Kanican (kanican) wrote,
Kanican
kanican

Enmity Testing (Part III)

Overview

In part II, we really started to get some nice results and have begun to really understand the basic system behind enmity - namely finding a method that quantifies actions, determining that there are 2 types of enmity, and that at least one of these types caps.  In this post, I'll go over more testing which attempts to help further grasp the overall concept of enmity.  I'll also go over methods of testing and obtaining enmity values for anyone interested in trying to find unique numbers yourself.

General ideas that I'll go over in this post for anyone looking for a brief overview...

      -  The exact unit value of the cumulative enmity (non-decaying hate) cap is found
      -  A test to show that cumulative (non-decay) and volatile (decay) hate act seperately is shown
      -  A test to show that logging out or zoning does not immediately reset all enmity is shown
      -  A test is performed to help suggest that enmity per action simple stack or add upon each other is done

      -  A few examples of how to go about obtaining cumulative enmity values are determined

I guess the key to this post will be to show that the 2 types of hate act seperately - that is, the value of one does not effect the properties or values of the other.  Knowing this, we can start to focus on obtaining values for various spells, especially those of particular interest to tanking.  I will go over a few methods for determining the cumulative enmity of any ability in case anyone wants to learn how to test his themself.

Once again, this topic doesn't lend much chance at pictures, so sorry!


Previous: 
Enmity Testing (Part II)

Next:  Enmity Testing (Part IV)

Enmity Table

____________________________________________________________________________________________________

Naming Convention

A lot of names are being thrown out right now for the types of hate.  In an effort to try to keep things consistant, I ask that everyone who posts here or refers to here use the following terms and definitions...

      - Cumulative Enmity (CE) - the enmity of an action that does not decay naturally over time
      - Volatile Enmity (VE) - the enmity of an action that does decay naturally over time
      - Total Enmity (TE) - the sum of Cumulative and Volatile Enmity; the highest determines who has "hate"
      - Hate - a term for who the enemy is currently focusing actions on; NOT a quantifiable term
      - Hate List - the list of players of which could take "hate"; "top of the hate list" is who has "hate"
      - "Enmity" (E) - the "base unit" of enmity; by definition - "the enmity generated using Cure 1 for 0 HP"

This naming convention is generally consistant with the JP Wiki except for the Total Enmity term.  I would prefer not to use the word "hate" too often when formally going over the system because SE chooses not to use that word - it's really just something the community has used, not what SE uses.  The "proper" term in the game is "Enmity", so I prefer to stick to that.  The unit I just called "Enmity", so it's easy to say - like "you need 3,000 more Enmity to reach the cap".  Some examples using this system...

      - Dispel's CE (Cumulative Enmity) is 320 E
      - Provoke's CE is 1 E, while its VE (Volatile Enmity) is greater than 1 E
      - The VE of any player given infinite time is always 0 E
      - Given two players at the "top of the hate list" have the same TE, the most recent action will determine "hate"
      - Short-hand method - "Dispel has 320CE and unknown VE"

I hope to maintain a pretty good nomenclature for this as soon as possible.  Please try to follow this so we can remain consistant and better show results!

____________________________________________________________________________________________________
____________________________________________________________________________________________________

Test  9

What is the exact "hard cap" of Cumulative Enmity (CE) in Enmity Units (E)?


Testing Setup:

- 1 puller character with Silence access (BRD/WHM)
- 2 level 75 members with ability in question (RDM/NIN, BLM/RDM)
- 2 characters must have 0 enmity merits (+/-), be stripped of all gear, and have access to "distance plug-in"
- Use a "level 0" monster - we chose Bumblebee in West Sarutabaruta

- All characters must have either Stoneskin or Phalanx buffed prior to starting to ensure 0 damage
- Utsusemi shadows must be off when this experiment takes place

- Puller pulls with Silence while the 2 test characters move to 0.7 distance (same distance)
- Test character 1 casts Dispel 32 times (reaches Cap: see test 8)
- Wait 2 minutes for any possible long term decay to occur
- Test character 2 casts Dispel 31 times (9,920 E: see test 7)

          At this point, character 2 will take hate momentarily but permanently lose hate after a few seconds.

