Page 2 of 2 FirstFirst 1 2
Results 11 to 19 of 19
  1. #11
    Player Motenten's Avatar
    Join Date
    Mar 2011
    Posts
    321
    A bit of a correction there. VE accumulates at three times the rate of CE, with respect to damage done (as well as damage taken and damage cured). On level 120 mobs (T6 VWNMs) every point of damage done is 1 CE and 3 VE. So VE will cap long before CE will, despite decay.

    Reaching VE cap is remarkably easy. A pld can cap it in 4 actions after getting on hate list: Sentinel, Provoke, Divine Emblem+Flash.

    CE, on the other hand, is far more difficult. It's essentially impossible to build CE without dealing with damage in some way (damage done or damage cured), and is also affected by damage taken. As noted above, you get 1 CE per damage done. In additional, assuming ~1800 HP, you lose 1 CE for every point of damage taken. Therefore for a pld to cap CE, they have to do 10k more damage to the mob than the mob does to them to maintain capped CE.

    Now, the damage done to the pld will be far less than the damage done to any other player (eg: when stacking defense the way SE apparently suggests), and that in itself reduces CE decay. The problem is that in putting up sufficient defense that way, you've removed pretty much everything offensive a pld can do. A pld all Defendered up, with Gallant's Roll and Minnes and Tacos or whatever else, is sacrificing opportunities to improve their damage output with things like Chaos Roll and Marches and various attack/accuracy foods. Therefore their damage output is miserably pathetic and they'll sit at pretty much 0 CE indefinitely.

    As an alternative, they can cure themselves, but that's a fairly weak option. Put in terms of DPS (ie: damage done per second, which directly correlates to CE generated per second), it tops out around 100 DPS using the current cure formulas. However, even with generous MP recovery conditions (including using 300 TP Chivalry), that can't be maintained for more than about one minute, which is still only halfway to the CE cap.

    A DD, on the other hand, is going offensive. Assuming the pld is sitting at 10k VE all the time, and given the 1:3 ratio of CE:VE, a DD matches the pld's total enmity when the DD hits 2500 CE/7500 VE. That amount of CE/VE is done with a mere 2500 damage. After that it's just a question of who's doing more damage: the DD or the mob. If the DD is doing more, their CE increases; if the mob is doing more, the player's CE decreases.

    In general, any decent DD will reach 2500 damage within ~30 seconds. In cases where they have TP built up before the fight (eg: Voidwatch), they can hit 2500 damage in 10 seconds or less. So that pretty much defines the window in which the pld can claim to be the tank.


    The second threshhold is when CE and VE are both capped. That essentially takes 10k damage minus whatever AOE damage the DD takes (aside from whoever is at the top of the hate list, who might take more). This is actually a slightly more difficult threshhold to hit. A VWNM might have ~190k HP. 10k HP is 5.3% of the NM's entire HP pool. Decent DDs will likely do at least ~10% of the mob's HP over the course of the fight, while extremely strong DDs might be upwards of 25%-30%. Typically, though, you could expect a DD to reach cap within a couple minutes.


    Essentially, what's needed is some way for a heavily defensive pld to be at some sort of advantage in terms of generating CE. At present you have an either/or situation: either you're good offensively and can gain CE, or you're good defensively and can't generate CE, which completely undermines the entire point of having a tank.

    People have clamored for native Provoke on pld for years, but that would never solve the problem. Pld doesn't have issues generating VE, which is all that Provoke gives. What it needs is sort of the complete opposite: a tool that can generate a significant amount of *CE* that can be used on a relatively frequent basis (30-60 seconds). I'd say that if it were constructed like Cannonball (which does damage based on your current defense), and generated CE based on your defense, things would be a lot more feasible.

    Consider how much defense is feasible to get: 565 base+gear; Protect V w/ring +65; Defender +157; Gallant's +157; double Minne +67+61; Tacos +150 = 1223 defense. Even removing Gallant's + Minnes, that's 937.

    I think an ability that could generate CE equal to twice the pld's current defense, or maybe equal to the pld's defense since it would also be affected by +enmity (and reaching double under the effect of Sentinel) would work. If it's equal to the pld's defense+enmity, probably have it on a 30 second timer. It would still take a few minutes to cap CE even under optimal conditions, but that's supplemented by any damage done plus cures.

    Maybe have the shield being used be an additional scaling factor: an extra 10% CE per size of the shield (so +10% for bucklers, +30% for kite shields, +50% for Aegis/Ochain).

