PDA

View Full Version : Balancing modes with and without gear stats - 20% more damage



We.the.North
08-27-2017, 05:31 PM
Armor penetration is a problem. With it maxed out, you can negate nearly 100% of an enemy who techs into "defense". But at the same time, you can still pick some +attack pieces and your attack at 180 gear level will reach a little higher than +20%.

Conclusion :
Against geared opponent, you will always take 20% more damage in 4v4 modes than you will in 2v2 or 1v1 modes, regardless if you tech or not into +defense pieces. Yes, everyone techs armor penetration, it's that good.

How can you hope to balance characters in both 1v1 and 4v4 when the attack values are buffed by 20% in a mode ??

Last Warrior's Den, one of the two Directors (I forgot which one and when) said their main focus was on Dominion. I'd like to know how that focus is being considered on a day to day decision making. How does defense penetration (which used to be a stat made so that gray gear players wouldn't be totally outmatched) can still be justified now that people have gear and now that it translate into a massive +20% damage across the board.

As a Shinobi player dying in 2 hits from a berserker ... (and kiling people in a single sickle rain with little help from a nearby ally) ... that +20% damage with no counterpart is a problem.

S0Mi_xD
08-27-2017, 07:18 PM
At first, stop this stupid bersreker example, same can happen to you with other heroes aswell, the top heavy combofinisher is not unstopable or uncounterable.

But let's get to the important part at here.

I agree with you, that Def penetration is a problem, but not the Stat itself, but the way it works.
Me and some other people saw this problem from the very start, with the introduction of this stat.

At the launch of Season 2, i put out the question, how does the Def Penetration works.

Also i presented a simple way how this stat could should work without being overpowered and still worth using.

http://forums.ubi.com/showthread.php/1670842-Ubi-i-need-an-answer-for-this-Defense-Penetration!

I'll post the way it should work in here again:

Currently DEFpenetration works this way:
Def - Def Pen = End Def
Def Pen can't make Defense Below Zero

But this leads to the problems, that...
A: It negates nearly all Defense stats if maxed out.
B: Negative Def Pen has no noticable effect.

How it would work well:
Def -(Def x Def Pen) = End Def


Some calculations as examples:

Variables:
● 2 pieces invested to max ATT, 1 piece to 2nd ATT = +17.1%
● DEFpen at 'MAX' = 16.9%
● target has MAX DEF = 14,1%

Basic ATK x (1 + (ATK Stat - (Defense Stat - (Defense Stat x Def Pen)))) = End DMG
50 x (1 + (0.171 - (0,141 - (0,141 x 0,169)))) = 52,7

Basic ATK x (1 + (ATK Stat - (Defense Stat - actuall Def loss through Def Pen))) = End DMG
50 x (1 + (0.171 - (0,141 - 0,024))) = 52,7

Basic ATK x (1 + (ATK Stat - actuall Defense after subtract Def Pen)) = End DMG
50 x (1 + (0.171 - 0,117 )) = 52,7

Basic ATK x (1 + ATK Stat after subtract Defense Stat) = End DMG
50 x (1 + 0,054) = 52,7 ~ 53dmg


Variables:
● 2 pieces invested to max ATT, 1 piece to 2nd ATT = +17.1%
● DEFpen negative = -16.9%
● target has MAX DEF = 14,1%

50 x (1 + (0.171 - (0,141 - (0,141 x -0,169)))) = 50,3
50 x (1 + (0.171 -(0,141 + 0,024 ))) = 50,3
50 x (1 + (0.171 - 0,165 )) = 50,3 ~ 50 dmg

Variables:
● 2 pieces invested to max ATT, 1 piece to 2nd ATT = +17.1%
● DEFpen at 'MAX' = 16.9%
● target has neg DEF = -18.9%

50 x (1 + (0.171 - (-0.189 - (-0,189 x 0,169)))) = 69,6
50 x (1 + (0.171 -(- 0,189 - 0,032 ))) = 69,6
50 x (1 + (0.171 + 0,221 )) = 69,6 ~ 70 dmg

Variables:
● 2 pieces invested to max ATT, 1 piece to 2nd ATT = +17.1%
● DEFpen negative = -16.9%
● target has neg DEF = -18.9%

50 x (1 + (0.171 - (-0.189 - (-0,189 x -0,169)))) = 66,4
50 x (1 + (0.171 -(- 0,189 + 0,032 ))) = 66,4
50 x (1 + (0.171 + 0,157 )) = 66,4 ~ 66 dmg


Edit:
Also a way could be to reduce the current Def Pen Stat to a fair amount, but there is still the negative Def Pen, and the " Def pen can't lower you def below Zero.