- Test character 2 casts Cure 1 for 0 HP until hate is taken
- Count the number of Cure 1 for 0 HP were required to take hate

          The Cumulative Enmity Cap will be 9,920 E + the number of Cures Counted


Number of Cure 1 for 0 HP Required80

          Cumulative Enmity Cap = 9,920E + 80 E = 10,000 E

If you perform this, you will actually get hate at exactly 80 cures, meaning the "cap" of Cumulative Enmity is indeed exactly 10,000 E - if you Cure 1 someone with hate for 0 HP 10,000 times in a row you will reach the Cumulative Enmity cap.  This is extremely good news to me as it suggests that the base units we are using "E" may indeed be the actual machine value unit used in this game.  It would seem like a huge coincedence otherwise.


Conclusion

The Cumulative Enmity Cap is exactly 10,000 E.


img219/3369/img20071108200808rv0.png 

____________________________________________________________________________________________________
____________________________________________________________________________________________________


Test 10

Is it possible to gain Volatile Enmity when Cumulative Enmity is capped?

      H0:  It is still possible to gain Volatile Enmity when Cumulative Enmity is capped
      H1It is not possible to gain Volatile Enmity when Cumulative Enmity is capped

Testing Setup:

- 1 puller character with Silence access (BRD/WHM)
- 2 level 75 members with ability in question (RDM/WAR, BLM/RDM)
- 2 characters must have 0 enmity merits (+/-), be stripped of all gear, and have access to "distance plug-in"
- Use a "level 0" monster - we chose Bumblebee in West Sarutabaruta

- All characters must have either Stoneskin or Phalanx buffed prior to starting to ensure 0 damage
- Utsusemi shadows must be off when this experiment takes place

- Puller pulls with Silence while the 2 test characters move to 0.7 distance (same distance)
- Test character 1 casts Dispel 32 times (reaches cap)
- Test character 2 casts Dispel 32 times (reaches cap)
- Character 1 uses Provoke
- Character 2 quickly follows using Blind

          - Provoke and Blind both produce fairly strong amounts of Volatile Enmity, but Provoke's is much higher

- Document who has hate immediately after character 2 uses Blind

      H0:  Player 1 will still maintain hate immediately after character 2 used blind
      H1Player 2 will take hate immediately after using Blind

This test shows that even if Cumulative Enmity has been capped Volatile Enmity can still be gained.  This supports the idea of 2 seperately run Enmity systems which are then pooled at the end (Total Enmity, TE).  If this were not true, than player 2 should have taken hate simply because they were both at cap and by virtue of casting the most recent action, should have taken hate.  The fact that player 1 still maintained hate suggests that Volatile Enmity was higher at the time and that played a role in the final hate determination.

This test cannot conclusively prove that these 2 hate systems act completely seperately; however, it does strongly support that claim.


Conclusion

Volatile Enmity can still be gained even once Cumulative Enmity has been capped.
Hate is decided by the sum of Cumulative and Volatile Enmity (Total Enmity).


____________________________________________________________________________________________________
____________________________________________________________________________________________________


Test  11A

Does "logging out" immediately reduce TE to 0 and remove a player from the "hate list"?

      H0:  Logging out does not immediately reduce TE to 0 and remove the player from the hate list
      H1Logging out does immediately reduce TE to 0 and remove the player from the hate list

Testing Setup:

- 1 puller character with Silence access (BRD/WHM)
- 2 level 75 members with ability in question (RDM/WAR, BLM/RDM)
- 2 characters must have 0 enmity merits (+/-), be stripped of all gear, and have access to "distance plug-in"
- Use a "level 0" monster - we chose Bumblebee in West Sarutabaruta

- All characters must have either Stoneskin or Phalanx buffed prior to starting to ensure 0 damage
- Utsusemi shadows must be off when this experiment takes place

- Puller pulls with Silence while the 2 test characters move to 0.7 distance (same distance)
- Test character 1 casts Dispel 32 times (reaches cap)
- Test character 2 casts Dispel 1 time
- Character 1 "logs off" with hate
- Character 1 immediately logs back on
- Character 1 casts Cure 1 for 0 HP on Character 2

      H0:  Player 1 will take hate immediately after that Cure 1 for 0 HP
      H1Player 2 will maintain hate after the Cure 1 for 0 HP

This test shows that logging off and returning is NOT a good way to "reset hate".  Obviously most players know from experience that logging off is a good method of enmity reduction; however, this test shows that the act of logging off does not immediately reset your enmity.  This test does not go into the possible methods of enmity decay that logging out produces; it only shows that the act of logging out is not an instantaneous "hate reset".

