PDA

View Full Version : Alu's guide to Stat Weights.



Alustar.
05-17-2017, 02:13 AM
***This will be edited as further tests are conducted.***

So, as there are likely new and returning players that will have questions, here are some things to consider about the new gear formats. I've had some time to look over the changes after work and I have to say, I am impressed! The new changes actually look like they benefit the balance as a whole, while still allowing for individual control over a specific hero.

Some things to be aware of before we begin:

1.) Understand that, if memory serves, Ubisoft has also implemented gear balancing during matches to further ease the disparity with lower GS heroes when matched against higher opponents. (if this is taking place I don't know, nor have I tested how it affects play)

2.) the stats weights themselves have changed. They likely no longer give the returns they did prior to this update.

3.) The final and most important aspect, we can now see the actual values returned from exchanging/upgrading a piece. This is important both because now we have a tangible value, and because anything else prior was a guessing game.

So, moving on.

Setting up the base line:

Every hero starts with two stats buffed, attack/defense with values of 3.0/2.3 respectively (anyone else see the subtle change to offensive meta?) everything else is unbuffed and unaffected.

Now that our baseline is established lets delve into the specifics of this new toy we have before us! The first thing you should notice is that each piece of gear is now redesigned to function as it is intended, I.E. Armor pieces now buff as stat related to defense, weapon pieces now buff something related to offense. **Also consider, it looks as though debuffing of a stat will occur.**(not tested)

Your armor will provide the following changes: Exhaustion recovery, debuff resistance / revenge mode defense, revenge mode duration / Stamina cost reduction, stamina regen
Your weapons will pride the following changes: Defense penetration, Block damage / execution health regen, revive speed / revenge gained, revenge mode attack.

So with all of these changes, you will find that you are going to have new holes in your builds that may need filling in. A person running full attack/defense will 6 stats that are taking hits, if you find the ones he is lacking, that will make you much more effective at countering them.

I will add to this as needed, things I will have on schedule to fill out (if any lurkers out there are willing to help) I plan to see how the stats scale with each level, and if there is any defuffing of stats taking place.

Have fun, and don't do anything I wouldn't do.

EDIT: no two pieces of gear are made equal. One piece may give slight increases/decreases where others will give larger gains/losses respectively. This is true for pieces of equal level.


***This will be edited as further tests are conducted.***

kweassa1917
05-17-2017, 02:19 AM
alu, I'm actually seeing an increase to block damage, or "chip damage" if you will.


I haven't thoroughly tested it out to confirm it, and I admit it could be just placebo -- but as it is, it seems that giving up one of the attack piece's ATT value and going for #1 Block Damage. #2 DEF penetration might be actually a viable option.

Before v1.07, the only instances I've ever seen viable amount of block damage that was visually confirmable (as in, seeing the "red chip" in enemy HP bar) was with the Shugoki, but now, I'm seeing it in other classes as well when I've set for primary Block Damage + secondary DEF penetration.


I think the new stats still require way more testing to be done to see what goodies it hides.

Alustar.
05-17-2017, 02:30 AM
alu, I'm actually seeing an increase to block damage, or "chip damage" if you will.


I haven't thoroughly tested it out to confirm it, and I admit it could be just placebo -- but as it is, it seems that giving up one of the attack piece's ATT value and going for #1 Block Damage. #2 DEF penetration might be actually a viable option.

Before v1.07, the only instances I've ever seen viable amount of block damage that was visually confirmable (as in, seeing the "red chip" in enemy HP bar) was with the Shugoki, but now, I'm seeing it in other classes as well when I've set for primary Block Damage + secondary DEF penetration.


I think the new stats still require way more testing to be done to see what goodies it hides.

I see that as well ,apparently it looks like they have two stats for that defense penetration and block damage, I think the defense pen is tied to how much damage you will do to someone that bypasses the stated defense attribute, where the block damage applies as it did before patch as the "chip damage". good catch, I was curious about that!

Alustar.
05-17-2017, 04:11 AM
I wonder if the gains from upgrading are uniform across all different weights?

kweassa1917
05-17-2017, 05:09 AM
I see that as well ,apparently it looks like they have two stats for that defense penetration and block damage, I think the defense pen is tied to how much damage you will do to someone that bypasses the stated defense attribute, where the block damage applies as it did before patch as the "chip damage". good catch, I was curious about that!

Yep. Basically, exactly in the way I predicted :D.

kweassa1917
05-17-2017, 05:49 AM
I wonder if the gains from upgrading are uniform across all different weights?

I actually have a theory in regards to that question, that I'm very confident of. The range of how the numbers are set up tells you a LOT of things.

I'll write it down some time later, too busy right now, but I believe I know how all the stats work into actual calculation.

Alustar.
05-17-2017, 06:06 AM
I actually have a theory in regards to that question, that I'm very confident of. The range of how the numbers are set up tells you a LOT of things.

I'll write it down some time later, too busy right now, but I believe I know how all the stats work into actual calculation.

We might be on the same wave length, I need to do some test upgrades on gear to be certain. I'm eager to compare notes! I'm actually kinda stoked about this new set up. I was anxious before I got home and saw what it was. I was prepared for a nightmare!

kweassa1917
05-17-2017, 08:50 AM
Experience usually tells us that there's a certain range of figures game developers like to use when setting up variables like "damage buff" or "damage resistance". Usually, a normal MMORPG game with stable mode of combat ranges in between perhaps 20~50% max for damage buffs, and similarly, also 20~50% for damage resistances. Stuff like time-limited active buffs may go over that range.. in some cases crazy high numbers like hundreds of percent, but when we're talking about constant/persistent buffs, the devs don't want to go up too high because higher figures tend to mess up initial design purposes when they made that class, and bring the combat between different classes into an area of uncertainty and randomness.

So, looking at the ATT and DEF figures maxing around 6~7% at lv24 with three pieces for a sum of (on average) 20%, this gives me a very good hunch that the figures for ATT and DEF are the usual, common type of damage/resistance buffs that use multiplicative percentage calculation, with ATT and DEF figures themselves being in a additive/subtractive calculation.

On the contrary, other figures such as Block Damage, Debuff resistance and such go over 50% easily, around 70% at maximum -- not unexpected IMO, since other games also have variables that don't directly relate to damage or damage resistance, tend to have much higher figures than ATT/DEF figures. DEFpen is also around 20% IIRC, which gives some interesting points to observe, but I'll explain that later.


From this, we can probably easily deduce the basic form of ATT/DEF calculation in FH follows basically the same base format any other RPG games use.



● FINAL DAMAGE = BASE DAMAGE x DAMAGE MODIFIER

● DMG MOD = 1 + (ATT - DEF)

▶ FINAL DAMAGE = BASE DMG x (1 + (ATT - DEF))



For example, let's suppose an attack with 20 base damage hits the target. The attacker has a total of 20% ATT (= 6.3% per weapon piece), while the target has no DEF at all. In reality the devs added in a 'base defense' value for "no gear" state, but for the ease of calculation let's just suppose that doesn't exist. In this case it's simply a straight-up 20% damage buff, so the damage will be 20 → 24.

Therefore, the DEF stats will be in direct subtractive relation with the ATT stats, so if the target has 20% DEF (=6.3% per armor piece), then the damage modifier will be "ATT minus DEF" = 20 - 20 = 0... so base damage 20 will result in same 20 damage.

If the attacker has 13% ATT, but the target has 20% DEF, then the damage will be reduced 20 → 18.6



But then, how does the DEFpen. stat work?

Some people seem to think the DEFpen is a direct, straight-up counter to DEF stat, as in "DEF minus DEFpen"... but I don't believe this is the case. The reason is, because if this is true, judging by the relative size of numbers a DEFpen stat on a single piece of the weapon is powerful enough to nullify the DEF sum of 3 armor pieces combined. The DEFpen figures can be around 15~20%. Individual DEF values are like 6~7%.

My guess is, the DEFpen is a multiplicative modifier to the DEF stat... as in...



