+ Reply to Thread
Page 1 of 12 1 2 3 4 5 11 ... LastLast
Results 1 to 15 of 167
Like Tree1Likes

Thread: Armor, Avoidance, Diminishing Returns, and You!

  1. #1
    Sword of Telara Fasc's Avatar
    Join Date
    Feb 2011
    Posts
    841

    Default Armor, Avoidance, Diminishing Returns, and You!

    So I'm a huge math geek and decided that rather than charge to the finish line with the current Beta, that I'd take my time, smell the roses, oh and do a lot of jotting down of numbers and such each time I level. I figured this open forum for all Classes, not just my preferred Class of Rogue, would be most appropriate for this information and thread starter.

    Some definitions/acronyms I commonly use

    RNG: Random Number Generator. This refers to any mechanic that is based upon a chance of happening, like Dodge or Parry or some procs. It can even be used in a phrase like a verb such as "Man I got RNG'd hard there..." which is to say that luck was not on your side and you got punished hard for it.

    TTL: Time-To-Live. This is a specific value based upon how much time would a Boss take to kill you with zero Healing from full Health to zero. This accounts for an averaged Dodge/Parry/Block as well as other nuances that may exist with a particular class.

    Worst Case TTL: Worst Case Time-To-Live. This is similar to TTL except you EXCLUDE all forms of Avoidance or any other RNG aspect that could possibly not happen. Essentially this is how long you can live without Healing based purely on your Health/Armor/Damage Reductions.

    EH: Effective Health. This is a specific value that takes into account just your Armor/Health/Damage Reductions and transforms it into a singular "Health" number. What this value indicates is how much "Health" you would have if you stripped all your Armor and Damage Reduction off and had JUST Health to soak damage with. This is a value you want to be high enough so that you can survive the punishment of a Boss long enough for your Healers to respond and actually save you.

    DR: Diminishing Returns. This acronym is sometimes used for Damage Reduction as well but I learned it first for Diminishing Returns. DR is the concept that for every point gained of a particular stat, the overall weight/value of that point is less than the previous. In other words, if you have 100 of Stat X, and gain 1 point, that point is worth more than if you have 200 of Stat X and gain 1 point. This phrase can be used in other contexts but generally speaking when DRs are brought up, it is in reference to some stackable/obtainable character stat that does not overtly scale in a linear fashion. More on the specifics of this later in the post.

    ------------------------------------------------------------------------------------------------------------------------------------------------------------------

    Now that that is out of the way, first: Armor

    Armor is known to have diminishing returns that is clearly evidenced by simply looking at tooltips and remembering in your head that you seem to be getting fewer and fewer % gains on your total damage reduction. This is essentially by design and is hardly uncommon with most MMORPGs. Along with Armor diminishing itself, another commonly known fact is that as you get higher and higher level, to achieve some X% of damage reduction, you need more Armor than you did previously. Both of these are rather uninteresting facts for any veterans of the game type.

    What IS interesting is the nature of Armor despite this seemingly diminished return. Even though I require more and more points of Armor to gain each percent gain, my net gain in Survivability or TTL is still progressing quite well, linearly in fact. This has to do with percentage decreases and how that changes as you affect the total amount you are diminishing. To gain 1% more damage reduction when you start at 0%, you need exactly 1%. However, to gain ANOTHER 1% damage reduction, you need to chip away 1% of the REMAINING damage that you could take, so 1% of 99%, or 0.99% in raw damage reduction. As you can see, as you approach larger and larger values of Armor, the actual percent gain needed to obtain the same increase is less and less, which corresponds to the more and more Armor needed per percent, hence linear scaling.

    I think I've pinned down a "close enough" equation to math out damage reduction from Armor in this game. I say "close enough" because I don't feel I have enough data samples to confirm but more on that in a moment.

    The current equation based on my testing is as follows:
    Damage Reduction = (Armor) / (Armor + 120*Level + 500)
    To create this formula I simply went back to the WoW based formula, and made a relatively safe assumption that RIFT's Armor formula would work similarly, but with different constants. Upon testing that assumption with simple algebra, it appeared to be that the general form of the equation held true, and sample data was all that was needed to confirm rough values.