I think it is interesting to note... It is obvious that Player 1 still has much higher hate than Player 2 in this test, but Player 1 must perform an action to receive hate back.  This is a bit odd because the player with the highest enmity on the mob's list is NOT the one with hate.


Test  11B

Does "zoning" immediately reduce TE to 0 and remove a player from the "hate list"?

Testing Setup:

Repeat the testing setup of 10A, except instead of logging out, simply zone out.  This experiment must be conducted right at a zone in order to limit the possible effects of Distance changes.

ResultSimilar to 11A - player 1 will retake hate upon re-entering the zone


Conclusion

"Zoning" or "Logging Out" does not immediately reset Enmity to 0.

____________________________________________________________________________________________________
____________________________________________________________________________________________________


Determining Cumulative Enmity

At this point in time, I would prefer to focus more on Cumulative Enmity as it is easier to maintain controls for these kinds of tests.  At this point, I believe that the knowledge gained from the first 11 tests is adequate to find the Cumulative Enmity of any action that does not have a damage or cured value associated with it.  These tests will run similarly or exactly like Test 7, the test used to determine that Dispel was 320 E.

Fortunately, you have more options than just casting Cure 1 for 0 HP - that would take an incredibly long time.  As with the enmity cap testing, it is possible to stack spells with known CE values like Dispel.  As knowledge of variouis abilities and their associated CEs are found, it should become easier to determine the CE of any ability.  I will go over 2 examples of how you can go about doing this; it does involve some basic math.

img230/169/img20071107184750nx1.png 


Example 1 - Finding Cumulative Enmity Loss of an Utsusemi Shadow

We know from Test 2 that losing an Utsusemi Shadow causes some form of Cumulative Enmity loss.  We can now find it using what I call a "Table Method".  The Table Method is used to keep track of 2 players' current state of Cumulative Enmity by documenting particular actions.  I think this method will become clear after seeing this example.

The table begins AFTER the 3rd player pulls the mob.  The third player is not included...

Actions Performed

Kaeko

Ashira

Hate

  Kaeko casts Dispel x1

320

0

Kaeko

  Ashira casts Dispel x1

320

320

Ashira

  Kaeko casts Cure 1 for 0 HP x1

321

320

Kaeko

  Ashira casts Utsusemi: Ichi

321

321

Ashira

  Ashira loses 1 Shadow

321

321 + X

Kaeko

  Ashira casts Cure 1 for 0 HP x25

321

321 + X + 25

Ashira



The final action in this Table Method should always leave both players at equal enmity.  To ensure this, try to step up with cures.  If the final action was not a cure, but you think both players are at the same enmity, you can still verify - have the player without hate cast cure twice in a row.  If you are at equal enmity, that player will take hate the first cure.  If the other player (the one originally with hate) then casts cure, he should not have hate back.  This is a simple verification process you can use.

Note that the enmity values in the table ONLY include the Cumulative Enmity (CE) of actions, Volatile Enmity (VE) is completely ignored.  This is a proper assumption as long as you wait a good 10 seconds between actions other than Cure 1 for 0 HP, which generates no VE.  So the final line, we can basically set an equal sign and solve for our unknown variable X...

          321 = 321 + X + 25  >>>  X = Cumulative Enmity of Losing an Utsusemi Shadow = - 25 CE

This is a simple example using just Cures and Dispel to determine the Enmity of losing an Utsusemi Shadow.  Now that we know this value for Utsusemi though, we can immediately use it to build on further determinations...


Example 2 - Finding Cumulative Enmity Gained from Hojo: Ichi

We'll apply the Table Method again, but this time, we can use the value we got in Example 1 for Utsusemi Shadow loss.  This gives us a huge leeway in not having to Cure a billion times - which speed up the process considerably.  I would definitely not advocate determining any of these values using solely Cures like we did for Test 7.  I did verify the numbers afterwards using this to make sure though - it does work out.

In this test, Ashira has pre-buffed Utsusemi: Ichi prior to the pull and beginning.  Once again, the table lists the actions completed right after the 3rd player has pulled the enemy in place...