● FINAL DEFENSE = BASE DEF x DEFENSE MODIFIER

● DEF MOD = 1 - DEFpen

● FINAL DEFENSE = BASE DEF x (1 - DEFpen)



So, the revised total equation when taking these facts into account, would be like this:


● FINAL DAMAGE = BASE DAMAGE x <1 + {ATT - (DEF x (1 - DEFpen))}>



So, let's assume the attacker has 12.6% ATT(2 pieces of weapon), 15% DEFpen... and the target has 20% DEF (3 pieces of armor)... and the attacker throws a 20 damage attack. The calculation will be...



● 20 x <1 + {0.126 - (0.20 x (1 - 0.15))}>

= 20 x <1 + (0.126 - ( -0.17))

= 20 x 0.956

= 19.12



So...

● when 20% ATT clashes against 20% DEF, the base damage 20 remains unchanged.

● when 12.6% ATT and 15% DEFpen clashes with 20% DEF, you deal 19.12 damage


Then, what's the use of sacrificing one ATT stat to invest in stuff like BLOCK DMG or DEFpen??

I believe it's because the DEFpen might be more than just one role. I believe the DEFpen stat also dubs as a BLOCKpen stat as well. In other words, I believe the newly introduced DEFpen modifier, also doubles as a BLOCKpen modifier for the purpose of increasing "chip damage" people have been asking for. This "DEFpen" is the only stat in game that didn't exist before. Other stats may be modified or integrated into one, but it all existed before. But the "DEFpen" is something new... and since we do seem to be witnessing a considerably more noticeable increase in BLOCK damage, my guess is the reason should be found with the DEFpen stat.

My theory is, as said, DEFpen also doubles as a BLOCKpen.

So, a 15% DEFpen, would also mean 15% of FINAL DMG will penetrate opponent block. Using the exampel right above...


● when 12.6% ATT and 15% DEFpen clashes with 20% DEF, you deal 19.12 damage


So what if the target BLOCKS the above attack?


● 19.12 x 0.15 = 2.87 damage will make it through the block.



I believe that in the versions before v1.07, this theoretical "BLOCKpen" variable existed in the game, but was unmodifiable and hard-wired. Only a special feat like the Shugoki T3 "Punch Through" would modify it. I believe, that with v1.07 they pulled out that "BLOCKpen" concept out in the open, renamed it "DEFpen" and set up a reasonable range of spec that only mildly effects normal hits, but in truth more profoundly effects blocks.

SO, the final piece of the puzzle... the BLOCK DMG stat, which goes up pretty high.



● Let's assume that someone has 12.6% ATT, 15% DEFpen, 60% BLOCK DMG, and hits a target with 20% DEF with a 20 DMG attack.

● As calculated above, on hit, this will result in 19.12 DMG, slightly lower than base DMG.

● But if the target BLOCKS this attack, then 2.87 damage punches through, and modified directly by 60% through BLOCK DMG stat

● FINAL BLOCK DMG = 2.87 x 1.60 = 4.59


This, is actually not half bad. Light attacks are like 10~15 in damage. So with a high DEFpen and maximized BLOCK DMG setting, a 20 DMG attack when blocked will deal 33~50% amount of a normal light attack hit. If you hit someone with a 40 DMG heavy attack and he blocks it under this setting, you will deal 9.6 BLOCK DMG = which is basically the amount of a weak basic light attack.


I'll need to check up on the numbers with real examples and see if it's similar to what I've laid down here, but IMO it's worth a shot.

Alustar.
05-18-2017, 01:38 AM
Experience usually tells us that there's a certain range of figures game developers like to use when setting up variables like "damage buff" or "damage resistance". Usually, a normal MMORPG game with stable mode of combat ranges in between perhaps 20~50% max for damage buffs, and similarly, also 20~50% for damage resistances. Stuff like time-limited active buffs may go over that range.. in some cases crazy high numbers like hundreds of percent, but when we're talking about constant/persistent buffs, the devs don't want to go up too high because higher figures tend to mess up initial design purposes when they made that class, and bring the combat between different classes into an area of uncertainty and randomness.

