Illusions Benefit Neza: +10% dodge

Discussion in 'Tanks' started by Ecchicon, Jan 26, 2018.

  1. Daedly Augur

    Ok, I am new to tanking and have been reading and rereading this thread over and over. When I got home from work last night, I decided to make a spreadsheet with the formula and put my SK's, Biodiesiel, numbers into it. I made sure to use truncated numbers and grabbed his stats off his level I.

    The formula was telling me that he should have a dodge rate of around 19 or 20 if the heroic agility bonus was completely negated. So I loaded of gameparse and combined a night of OT killing to see what it said. It showed a dodge rate of roughly 3% on around 630 hit attempts.

    So I combined some other exp sessions. In Lcea killing spiders I was seeing about 8% and on the PoWar anniversary mission it showed about 16% dodge rate.

    All of them lower than the formula is telling me it should be. Not huge sample sizes, but Lcea and OT were way off. Is their wording about the formula giving TOTAL dodge chance incorrect? There was mention of regular strikethrough but the devs post seems to imply that it doesnt get taken into the equation for dodge. What am i missing or where does my understanding lack?
  2. Lianeb Augur

    You seem to be glossing over the part of the devs post talking about heroic strike through. This number climbs with each new iteration of the game, as does your hdex to counter it, I would expect your dodge rate to increase on older content as you have shown in your digging and be at its absolute lowest in current content. Raid mob having the highest amount of strike through and eliminating PC dodge chances.

    On a side note: take 110 monk and go start some older raids. With and without their dodge aura on (which is currently somehow broken on older content)
  3. Daedly Augur

    Going only by the formula, as I understand it, the amount heroic strikethrough only plays a role in wether or not I recieve a bonus to my base dodge. Using the devs last example, you could give the NPC a Hstrikethrough of 30,000 and the last step would still come out to 25.

    So I have to assume that there is more than just that formula that goes into wether or not we dodge an attack, the formula is wrong that they posted, the formula that the game is using was accidentally messed up, or I suck at life.
  4. menown Augur

    The formula works perfectly fine. I plugged it into excel and found at a certain level of NPC hStrikethrough, all avoidance drops to zero.
  5. Lianeb Augur

    Using the dev example.
    Step 6 25 + (40 - 30,000) using your number
    Is 25 + -29960 = -29935 or 0

    Or I am reading it wrong.
  6. Daedly Augur

    Step1 = HeroicStrikethrough(NPC)
    Step2 = HeroicAgility / 25
    Step3 = MIN( Step1, Step2 )
    Step4 = DodgeSkill + 100
    Step5 = Step4 + ( DodgeSkill * DodgeSPA ) / 100
    Step6 = Step5 / 45
    DodgeChance = Step6 + ( Step2 - Step3 )

    Final chance would look like this,

    Step6(25) +(Step2(40)-Step3(40)) = 25

    As step3 is the Min of step1 and step2.
  7. p2aa Augur

  8. p2aa Augur

    I agree with you. If you read the parry and dodge formulas, you have in theory a min parry and dodge value that are independent to how many Hstrikethrough the mobs have.
    A kinda "fixed" value.
    So if in raid we see 0 parry and 0 dodge of raids mobs, there is another explaination than insane heroic strikethrough of raid mobs. Unless the random number rolled by raid mob is set to be always higher than this "fixed" value ? For example, If the max "fixed" value obtainable is 25, the random number is rolled into [30-100] ?
  9. p2aa Augur

    Per "
    • Active defensive skill order (succeeding an earlier skill will mean the later ones won't bother checking):
      • Riposte
      • Block
      • Parry
      • Dodge
      • Shield Block
      • Staff Block
        • Then regular strike-through is checked."
    Regular strikethrough enters in action after the Parry, Dodge and co skills formulas. So in theory should have no impact on Parry / Dodge %.
  10. menown Augur

    I rechecked my excel formulas after I got home last night and you are correct. There is still a base dodge skill no matter how high the NPC's hStrikethrough.

    I am wondering if there is either a different check that raid NPCs have that we don't know of or the hStrikethrough is checked twice. If it is checked twice, then it could be that your final dodge chance has to be higher than the NPC's hStrikethrough if you want any chance at all. For example, step 1 can still be greater than step 2, but step 1 could still be below your final dodge chance calculation, allowing you to dodge against the NPC. But once the hStrikethrough gets higher than your calculated dodge chance, you lose all chance to dodge.
  11. Brohg Augur

    No, none of that. It's just strike through.
  12. menown Augur

    So it is just the regular stikethrough values that are so high against raid NPC's?
  13. Wulfhere Augur

    We can't know or count defense types vs raid mobs because the strike through message supersedes the defense messages.