We.the.North
08-27-2017, 10:42 PM
At first, stop this stupid bersreker example, same can happen to you with other heroes aswell, the top heavy combofinisher is not unstopable or uncounterable.

I chose the top heavy finisher of the berserker for a single reason : I can't count the amount of time I died to it. As for it being uncounterable, keep in mind I play Shinobi and that character recovery frames are SO BAD, you cannot dodge out of a berserker's range between 2 attacks. In other words, I can't count the amount of time I got hit by a berserker dodge attack, then dying WHILE DODGING AWAY from a zooming hyper armor 66 damage top heavy that finish me off.

2 attacks, the only counter would have been to "not panic" and "parry the top heavy" ... which is hard (my parry timing ain't the best) and could always be feinted ...

I think it's ******** I can't dodge between 2 of berserker's attacks, but that's me. Shinobi recovery frames are awful.

UbiNoty
08-28-2017, 12:09 AM
I can ask them to provide some insight into their process and thoughts on def pen in the 4v4 modes - and maybe that will also perhaps prompt a re-evaluation of it in it's current state and across different modes.

CrunaCross
08-28-2017, 12:19 AM
That is why I'm keeping defense at about 0% While focusing on other stats. Everybody will use defense penetration even if it's not maxed. Max def will give me no advantage, or only a small advantage which is almost not noticeable. I agree that this creates a problem in build diversity making some build ideas completely pointless to try.

Mia.Nora
08-28-2017, 12:25 AM
I can ask them to provide some insight into their process and thoughts on def pen in the 4v4 modes - and maybe that will also perhaps prompt a re-evaluation of it in it's current state and across different modes.

I think the main problem is that developers are basing every decision they have based on data, but they don't have actual data scientists who can interpret data instead of reaching silly correlations. Roman's answer to Def Penetration issue on a Warrior's Den was a very clear indication of that. He said Def Pen was in a good spot, because the people with more Def Pen were not winning more than people with less of it. That's not how data science works. If a data variable that has indirect effect on other variables through external means, you need to eliminate the situational conditions those effects occur. Namely in this case, the external means Def Pen effects Def is by people not getting anymore than %0-5 Def because it becomes obsolete. Therefore in majority of battle situations there is no practical difference between a player with %15 Def Pen or %5 Def Pen. The only data that warrants looking at is the comparison of W/L conditions of a range of opponents with varying Def Pens vs players who make the mistake of investing into Def beyond %15 mark.

Also what we.north explained in OP is very valid, regardless of whether it may or may not be balanced as a stat itself, its effect on the game is %20 more damage on gear based modes. If that is intended to make them last faster, then it works. If that is the case because devs are oblivious to the way their game works, then they need to ask themselves the right questions.

S0Mi_xD
08-28-2017, 12:43 AM
I think the main problem is that developers are basing every decision they have based on data, but they don't have actual data scientists who can interpret data instead of reaching silly correlations. Roman's answer to Def Penetration issue on a Warrior's Den was a very clear indication of that. He said Def Pen was in a good spot, because the people with more Def Pen were not winning more than people with less of it. That's not how data science works. If a data variable that has indirect effect on other variables through external means, you need to eliminate the situational conditions those effects occur. Namely in this case, the external means Def Pen effects Def is by people not getting anymore than %0-5 Def because it becomes obsolete. Therefore in majority of battle situations there is no practical difference between a player with %15 Def Pen or %5 Def Pen. The only data that warrants looking at is the comparison of W/L conditions of a range of opponents with varying Def Pens vs players who make the mistake of investing into Def beyond %15 mark.

Also what we.north explained in OP is very valid, regardless of whether it may or may not be balanced as a stat itself, its effect on the game is %20 more damage on gear based modes. If that is intended to make them last faster, then it works. If that is the case because devs are oblivious to the way their game works, then they need to ask themselves the right questions.

Yeah, thats right - balance based purely on such simple "data" like win/loss, isn't professional in any way.

It is important to have different data.
In game numbers like kills/deaths, win/loss, how often a hero is picked.
But more important is a data called experience, and that's actually the part the player can bring tons of them, as well as the devs, because they play the game too.
But i don't think that the devs play as much as competetive or hardcore player, i can estimate them on average lvl.

But more important they need the people who can use this data properly, analiese it and give the right solutions.