Combat Mechanics, 3.0 - Page 11

dfinberg · 06/02/06, 4:23 PM

That's the point wodin. How often are you getting crit/crushed? Crit and dodge should be the only things in the table. If you can push your dodge past 85% (is that reasonable, I don't have a rogue), either dodge or crit is getting dropped from the table. Unforunately, record keeping for 19 seconds every 5 mins is going to blow building up a large sample, unless you can get up to 95% dodge or so. Take 5-6 rogues to mag and let them each take shots and get rezzed for 15 minutes one night.

Deathwing · 06/02/06, 4:27 PM

Thorb, that system you're using is somewhat similar to what I use(see the bonescythe thread), but I think you're getting a little lost. First, I think calculating attacks/flurry is going to lead you in the wrong direction because the two affect each other directly. Basically, I went on the principle of "did I crit in my last three swings?". Here's what I did: there's your basic flurry chance formula, which you already posted. That formula is independant of weapon speed, so no need to recalculate it. That takes care of all your crits from melee swings, leaving just your instant attacks. I'm not 100% sure on this model, but the flurry speed of my weapons from the previous calc and then applied to see how many attacks there would be between the cooldown of each instant attack(BT, WW, and hamstring). So, if you have 8 swings between each WW, and a crit chance of 25%, I took that to mean each WW would add 3/8*.25 to your flurry chance.

I add the flurry chance from the instant cooldown method to the flurry chance from previous, and that's your final flurry chance and your final weapon speed to use in other calcs(like overpower modeling). The biggest flaw I see in this method is that it doesn't account for flurry resetting itself from the cooldowns.

Twid · 06/02/06, 5:03 PM

It's similar to calculating improved shadowbolt for warlocks, in that when we crit, the next 4 sources of direct shadow damage are increased by 20%. Using the forumla that Moleva suggested in my Caster DPS thread: (1 - (1-Crit Rate)^4)% of attacks benefit from the increased attack speed.

So with a 25% crit rate, you'd have 68% of attacks flurried.

Deathwing · 06/02/06, 5:37 PM

Originally Posted by Twid

It's similar to calculating improved shadowbolt for warlocks, in that when we crit, the next 4 sources of direct shadow damage are increased by 20%. Using the forumla that Moleva suggested in my Caster DPS thread: (1 - (1-Crit Rate)^4)% of attacks benefit from the increased attack speed.

So with a 25% crit rate, you'd have 68% of attacks flurried.

Yes, we have the part covered already though. It would be 1 - (1 - C)^3, btw. The hard part is how much instant attack crits would help.

Kenco · 06/02/06, 9:27 PM

You could get crit / crushed while shield block is up if
1) Your shield isn't out (even if it's equipped). This can happen e.g. after some emotes
2) You aren't facing the mob. You can only block from infront / 180 degrees.

The reason why block has a lower priority than miss / dodge / parry is that your m/d/p rates are unaffected by shield block. This isn't too hard to test, give it a try.

Thorb · 06/03/06, 3:12 AM

One easy way to see statistic is to start the combat from 0. What we are looking for is the average % of uptime once it get going.

The chance that flurry get off for a specific second is the number of attack I have in that second * crit rate or 1-(1-c)^A. A flurry "cycle" will last between 3 times the weapon speed and 3 times the hasted weapon speed. The exact time for that cycle will actually be 3 times the weapon speed * average haste, which is the uptime I'm looking for. An easy solution to this is to take the max haste as the weapon speed for a cycle. This tend to hinder the uptime a bit, since this doesn't affect all instants. The thing I'm worried and can't really calculate atm is the overlap, same problem we have with crusader. I don't think it's that bad but probably account for a lost of 10 to 20% efficienty.

That's why I think using a cycle is more generic then using the last 3 attacks. I'll probably apply the haste on normal attacks at a disadvantage to flurry to escape my circular reference but that could account for overlap and I'll call it even.

I could so a combat simulator but that's even more time then doing it in excel, probably not worth it.

I still have to do deep wounds, heroic strike and excutes and somehow "easily" allow the user to prioritize attacks based on amount of available rage.

Vanick · 06/06/06, 3:59 PM

So if the priority indeed looks like Block > Crit/Crush > Dodge/Parry, what happens if you shield block but do not have 100% block chance? On a 63 mob if you do not block it would appear that it would be a guaranteed crit/crush, assuming you have ~85% block with shield block up.

Beliandra · 06/16/06, 2:40 AM

Heya,

Was just taking a dig through this thread after having a discussion with a guildie who swears that melee crits generate more threat than non-crits for the same amount of damage. I didn't think that was the case. Anyway, as far as I can gather, there was something about that in the original post in this thread, but then it was removed?

What is the current state of research into this? Does a crit for x damage generate more threat than a hit for x?

• Hamlet · 06/16/06, 3:05 AM

Originally Posted by Beliandra

Heya,

Was just taking a dig through this thread after having a discussion with a guildie who swears that melee crits generate more threat than non-crits for the same amount of damage. I didn't think that was the case. Anyway, as far as I can gather, there was something about that in the original post in this thread, but then it was removed?

What is the current state of research into this? Does a crit for x damage generate more threat than a hit for x?

That was true (or at the very least, thought to be true) for a while, but I believe it's currently known not to be the case.

james · 06/29/06, 3:03 AM

One of our guild warriors is of the belief that dodge + parry + block is effectively "capped" at 60% and that any additional dodge + parry pushes your block down, except when shield block is used. Is this accurate ?

Kalman · 06/29/06, 9:28 AM

James: No, any rogue can tell you that's false. See: Evasion.

Maledict · 06/29/06, 9:53 AM

One of our guild warriors is of the belief that dodge + parry + block is effectively "capped" at 60% and that any additional dodge + parry pushes your block down, except when shield block is used. Is this accurate ?

No, this is an old rumour that was proven false by various tests. Unfortunately, I cannot remember the link that showed how it was false - but rest assurred, it's a dead horse issue. They aren't capped.

sp00n · 06/29/06, 5:26 PM

Yeah, tested sort of this with a recklesness vs evasion test.
A warrior popped recklesness, I popped evasion.

Not one single hit (or even a crit) ^^

Can post the screens if anyone wants them (and if I find them).

Skat · 07/26/06, 6:14 AM

I've taken the liberty of copy-pasting you entire post into this wiki:
http://tpr-wow.wiki.xs4all.nl/index....mber_Crunching

I hope you don't mind. :S

novasphere · 07/26/06, 7:04 AM

Quick addendum for "Healing Aggro":

In addition to Dark Pact, Warlocks generate continuous "Healing" threat through each individual tick of Fel Stamina or Energy through Demonic Sacrafice.