    A plain miss looks like this:

    [Mon Apr 09 19:46:38 2018] A surreal succulent tries to smash YOU, but misses!

    A defended attack that was overridden, and then still missed:

    [Mon Apr 09 19:46:45 2018] Your opponent strikes through your defenses!
    [Mon Apr 09 19:46:45 2018] A surreal succulent tries to smash YOU, but misses!

    A defended attack that was overridden and hits:

    [Mon Apr 09 19:46:45 2018] Your opponent strikes through your defenses!
    [Mon Apr 09 19:46:45 2018] A surreal succulent smashes YOU for 18747 points of damage.

    Besides strike through, current raid mob accuracy is very high. We don't know how high our avoidance skill needs to be to improve miss rate. Raid tanks are stacking hAgi in hopes of seeing a break with avoidance. The highest Magelo is just over 2800 agility with about 1800 in hAgi. We do know that "SPA 184 - Accuracy %" debuffs of 10% are far more impact-full (and a must on raids) then any amount of hAgi so far reported.

    It's interesting to note that Enchanter gear has the highest values of hAgi (not counting augs). The devs intend for them to be "missed" more then other classes, presumably so they can survive their role of CC.
  14. p2aa Augur

    Strikethrough has no impact on dodge %. Strikethrough is the last stage before Accuracy Versus Avoidance roll, and happening after the game checked if you dodged or not.
  15. p2aa Augur

    Shield Block : A random 1-100 is made by the mob and checked against yours.
    Shield Block AA is 29 % at max AA. As we shield block 0 in raid too, it means that raid mob random something like [30-100] ?
  16. p2aa Augur

    I guess the values are out of reach per expansion.
  17. josh Augur

    I think you are misunderstanding what that dev was saying in the post you linked. Strikethrough is applied after the defensive ability is determined. so, for instance, i roll riposte and it fails, I don't have block so skip that, i roll parry and it succeeds, i dont bother rolling dodge shield block or staff block, the mob checks to see if he strikes through my defenses, if he doesn't than i parry, if he does than i roll accuracy vs avoidance.

    You and wulfhere are both incorrect, I pointed this out in another thread already and both of you were there so i don't know why I'm saying this again, but raid mobs have LESS accuracy attack power than group mobs. And after abilities like somnolence and gut punch are applied they have even less than that.

    More avoidance ac, means less chance to be hit, every parse you've ever done on a raid is pointless because there's no way to know how often things that decrease a mobs accuracy attack power were on, how frequently your enchanters used somnolence or how frequently your warriors were gut punching.
  18. p2aa Augur

    That's another interpretation of the post , but it could be understood the way I did too. As he put the strikethrough and accuracy things at the complete bottom after the full list of avoidance skills.
    We would need a more clear answer on it.

    Parses of others are wrong but yours are right ?
    Other conclusions are wrong but yours are right ?
    The ego is strong with you, isn't it ?
    Lol. I think you are the only one to think raid mob accuracy is less than group mob. Makes no sense at all. Same to think warrior gut punch is impacting raid mob accuracy so much.
  19. josh Augur

    Gut punch is affecting it that much, It's actually a bit unbelievable how much things like somnolence and gut punch affect it. Looking at parses I'm a bit upset that paladins are the only tank class that don't have any form of accuracy decreasing debuff.

    It does make sense that its less for raid mobs because they removed the ability to defend against them, their goal appears to be to simplify the process of balancing raid mobs. all they have to worry about is balancing avoidance AC vs attack accuracy. They remove the ability to defend but than lower the accuracy so that we still avoid damage at a reasonable rate.

    And it isn't arrogance, I've done a lot of parses and mine isolate variables more than the ones that you posted, for instance, you have parses that show your avoidance is better with HSta, and that is definitely wrong, which basically proves my point that you can't parse avoidance during a raid because there are too many factors that impact it.

    Here are some recent parses i did that show how accuracy vs avoidance works in PVP. I dueled a bot and lowered its avoidance AC until it matched my accuracy attack power. 1065 ac vs 1066 attack power.

