CP Calculation

Hey all, I’m still collecting hero data to try and refine the equation. I will hopefully start to have more time on my hands to try and figure one out, but in the meantime, I have tabs on my original Hero Data spreadsheet for filling in data points if anyone would like to add to it.

Guys…I think I finally did it!
@Liliac @Garanwyn

Final equation for 4*:
FLOOR(((((CEILING(((AttributeMax-AttributeMin)/EventsTotal*20)+AttributeMin,0.1))-AttributeMin)/20)*EventCurrent+AttributeMin),1)

Final equation for 5*:
FLOOR(((((CEILING(((AttributeMax-AttributeMin)/EventsTotal*25)+AttributeMin,0.1))-AttributeMin)/25)*EventCurrent+AttributeMin),1)

I am going to assume for 3* it’s 15 (instead of 20/25). Think it might be a multiple of rarity, so Rarity*5. Haven’t verified but if that’s the case, then the general equation would be:

FLOOR(((((CEILING(((AttributeMax-AttributeMin)/EventsTotal*(Rarity * 5))+AttributeMin,0.1))-AttributeMin)/(Rarity * 5))*EventCurrent+AttributeMin),1)

This has taken so long, but I am certain this is it. If you check the spreadsheet I’ve posted, then you will see the equation in action.

edit: does not apply to 3* heroes. I’m so disappointed, but it’s back to the drawing board for those.

3 Likes

I’m wondering if anyone has updated this given some of the new heroes.

I’d say there’s something off when we compare Noor and Telluria max stats
Noor Telluria
Power 774 772
Atk 704 613
Def 782 832
Health 1314 1425

Specials… well enough said that it greatly favors Telluria.

How can Noor have a higher power ranking when the only advantage is a +91 Atk?

Because ATK factor is greater than DEF and HP ones. Did the math and it’s correct.

Interesting, but it makes the Power calculation very biased towards attack when a +91 Atk outweighs +50 Def and +111 Health. The fact that the type of special doesn’t have any consideration is a bit of a blind spot.

Discussion is here

There’s actually another topic where that has been discussed. Power doesn’t really reflect efectiveness of the hero, and doing that would be very difficult to calculate, I feel.

1 Like

I just used the criteria explained to calculate a 5 star domitia hero in the first level, that is, 0.
have A388 D363 H 723
According to the calculations the result gave me 373.66 nothing is further from reality the hero’s letter says that his power is 433 query where is the fault ???

Your calculation seems to be missing the “base CP for stars”.

I get your result of 373 if I leave the base CP out:
0.35 x 388 + 0.28 x 363 + 0.14 x 723 + (8-1) x 5 = 373

The correct calculation should give 463:
90 + 0.35 x 388 + 0.28 x 363 + 0.14 x 723 + (8-1) x 5 = 463

Is the 433 a typo?

I happen to have a Domitia at 1-1, she has 397 CP. The formula gives the same result:
90 + 0.35 x 352 + 0.28 x 329 + 0.14 x 655 + (1-1) x 5 = 397

So your Domitia seems to be at something like 1-25, right?

for the formula, what are the values 0.35,0.28,0.14…what do they mean?

As we’re getting a balance update that is supposed to improve the situation with heroes being too powerful, it might be a good time to revisit the CP formula and attempt to improve it.
(Actually, I thought about this only after I got matched with a top defense team in the raid tournament this Wednesday and lost miserably against the board AND that defense team. In normal raid I would totally skip such a team…)

Since the beginning of time (when there were only S1 heroes) the formula seems to have been:
CP = (base of CP for stars) + floor(attack x 0.35 + defense x 0.28 + health x 0.14) + (special level – 1) x 5 + (Talent tree nodes) x 5
(Ok, the talent tree node thing only was added with the arrival of emblems, but the rest of the formula should have been valid before that)

Nowadays, when looking at a hero’s power, we have
a) stats - already included
b) special skill level - already included
c) combination of speed and special skill - already included, but needs to be updated
d) family bonuses - missing
e) passive skills - missing
f) mana generation bonuses from costumes - missing
g) synergies with other heroes - unclear on how to include
h) board luck and tactics - can’t include
i) lb2 passive - missing
j) super class talent - missing
k) relevant things missing in this list - who knows?

a) Stats are already considered in the formula. The values of 0.35 for attack, 0.28 for defense and 0.14 for HP seem kind of arbitrary, but probably were selected by SG based on their initial raid battle simulations with S1 heroes.