    So some sort of "Gallant Charge" or whatever, with emphasis on the shield. Doesn't do damage or stun or anything; just generates CE. The stronger your defense, the more effective it is. I think it might actually be a workable solution, at least on the pld side...


    Then if they just added some way for thf to dump some of the enmity they gain from Accomplice/Collaborator, the tools for a more controlled hate structure would mostly be in place.
    (4)

  2. #12
    Player Babekeke's Avatar
    Join Date
    Mar 2011
    Location
    Windurst
    Posts
    2,273
    Quote Originally Posted by Motenten View Post
    What it needs is sort of the complete opposite: a tool that can generate a significant amount of *CE* that can be used on a relatively frequent basis (30-60 seconds).
    I think this is what SE's interpretation of Trick Attack is for. PLD doesn't get an ability to deal damage itself, but while stacking up on DEF/-DT, a THF or /THF can pile up large amounts of CE on them.
    I'm not saying that they're right, or especially that it's worth bringing a THF or /THF to VW for, but this is their view, and why nothing is being done about it.

    Quote Originally Posted by Motenten View Post
    Then if they just added some way for thf to dump some of the enmity they gain from Accomplice/Collaborator, the tools for a more controlled hate structure would mostly be in place.
    Trick Attack >>> Accomplice^^
    (1)

  3. #13
    Player saevel's Avatar
    Join Date
    Apr 2011
    Posts
    2,350
    Character
    Saevel
    World
    Asura
    Main Class
    RDM Lv 99
    I always assumed it was a 1:1 split between CE and VE on damage done. At 1:3 the PLD is totally screwed as DD's can build CE via damage quicker then the PLD can, and everyone has VE capped already.

    Giving the PLD the ability to generate more CE doesn't change anything, a MNK hitting the monster will generate more CE/VE then a PLD ever could hope to. The actual hate generation formulas need changed, hate gained via damage needs to be reduced period. Otherwise it will always be a better idea to hit the monster more then to turtle up. And the defense argument fails because monster spam AoE's that kill everyone, or deal debilitating status ailments. Charmga, Deathga, or my favorite from Aello, Amnesia + Silence + Paralyze + Encumbrance + Muddle + high damage. Your sitting there naked unable to cast spells, just job abilities, put on gear nor use items. Mind as well go take a WC break or something. Chainspell Meteor spam, 5-count Doomga, or just ridiculous amounts of damage while being nigh invulnerable.
    (0)

  4. #14
    Player Motenten's Avatar
    Join Date
    Mar 2011
    Posts
    321
    Quote Originally Posted by Babekeke
    I think this is what SE's interpretation of Trick Attack is for. PLD doesn't get an ability to deal damage itself, but while stacking up on DEF/-DT, a THF or /THF can pile up large amounts of CE on them.
    I'm not saying that they're right, or especially that it's worth bringing a THF or /THF to VW for, but this is their view, and why nothing is being done about it.
    Perhaps. However going by that, the lack of a thf cripples the pld's ability to tank. While there's nothing wrong with needing support from other classes to reach your best potential (eg: brds, cors, healers), pld's limitations are more fundamentally tied to basic game mechanics.

    A pld can be buffed to improve their greatest strength -- mitigating damage -- and gain progressively greater survivability. That aspect is normal, and expected from support jobs. However that mitigation is worthless if the pld doesn't have hate, and choosing to buff a pld's strength weakens his ability to hold hate.

    Having another job assist you in your primary function should move you from good to great, not from abysmal to maybe passable.

    Quote Originally Posted by Babekeke
    Trick Attack >>> Accomplice^^
    Trick Attack and Accomplice are for opposite functions. TA adds hate to the target; Accomplice adds hate to the thf (and removes hate from another target). However, with no means of dumping her own hate, the thf hasn't really fixed the problem, but only changed which person the problem revolves around.


    Quote Originally Posted by Saevel
    Giving the PLD the ability to generate more CE doesn't change anything, a MNK hitting the monster will generate more CE/VE then a PLD ever could hope to.
    Actually it does (or at least can, theoretically) change the overall balance. If the pld can build up CE at a decent pace at least vaguely on par with the DDs, then the CE reduction due to damage begins to heavily favor the pld. A 1000-damage AOE that hits the pld for a mere 600 is effectively a net gain for the pld for the duration of time it takes the mnk to regain that CE. Likewise if the mob turns and hits the mnk for 1000 damage that quickly drops them below the pld who has been taking less damage in the interim.

