Tank Parses

Discussion in 'Tanks' started by Sirene_Fippy, Apr 11, 2016.

  1. Sirene_Fippy Okayest Bard

    How do you determine tank performance via gamparse? (inspired by the heal parse thread)

    I use gamparse frequently, but almost never to look at tanks. I don't know how to look for when tanks use their abilities, and I don't know what your tanking abilities are (except for 1 or 2).

    After playing a cleric for a while, I tend to gauge tanks by "feel" alone. I'm curious to how others judge tanks, and how tanks measure their own performance.
  2. sojero One hit wonder

    We don't parse tanks, you are not allowed to parse your tanks, forget this idea, all tanks are gods!

    OK now that I got the usual banter out of the way, and you though clerics were drama queens, meet the warriors.

    I generally look at a couple different things when looking at the tank parses. First I look at the graphs for the various tanks.
    I look at who was taking the biggest spikes. Big spikes usually mean that they didn't transition well or that there was server lag that got them.
    I look at their average hit vs the other tanks.
    I look at total damage taken over time.
    I look at spells/disc cast, but this is something that gamparse does not pick up well, so you really have to know what you are looking at, it usually picks up the procs.

    There are some things that you have to know prior. Tank AC/abilities. you need to know who your more geared tanks are and the ones that have max abilities. Comparing your #1 ac/hp tank to a new guest can be done, but not without great understand of how many things work.

    for knights, you need to look at what they are casting, are they self healing, are they generating enough agro? Agro is a big one for tanks, if they don't generate enough, they they are holding back the rest of the raid.

    The last thing and also cannot be fully looked at through gamparse is how well they communicate and transition to tanking. Do they tell the raid that they are taking over (I am bad at this) and are they using short duration abilities, shield flash/flash of anger to give the healers time to transition.

    These are just my opinions and how I look at other tanks and try to help them.
  3. Mistatk Augur

    I think it is very hard to tell much from the parse. The primary thing would be, what did the tank cast, but even then, being .5second late on that cast (like deflect) could be the difference between dieing or not dieing. You could maybe look at the list of casts, and determine someone didn't use epic or anguish bp or certain other discs. Second thing is, a ton of stuff that affects how the tank does is casted or not casted from outside the group, so it is hard to tell if the tank is given it or not. most of tanking is being a good player, and thats really hard to tell from just avg hit, max hit, etc type info
  4. Beimeith Lord of the Game

    Just make sure you are using gamparse or above. All earlier versions do not correctly parse tanking.
  5. Zaknaffein Augur

    You can't determine a tanks ability with gampars.. You can however see if they are more than a one dimensional player.

    You should know though after playing with me what a good tank looks like.

    I have a high standard for tanks, and most don't meet them. But here is just a few things I look for:

    • I look for the ability to keep agro when desired and have a quick reaction speed with the ability to adapt.
    • Situational awareness is important. I play in third person which helps tremendously with this.
    • You want tanks to be confident in thier own ability.
    • Being able to maintain a Tank order is important. There have been so many times when I am third or later in the MTO and the first tank dies, and the 2nd tank isn't paying attention and I am standing there watching DPS die and thinking to myself, "Well I could snap agro and save some DPS, but odds are the 2nd guy in the MTO will wake up and yank it off me after I've hit defensive now, and chances are he will get himself and myself killed now because healers are gonna be split healing two tanks or busy cussing at the tanks for ping ponging agro"
    • Know when to use the right discs.
    • General knowledge of every class and every raids mechanics, not just what applies to the tanking aspect of the raid.
    • Intelligence.
    • Good attitude
    • Ability to communicate well.
    • Ability to feel the pulse of the raid and know the play-styles of guildmates so you can anticipate how the raid is going and have foresight to save discs/use special once a day like abilities at the right time to save a raid from wiping etc.
    • I don't care what a tanks gear is as I don't judge thier ability based on thier gear.
    Kleitus_Xegony likes this.
  6. Zaknaffein Augur

  7. Triconix Augur

    1) Average hit vs length of time tanking. This is a good indicator of what the warrior does to twist his abilities into maximizing mitigation over extended periods of time. Almost all parses are going to look identical when LS is running for 3 minutes. What does the other 7 minutes look like?