Actions Performed

Kaeko

Ashira

Hate

  Kaeko casts Dispel x1

320

0

Kaeko

  Ashira casts Dispel x1

320

320

Ashira

  Ashira casts Hojo: Ichi x1

320

320 + X

Ashira

  Ashira loses 3 Shadows

320

320 + X - 3(25)

Ashira

  (All Shadows Removed)

320

320 + X - 3(25)

Ashira

  Kaeko casts Cure 1 for 0 HP x5

325

320 + X - 3(25)

Kaeko



We can now set the 2 terms equal to each other given that we assume CE for both players is currently equal.  Again, we can run that small Cure test to verify this.  If you are using Cure 1 for 0 HP as your final action and just stepping up 1 unit at a time, you shouldn't have to do this, though.

          320 = 320 + X - 75  >>>  X = Cumulative Enmity of Hojo: Ichi = 80 CE

Now that we have 3 Cumulative Enmity values greater than 1, things should be much easier in testing.  At this point, it's really about just using the proven and tested data and building upon it.  You kind of have to mix and match and determine the values of various spells that way.  Again, this does not have anything to do with Volatile Enmity in any way.

My hope is that an extensive search for all abilities, actions, and spells that do not involve Curing or Damage can be done to find all of them that have a Cumulative Enmity not 1 CE.  Using this method, it should actually be rather quick to find the CE of any particular action.


____________________________________________________________________________________________________
____________________________________________________________________________________________________

Conclusions

I think a lot of great strides were made in part II of the testing on ths blog.  At this point, I feel we are ready to do an extensive search of all abilities with a Cumulative Enmity greater than 1 CE.  Any help in performing this would be greatly appreciated.  At this point in time, we have started a small list of known Cumulative Enmity values from certain actions, abilities, or spells.  This does not include ones where CE was equal to 1.

               ________________________________________________________________________________

Action Performed

Cumulative Enmity (CE)

    Aspir *

320

    Cure V **

400

    Dispel

320

    Flash

180

    Hojo: Ichi

80

    Sleep

320

    Sleep II

480

    Sleepga ***

180

    Sleepga II ***

180

    Stun

180

    Utsusemi Shadow Loss

-25

    Blink Spell Shadow Loss

0


      *  Aspir was tested using a 0 MP Aspir - whether MP drained affects hate has not been tested
      **  Cure V is the same amount of Cumulative Enmity regardless of the amount of HP cured
      ***  Sleepga spells grant the same hate regardless of number of targets hit (tested)


These actions have Volatile Enmity as well!  At this time, we are just looking at the Cumulative Enmity!

               ________________________________________________________________________________

I hope others can use this data to try to find more data values.  Feel free to post any results you get if you're interested in contributing.  As I have done previously, I will cite when the information is used on here if you happen to give your FFXI Name and Server.  Ashira and I are really excited about this and I hope others will help contribute as well.  There are way too many spells to do a full set of testing by just the 2 of us, so we do need some help!  In particular, neither of us has a BLU, so if there are any willing testers with 75 BLU, that would be great.

Here is a summary of findings for those too lazy to disect the tests...

      -  The Cumulative Enmity Cap is exactly 10,000 CE
      -  Cumulative and Volatile Enmity act as seperate entities
      -  An Action can produce both Cumulative and Volatile Enmity, but it MUST produce at least 1 CE
      -  Zoning and Logging Off do not immediately "reset hate"


I know these posts started out pretty dry, but I hope others see the breakthroughs we've made on the subject of Enmity now.  Already, I can think of a number of useful strategies - both tanking and non-tanking which can be developed with just the limited data we have right now.  I really hope that this list above can be expanded quickly with the help of the endgame community.

____________________________________________________________________________________________________

Pic of the Day

While Saku and I go 0/40+ on X's Knife...

Tinkly gets it on the 5th attempt!

img230/6743/img20071103230511cy5.png



Drama Thread of the Day

http://killingifrit.com/forums.php?m=posts&q=146317

Diabolos Server Drama - botting, kickings, screenshots, and fairly high emo level from OP.



____________________________________________________________________________________________________
____________________________________________________________________________________________________
Tags: enmity testing!
Subscribe
  • Post a new comment

    Error

    default userpic
    When you submit the form an invisible reCAPTCHA check will be performed.
    You must follow the Privacy Policy and Google Terms of use.
  • 67 comments
Previous
← Ctrl ← Alt
Next
Ctrl → Alt →
Previous
← Ctrl ← Alt
Next
Ctrl → Alt →