    Of course there's still the issue of how quickly the mnk can rebuild CE after that. If it's too fast, you still have a bit of a spinning top effect going.

    The progression of enmity gain via damage is a decreasing scale as mob level increases. That is, as we get higher level and fight higher level mobs, doing more damage simply due to the nature of level progression, we gain less enmity for each point of damage done. This is mostly fine, but the overall balance has broken down a bit because it's a 1/N scale, which means the rate of decrease drops as levels get higher. At the same time our damage output increases in a more-or-less linear scale.

    The result is a net increase in the enmity accumulated per unit time. D*E (where D is damage and E is enmity factor), with D increasing normally and E decreasing slowly yields a total result with the product increasing slowly.


    So, in addition to the original suggestion, you also need a rebalancing of the enmity accumulation factors, such that damage taken results in a relatively more significant CE loss (perhaps increasing exponentially rather than linearly), while damage done does not accumulate CE quite as rapidly (a logarithmic increase?). And it has to be a formulation that works all the way from level 1 to 99, and across a broad range of individual survivability, along with the means of keeping hate off of the back line with a fair bit of reliability. Quite a bit of work.

    Going to poke around with the idea of an exponential/logarithmic scale..
    (2)

  5. #15
    Player saevel's Avatar
    Join Date
    Apr 2011
    Posts
    2,350
    Character
    Saevel
    World
    Asura
    Main Class
    RDM Lv 99
    Lets not kid ourselves, 10,000 CE isn't that hard to cap when your spamming big WS's and hitting like a machine gun. A monster hitting for 1,000 damage will just kill the DD's. Outside of Abyssea we simply do not have the HP to survive damage on that scale, hence fanatics spam in voidwatch. If we don't have Fanatics we use SMN PD zerg to mitigate that damage. Anything that can do back to back 1K+ AoE's has the potential to wipe you, it's SE version of a "I Win" button.

    DD's are looking at 2~3x the damage output of a Sword/Shield PLD, there is nothing the PLD can do to hold hate. You could give them a JA that gives them 5,000 VE and 5,000 on a 30s timer, it wouldn't matter. The PLD would cap hate nearly instantly, and shortly afterwards the super buffed DD's would cap hate, then whomever is hitting the most tanks. A monster hits you for 500, it turns to the next DD and smacks them for 500, in that time you've hit the monster two to three times and regained that enmity you lost. Only way for a PLD to "tank" is if there are no super buffed DD's, aka 2004/2005 DD/THF or kiting Sky NM era.

    Like I said, hate from damage needs a dramatic reduction, on the order of half or more. Only way a non DW 1H is going to hold hate vs everyone else.
    (1)

  6. #16
    Player Motenten's Avatar
    Join Date
    Mar 2011
    Posts
    321
    Yes, I can see that. Worked out some possible revisions, though it's mostly pointless since SE would never actually use it. Plus I've only looked at a few of the implications; it would also require a complete reconsideration of the hate from curing, and probably all JA/spell use. Not to mention making sure it scaled across mob levels (I've poked at it a bit, but certainly haven't considered all possible configurations).

    Basic idea focused on two principals: That threat should be based on the scale of damage done rather than a linear comparison of values; and that enmity should be proportional to the actual threat the player poses.

    Raw enmity from damage = square root(damage done). This slows growth of enmity with damage dramatically. 500 damage is far more noticeable to the mob than 10 damage (going from ~3 to ~22), but doing 1000 damage doesn't get as much additional attention (going from ~22 to ~31), and so on. Basically, focus more on the scale of damage than on the actual value.

    Multiplier = 1000 * damage/mob max HP. Minimum of 1. The greater the percentage of the mob's health done by the player, the greater the threat. A 3000 damage weaponskill on a mob with 6000 HP would be vastly more threatening than the same 3000 damage weaponskill on a mob with 180000 HP.

    Then give pld a means of generating CE directly (proportional to his defense, preferably).

    Suppose the pld can generate 1000 CE with this JA, compared to a 2-handed DD vs a tough mob:

    Middling war does five 100-damage hits and a 1500-damage weaponskill vs Pil. sqrt(100) = 10, raw enmity per hit. Damage is less than 360, so will hit the minimum multiplier of 1. CE gained: 10 per hit, 50 total CE. sqrt(1500) = 38.7, with a multiplier of 1500/180000 = 8, total of 309. 359 CE altogether. Pld easily maintains lead.

