pagansaint

But, only 51% of the listed effectiveness came through for him while 82% of the parry listed effectiveness came through.

The dodge% effectiveness, 51% would be what I expect, with that small of a test set, after being transferred a 60% lending buff. But the parry effectiveness seems inordinately high considering it is going through the same lending buff. But that is just a limited test set that could be influenced heavily by random luck, for lack of a better phrase, even though it is a trend of some avoidance types being more effective, even in uncontested forms, than others.

What I would really like to know, for peace of mind if nothing else, is whether or not the differing avoidance% items and abilities are added into like groups then applied as avoidance or if they are applied each separately. This is what I think causes the most confusion to the actual worth of some items compared to what could be used instead.

__________________

Enoa

I have to appologize there was a fundamental error in my calculation. the 82% caculation of "effectiveness" was based on a 6.71% of AVOIDS parried. If instead i would have used the 5.26% of SWINGS parried you would get 64.1% effectiveness (5.26/8.2). The reason the (l"ater checked") dodge appears < 60% is because it's not seeing as many swings as parry (if it is indeed checked after parry). Assuming my dodge is checked last it was only "exposed to" 205 swings (912 swings - 715 avoids + 8 dodges) = ~ 3.9% passed (8 dodges/ 205 swings) -> almost 200% effctiveness. Of course my numbes are small and the assumption that my dodge is checked last might not be correct. I'm just illustrating that the %'s displayed in the avoidance calculators are based on TOTAL Swings or avoids not what that skill is checked against depending on the sequence of checks.


In my sleep deprived state I entertained the idea that parry might be better if it's checked first but in retrospect Pinski is 100% correct. (i've always used dodge drink over parry cause it has stats and i assumed the skills were equivalent)

Ignore SA for a sec and assume you are a figher with only parry and dodge skill
Scenario 1: 50% parry and 25% dodge skill with 100 swings
= 50 swings parried -> 50 swings go to check against dodge -> 12.5 swings dodged = 37.5 swings HIT
NOTE: ACT would report this as 50% of swings parried and 12.5% of swings dodged
Scenarior 2: 25% parry and 50% dodge with 100 swings
= 25 swings parried -> 75 swings checked against dodge -> 37.5 swings dodged = 37.5 swings HIT

Talk about beating the dead horse.

Enoa

__________________
Enoa - Templar - Eevo

pagansaint

Order of application only creates an appearance of one type being more effective, like Pinski posted with the 90% and 90% example.

And its my bad for not going back and actually running the numbers myself, just quoting you on them, I should have caught that, but at that point I was rather sleep deprived also.

But... I take it no one has found anything that would suggest avoidance% pieces are added into like groups then applied or if each piece is separate?

I get roughly a 5% difference in what the possible uncontested avoidance is for tank classes between the added then applied and applied separately methods. The added then applied method gets the higher result on paper.

__________________

Tynx

I am not going to act like I know it all but IMHO what I gather on the subject is from ACT avoidance reports and what I seen in the following:

Game Update #29: November 14, 2006 (Echoes of Faydwer)

Avoidance in
Combat

- Display: The way avoidance is
displayed has changed. A value similar to mitigation is now used to show your
overall avoidance. A tool tip displays details for those who desire more
information.

- Parry Check: (Base 5%, Chance increases with additional parry skill.) The chance to parry an incoming attack is modified by level and offense skill of the mob you are fighting. Note: A base 20% of parries turn into ripostes. (Modified by certain achievements)

- Block Check: (Shield Required). The quality of the shield determines the block chance. Block is only modified by mob level and is not a contested roll vs. mob skill.

- Deflection Check: (Monk / Bruiser only) Base 25%, Chance increases with additional Deflection skill. The chance to deflect an incoming attack is modified by level offense skill of the mob you are fighting.

- Dodge Check: (Everyone) Base chance is determined by type of armor worn. Chance increases with additional Defense Skill and Agility. The chance to dodge an incoming attack is modified by level and offense skill of the mob you are fighting.