For the calculated power this then means that 400 attack is equal to 500 defense or 1000 HP (140 = 400 x 0.35 = 500 x 0.28 = 1000 x 0.14) and 800 attack is equal to 1000 defense or 2000 HP.
What does that mean? The answer is: yes :man_shrugging::wink:
These values seem to work ok, though.

b) Same situation for the special skill level. The 5 CP per skill level seem to work ok, and add a total of 35 CP

c) The combination of speed and special skill is included in the fixed value of “base CP for stars”, which means that the speed and specials skills of the same star are considered to be similarly powerful. This may indeed have been the case when there were only S1 heroes, but things have changed a lot since then.

1* Base CP: 0
2* Base CP: 10
3* Base CP: 30
4* Base CP: 50
5* Base CP: 90

A regularly maxed Kelile’s special has an attack power of 2160 (= 675 attack x 320%), whereas a regularly maxed Marjana’s special has an attack power of 3064 (= 669 attack x 458%), which is 41.85% stronger. The base CP for 5* is 90 CP, for 4* it is 50 CP.
=> So SG deemed approx. 40% more special skill power to be worth approx. 40 CP.

There seem to have been no suggestions on how to take the increased power of speed and special skill of newer heroes into consideration. So up to now, this increased power is completely missing for every non-S1 hero.
=> any improvement will be an improvement

=> Without any better idea we could consider the additional CP for the stronger speed+special skills by applying the same CP amount as from the stats increase compared to the average classic S1 CP. Not accurate, but imho better than nothing.
=> A regularly maxed hero with 795 CP has 40 CP more than the average classic S1 (CP=755), so with this suggestion the hero would get another 40 CP for their special and then have 835 CP_new (open for fine-tuning)

d) The family bonuses add useful power like increased stats, mana generation, healing, etc. Some families have 2 bonuses, like slayers getting a stack (1) for HoT and (2) for mana generation.
We could apply a certain CP number per bonus, for example 10 CP per family bonus (open for fine-tuning)

e) Passive skills add considerable power to heroes. Some heroes have more than 1 passive skill. Some of them are undispellable, which makes them even more powerful.
We could apply a certain CP number per passive, and double this number, if a passive is undispellable, for example 15 CP per passive and also per undispellable passive. (open for fine-tuning)

f) The mana generation bonuses from costumes allow heroes to potentially charge faster, which can make a considerable difference in power. Some heroes get +1%, some get +5% mana generation.
We could apply a CP number by multiplying the mana generation bonus by 3, for example. (open for fine-tuning)

g) Synergies with other heroes depend on the available team members and are different for each player. Imho this can’t be put into a number.

h) Board luck and tactics depend on RNG and vary from player to player. No CP number for this.

i) The lb2 passives could be treated like the other passives, so we could add 15 CP. (open for fine-tuning)

j) The superior class talents could be treated like the other passives, so we could add 15 CP. (open for fine-tuning)

k) Nothing here, for now.

For the classic S1 heroes the current CP formula should remain.
For all other heroes the new CP formula could then be something like this:

CP_new = (base of CP for stars) + floor(attack x 0.35 + defense x 0.28 + health x 0.14) + floor(attack_at_regular_max x 0.35 + defense_at_regular_max x 0.28 + health_at_regular_max x 0.14) – (average power of S1 for stars) + (special level – 1) x 5 + (Talent tree nodes) x 5 + 15 x (number of passive skills + number of undispellable passive skills) + 10 x (number of family bonuses) + 3 x (mana generation bonus % from costume)

average power of S1 for stars = 378 for 3*, 557 for 4* and 630 for 5*
(that’s the average CP values minus the 35 from special skill level and minus the base CP for stars:
630 = 755 - 35 - 90
557 = 642 - 35 - 50
378 = 443 - 35 - 30)

At present, the suggested factors are just a gut feeling. Maybe somebody can do some fine-tuning?
Also, things might still be missing in the new formula.

Big question 1: Is this actually a better approach for the CP calculation and should we continue with this?
Big question 2: If this or something similar is a better approach, how do we convince SG to implement it?

edit: added i) lb2 passive and corrected the new formula for the CP for speed+special skill to be fixed to the stats for a regular maxed hero.
edit 2: added k) the superior class talent

1 Like

Apologies for resurrecting an old thread but…

…was really useful but these days things have moved on (boy, they really have moved on!)

I understand the mechanics of costume bonuses and emblems but haven’t yet figured how to calculate the CP of limit broken heroes.

Anyone out there can give an insight as to how the additional CP is calculated after limit breaks?

TIA.

EDIT: I think I’ve figured most of it out using info from

It’s complicated due to the order things are applied and some calculations seem to be truncated(FLOOR) and others are rounded. Also seemingly arbitrary constants which are not whole number values i.e.+3.2%.

Finding info is made harder by Limit Break or Level Break, one word or two, hyphenated or not and even the occasional phonetic spelling of brake.

Just have to see what happens when I actual gain one of these new levels rather than just activating them.

2 Likes