So, looking at the ATT and DEF figures maxing around 6~7% at lv24 with three pieces for a sum of (on average) 20%, this gives me a very good hunch that the figures for ATT and DEF are the usual, common type of damage/resistance buffs that use multiplicative percentage calculation, with ATT and DEF figures themselves being in a additive/subtractive calculation.

On the contrary, other figures such as Block Damage, Debuff resistance and such go over 50% easily, around 70% at maximum -- not unexpected IMO, since other games also have variables that don't directly relate to damage or damage resistance, tend to have much higher figures than ATT/DEF figures. DEFpen is also around 20% IIRC, which gives some interesting points to observe, but I'll explain that later.


From this, we can probably easily deduce the basic form of ATT/DEF calculation in FH follows basically the same base format any other RPG games use.



● FINAL DAMAGE = BASE DAMAGE x DAMAGE MODIFIER

● DMG MOD = 1 + (ATT - DEF)

▶ FINAL DAMAGE = BASE DMG x (1 + (ATT - DEF))



For example, let's suppose an attack with 20 base damage hits the target. The attacker has a total of 20% ATT (= 6.3% per weapon piece), while the target has no DEF at all. In reality the devs added in a 'base defense' value for "no gear" state, but for the ease of calculation let's just suppose that doesn't exist. In this case it's simply a straight-up 20% damage buff, so the damage will be 20 → 24.

Therefore, the DEF stats will be in direct subtractive relation with the ATT stats, so if the target has 20% DEF (=6.3% per armor piece), then the damage modifier will be "ATT minus DEF" = 20 - 20 = 0... so base damage 20 will result in same 20 damage.

If the attacker has 13% ATT, but the target has 20% DEF, then the damage will be reduced 20 → 18.6



But then, how does the DEFpen. stat work?

Some people seem to think the DEFpen is a direct, straight-up counter to DEF stat, as in "DEF minus DEFpen"... but I don't believe this is the case. The reason is, because if this is true, judging by the relative size of numbers a DEFpen stat on a single piece of the weapon is powerful enough to nullify the DEF sum of 3 armor pieces combined. The DEFpen figures can be around 15~20%. Individual DEF values are like 6~7%.

My guess is, the DEFpen is a multiplicative modifier to the DEF stat... as in...



● FINAL DEFENSE = BASE DEF x DEFENSE MODIFIER

● DEF MOD = 1 - DEFpen

● FINAL DEFENSE = BASE DEF x (1 - DEFpen)



So, the revised total equation when taking these facts into account, would be like this:


● FINAL DAMAGE = BASE DAMAGE x <1 + {ATT - (DEF x (1 - DEFpen))}>



So, let's assume the attacker has 12.6% ATT(2 pieces of weapon), 15% DEFpen... and the target has 20% DEF (3 pieces of armor)... and the attacker throws a 20 damage attack. The calculation will be...



● 20 x <1 + {0.126 - (0.20 x (1 - 0.15))}>

= 20 x <1 + (0.126 - ( -0.17))

= 20 x 0.956

= 19.12



So...

● when 20% ATT clashes against 20% DEF, the base damage 20 remains unchanged.

● when 12.6% ATT and 15% DEFpen clashes with 20% DEF, you deal 19.12 damage


Then, what's the use of sacrificing one ATT stat to invest in stuff like BLOCK DMG or DEFpen??

I believe it's because the DEFpen might be more than just one role. I believe the DEFpen stat also dubs as a BLOCKpen stat as well. In other words, I believe the newly introduced DEFpen modifier, also doubles as a BLOCKpen modifier for the purpose of increasing "chip damage" people have been asking for. This "DEFpen" is the only stat in game that didn't exist before. Other stats may be modified or integrated into one, but it all existed before. But the "DEFpen" is something new... and since we do seem to be witnessing a considerably more noticeable increase in BLOCK damage, my guess is the reason should be found with the DEFpen stat.

My theory is, as said, DEFpen also doubles as a BLOCKpen.

So, a 15% DEFpen, would also mean 15% of FINAL DMG will penetrate opponent block. Using the exampel right above...


● when 12.6% ATT and 15% DEFpen clashes with 20% DEF, you deal 19.12 damage


So what if the target BLOCKS the above attack?


● 19.12 x 0.15 = 2.87 damage will make it through the block.



I believe that in the versions before v1.07, this theoretical "BLOCKpen" variable existed in the game, but was unmodifiable and hard-wired. Only a special feat like the Shugoki T3 "Punch Through" would modify it. I believe, that with v1.07 they pulled out that "BLOCKpen" concept out in the open, renamed it "DEFpen" and set up a reasonable range of spec that only mildly effects normal hits, but in truth more profoundly effects blocks.

SO, the final piece of the puzzle... the BLOCK DMG stat, which goes up pretty high.



● Let's assume that someone has 12.6% ATT, 15% DEFpen, 60% BLOCK DMG, and hits a target with 20% DEF with a 20 DMG attack.

● As calculated above, on hit, this will result in 19.12 DMG, slightly lower than base DMG.

● But if the target BLOCKS this attack, then 2.87 damage punches through, and modified directly by 60% through BLOCK DMG stat

● FINAL BLOCK DMG = 2.87 x 1.60 = 4.59


This, is actually not half bad. Light attacks are like 10~15 in damage. So with a high DEFpen and maximized BLOCK DMG setting, a 20 DMG attack when blocked will deal 33~50% amount of a normal light attack hit. If you hit someone with a 40 DMG heavy attack and he blocks it under this setting, you will deal 9.6 BLOCK DMG = which is basically the amount of a weak basic light attack.


I'll need to check up on the numbers with real examples and see if it's similar to what I've laid down here, but IMO it's worth a shot.

I want your babies... that was beautiful

Alustar.
05-18-2017, 01:50 AM
SO, I did some checking intot the vlaues as well and have noticed that there is a definite pattern to the gear weights as well.

It looks like you have 4 different variables to the gear stats now,

A base line that increases all by a moderate amount example would be a head piece. at level 22 defense of +1.9 Exhuast. Recov. +37.2 and Debuff Resist of +17.6. The remaining three types would all have a single stat debuffed while the others are given greater buffs.

Also it's important to not that I don't think the same tug of war values will be applicable.

More areas I want to watch are the way the pieces scale the stats individually, and if it's uniform across variant gear pieces, I.E. will Exhaustion Recovery always increase and decrease by a particular rate, regardless of what stats are favored.

Side note, some of the new Gear skins for PK looks so killer....

UbiNoty
05-18-2017, 01:59 AM
I've asked the team to see if they'd be willing to give some more clarification/insight into the exact algorithms they're using for the stat system.
Also - this is how you look to me right now:
https://media.giphy.com/media/AXorq76Tg3Vte/giphy.gif

kweassa1917
05-18-2017, 02:02 AM
I've asked the team to see if they'd be willing to give some more clarification/insight into the exact algorithms they're using for the stat system.
Also - this is how you look to me right now:
https://media.giphy.com/media/AXorq76Tg3Vte/giphy.gif

It's a pity some of the theories are already proven wrong tho' :D

I'm pretty confident about the basic damage calculation, since like, almost EVERY game uses the exact one, or variations of it.... but DEFpen seems to be a lil more complex, and how Block Damage works is entirely guesswork, which baffles me the most.

Alustar.
05-18-2017, 02:11 AM
I've asked the team to see if they'd be willing to give some more clarification/insight into the exact algorithms they're using for the stat system.
Also - this is how you look to me right now:
https://media.giphy.com/media/AXorq76Tg3Vte/giphy.gif

Lol, Kweassa wins my interwebs

Regardless of the exact numbers it does seem like you have a greater amount of control over your character's attributes. This is bound to make players both very happy and very upset. I like it




Also, I just really wanted to have 3 of these on one page in honor of Kweassa. Much deserved!

Alustar.
05-18-2017, 02:16 AM
Thread title needs to be changed to KweAlu's Guide to Gear.?

mrmistark
05-18-2017, 05:18 AM
This made my head hurt trying to follow the math and equations 😲