Damage Calculation

You just offered a data scientist more data. Of course I would like more data.

In fact, here’s a Google Form you can fill out to make data entry easier. I’ll try to put together a thread on this when I get a chance, to get as many people adding data as possible.


Awesome, thank you. I will try to fill whenever I am a little free. I am sure every data counts :smiley:

1 Like

Here is some data

([Raw Data] Data log for tile damage mono yellow 4,185 ATT attack team versus mono purple 1* to 4* defense team)


Attackers: Kiril & Grimm (Att=(725 + 844)x1.1=1726)
In the example uses 02 blue heroes, Grim and Kiril x1.1
If I use 5 heroes of the same color does the formula continue x1.1?

1 Like

Perhaps I’m missing something, but I don’t see the Special Multiplier… is this Tile damage?

Yes. This is tile damage.

1 Like

Thank you, but the task I’m working on is Special damage, to determine what the Special Modifier actually does.


Would someone please explain the exact mathematical combat damage algorithm to me please? I’m constantly encountering situations where I’ll have a special attack ready and so will one of the enemy who has only a sliver of health left. So I figure I’ll use my special attack on a healthier enemy and then finish off the charged up enemy with a puzzle attack. Only for that bastard to survive the puzzle attack with <20 hp and trash my party with his special attack all because I didn’t understand how to mathematically take my attack rating and troop rating and his defense rating and estimate the resulting puzzle damage! I’m sick of E&P taking advantage of my ignorance!

1 Like

What guvnor posted.

Not exact, but it’s pretty close.


The reason why I use tile damage, is more data points.

@Garanwyn, @Kerridoc, myself and others strongly suspect a soft cap on the attack stat. Without the mechanics of the soft cap, the actual random variable is unknowable since variation can be attributed to both the change in soft cap and the random damage.

(Search results for 'Attack stat soft cap ' - Empires & Puzzles Community Forum)

No, I get that, but there’s a meaningful (and significant) difference between whether a special multiplier goes inside or outside of the exponent. For a given Attack, Defense, and θ, if the special multiplier modifies the attack, Lianna’s special will do 906.8% the damage that three tiles would, while if it modifies the damage (outside the exponent), it only does 512%, which is quite a significant difference.

And, because it is such a significant difference, it shouldn’t require as many data points to demonstrate the skew.

1 Like

Idk if this is the correct place to post this but I can’t help but wonder what’s happening with the algorithms when raiding. I just went into the tournament with 3 green to take down 2 blue. I counted & got roughly 5 tiles of green so in essence I couldn’t light up my green hero’s & they took me down :woman_facepalming:t2: But I’ve noticed a change in tiles when attacking the Titan as well. Are there changes being made? I can’t get tiles in colors I need anymore making it nearly impossible to beat my opponent, or at least, for the most part, close to impossible. What’s happening? Anyone know?

This is a good point.

If you can figure it out, awesome.

But there are many other factors that get really weird you might have too many hidden design decisions ( see notes)

I really really really wish SGG was more transparent about damage calculation. It would make checking for bugs, and planning team tactics, much easier.


1 Like

This comes up often.


(Color Stacking Fairness Project)

and 44 other linked topics.

Not even sure what the current merge topic.


Nope, you simply add their attacks together. The reason this 1.1 appears in this specific case is because these tiles are part of a combo. You can see in the picture “combo 2; attaque x 1.1”

1 Like

That’s great, I’ll try to check my videos here to contribute with some data.

1 Like


Can you please discard my last entry (892 Atk, 904 Def, 131%, 224 damage)?
Reason: I checked the wrong Leonidas :sweat:

I’m working on it, doing some tests.