    2) Spike damage and the frequency of these spikes. If a tank is experiencing more spikes than the norm, it's usually an indication that he's missing something or another class is missing something (enc mental contortion for example).

    3) The rolling dps the tank is experiencing. You want this to stay as close an even line as possible without periods of massive increases in damage intake.

    4) Cast parses. Many discs don't show up for a warrior, but the odds are if they are spamming stuff that does appear, then they are spamming stuff that doesn't IE insult. You'll also want to see when things were cast and if they had anything that didn't stack with each other running at the same time.

    5) I also look at the parry/riposte %s. The higher the percentage, the more he used flash and vise versa. I'll also attempt to look at his tank graph and try and match up when he used flash to see if he was using it properly. For example right after a big hit spike is the prime time to use it to prevent untimely deaths.

    Then there is stuff that gamparse will never show: Aggro control, reaction speed, bandolier switches, disc/aa rotations, etc. Tanking, unlike anything else in EQ, is all action and reaction. How are you going to react to a certain situation at any moment of time under less than optimal setting?
    Orbital101 and Kleitus_Xegony like this.
  8. Orbital101 Augur

    Result of 2 my 2 latest parses from best hsta setup to almost best hagi (minus 2 avoidance since i woke up late for work and dint had time to switch augs properly)
    12 hrs parse x2 buffless on the same mob with 2 different setup


    Huge difference from my old parse from test for only 15 avoidance difference. Though parses on raids boss are far from showing simular result with 0% dodge and barely any if at all in misses. Which where loosing 10k hp 1 shielding may be overlooked.
    Sirene_Fippy likes this.
  9. Mistatk Augur

    ye none of the test dummy even on test i dont think can be ramped up to anything like a real raid boss. they may have tuned raid boss from 100% strikethrough to 98% strikethrough or w/e, but its still hard to make an arguement for gearing augs toward defending the hits with parry/riposte etc. i.e dex vs agi only 2 choices, and only agi adds to AC which slightly help actually getting hit less (by current raid boss).(knights OP cuz stance pls fix).
  10. Sirene_Fippy Okayest Bard

    I updated my spreadsheet but I have a couple questions before I post it, cause I think some of my #s are wrong.

    Do you have Tears of Alaris, and did you use it on your older parses also?

    I appreciate and am happy you did these parses, but can you please upload to a different website next time (imgur perhaps) I almost can't read those 2 you posted. :p
  11. Orbital101 Augur

    yes, I do have the tears and used it on all my parses and I will check out imgur for next time. ATM is almost better to download the SS and use your own picture viewer to zoom in heh.
  12. Sirene_Fippy Okayest Bard

    Here is updated chart, new parses are the boxed ones:


    Here is updated graph:


    I don't think the shielding bonus is worth it from HSTA. Overall the HAGI setup takes less damage per attempt, due to miss/dodge. The difference in average hit is only 52 dmg.

    In raid content things might be different. Let's guesstimate Anashti Sul, Enslaver of Souls has a DB of 23000 and DI of 1750 (so she'd hit a warrior for ~15-50k). +1% shielding gives you -230 damage per hit.
    On my last Enslaver fight, the MT took 110 mil damage over 450 hits (25k average hit). Assuming 23k DB is correct, if they had 1% more shielding they would take 450 * 230 ~= 100k less damage total, or a total damage reduction of ~1%.

    The alternative is HDEX or HAGI - to offset the difference in shielding we need to avoid 5/450 total additional hits, or > +1% avoidance. If it's over +1% for +400 HDEX/HAGI, it's a better payoff than shielding. The question is, how much avoidance does 1200 HDEX/HAGI provide vs raid mobs.
  13. p2aa Augur

    I understand that the tank would have taken less dmg overall thanks to the 1% Shielding.
    But imo what kills tanks against raid bosses is more current dmg rounds the warrior takes at one time than the total damage taken of the fight
    So if you get hit by 230 less on each hit, does it save you from death, compared to the added avoidance and AC bonus of Heroic Agility applied to each hit too ? By AC Bonus, I mean, correct me if i'm wrong, that Heroic Agility plays also a role in the "AC Sum" part of the equation (Step 17 in the Dzarn thread), which should be the part of AC that reduces the DI part of mobs ?
  14. Sphynx_CT_SR Elder

