AC issues are starting to be really problematic.

[pasi]

The 5% was a damage interval only mitigation added when they bumped defensive down to 45% (2003). The net effect being a buff to non-discipline tanking while keeping discipline tanking the same.

The monk mitigation thing was always pretty overblown. Top of the line geared monks might have taken less overall DPS than top of the line warriors, but the conclusion that this made them better tanks ignores the reasons why warriors have always had a monopoly on tanking. Namely, HP returns and Defensive. It was mostly warriors bitching about monks soloing. Warriors complaining while being the most powerful class in the game - nothing out of the ordinary.

HP is the most important tanking statistic in the era of CHeal.

[Danth]

Originally Posted by Rogean [You must be logged in to view images. Log in or Register.]

Excuse me for getting annoyed that you guys seem to think that these values are well known or something and that it's our code that's broken.

Your position is a difficult one, but it's only natural that people are going to wonder if your code is broken when you can go in-game, take off most of your armor, and not get hit for visibly more damage per minute than you do with full AC gear on. Which, incidentally, I did last night against a variety of monsters ranging from high 50's cliff golems to mid dark-blues to light blue trash. Regardless of the cause, something seems very wrong with respect to the end-result of armor function (or rather, lack therof) in game.

I know you can't perfectly replicate classic in this case due to the necessary information simply not existing. That's fine. Most of us players probably don't care overmuch about the workings behind the scenes. We just want to be able to put on armor and feel like it has some noticeable effect. You want your systems to be tuned as ideally as possible, too, so we all have the same goal here.

You said useful feedback would boil down to "Soandso mob should be hitting harder or softer, spiking more or less, etc." That feedback turns into the "fix your AC" comments because in effect the problem is with the great majority of the many hundreds of monsters in the game, from low levels through cap. Simply put, there are precious few areas where armor seems to have a beneficial effect for a level-appropriate tank. I mean, I could go to the bugs forum and start listing almost every last monster in the entire game as hitting for nearly the same damage regardless of target AC....but that wouldn't really be too useful, would it? It's more efficient simply to discuss it under the "AC is whack" subject since it's so widespread.

For the sake of testing, is there a zone or monster--particularly one already in-game (old world/kunark)--where you feel monsters are tuned correctly against high-level tanks, and where going from say 950 to 1100 displayed AC will have a noticeable and significant effect? Knowing such would be immensely useful as a "control" to compare against other zones in testing.

Danth

[Haynar]

When a mob is way below your level, there is an adjustment due to level difference.

This is probably overcompensating when you ac is low. It was tuned for higher ac. It should be scaled.

Will look at it. I know even cons totally will beat the crap outta u if ur naked. That needs adjusted for blue and light blue. Green, maybe not so much.

H

[Rogean]

Originally Posted by Danth [You must be logged in to view images. Log in or Register.]

Your position is a difficult one, but it's only natural that people are going to wonder if your code is broken when you can go in-game, take off most of your armor, and not get hit for visibly more damage per minute than you do with full AC gear on.

Again, you're missing the point of the entire argument. We're trying to get you guys to understand the difference between incorrect code and incorrect values. Both of which are needed in order to have a fully working system. The fact that you're not seeing AC working properly is not automatically a code problem. I'm saying that it's more likely the mob's are not tuned correctly (in the case of most kunark mobs, they aren't tuned at all).

The problems don't just all magically go away with correct code if the mobs don't have correct stats in the first place.

[Danth]

No lack of understanding Rogean; I know exactly where you're coming from. It's simply the difference of perspective between that of the developer versus that of the end-user. As a developer, you can look at the back end and see where the issues lie (in this case, tuning). As a user, all I can do is log on and see that the result of the system as a whole is not as expected. In so many words, when someone like me says "AC is broken," we don't necessarily assume your specific AC code doesn't work in the literal sense. We simply mean putting on more armor in-game has no appreciable effect--for whatever reason.

Frankly it'd be better if the code itself was the issue because it'd probably be easier to fix. Going back and changing the attack values of every last monster in the entire game--or at least the great majority of them--seems like an awful lot of work. Is there anything that can be done globally?

-----------------------------------------

Haynar: The tuning you mention seems like it's potentially on the right track in terms of the behavior I observed. The funny thing I noticed was that while I wasn't taking appreciably more damage with my gear off, the peculiar thing was the rate of max hits was about the same, and much less than every round. Tanking high level monsters near to my level with practically no armor on, the expected result was to get hit for max damage nearly every time. That did not happen; rather the removal of well over 100 worn AC worth of armor, upwards of 200+ at times, had no appreciable effect on the rate of max hits (or damage per minute in general).

Danth

[wycca]

Originally Posted by Rogean [You must be logged in to view images. Log in or Register.]

Vindi's chance to hit a level 60 warrior, after level and skill calculations, is currently around 68%. Evasive is a 50% reduction to chance to hit, which makes it 34%. Are you saying Evasive shouldn't be dropping it that low? It's possible evasive is not being applied at the correct spot. Perhaps only to the mob's base chance before those extra calculations are applied (level and skill).