    The general issue with coming up with a high degree of confidence is that the game does not show us all the digits that it actually uses to tally up its numbers. We see up to the hundredths, but the game is likely going at least several decimal places beyond that. I am unaware of any commands that let us see what the game "sees" and if such a command exists, I would be very happy to use it and retest.

    Also note that I'm not level 30 or level 40 and certainly not an Alpha tester with access to level 50. As is well-known at least in WoW, the Armor formula CHANGES when you reach level 60. It doesn't change drastically, but it certainly changes.

    Take the above formula with a grain of salt and know that it may very well not be valid at higher levels. Furthermore, artificial caps on damage reduction via Armor are unknown as well.
    ------------------------------------------------------------------------------------------------------------------------------------------------------------------

    Second item on the menu: Avoidance Stats

    Avoidance stats are those that allow you to completely avoid taking any damage at all, in other words Parry and Dodge. These stats have both Rating and secondary sources, that is to say you can directly improve Dodge by wearing an item with Dodge Rating on it, and you can improve Dodge by wearing an item with Dexterity on it, due to Dexterity providing Dodge Rating. One thing that isn't always true of Avoidance stats is the idea of diminishing returns. Unlike Armor, there have been MMOs in which Avoidance stats have had diminishing returns, others that have had none, and still others that began with none and later implemented them.

    Why implement diminishing returns at all? Doesn't it defeat the purpose of high levels of Dodge or Parry if it just becomes insanely hard to improve?

    Well the reason most game designers put diminishing returns on something like Avoidance stats is because of how they scale. If undiminished, Avoidance will grant exponential returns the higher and higher you go. That is to say, at 50% Dodge, another 1% will be MUCH more valuable than if you start at 25% Dodge. The reason being is this:

    At 25% Avoidance you are taking 75% of the damage, adding 1% means you now take 74%. This yields a damage reduction of 1.333%.

    At 50% Avoidance you are taking 50% of the damage, adding 1% means you now take 49%. This yields a damage reduction of 2.000%.

    At 98% Avoidance you are taking 02% of the damage, adding 1% means you now take 01%. This yields a damage reduction of 50%.

    Now you're probably saying "Fasc, this is ridiculous, they aren't going to allow stats to get THAT high! And besides, they likely already have diminishing returns in place to prevent that."

    To the first part, probably so, most games don't allow extremes to occur easily, but games of the past have allowed it to happen or nearly happen, causing severe balance issues.

    To the second part, Trion has simply not done so, at least at the lower levels.

    In recording values at each level gained, I've checked varying amounts of Dexterity and Strength on my Rogue character. It wasn't until higher that I even had access to Strength values to alter my Parry much but I've had larger and larger ranges of Dexterity to toy with as I've gone up in level. So far, there is ZERO evidence of any sort of diminishing return on Dodge or Parry.

    As an example, at lvl7, a low level, every point of Dodge Rating (and consequently Dexterity) granted exactly 0.3333% Dodge, from a range of 9.00% Dodge at 27 Rating to 13.00% Dodge at 39 Rating. Now again, just like with Armor, this could very well change at the upper registers of levels, but here nearly at lvl20, it shows no signs of changing. Only LEVEL seems to affect the amount gained per point, but no other diminishing returns seems to exist whatsoever. Every graph of Rating vs Dodge, assuming no Soul skills or abilities that artificially inflate Dodge, has been a perfectly straight line with an R-squared value of practically 1.000 every single time (most hover around 0.99995...)

    I've held off on any sort of precise formulation of Dodge Rating Value vs Level for now. There does not seem to be any clear/consistent relationship between the coefficients for Dodge/Parry and Level at this time, particularly at the lower levels where the changes make almost no sense at all. More data here is needed.