    Umm. 100k out of 110mil is .1% not 1%
  15. Sirene_Fippy Okayest Bard

    Hi Sphynx :p You're right, my rounded total damage value is missing a 0. So shielding is pretty bad.
  16. Sphynx_CT_SR Elder

    In my guild, the difference between a tank parse with dex gear vs agi gear means little. With more being determined by the timing and use of discs and abilities then the stats. Skill still greater than gear.
  17. Zebarathe New Member

    Shielding grows the harder the mob hits for, 1% shield for minatk - 100 to maxatk - 1000 damage vs 1% shielding for minatk - 10000 to maxatk - 100000 is a massive difference. So with that being said shielding is for end game where if you do get hit 4 times in a row you need to be able to survive the attack. With high hagi/hdex you do get an offset of being hit but that is based on a chance shielding is not chance it completely alters the attack range of the mob you are up against.

    pretty much you are comparing apples to oranges avoidance to mitigation and for a tank mitigation is key not avoidance, you are supposed to take attacks, its going to happen, when it does you need to be prepared to get the best result, which would be the lowest possible damage and shielding will reduce the max attack and min attack of the attack range of the mob you are tanking.

    One more thing is AC exactly the same between test and live? are you just changing type 5 augs? do you have the 35 ac type 5 aug in your shield? Also notice the difference between dodge with hagi and hsta, now compare the difference for parry, block, misses, repo etc... you make up the %. The chance for dodge and miss is applied at a different time then block, parry, repo.
  18. Sirene_Fippy Okayest Bard

    Here's an explanation of how Shielding works.
    Mobs have a DI value (damage interval) and DB (damage bonus). The range of values a mob can hit for are DB + DI*(1-20). For the dummy values I used in my example, I made the assumption that Anashti Sul has a DB of 23000 and a DI of 1750. This would give her a hit range on someone with 35% Shielding of 15-50k like so:


    If we have 800 HSTA we get 36% Shielding, at 1200 we get 37% Shielding. Each % reduces the DB by 1%, which ends up not being that big overall, like 1/2 of a % difference for max hits (230 dmg total).

    You're right that mitigation and avoidance aren't directly comparable, but in the long run of a fight, if you shield a certain # of damage by stacking HSTA, you can compare that to the number of hits you'd need to avoid to outweigh the difference.

    For example, a MT in my guild took 110 million damage over 450 hits vs. Anashti Sul, or 25k average hit. If they added 1% shielding, at 230 damage reduction per hit, they'd avoid about 100k extra damage, or ~0.1% total damage reduction. If they instead went with 1200 HDEX or HAGI, the only question is whether they successfully avoided 5 additional hits or not vs the 1200 HSTA avoidance rate. If they did, it makes stacking avoidance stats overall a better choice.

    As for Shielding being a wiser choice due to total damage reduction, at this time in EQ I think avoiding 1 hit is more likely to save a tank from a potentially deadly round than reducing all hits by 230 (for example). I really want Shielding to be better than it is, but I think mob stats and defensive disciplines have outgrown its usefulness. Due to heroic strikethrough, HAGI and HDEX don't have overwhelmingly powerful advantages in the raid scene, either, so either way you're not going to be getting much of a difference (sadly).

    I didn't do these parses, Orbital101 did, you can check his magelo in his signature.

    I wouldn't directly compare the other avoidance stats in these parses. In the HAGI vs HSTA parses, there were ~6-7k dodges, but what we were really trying to parse is average hit and miss %. There are 14-15k misses and over 80k total attempts, quite a larger bit of data than 6-7k. You'll notice that the riposte/dodge etc rates are inconsistent across different parses, even for the same exact heroic values, suggesting there is a need for a huge amount of effort in terms of time spent parsing to collect valid data.
  19. Orbital101 Augur

    dont forget that 8-12k extra HPs is 1/2 to 2-3 hits extra per round(depending raid/group) allowing gap between heals to be more efficient constantly. I personally get much better result with more hps then going full hagi/dex in raid content.
  20. Mistatk Augur

    in going for hagi, i was only looking at the AC value of the agi/hagi. According to that formula posted abit back if figured i was getting between 1-2ac per hagi added. Still all of a sudden i feel i need to change all the hsta.