A long time ago when I was still in medical school and hopelessly addicted to EQ I had some strong opinions. I think the specific expansion was Depths of Darkhallow. I was an a committed AC junkie. Bear in mind that the oldschool line of thought was still that hp was king for tanks. This was a holdover from the era when complete heals were actually complete and without upper limits - everything in the early days was built around stacking as much hp as you can and then having a complete heal chain rolling at set intervals to compensate not just for mob average dps but also for damage spikes.
Well, the game changed. Complete heals were not longer complete. All values for everything possible continued to climb and alternate abilities started to have a progressively bigger impact on the dynamics of tank survivability.
Point is I was an AC whore. I was committed to it by my understanding of the game and pursued it viciously. This was not well received on thesteelwarrior.om - it was much more appreciated on the SK forums Evilgamer.net. There were endless debates. I was consistently in the top 10, then the top 5 ... then finally I had the highest ac of any warrior across any server despite not having the best actual gear at any point in time.
Fast forward to around depths of darkhallow expansion and the Demiplane of Blood raiding zone. Despite having several thousand less HP than the fattest tanks in our guild I was typically placed as point because inexplicably (to the rest of them) - I seemed to not have difficulty single tanking every mob in the zone. All of us could do it easily while discipline was rolling. The rest of them ping-pong'd a lot and tended to die and need a tank transition at some point after that. We were always sort of light on clerics like that. For my warrior, damage intake was smooth and predictable. What few clerics we had loved it. I did parsing back then, but not nearly as much of it. Most of what I was doing was the equivalent of napkin math and theorycraft.
I noticed a pattern. The internet and other tanks in my guild were citing each mob had a max hit of X. I found this confusing because I never really saw X ... or X-1 ... or X-2 (DI intervals). I scrolled back through my logs and they just were not there. I downloaded a good parsing program that would graph out my damage intake and break down with bar graphs the incoming hit distribution. The upper DIs never happened.
What did the young me do? I ran back to thesteelwarrior to show off my findings. I made the claim that with enough AC you could completely eliminate the upper DIs. I wish I could remember his name now but there was a smart guy who pointed out that though my findings were impressive and consistently reproducible, they were of insufficient size to make such bold claims with absolute confidence. He was both right and wrong. He was right in the sense that I simply did not have sufficient data to make the claim that it was possible to fully eliminate all the upper DIs. Sure I'd tanked Mayong Mistmoore a dozen times - had the logs to prove it - and saw zero DIs for 18, 19, and 20 (the 3 highest). He insisted that though it certainly seemed that I had gotten to a point that it was statistically improbable - I had not proven it was impossible. I needed larger sample sizes.
Now, you can't just tanking Mayong Mistmoore for 12 hours straight to collect this data. I did go out and test my hypothesis on the biggest, nastiest, highest level and attack groupable mobs or raid trash I could find. At one point I dragged one of my buddies on his cleric as well as boxing my own cleric on the side to collect some truly massive parses (10s of thousands of received hits each). I did gather that data. I had actually eliminated DIs 16 and up for xp group trash. I had more or less proven that the math and concepts of my theory were correct. Did that carry over to Mayong Mistmoore, the end boss of the expansion? Possibly. I believed then and still do not that I had done it, but I acknowledge that despite all the times I tanked him the data pool was still too small to say so with absolute confidence. Raid mobs at raid level (compared to my raid level) with raid level attack are not the same as xp group trash or raid trash at their levels and attack values.
But, with that kind of data (and the theorycraft and napkin math applied after the case) I was able to make the strongest case possible that the goal of surviving wasn't simply in decreasing average incoming dps or padding hp pool to be higher ... but rather in the near-complete elimination of the types of damage spikes that kill tanks. Tanks do not die to mob average dps. They die when they don't get heals at all or during the moments of unexpected surges in instant or back to back damage that happens quicker than your healers can adjust for.
The playerbase actually did more broadly start to change their approach. Armor class on gear was prioritized more heavily and you started seeing more tanks drop their raw hp augments for ac or ac/hp hybrid augments. Unfortunately our efforts (Brael and myself mostly) resulted in me being pretty quickly un-seated as the ac king. Another warrior on another server with better gear adopted my approach to augments and I was now the 2nd highest ac ... before being displaced to the 3rd highest and ultimately kicked out of the top 10. Personal victory and devastation to my ego at the time lol.
SoE had not intended this and you could see this in subsequent expansions such as Secrets of Faydwer. Mob attack was ramped up significantly and I could no longer eliminate upper DIs. I don't think any of us could at that point.
Many years later, after medical school ... and residency ... as a young doctor I came back to the game. I didn't have time to raid or play much so I did the grouping game. In the year leading up to me finding p99 and joining these forums I test-copied all my toons to the test server so I could play for free. I found that for tanks that DI eliminating AC issue was still no longer a thing - but it did seem to exist for pets (mage pets especially) once you had all the correct mage pet AAs. With full sets of mage summon pet armor and full buffs those pets could still do what tanks used to be able to do - a terribly unbalanced situation. Mercs were now in game so gathering data was simple. Aggro the mob, /pet hold and let 2-3 merc clerics keep that pet alive for the 8-10 hours afk while I was at work. Data was gathered, point was proven, and I got a whole heck of a lot of hate on the SoE forums when I published my findings. Ultimately there was a patch somewhere in there that brought pet mitigation levels back down to reality ... again resulting in a lot of hate.
Then I came to p99 in 2014.
So yeah, 2-3 minute parses are really short. They are evidence only of what happened during that 2-3 minutes with your character and against that specific mob. In general, any data is good as long as you acknowledge what it is, what it isn't and resist the temptation to read too deeply into it. More data is always better.
Hope that helps.
Disclosure: p99 is not EQ live in PoP, GOD, DoDH, TSS, SoF etc. We are still in the era of complete complete heals and there are mobs such as AoW that it is near-impossible to stack enough hp to survive the worst of the worst round. HP (and resists to a lesser degree) are and forever will remain king here.
|