    Total Melee --- Damage: 1637067 --- Avg hit: 338 --- Attempts: 9536 --- Missed: 4686/9536 [49.14%] --- Hits: 4850/9536 [50.86%] --- Real Hits: 4850/4850 [100%]

    at equal values my miss chance is 50% basically. so i raised it by about 200 to 1263

    Total Melee --- Damage: 1126974 --- Avg hit: 307 --- Attempts: 8717 --- Missed: 5052/8717 [57.96%] --- Hits: 3665/8717 [42.04%] --- Real Hits: 3665/3665 [100%]

    this follows a formula, 50 * 1066/1263 = 42, if this trend follows than if i have double the AC i should have 25% chance to hit, so, at 2138 AC

    Total Melee --- Damage: 626009 --- Avg hit: 276 --- Attempts: 8940 --- Missed: 6672/8940 [74.63%] --- Hits: 2268/8940 [25.37%] --- Real Hits: 2268/2268 [100%]

    so, for PVP it is literally this simple. your level doesn't matter, your other stats don't matter, it is literally their attack power vs your avoidance AC. I'm not saying that's the actual formula, because if you wanted that to continue working on values where your attack power is higher than your opponents AC you would have to swap them to be avoidance/accuracy but still, it illustrates my point that there is no threshold or anything like that, if you have more avoidance AC, you will get hit less, period.

    Now, i didn't stop there, i applied this to an actual mob to try and figure out its attack power. I did a parse with 3412 avoidance AC against a mob in TBM. I have to admit, these parses just aren't long enough to prove anything, but i think they are interesting nonetheless

    Punch --- Damage: 29905596 --- Avg hit: 6537 --- Attempts: 6627 --- Parried: 2/6627 [0.03%] --- Dodged: 1/6625 [0.02%] --- Defended: 3/6627 [0.05%] --- Missed: 2049/6624 [30.93%] --- Hits: 4575/6624 [69.07%] --- Real Hits: 4575/4575 [100%]

    Not long enough to get it accurate down to the decimal value but if we take the simple formula and apply it to this guys chance to hit me, we get an attack power of 5503. Now, to try and confirm this i debuffed him with feralisata. an 8% attack power debuff, and this is what that actually does, to pc's at least, when you cast it on you, your accuracy attack power is decreased by roughly 8% after rounding. so, if it did decrease it by 8% i should have a miss chance of 33.6%

    Punch --- Damage: 28756560 --- Avg hit: 6422 --- Attempts: 6669 --- Parried: 3/6669 [0.04%] --- Dodged: 1/6666 [0.02%] --- Defended: 4/6669 [0.06%] --- Missed: 2187/6665 [32.81%] --- Hits: 4478/6665 [67.19%] --- Real Hits: 4478/4478 [100%]

    so a 1.88% increase. when i was expecting about a 2.6% increase. it's not super far off, its possible the parse just wasn't long enough but it also might suggest that there is slightly more going on when you are fighting an NPC vs fighting a player. More parses with this methodology would confirm or deny it but they are pretty tedious to do.

    This parse is the one that illustrates how much gut punch and somnolence do, gut punch is 50% of the same SPA as feralisata. I assure you, this has a MASSIVE impact on a mobs chance to hit you.

    Even if there is more to it when fighting an NPC the basis is still what i observed in PVP. meaning, more avoidance AC means more misses. In addition, the closer you get your avoidance AC to the mobs accuracy the larger the affect an increase in avoidance AC will have on your chance to miss, for example,
    let's take a hypothetical raid mob with a 16% chance to miss against you with 3200 avoidance AC. This means he has 10000 attack power using my simple formula

    50 * 3200/10000 = 16.

    if we raise our avoidance ac to 3600 we get a miss chance of 18%, 2% increase so not bad, but lets throw in somnolence with 3200 AC

    50 * 3200/(10000 * .75) = 21.3%.

    than, with 3600 AC

    50 * 3600/(10000 * .75) = 24

    2.7% increase in miss chance over 3200 AC. and it just keeps getting better, the closer you get to a 1/1 ratio of avoidance AC to accuracy the better an increase to avoidance AC becomes.

    It's interesting, these accuracy debuffs that make it impossible to parse the affect of HAgi on raids are actually what make HAgi useful on raids.
    Wulfhere likes this.
  20. josh Augur

    so, a thought just occurred to me, EQ doesn't do floating point values. so it's possible that in my parse against the TBM mob, my miss chances are truncated to 31% and 33%. so while i should get a 2.6% increase i actually got a 2% increase and the parse is off by only a small amount. this makes sense and does institute a sort of threshold. This would make avoidance AC darn near impossible to optimize for, which sucks.

Share This Page