I Can confirm that turn 3 skills add together fully so 1.2 + 1.2 = 2.4x
I Can confirm that base skill ups add by multiplying 1.2 + 1.2 = 1.44x
Based on that i'd say 3 1.2x girls would be 1.73x (1.728 rounded up).
I can also say that having two 1.2 all the time girls and 2 1.2 at turn 3 girls in the same squad = close to 75% activation of skills for turns 1-2 and around 95% for turn 3.
Had a team setup like that for around 50 runs and that was the most common pattern.
I do not have Cymbidium to see how her 2x turn 1 skill effect others tho.
At the start of combat, Cymbidiums' skill is announced seperately from other activation rate effects, so I get the 1.2 activation rate anouncement from Geranium, and a seperate 2x activation rate announcement from Cymbidium. How those two combine "under the hood" I really can't say. Could be either 1.2 + 2 = 3.2 or 1.2 * 2 = 2.4. But the result is generally 4 or 5 skill activations on that first turn for me.
If it's pretty much always 4 or 5 activations, then it's likely the 3.2
Assuming average skill activation chance base is 28%:
3.2 * 28 = 89.6%
2.4 * 28 = 67.2%
If it was 2.4, you'd generally have 3 or 4 activations on the first turn:
2/3 rolled 5x and getting 5 activations would be 32/243 = 13.17% chance of it happening, roughly 1/7. For it to be a regular occurrence, you'd have to be absurdly lucky.
2/3 happening 4/5 times is about 32.9% chance of happening, still unlikely for it to be that frequent.
This is my alt, pretty much i experiment with a different management style in this one, ultimate hipster stuff but this is around my progress in 2 months (around geranium's event last week) or so, right now i am at the point where i only need to evolve and level my 5* stars, the order is not optimized by no means, its just ordered by total power (310,758), right now it can deal with ultimate stages to get the 3 medals, stronger than the numbers say definitively.
Last edited by HakuYowane; 10-07-2016 at 04:44 PM.
FKG: 053232470 Strength: 479,458
Confirmation here that Geranium and Cyrtanthus multiply their modifier. (Make sure both girls are showered with presents first, since it's based on affection).
Screenie:
I've just obtained evidence that says otherwise.
If it was multiplying, then the skill rates we should have seen would have been either 1.72 or 1.73 to round up the value. Instead, it appears that our individual 20% modifiers are adding on top of each other to get a final skill modifer of 1.6, or +60%.
Further and many more tests will have to be done to see how this will actually stack with Moonvine and Tachibana's own modifier skills. If it stacks additively (1.6 + 2.4) like Moonvine and Tachibana's, then a 25% skill rate will be enough to land 100% of the time. If it's multiplicative (1.6 * 2.4), then 26% would be enough. However, if it stacks additively like Justicia/Cyrtanthus/Geranium (0.6 + 2.4), then 33% will be needed for a 100% activation on turn 3.
As such, I shall be taking chunks of 100~200k damage out of your raid bosses in order to test this theory
Edit: Large additive (1.6 + 2.4) was busted ages ago, ever since Geranium + Moonvine/Tachibana combo was possible, forget that. Small additive (0.6 +2.4) and multiplicative (1.6 * 2.4) are what's left to test.
- - - Updated - - -
After 14 Raid Battles and 56 skill activations, I can say with at minimum, 98% certainty (it's confirmed, no way it's not after that) that the skill modifiers of Cyrtanthus/Geranium/Justicia multiply on top of additive bonus of Moonvine/Tachibana's x2.4 modifier.
30%(x1.4 * x2.4) was the used setup here. If the modifiers multiplied on top of each other, then all units with a 30% activation rate would have a guaranteed 100% rate on the third turn. If it followed the low, additive rule of stacking Cyrtanthus, Geranium and Justicia on top of each other and consequently, other skill modifiers, the 3rd turn probability would have been 84% [30%(x0.4 + x2.4 =2.8)].
Taking this established fact into account, it's safe to assume and hypothesis that Skill Modifiers of the same type will stack in an additive fashion with each other and multiply with modifiers of differing types. However, it'd still be very nice if we can finally put this matter to rest by testing how Cymbidium and Lindley's stack with each other, along with the other types.
Last edited by Corintis; 10-10-2016 at 06:33 PM.