    Now make it a super-buffed Ukko war doing 250 per hit and 3000 WS. Melee enmity per hit is 15 each, so 75 there. Weaponskill generates 876 enmity, for 951 total. Pld is still ahead, but could easily slip behind depending on how quickly each performs their actions.


    This idea would require a lot of work just to make sure the basic math is well balanced relative to the mobs in the game, across various levels, and would require a complete overhaul of spells and JAs. Doubt the devs want to devote that much energy to this problem, as it essentially amounts to a complete re-imagining of the entire battle AI.

    As secondary issues, I'd probably do things like increase the enmity of certain debuffs based on the mob type. A blm should find Silence being cast on it (whether or not successful) to be far more of a threat than a pld mob would, while a mnk mob would consider it inconsequential, but might consider Slow or Paralyze to be far more of a threat.
    (3)

  7. #17
    Player Motenten's Avatar
    Join Date
    Mar 2011
    Posts
    321
    Extending the idea into enmity from curing...

    Unlike damage, which scales more slowly than linearly, amount healed should maintain its current linear growth. However, unlike at present, enmity generated should increase with the level of the target healed rather than decrease.

    Basically, the threat generated by healing a target should be proportional to the threat the target actually poses. That threat can grow as square(player level/mob level).

    Example:

    Lvl 100 player (theoretical) vs lvl 100 mob: square(1/1) = 1.0 multiplier. Essentially, it's 1.0 when target player and mob are the same level.
    Lvl 90 player vs lvl 100 mob: square(90/100) = 81%
    Lvl 80 player vs lvl 100 mob: square(80/100) = 64%
    Lvl 50 player vs lvl 100 mob: square(50/100) = 25%
    Lvl 10 player vs lvl 100 mob: square(10/100) = 1%

    The lower the player level vs the mob, the less of a potential threat they are, so the less concerned they should be about that player getting healed.

    This grows in the other direction, too. A lvl 99 player vs a lvl 90 mob would have a scaling factor if square(99/90) = 121%. Should probably cap it at 200% (eg: lvl 84 player vs lvl 60 mob).


    Currently, a lvl 99 player has a ~58% scaling factor in enmity from cures. With the above, it would be 81% vs lvl 110 mobs, and 68% vs lvl 120 mobs, making healer enmity a bit higher of a concern (or a bit more of a benefit, for a pld). However a theoretical lvl 11 whm healing a lvl 11 tank vs a lvl 15 rabbit in the dunes would have just a 54% scaling factor instead of the current 200% scaling factor.
    (2)

  8. #18
    Player Babekeke's Avatar
    Join Date
    Mar 2011
    Location
    Windurst
    Posts
    2,273
    Originally Posted by Babekeke
    Trick Attack >>> Accomplice^^
    Quote Originally Posted by Motenten View Post
    Trick Attack and Accomplice are for opposite functions. TA adds hate to the target; Accomplice adds hate to the thf (and removes hate from another target). However, with no means of dumping her own hate, the thf hasn't really fixed the problem, but only changed which person the problem revolves around.
    Sorry, I had to leave for work and so just quickly typed something simple. What I was trying to say was that we should have the ability to stack trick attack with accomplerator to steal a fellow target's hate and transfer it to the tank, instead of to ourselves. That said on a very limited number of mobs (non-true sight, sight-only tracking mobs) we can accomplerator > hide to steal another players hate then shed it all. I assume that anything that matters (VW/Legion) will not be hideable though.
    (1)

  9. #19
    Player Llana_Virren's Avatar
    Join Date
    Apr 2012
    Location
    Okinawa, Japan!!
    Posts
    491
    Character
    Llana
    World
    Lakshmi
    Main Class
    WHM Lv 99
    Quote Originally Posted by Motenten View Post
    Extending the idea into enmity from curing...

    ...making healer enmity a bit higher of a concern (or a bit more of a benefit, for a pld).
    While I agree with your math, I think that this would be detrimental outside of a "normal" EXP battle. A tank (or any front-lined DD) recieving a severe AOE effect (which is essentially a given, nowadays) would be outright chaos. In those instances, Cure Potency is far more relevant than Enmity down is.

    Conversely, this would mean that WHMs (or any primary healing role) would have enmity potentially exceeding other DDs (BLMs included).

    Unless of course, to keep WHMs on top of the "healer's ladder", you enriched them with several instances of "Reduced Enmity" Job Traits (because Enmity Down merits will NOT be enough).
    (1)

Page 2 of 2 FirstFirst 1 2