    Also note that some have reported a hard cap on some stats like Parry. A hard cap would imply that Trion has artificially stopped any chances of exceptionally high levels of Avoidance from happening, and if any character were capable of achieving these levels, further points towards Avoidance would be entirely wasted. These caps, and whether or not they exist, is not yet confirmed.
    ------------------------------------------------------------------------------------------------------------------------------------------------------------------

    Third, because I don't hate Clerics and Warriors: Block

    This section will be a tad bare for now as I'm starting my RIFT experience with a Rogue, not a Cleric/Warrior, and Rogues cannot Block. I don't rightly consider Block to be Avoidance either, although some tend to lump it in. It does not prevent 100% of the damage, it chips off a portion to lessen the blow, so I consider it separate. Furthermore, there is anecdotal evidence suggesting that the combat table is a tad more complicated than a single roll system. Testing this is a huge pain and is not something I am ready to delve into just yet, not while I'm still leveling. If there is some sort of 2-roll or more system, how this affects Block and possibly even the other stats of Parry and Dodge, could be very important for theorycrafting and gear planning.
    ------------------------------------------------------------------------------------------------------------------------------------------------------------------

    So what's this have to do with the lot of you? Well I don't necessarily need a bunch of mathematicians to debate with, although that's always fun. What I would love though is some more field testing and data samples.

    Do you have Armor numbers that cause the above formula to not even be close at all? Post them up, provide some screenshots and give any other info you can, that is stuff I can use!

    Do you have numbers to show evidence of diminishing returns on something like Parry or Dodge? Same as before, post them up!

    Do you not care about all this math mumbojumbo but you want to help out? Just plain post up numbers with level and everything and I can add it to the batch!

    Wall'o'text I know but something I'd figure I would share.

    Comments, concerns, criticisms, questions, and queries are welcome!
    Last edited by Fasc; 02-25-2011 at 08:14 AM. Reason: Update 01

  2. #2
    Banned
    Join Date
    Feb 2011
    Posts
    81

    Default

    he knows too much. call in the hit.

  3. #3
    Soulwalker
    Join Date
    Jan 2011
    Posts
    2

    Default

    I have a lvl 22 cleric with 2017 armor and what you say seems to be correct for the return. I dont have any bonuses to block or parry except for the base line.

  4. #4
    Telaran
    Join Date
    Jan 2011
    Posts
    76

    Default

    lvl 42 warrior
    Armor 6435, dmg red 53,74%
    Seems right.

  5. #5
    Plane Touched
    Join Date
    Jan 2011
    Posts
    199

    Default

    lvl 42 rogue specced 8 pts into bd.
    225 Dex
    232 dodge=12.07%dodge (+5% from false blade)
    140 parry=2.92%

  6. #6
    Rift Disciple bodr's Avatar
    Join Date
    Nov 2010
    Posts
    135

    Default

    tnx, very interested

  7. #7
    Sword of Telara Fasc's Avatar
    Join Date
    Feb 2011
    Posts
    841

    Default

    Quote Originally Posted by Isomalt View Post
    lvl 42 rogue specced 8 pts into bd.
    225 Dex
    232 dodge=12.07%dodge (+5% from false blade)
    140 parry=2.92%
    Is your 12.07% with or without the 5% Dodge from Bladedancer?

    232 Dodge = 7.07% Dodge + 5% (Quick Reflexes) = 12.07%?
    232 Dodge = 12.07% Dodge?

    I know servers are down now so you might not remember but just wanted to clarify.

  8. #8
    Prophet of Telara Skryth's Avatar
    Join Date
    Jan 2011
    Posts
    1,018

    Default

    Paladins are crying over the leaving out of block. You are making holy warriors sad...for shame...
    Jolly fat people are living proof that it is not possible to "laugh your *** off", though we're trying...

  9. #9
    Sword of Telara Fasc's Avatar
    Join Date
    Feb 2011
    Posts
    841

    Default

    Quote Originally Posted by Skryth View Post
    Paladins are crying over the leaving out of block. You are making holy warriors sad...for shame...
    Hahaha... sorry sorry. I just had limited time to play this Beta (was my only solid play experience) so I'll get around to Block at some point.

    For those that might know or recognize the name from WoW, it will come as no surprise that the Warrior archetype of Tanking will be the LAST one I try, although Clerics Block so the numbers won't be hard to come by once the 24th hits.

    Another thing that was claimed elsewhere that I am VERY skeptical on is the idea that all the mechanics we see and read on our skills and abilities and such are additive, and not a mixture or purely multiplicative. Some skills, particularly those that improve others, it could go either way since the wording in ambiguous and you get nice round numbers either way, but that's simple enough to test usually.