A level 60 warrior also has around a 6% chance to trigger either Dodge, Parry or Riposte. However, these abilities are all checked separately. That's three separate 6% chances to completely avoid an attack. I'm not convinced this is working as it should.

Let me know your thoughts. This is one of the systems we have to best guess on as we have no idea of the real sequence of processing and math behind the eqlive system.

I wanted to jump in on the defensive skill thing. I was in a pretty unique position to discuss these things with Maddoc while he was working on skill revamps/increases and in regards to various AA. I also did extensive parsing for years. Most of this is from memory, since I long since deleted the emails and I don't think my old SoE PM's are around since they changed forum software. I can try to dig for some self-quotes on Monkly Business, but I know I never talked very much (or at all) publically about fire rates because over time, the monk fire rate for block became quite insane (Maddoc wanted to nerf it).

Check order was -
Block
Parry
Dodge
Riposte

Each check obviously reduced the next skill's overall fires a bit, since a block rate of 10% meant you were checking Dodge on only 90% of the attacks you were checking Block on.

Each one fired the same % for skill, except for Block, which fired exactly 2x that of Parry (and the others).

The fire rate of 6% for Parry/Dodge/Riposte may be about right - at least for a Monk (not sure how different Warrior values are offhand). This would give block a rate of 12% - also seems about right, maybe a tad bit high. I seem to remember alot of sub 5% Riposte parses, while others were higher due to the nesting and Riposte being checked last. They SHOULD NOT be checking independently, the checks are always done in order. Ie, 3x checks at 6% fire rate should not = 18% mob melee hit reduction, it's -6%, -(6% of 94%), -(6% of (94% of 94%).

The neat thing is that this is probably easy to test on live still since (even tho he wanted to) Maddoc never changed how these fired (and I doubt devs post-2008 mucked with those systems). The skill levels did not change much by level at least thru 2008 when I quit, so it should be easy to remove the +% items and Improved focus on live from this era to test values that you can extrapolate comparable values from in your system. I think I still have a lvl 70 monk with 30million hp and GM regen on test leftover from assisting Rashere one time. This character could also be a good option for testing resists (which I think are off here vs. on live, altho there have been resist revamps since then on live, including ones that I think may have made resists less effective than they were in Velious (bards ruled Velious IMO)). Is there any interest in this? Obviously can't quite create the same test on live with a lvl 60 toon as we could here, but can do the reverse - ie cloning the live toon's level/skill values/gear. If we're lucky, maybe the AA was reset at some point since I was on test or maybe I can get a GM to reset it, so may even be able to remove that variant (if its not in code for some reason). Given the buffs on that character, it would be an easy matter to test about any mob in Classic-Velious for defensive skills and resists. If the numbers match at that level in a very long parse, then they should tell you if the system/npc values are fairly close even if we can't test lvl 60's. I just don't know how I might de-level the character without removing the GM buffs or having to explain to a GM that I have some buffs I shouldn't have.

This may also work well for naked AC parsing if that's helpful, altho due to the many AC/returns revamps on live, it won't show anything for geared mitigation past your softcap levels. However, if I went down to say, 0/100/200/300 raw ac, it may be comparable. Hopefully they haven't removed classic mobs or wiped test, since I haven't been on that server/live in forever.

[Splorf22]

Kael Parses

veteran hjrek: shissar (1089 displayed AC)
Accuracy: 56.1%
Hits: 101
Min: 46 4.9%
Max: 168 16.8%
Average Hit: 109

veteran sjrelt: shield of elders (1155 displayed AC)
Accuracy: 60.1%
Hits: 134
Min: 14.9%
Max: 12.7%
Average Hit: 94

a protector of zek: shissar (1099 displayed AC)
hits: 72
Min: 5/72 = 6.9%
Max: 13/72 = 18.1%
Average: 290

a protector of zek: shield of elders (1165 displayed AC)
Hits: 129
Min: 20/129 = 15.5%
Max: 13/129 = 10.1%
Average: 261

looks pretty good here I would say . . . maybe even too good? Anyway, AC clearly works on these two mobs.

is there any way we can copy over the Veteran's atk/str to say the cliff golem and such?

[Sylexis]

NT

[Rogean]

Wycca,

I'm not sure I'm following you. You said that it's not suppose to be checking each of those skills separately.. I agree but I don't understand the method you're suggesting it does check.

[Treats]

I think this is what he means:

Wrong way (%'s made up)
NPC attacks 100 times:

10% Block
10% Parry
10% Dodge
10% Riposte

10 Blocks - 100 attacks
10 Parrys - 100 attacks
10 Dodges - 100 attacks
10 Ripostes - 100 attacks

Right way (%'s made up)

10% Block
10% Parry
10% Dodge
10% Riposte

10 Blocks - 100 attacks
9 Parrys - 90 attacks
8 Dodges - 81 attacks
7 Ripostes - 73 attacks

Pretty sure Kanras fixed this though awhile back.