In the current game, targets higher level than you gain 8 resistance to all per level. For a level 60 versus a level 63 boss, this is 24 resist to all.

This is my current understanding of this feature/bug in the current game:

This resistance can be removed with -resist gear but ONLY for binary spells. For non-binary spells, no ammount of -resist gear or warlock curses removes the 8*levelDiff resists and thus, you always get some partial resists. Furthermore, DoTs are binary spells, and the -resist can increase the chance for the DoT to land but the DoT will still get partial resists on the DoT ticks. Thus another way to explain it is that -resist gear can affect the chance to hit check on binary spells due to this resistance, but not the partial resists.

There is a topic about this with respect to the current way it works in game. I searched around but could not find a link to the topic. In any event, it is long and not really the subject I want to talk about, it is simply reference for the current mechanics.

--->
#1 Has this "unremovable resist" mechanic changed in TBC?
#2 Does it still affect binary and non-binary spells differently?
#3 Now is the time! If we as a community want to attempt to change this or make sure that at least binary and non-binary spells are consistent, lets gather the data and post on the beta forums in a constructive way.

I just got a TBC beta account, and I have a (currently) level 60 warlock with up to -60 to resist gear to experiment with. What tools are best to parse the combat logs and get some good data? Where are some level 63/64 ish mobs to experiment on (I've played beta for about 45 minutes so far)? Anyone else interested in this?

#1 still exist as far as I can tell
#2 I don't have enough data to answer this
#3 I just released a tbc version of my resist tracker this weekend. It's still in beta phase until I'm sure everything works correctly. You can get it from curse: http://www.curse-gaming.com/en/wow/a...sttracker.html

From what I recall, all of the high precision (many thousands of casts) measurements have turned up with a 6% reduction in damage for a level 63 boss due to partial resists no matter how much penetration gear you are wearing, which is 24 resistance.
Data on level 61 and 62's is not as solid with fewer casts, but seems linear. My memory of this is from posts both on these forums and on the WoW forums.

I've dug around, and it sems that the EJ forums don't have as much info as elsewhere. Lets start here:

http://forums.worldofwarcraft.com/th...12051812&sid=1

Several useful links on that.

Included is how these resistances CAN be penetrated by frost mages, though -resist gear is still useless for them since Curse of Elements is plenty to cover the 24 resists on a raid boss.

Level-based resist is definitely still there in Beta (tested with my Paladin over few hundred hits). I can't test to see if spell penetration removes it, because I don't have any items with that stat. I think it's intentional though, since the numbers now correspond with melee glancing blow mechanic (which is also unremovable).

If you want, Ango'rosh Ogres in Zangamarsh make good test subjects. No annoying abilities and quick spawns. As for a tool to analyze the data, /combatlog and I can write up a quick script to parse the data if you don't find anything else more appropriate.

Kavan obtained quite a bit of data and posted his results here.

Obviously, that data is likely obtained across various game versions, but seemed reasonably sound and corresponded to the small amount of data I collected on various level 61-63 level targets.

Presumably the various tests will have to be redone in the expansion. I think Kavan wrote himself a personal add-on for testing purposes, so maybe if we get a bunch of casters to use it we can pool the data.

Could you quote your source for this please?

Every parse I have ever done implied that frostbolt simply didn't suffer this effect at all.

It's 5 resistance per level, cannot be penetrated by any spell or gear. 5 resistance/level makes sense, since it requires 5 resistance per level of caster to gain maximum damage reduction against their spells. (300 resistance vs 60, 315 vs 63)

Chances are, it will stay like this in the expansion.

"159 samples"

"Average partial resist: 3% (expected 4% from 16 level based resistance) 475 samples"

"Average partial resist: 3.9% (expected 0% from 16 level based resistance and countered by penetration)"

Shoot many, many more bolts and stuff, you're fudging your math to meet 8 resistances per level.

And the binary spell thing is false too IMO, look at the numbers you've shown yourself.

"In addition you have this value converted into respective resistance score, which is computed as (1 - value) * 400."

Since your formula is based on damage, and the 2 Curses add 10% damage (along with other debuffs on the mob), you really sould factor the damage increase in your formula.

@Maels:

If your proof for 5 resistance per level is that it makes more sense, then you'll have to do better than that.

The point of the Core Hounds example that you are quoting is not in quantity. It is qualitative example showing that these kind of resistances cannot be countered. The result of 24 resist for 3+ levels is based on 10k+ samples and cross verified with others collecting the data. It doesn't take that much data to distinguish between 24 and 15.

We know binary spells can counter the level-based resistances. The only thing that we don't know is whether they don't suffer from it in the first place or if penetration works for them.

And the formula I gave doesn't work on damage directly, but on partial resist counts.

As far as I know this conversion of collected counts to resistance in absence of full resists is correct.


@Zoner:

It definitely applies to non-elites and non-instance mobs. We also know that the mobs actually don't have any resistance by use of Beast Lore. So it might be that it only applies to 60+. But I haven't seen anyone actually collecting resistance data at lower levels. If you have some numbers as to what number of samples you have without any partial resists that would be very interesting to see.

Nah, From what I understand, the combat log clearly indicates whether it is a 25%, 50%, 75%, or 100% reisist. No effects from Curses or other amplification affect the calculation at all.

The 10% boost from curse of elements/shadow does not affect the % resisted. It is still in one of 5 buckets of damage, with a simple formula to add them up and get a total resist % independant of such damage boosters, the gear of the player or any knowledge of how much damage a spell should do in the first place.

% lost to partial resists is:
Num casts no resist A
Num casts 25% resist B
Num casts 50% resist C
Num casts 75% resist D
Num casts 100% resist E
num casts N = A+B+C+D+E

100% resists can be due to the % miss caused by level differences. Since at low values of resist (below 75 resist at level 60) these are never due to parial resists, you can determine the % lost to partial resist like this:

%lost = (B*25 + C*50 + D*75)/N

For large values of N this trends to 2 % per target level above you.

An open question is whether it is actualy 8 resistanceper level, which is 2% mitigation at 60, or if it is 2% mitigation per level.
Since at level 70 8 resistance will be less than 2% mitigation, it may simply be 2% rather than 8 resists.
In addition, at very low levels 8 resist is a lot. at level 20, 8 resistance is 6% mitigation.

Actually the combat log has the resist values displayed before amplification and crit bonus (which is an indicator on the ordering of computations they are doing).

You'll see entries like crit for 1650 (250 resisted). This could be for example shadowbolt with 2x crit multiplier and COS up. Normal hit would be 1000, but 25% got resisted. The remaining 750 gets amplified by x2 from crit and x1.1 from COS adding up to 1650.

If you're actually very precise you'll notice that numbers are often off by 1 or 2 damage, depending on number of amplifiers. If you calculate everything out it looks like they apply each amplifier separately and round to integer after each one.

ResistTracker knows your talents, and tracks all debuffs on mob, so it knows the current amplification and crit multipliers which it uses to compute the correct %reduction. After tons of data collected I've never seen an entry in combat log where resist wouldn't be in increments of 25%.

As far as 2% vs 8 resistance question goes, the mod keeps data separated based on what level you collected it at and uses an appropriate conversion from percentages obtained to resistance score. So it will be possible to compare data for the same mob obtained at different levels.