According to SOE it leads one to believe that avoidance is checked in this order. Parry/block/deflection/dodge. It gives a base to certain avoidance skills that leads one to believe there is a guarantee to that skill.
( IE: base of 5% parry is like a for sure chance if the skill is above the base of 5% )
Which in return makes dodge < parry < deflection < block. Why I put block at top? Those who use shields usually get their greatest source of avoidance from block and block is not contested to mob skill. For avoidance reports not showing this order is beyond me. Dodge showing up before parry you got me there again but according to SOE parry > dodge. Also without %dodge items/food/drink dodge will not show up for SA or fighter avoidance in my experience.

__________________

Tyla

I may have misunderstood this, but since Templars don't have parry skills without using % based provisions/adorns etc. this doesn't count for us - right? On that basis where does this leave the parry vs. dodge argument? I mean, if both skills give you a % chance to completely avoid being hit, I don't see how one can be called > than the other? Can it?

__________________

gorlab

Can you have two templars in a group and both have SA going on the MT? I realize they wont stack, but will the check hit first one than check against the other? Thanks for your time.

jdeborde

So, at what point does healer stats (+heal crit) outweigh SA avoidance stats.

Example: Mastercrafted that gives %reposte vs Vyxx crest locket

EQAditu

If you're lazy and just look at the straight numbers, they won't... ever. The calculated average amount that Shield Ally avoids is always greater than my total healing amount(our defiler doesn't suck and I take SA seriously).

But if you want lazy numbers that don't really mean anything... the amount of damage ACT calculated riposte avoiding was equal to +36% critical healing in a full VP run. That is if I critical healed 100% of the time and no critical healing was ever wasted(I'm not a shaman). So in super crazy impossible land, 2% riposte was equal to 36% of the total critical healing.

Now to explain why part of that is lazy. When you do avoidance for someone successfully, it wasn't assured that they were going to be hit. The mob may still have missed them, so you cannot claim 100% responsibility for the "calculated average damage avoided" that ACT generates. When an avoidance check is early in the list, it will be successful more often, but it doesn't mean it is better or that the later checks are worse. Since ACT does not know the avoidance check order, it does not re-balance the values.

In summary, Sheild Ally is extremely powerful as long as there is still an amount of avoidance that can still be added to the target. We can only reach 100% avoidance after all.

Avoidance > Healing.

TheOtherDude

But weren't you counting the trash fights where you weren't healing at all (or very, very little)?

While I have found that the riposte necklace is better than anything else for the MT group, I'm also pretty sure that the Jade Restoration Pauldrons are better than the Mastercrafted shoulders with +1% block, and that the VP-set and PR bracers are better than the +1% block/riposte you can get for that slot.

My guild's MT doesn't have as high avoidance as the top-end guilds, but by pure math coincidence, it seems that 1% uncontested avoid for the cleric is roughly equal to 1% less damage for the tank, using ACT's approximations. If anything, that's more in favor of Shield Ally than it actually is, and as our MT's avoidance increases, Shield Ally will have less of an effect.

Typically, the MT takes something like 1500 incoming DPS on named fights, and sometimes it gets up to around 2K or something. So that means those 1% uncontested Avoidance items are gonna be worth roughly 15-20 HPS (by reducing the damage the tank takes by about that much DPS). The Runic Cover proc on the Jade Pauldrons are generally worth at least about 30 HPS on the tank, so unless our tank is taking ~3K+ DPS, the Jade Pauldrons would be better, right? The bracers seem maybe a little more debatable since you can't get a clear HPS number for them, but they definitely should be close, and they both have much better HP and stuff than the Mastercrafted.

Just trying to check my math...

quasigenx

The math is interesting, but besides the point IMO. The reason I wear those items over anything else is not because they reduce the incoming DPS by 1%, but because they can completely block one big hit that can kill the tank. It's the variability in the incoming DPS that I'm trying to reduce, not the average.

Right now, this is only worth-while in VP and against the final named in T2 and T3. Otherwise, I wear more DPS oriented gear.

__________________