    What is less so is how things like Armor buffs and Damage Reductions stack up. There are those that claim that if I have a 30% damage reduction CD up along with some skill that also reduces damage by 6%, that I'd have a total of 36% damage reduction (additive) as opposed to my original assumption of 34.2% damage reduction (multiplicative). Same goes for buffs to skills like all the Armor bonuses that a Rift Stalker gets. Additively our Armor would be increased by 110% once fully specced and you put Guarded Steel up, but multiplicatively you'd actually have 157.4%.

    These are things I'll investigate and confirm later once things go Live. I'm bothered enough by the lack of diminishing returns on Avoidance stats just out of experience with the end of TBC in WoW where you had Feral Druids running amok with 80%+ total Avoidance making light of even Bosses like Brutallus. If damage reduction effects stack additively, you could actually have the issue (not from a player perspective, but a balance/developer one) of a Tank juggling numbers so he doesn't waste itemization on too much Armor because "he's already at 100% physical damage reduction." Doesn't take a genius to see the issues of that happening, especially if its just one Tank class that's doing it.
    Last edited by Fasc; 02-21-2011 at 08:34 PM. Reason: Clarity

  10. #10
    Rift Disciple
    Join Date
    Feb 2011
    Posts
    146

    Default

    Great post this should be stickied somewhere

  11. #11
    Sword of Telara Fasc's Avatar
    Join Date
    Feb 2011
    Posts
    841

    Default

    Bumping this back up just for visibility and a reminder that anyone and everyone is welcome to submit data and information for me to use, particularly if you can prove anything I've said or written as incorrect. Critical peer review is often the best way to learn and improve upon current understanding.

    See you guys in-game soon!

  12. #12
    Plane Touched Boondocker's Avatar
    Join Date
    Jan 2011
    Posts
    229

    Default

    Interesting, thanks for posting this! 5*
    Shout out to all our warfighters: veteran, active duty, reserve, and retired!

  13. #13
    Shadowlander
    Join Date
    Oct 2010
    Posts
    38

    Default

    I read someone on the forums state there was a hard cap on parry at 20%. I know I have seen my tooltip for parry say higher but perhaps anything above 20% just simply doesn't do anything.

    I don't know if anyone could dig up a post to confirm or if anyone has tested this. If it were true then any fears you have about the lack of diminishing returns on avoidance stats would be unfounded.

  14. #14
    Rift Master
    Join Date
    Jan 2011
    Location
    VA
    Posts
    674

    Default

    Quote Originally Posted by wershlak View Post
    I read someone on the forums state there was a hard cap on parry at 20%. I know I have seen my tooltip for parry say higher but perhaps anything above 20% just simply doesn't do anything.

    I don't know if anyone could dig up a post to confirm or if anyone has tested this. If it were true then any fears you have about the lack of diminishing returns on avoidance stats would be unfounded.
    Yes, but hard caps that were reachable during the beta would be even worse than not having diminishing returns. Unless the rating/point increase equation goes through the roof from 40-50 there would be little point in gearing up as it just pushes you further past the cap.

  15. #15
    Plane Walker Zayrinoke's Avatar
    Join Date
    Feb 2011
    Location
    Pennsylvania
    Posts
    460

    Default

    Thank you for this. Will you be updating this with available data following release? If so I'll be sure to chime in with my Riftstalker's stats and percentages as my gear changes.

    Effective Health: 51 Riftstalker/7 Bard/8 Ranger - Avoidance: 36 Riftstalker/8 Bard/22 Bladedancer - Healthy Mix: 51 Riftstalker/7 Bard/8 Bladedancer
    Riftstalker Tanking - Key Abilities and Mechanics 2.0
    Zayrinoke, Level 50 Bahmi Rogue - Zerstorer, Level 43 Bahmi Warrior - Asche, Level 20 Kelari Cleric - Haifisch, Level 32 Eth Mage
    Multiple Shards

+ Reply to Thread
Page 1 of 12 1 2 3 4 5 11 ... LastLast

Bookmarks

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts