buffs for tiles would be more effective than attack buffs for specials.
4 Likes
Sure, but that doesn’t address defense debuffs vs. attack buffs. My point is, a defense debuff applies to damage done by both tiles and specials, using the same % (right?). An attack buff applies to tiles more heavily than it does to specials.
G, sorry for not moving this over to the damage calc thread. I don’t know how to quote someone’s comment from one thread in another, or that’s what I’d have done here. =-[
2 Likes
I’ve shifted us over. Apparently I failed to do it correctly myself earlier 
2 Likes
It also means sniper will get less benefit from attack buffs.
When is the next AMA?
1 Like
I have not thinking of this mattter at the last AMA… the OP of this thread who originally research the damage formula seems no longer actively researching it so I guess I have to wait for the next AMA to get some clarity…
Special Skills =?
Tiles =?
Hi, can someone help me find the damage calculator sheet?
@KillKing050 Thanks, that was translated almost 200 posts ago. 
@ahriman512 The damage calculation is found in the very first post: