Elitist Jerks
Register
Blogs
Forums


Go Back   Elitist Jerks » Class Mechanics

 
 
LinkBack Thread Tools
Old 01/10/08, 1:02 PM   #76
Moogul
Soda Popinski
 
Moogul's Avatar
 
Human Rogue
 
Argent Dawn (EU)
As far as I can tell, from the way the mechanics have been explained, BrTarolg is completely right.

The idea seems to be that to get a 'free ignite' tick, you have to have it tick during the flight time of the next fireball - that way, it ticks AFTER the new ignite damage X has been calculated ('rolling' in the damage from the last one) but before the current ignite damage has been reset to X.

Ie.
first fireball finishes casting, calculated to crit for N. Ignite damage X = N * 0.4
first fireball actually arrives, ignite starts countdown to first tick
second fireball finishes casting, calculated to crit for M. New Ignite damage will be M * 0.4 + X
Ignite ticks, for 1/2 of X
second fireball reaches target, ignite starts countdown to first tick, damage set to M * 0.4 + X

etc.

So, the ignite will tick for the first time at (travel time + 2) seconds after the first fireball finishes casting. You therefore need to finish casting the second fireball before (travel time + 2), but after (travel time + 2 - travel time) = 2 seconds.

Ergo, 2 < (lag + cast time) < travel time + 2.

In a hypothetical 0 lag scenario, you want a cast time just under 2 + travel time. If travel time was 1s, then a 2.9 cast fireball would be enough to roll. If travel time is only 0.5s, then you'd need ~2.4

Great Britain Offline
Old 01/10/08, 1:07 PM   #77
drowsy
Von Kaiser
 
Human Mage
 
Dark Iron
If the theory is that this will occur when a fireball is in the air during an ignite tick, why not shoot for the second tick for testing purposes? Perhaps macro a G15 to try and time casts to happen around the second tick to get a better sample size than what heroism/icy veins allow.

Offline
Old 01/10/08, 1:32 PM   #78
Cardynal
Piston Honda
 
Orc Warrior
 
Mug'thol
Originally Posted by Pallandor View Post
text
Hrm...i'll look into it. Maybe I missed a crit above it somewhere that hadn't finished it's ignite. Please disreguard the log then until I can look into it further.

Offline
Old 01/10/08, 1:42 PM   #79
manly
Soda Popinski
 
manly's Avatar
 
Troll Mage
 
Mal'Ganis
BrTarolg is correct, and that is what I came to conclude in the post #62.

Let me point something out. This is the same thing I was saying in post #1. But I think some people will find it easier to understand.

At the time your fireball is done casting, the game calculates the ignite damage. Lets say 400 ignite damage (2x 200dmg ticks) are left from the previous ignite. Your fireball that you just finished casting will crit for 1000 dmg. What my theory states is that basically, at that point in time, the game would calculate that for the current fireball, the total ignite should be 400 (prev ignite) + 400 (current fireball ignite) = 800 dmg. Try to think of it like 'when the fireball lands, it will totally overwrite whatever ignite was in there, and apply whatever the current fireball contains (in this case, 800 dmg)'. If you get an ignite tick while your fireball is underway, then that is just free damage, the game doesn't care about that. When the fireball left your hand it made perfect sense to conclude that you were meant to get a 800 dmg ignite total when that one will land.

Now imagine my theory is correct. It explains the current 4th combustion crit behavior. Think about this case:

You have 10x combustion. You have 100% crit rate for your next '1' spell. You cast a fireball. That fireball will crit.
When you casted your fireball, according to my theory, the game concluded the fireball ignite should be 400 (previous ignite = 0dmg, current fireball = 1000 dmg). When that fireball lands, it will apply a 400 dmg ignite (2x 200 dmg ticks). But while your fireball is in mid-air, you fireblast, because combustion is still up - you haven't consumed it yet, because fireball has not landed. Your fireblast, when casted, has 100% crit rate. Your fireblast crits for 500 dmg. That fireblast applies a 200 dmg ignite (2x 100 dmg). After your fireblast lands, combustion fades. Then your fireball hits your target. Since the fireball damage was calculated when it was finished casting, you will still get 100% crit rate from combustion despise the fact it isn't there anymore (same as shatter combo). So your fireball crits for 1000 dmg.

Here is the key part: According to my theory, the fireball contains a 'payload' of 400 dmg ignite. You might currently have a 200 dmg ignite going on from the fireblast. But the game doesn't care. At the time you casted your fireball it was not there, therefore it is not calculated as part of the ignite that will be applied (and will override whatever ignite was there before).

The end result:
-> fireball is done casting, leaves your hand.
-> fireblast crits for 500 dmg.
-> fireball crits for 1000 dmg.
-> ignite 200
-> ignite 200
(oops fireblast ignite disappeared)

Using the very same principles, you can abuse the same mechanic in the reverse as I pointed out in the first post. If you get an ignite tick on a mid-air fireball, that one is free damage. If you fireblast during a mid-air fireball, then , assuming both crits, you will always get no ignite damage from your fireblast.

Last edited by manly : 01/10/08 at 1:50 PM.

<Eej> YOU"RE GONNA PULL
<Eej> IF YOU SQUEEZE OFF ANOTHER ARCANE BLAST
<Spectear> You've obviously never played with Manly.
<Spectear> That's hardly a reason to stop DPS.
Very Manly Staff

Canada Offline
Old 01/10/08, 2:52 PM   #80
Kaimani
Glass Joe
 
Troll Mage
 
Feathermoon
Not sure if this has been pointed out already, but this whole theory also explains why rapidly critting with a fireball and fireblast simultaneously will often 'erase' one or the other's ignite damage based on precisely when they land. (edit: durr, pointed out in post right above this one)

I'm still waiting on our raid leader to post our WWS report form last night so I can plumb through the logs, but on a couple of fights I was seeing some truly outrageous ignite ticks during IV + troll zerking. I'm pretty certain I was getting some free damage even without getting close to 2 second fireballs - in fact, unless my logic is flawed, it seems like rolling ignites should be possible provided your cast time is right around 2 seconds + fireball travel time. Take for instance a 2.25 second fireball cast from 0.25 seconds away. In this theoretical timeline I'm adding an extra tenth of a second due to the tiny delay between a fireball crit and when the ignite is actually applied)

0.00 - fire fireball leaves hand with ignite calculated
0.25 - fireball crits
0.35 - ignite debuff is applied (there's always a slight delay between when a spell crits and the ignite actually appears and starts counting)
2.25 - next fireball crit leaves hand
2.35 - ignite tick (free damage yay!)
2.50 - fireball crits with both ignites added together
2.60 - ignite is refreshed
4.50 - next fireball crit leaves hand
4.60 - ignite tick(even more free damage!)

etc.

Offline
Old 01/10/08, 4:37 PM   #81
Rudi-CO
Glass Joe
 
Gnome Mage
 
Baelgun
Originally Posted by Moogul View Post
So, the ignite will tick for the first time at (travel time + 2) seconds after the first fireball finishes casting. You therefore need to finish casting the second fireball before (travel time + 2), but after (travel time + 2 - travel time) = 2 seconds.

Ergo, 2 < (lag + cast time) < travel time + 2.

In a hypothetical 0 lag scenario, you want a cast time just under 2 + travel time. If travel time was 1s, then a 2.9 cast fireball would be enough to roll. If travel time is only 0.5s, then you'd need ~2.4
So... what we need is a mod that will display the required cast time needed for Fireball based on the distance to the target mob and your lag. Or a mod that will display the required distance to the target mob based on your haste rating and lag.

Offline
Old 01/10/08, 4:39 PM   #82
Cardynal
Piston Honda
 
Orc Warrior
 
Mug'thol
Casting fireball at 2.0 seconds w/ 0 latency

Assuming it takes 1 second for fireball to land once casted.

00:00 casting fireball a
02:00 fireball a finished casting
02:00 casting fireball b
03:00 fireball a crits target
04:00 fireball b finished casting
05:00 Ignite (20% of a)
05:00 Fireball b crits target
07:00 Ignite (20% of a + 20% of b)
09:00 Ignite (20% of a + 20% of b)

Assuming it takes 1.5 seconds for fireball to land

00:000 casting fireball a
02:000 fireball a finished casting
02:000 casting fireball b
03:500 fireball a crits target
04:000 fireball b finished casting
05:500 Ignite (20% of a)
05:500 Fireball b crits target
07:500 Ignite (20% of a + 20% of b)
09:500 Ignite (20% of a + 20% of b)

Assuming it takes 0.5 seconds for fireball to land

00:000 casting fireball a
02:000 fireball a finished casting
02:000 casting fireball b
02:500 fireball a crits target
04:000 fireball b finished casting
04:500 Ignite (20% of a)
04:500 Fireball b crits target
06:500 Ignite (20% of a + 20% of b)
08:500 Ignite (20% of a + 20% of b)


To me, it looks like the distance you stand is irrevelant to this theory. This is because there are only 2 things that matter to this theory

1) The delay between hits...which is only affected by the cast time and not the flight time (assuming you're standing still).

2) That your second fireball has finished between your previous fireball critting and the first ignite tick of that fireball.

I'm sure there is some sort of limit to this as i'm not very comfortable with the 2.5 second flight time...but you're not going to see that long of a flight time anyway at 41 yards.

Last edited by Cardynal : 01/10/08 at 5:01 PM.

Offline
Old 01/10/08, 5:44 PM   #83
Kewangeder
In the Rafters
 
Night Elf Druid
 
Gilneas
I think you're right to an extent, Cardynal - distance to mob is irrelevant, but only in the situation you outlined: the mob is stationary, and the mage chaincasts 2.0-second fireballs (reasonably) perfectly.

If the cast time is the usual talented, unhasted 3 seconds, distance to mob matters very much. I'll even delay B a little:

(1.5 second travel time)
0.00 fireball A cast starts
3.00 fireball A cast ends (igniteTic = 0.2N, where N=normalFireball)
3.01 fireball B cast starts
4.50 fireball A lands
6.01 fireball B cast ends (igniteTic = 0.2N + 0.2N)
6.50 ignite A tics, 0.2N damage
7.51 fireball B lands
9.51 ignite B tics, 0.4N
11.51 ignite B tics, 0.4N
total ignite = 1.0N

(0.5 second travel time)
0.00 fireball A cast starts
3.00 fireball A cast ends (igniteTic = 0.2N)
3.01 fireball B cast starts
3.50 fireball A lands
5.50 ignite A tics, 0.2N
6.01 fireball B cast ends (igniteTic = 0.2N + 0.1N)
6.51 fireball B lands
8.51 ignite B tics, 0.3N
10.51 ignite B tics, 0.3N
total ignite = 0.8N

See the difference? An ignite from A squeaked in during B's travel in the first example. Suppose the mage is moving (laterally, let's say, to flee a nearby rooted mob), and doesn't chaincast as a result:

(0.5 second travel time, caster isn't chaincasting)
0.00 fireball A cast starts
3.00 fireball A cast ends (igniteTic = 0.2N)
3.50 fireball A lands
4.10 fireball B cast starts (very late!)
5.50 ignite A tics, 0.2N
7.10 fireball B cast ends (igniteTic = 0.2N + 0.1N)
7.50 ignite A tics, 0.2N
7.60 fireball B lands
9.60 ignite B tics, 0.3N
11.60 ignite B tics, 0.3N
total ignite = 1.0N

To me, what's interesting here is that slow casts don't even matter, even when the first ignite used up a tic - all that matters above is whether you got ignite to tic during flight time. Which is what Manly is saying.

Note, however, that in the latest example, you're in worse shape when it comes to fireball C. The best ignite tic C can have is 0.5N; in the top example, it could have 0.6N. To put it another way, example 3 can only give you 0.3N extra damage at best, vs. 0.4N from example 1.

(I -think- I got all the numbers right...)

Offline
Old 01/10/08, 5:54 PM   #84
Kewangeder
In the Rafters
 
Night Elf Druid
 
Gilneas
Drawing further tentative conclusions:

- Max range unhasted fireball casting should always guarantee you a free ignite tic provided an ignite is present.
- The faster your cast time compared to travel time, the greater the chance that you will add both ignite tics to your next one, instead of just the second one, which grows your free tics twice as fast. (I.e., spell haste is very useful.)
- The closer you are to a mob, the longer and more precisely you must wait to time a cast so that an ignite tic falls within the travel window, which in turn lowers your overall DPS.
- The higher your crit%, the bigger an ignite you can grow before your chain ends and ignite ultimately falls off (and resets), just as it was for Naxx era rolling ignites.
- Crit rating is slightly more valuable relative to other stats than previously modeled, when free ignites are considered, since only crit% will increase the uptime of ignite. Exactly how much is unknown at this writing.

Last edited by Kewangeder : 01/10/08 at 5:59 PM. Reason: accuracy

Offline
Old 01/10/08, 7:42 PM   #85
houlia
Glass Joe
 
Human Mage
 
Kirin Tor
I did some testing, at 35 yards, fireball travels about 1.7 seconds in the air, at max range it seems to go to almost 2 seconds (although not quite).

That said, using Dr. Boom I was unable to get this to actually happen - perhaps it only occurs when the server is under a heavy load. Has anyone actually been able to create this in a controlled environment?

Offline
Old 01/10/08, 7:57 PM   #86
manly
Soda Popinski
 
manly's Avatar
 
Troll Mage
 
Mal'Ganis
Well, from the limited testing I did yesterday, it seems like the travel time we see does not match the 'real' travel time. It seems like you have a 100-200ms window of opportunity, if you get much above that you don't get anything. I know this sounds vague, but basically it seems like the travel time we see client side does not match the travel time on the server.

<Eej> YOU"RE GONNA PULL
<Eej> IF YOU SQUEEZE OFF ANOTHER ARCANE BLAST
<Spectear> You've obviously never played with Manly.
<Spectear> That's hardly a reason to stop DPS.
Very Manly Staff

Canada Offline
Old 01/10/08, 8:17 PM   #87
Mahonri
Glass Joe
 
Mahonri's Avatar
 
Gnome Mage
 
Perenolde
Grecasi's log here should be enough for us to calculate fireball travel time at whatever his distance was. Every time Grecasi's successfully casts fireball he gains combustion. There is lag between the server conceding that he cast and the client registering it, but the same lag will apply to the fireballs hitting Mother.

23:43'21.061	Grecasi's Fireball dots Mother Shahraz for 43 Fire damage
373	Grecasi's Fireball crits Mother Shahraz for 6553 Fire damage
23:43'22.404	Grecasi gains Combustion
404	Grecasi gains 128 Mana from Master of Elements
967	Grecasi gains 20 Mana from Mana Spring of Mana Spring Totem V
23:43'23.359	Grecasi's Fireball dots Mother Shahraz for 43 Fire damage
936	Grecasi's Fireball hits Mother Shahraz for 4443 Fire damage
23:43'24.061	Grecasi's Ignite dots Mother Shahraz for 1311 Fire damage
826	Grecasi gains Combustion
920	Grecasi gains 20 Mana from Mana Spring of Mana Spring Totem V
23:43'25.875	Grecasi's Fireball dots Mother Shahraz for 43 Fire damage
23:43'26.125	Grecasi's Ignite dots Mother Shahraz for 1310 Fire damage
328	Grecasi's Fireball crits Mother Shahraz for 6782 Fire damage
953	Grecasi gains 20 Mana from Mana Spring of Mana Spring Totem V
23:43'27.279	Grecasi gains Combustion
281	Grecasi gains 128 Mana from Master of Elements
23:43'28.203	Grecasi's Fireball dots Mother Shahraz for 43 Fire damage
391	Grecasi's Fireball crits Mother Shahraz for 5418 Fire damage
920	Grecasi gains 20 Mana from Mana Spring of Mana Spring Totem V
23:43'29.264	Grecasi gains 128 Mana from Master of Elements
23:43'30.484	Grecasi's Fireball dots Mother Shahraz for 34 Fire damage
922	Grecasi's Ignite dots Mother Shahraz for 2440 Fire damage
951	Grecasi gains 20 Mana from Mana Spring of Mana Spring Totem V
23:43'32.422	Grecasi's Fireball dots Mother Shahraz for 34 Fire damage
920	Grecasi's Ignite dots Mother Shahraz for 2440 Fire damage
969	Grecasi gains 20 Mana from Mana Spring of Mana Spring Totem V
23:43'34.451	Grecasi's Fireball dots Mother Shahraz for 34 Fire damage
Pertinent Info is:
23:43'22.404 Grecasi gains Combustion
23:43'23.936 Grecasi's Fireball hits Mother Shahraz for 4443 Fire damage
23:43'24.826 Grecasi gains Combustion
23:43'26.328 Grecasi's Fireball crits Mother Shahraz for 6782 Fire damage
23:43'27.279 Grecasi gains Combustion
23:43'28.391 Grecasi's Fireball crits Mother Shahraz for 5418 Fire damage

The differences are:
1.532
1.502
1.112

This should be the travel time.

Offline
Old 01/10/08, 8:55 PM   #88
PSGarak
Bald Bull
 
PSGarak's Avatar
 
Undead Warlock
 
Hyjal
Originally Posted by manly View Post
Well, from the limited testing I did yesterday, it seems like the travel time we see does not match the 'real' travel time. It seems like you have a 100-200ms window of opportunity, if you get much above that you don't get anything. I know this sounds vague, but basically it seems like the travel time we see client side does not match the travel time on the server.
Some observation on missles, speed, and animation:
When a mob casts a missle spell at me as I'm riding past, I register the damage when the missle has reached where I was when he cast it. The missle continues to travel though, and does not display it's impact animation until the spell has reached my actual location. Based on accounts of people outrunning spells on epic mounts, it's possible this is a Mob-On-PC interaction only and that mob and PC spells obey different mechanics.

Therefore, I'm going to pull out of my ass a couple of possible situations.
1. Constant travel speed, animation matches combat log - observable empirically
1a. Travel time locked in at cast - Missle will hit after a certain amount of time regarless of if the target moves afterwards
1a.a Synchronized graphic - Moving target results in faster missle
1a.b Asynchronized graphic - combat log and impact animation do not happen at the same time.
2. Constant travel time - probably untrue because point-blank missles travel instantly
2a. Constant travel time beyond range XX (constant speed until range XX) - meaning that there are no time gains to be made by standing beyond range XX
2a.a Graphic missle travels faster
2a.b Graphic travels at constant speed, combat log event is not synchronized to impact animation
3. Nonlinear scaling - standing further away results in a faster missle, resulting in nonlinear distance-time relationship
3a. Nonlinear piecewise scaling - like 2a above but nonlinear.
4. Phase of the Moon Theory® - missle speed and combat log synchronization depends on server load, ping, nagle's algorithm, queue, complexity of combat (number of participants), zone and continent (and populations thereof), in/out of instance/group/raid, and spell detail settings on your personal computer, as well as other factors potentially including your zodiac sign and what hillary clinton had for breakfast yesterday.

Empirical evidence would seem to indicate that a mob firing at a player is situation 1a.b, and a player firing at a player is situation 1, but neither of them preclude situations 3 or 4, or piecewise variations. Situation 4 will require finding an instance, and some of the finer distinctions are easiest to make with a blinking or teleporting mob, although a feared sprinting mob might work well.


Offline
Old 01/10/08, 11:15 PM   #89
Searix
Piston Honda
 
Searix's Avatar
 
Human Mage
 
Stormreaver
Hate to throw in a curveball, but needless to say it has a big effect on these theories:

The first ignite doesn't start ticking immediatly upon crit.

Take this simple sample i took from my combat log (casting some haste geared non cooldown fireballs aka 2.8 seconds per fireball):

12/9 01:58:45.453  Your Fireball crits Dr. Boom for 3113 Fire damage.
12/9 01:58:46.109  You gain 127 Mana from Master of Elements.
12/9 01:58:46.531  Dr. Boom is afflicted by Ignite.
12/9 01:58:46.531  Boom Bot dies.
12/9 01:58:47.234  You fail to cast Fireball: Not yet recovered.
12/9 01:58:47.453  You fail to cast Fireball: Not yet recovered.
12/9 01:58:47.531  Dr. Boom suffers 24 Fire damage from your Fireball.
12/9 01:58:48.203  Dr. Boom suffers 622 Fire damage from your Ignite.
12/9 01:58:48.515  Your Fireball hits Dr. Boom for 2047 Fire damage.
12/9 01:58:48.734  Boom Bot dies.
12/9 01:58:50.171  Boom Bot dies.
12/9 01:58:50.171  Boom Bot dies.
12/9 01:58:50.171  Dr. Boom suffers 623 Fire damage from your Ignite.
Basically, the first ignite takes a little longer to initially register, after the first ignite though every ~2 second fireballs with their subsequent ignites works like normal. But unless you compensate for the initial ignite starting lag you're not getting bonus damage off the first ignite

Offline
Old 01/11/08, 2:27 AM   #90
manly
Soda Popinski
 
manly's Avatar
 
Troll Mage
 
Mal'Ganis
That could actually explain why I kept hitting my fireballs too early (ie: before an ignite tick). I thought the 'delayed events' you get from like pretty much anything similar in the combatlogs were just a delay for you to receive the message from the server, but it could well be that the ignite is processed later on as you mention.

<Eej> YOU"RE GONNA PULL
<Eej> IF YOU SQUEEZE OFF ANOTHER ARCANE BLAST
<Spectear> You've obviously never played with Manly.
<Spectear> That's hardly a reason to stop DPS.
Very Manly Staff

Canada Offline
Old 01/11/08, 3:31 AM   #91
Hate Monkey
Don Flamenco
 
Hate Monkey's Avatar
 
Troll Mage
 
Arthas
Originally Posted by Searix View Post
Basically, the first ignite takes a little longer to initially register, after the first ignite though every ~2 second fireballs with their subsequent ignites works like normal. But unless you compensate for the initial ignite starting lag you're not getting bonus damage off the first ignite
Could explain why I kept seeing rolling ignites with Bloodlust+Berserking and no one else did, and was unable to even know why I was getting them.

Also, hate to say this too, but if the mob your targeting is stationary, and you're stationary, the travel time of the projectile varies depending upon the model of the mob, and the attack animation of the mob too.

Offline
Old 01/11/08, 6:39 AM   #92
Chira
Von Kaiser
 
Human Mage
 
Crushridge
Thought I would throw in some data with a small observation. This is chain casting pyroblast with the haste buff, taken a few weeks ago.

04:24'02.078	Chira's Pyroblast crits Amani'shi Guardian for 4587 Fire damage
04:24'03.578	Chira's Pyroblast hits Amani'shi Guardian for 3257 Fire damage
04:24'04.718	Chira's Ignite dots Amani'shi Guardian for 918 Fire damage
04:24'05.203	Chira's Pyroblast crits Amani'shi Guardian for 4548 Fire damage
04:24'07.031	Chira's Pyroblast hits Amani'shi Guardian for 2900 Fire damage
04:24'07.890	Chira's Ignite dots Amani'shi Guardian for 1368 Fire damage
04:24'08.640	Chira's Pyroblast crits Amani'shi Guardian for 2135 Fire damage (1423 resisted)
04:24'10.234	Chira's Pyroblast crits Amani'shi Guardian for 4220 Fire damage
04:24'12.031	Chira's Pyroblast crits Amani'shi Guardian for 4597 Fire damage
04:24'12.656	Chira's Ignite dots Amani'shi Guardian for 1956 Fire damage
04:24'13.750	Chira's Pyroblast crits Amani'shi Guardian for 4654 Fire damage
04:24'15.375	Chira's Pyroblast hits Amani'shi Guardian for 2989 Fire damage
04:24'16.265	Chira's Ignite dots Amani'shi Guardian for 3805 Fire damage
04:24'16.875	Chira's Pyroblast crits Amani'shi Guardian for 5304 Fire damage
Everything is acting normally until the bolded lines. The pyroblast crit occured, and a whole 600ms later, an ignite ticked. The crit was obviously not counted, and this ignite tick can be considered "free" as the next ignite is larger than it should have been by exactly 1956/2.

Offline
Old 01/11/08, 7:05 AM   #93
Hate Monkey
Don Flamenco
 
Hate Monkey's Avatar
 
Troll Mage
 
Arthas
Why is your first ignite tick a full 1.7s later than it should be? 2.7s after the first hit. Then you also have the 2nd ignite tick being 500 less than it should be.

Offline
Old 01/11/08, 7:36 AM   #94
Chira
Von Kaiser
 
Human Mage
 
Crushridge
Run the numbers again, I believe the second ignite tick is hitting for the correct amount. 4548*.2 + 917.4/2 = 1368.3

As far as the timing errors are, I'm not sure what to say. Are the timestamps placed on the server and then transmitted, or are they added when the message is received from the server? Would that even make a difference? Also, I went back to page 1 and looked at Grecasi's sample, and the time between the first reported crit and ignite was 24.061-21.373 = 2.688. Cardynal's from page 2 was reported as 2.469 seconds. It doesn't seem to be uncommon.

Offline
Old 01/11/08, 7:41 AM   #95
Searix
Piston Honda
 
Searix's Avatar
 
Human Mage
 
Stormreaver
Originally Posted by Hate Monkey View Post
Why is your first ignite tick a full 1.7s later than it should be? 2.7s after the first hit. Then you also have the 2nd ignite tick being 500 less than it should be.
Pretty sure i've seen this before.

Unless i'm mistaken, another unmentioned attribute of ignite (heh), if you chain crit a lot (as you did there, not letting the ignite tick for ~5 seconds), ignites occasionally spontaneously get "exhausted" and toss out a free tick for the heck of it.

Would be nice to know the rest of the log though

Originally Posted by Chira View Post
Run the numbers again, I believe the second ignite tick is hitting for the correct amount. 4548*.2 + 917.4/2 = 1368.3

As far as the timing errors are, I'm not sure what to say. Are the timestamps placed on the server and then transmitted, or are they added when the message is received from the server? Would that even make a difference? Also, I went back to page 1 and looked at Grecasi's sample, and the time between the first reported crit and ignite was 24.061-21.373 = 2.688. Cardynal's from page 2 was reported as 2.469 seconds. It doesn't seem to be uncommon.
As i said, it's not uncommon, and if you pop cools to get to 2.0 seconds you will almost always lose the first beneficial ignite because of this. I'm also 90% sure the combatlog is server responses transmitted.

Another way to test this "strange" mechanic is pay close attention to how ignite ticks next time you fire crit, assuming normal eye response latency of ~200-500 ms, if it were to tick a direct 2 seconds after a crit the ignite should be 1/4-1/3 the way to first tick. Observation demonstrates this is not the case, you can always see (and notice a slight server delay) between the crit, and when the ignite timer actually starts ticking down (and the fact it starts at 12 o'clock). This is the +.469-+.688 increases we are observing.

Edit: @Hatemonkey, +.7, not +1.7
@Chira, look at the 3rd ignite, not the 2nd

Edit2: I just thought of a really simple way to test the accuracy of the combatlog, have 2 people observe the same event, check the record of it in the WoWCombatLog, and see if they match up

Last edited by Searix : 01/11/08 at 7:52 AM.

Offline
Old 01/11/08, 7:46 AM   #96
Searix
Piston Honda
 
Searix's Avatar
 
Human Mage
 
Stormreaver
Edit: double post

Offline
Old 01/11/08, 7:54 AM   #97
Chira
Von Kaiser
 
Human Mage
 
Crushridge
Still, the third checks out, considering the server recognized the bold crit after the bold tick and didn't include the damage in it. (2135+4220)*.2 + 1368/2 = 1955. There is no more to the log, since the mob was unfortunately dead with that last pyro crit.

Edit: I'm almost positive these timestamps are based on server time as well. I went through another log of which I had a screenshot of, and the timestamps in the screenshot (added by Prat, using local computer time) are 10 seconds ahead of the one from WWS.

Last edited by Chira : 01/11/08 at 8:14 AM.

Offline
Old 01/11/08, 11:27 AM   #98
Cardynal
Piston Honda
 
Orc Warrior
 
Mug'thol
Originally Posted by Kewangeder View Post
text
I see your point with the un-hastened fireball where cast time would matter. However, since this is just coming around since it was possible to get under 2.0 second fireballs...I question if the theory needs a bit of slight revision.

Don't you think we would have seen this "rolling" affect sooner as we have been casting 3.0 second fireballs for the past year with the ignite mechanic (can't remember exactly how long it has been).

I would propose that there is a secondary check that is happening for the ignite mechanic that has to do with adding the ignite debuff itself, not the hit of the fireball. As we have seen, the mob will gain ignite as a second affect of the fireball hitting the target. What if it is at that time there is a check done. Maybe the "rolling" affect has something to do with the calculation done at the time of the cast finishing, ignites ticking, and the debuff being applied.


As pointed out above...the ignite being applied later.

Another mage at the start of the Teron fight last night. No other mages crit him before or during this second.

03:23'06.343 Evilangus's Fire Blast crits Teron Gorefiend for 2677 Fire damage
03:23'07.328 Teron Gorefiend is afflicted by Ignite

Offline
Old 01/11/08, 12:07 PM   #99
naturen
Glass Joe
 
Undead Mage
 
Boulderfist (EU)
I got a question if anyone have done some calculation on if its worth saveing the Skull of Gul´dan one extra min to pop it together with Icy Veins and Combustion to be able to get the ignites rolling.
I gess it depends on your crit rate but atleast if you are abit lucky it might be.

And btw this is pretty funny when your fireballs are in the crit mood
Loading...
=)

Offline
Old 01/11/08, 12:28 PM   #100
manly
Soda Popinski
 
manly's Avatar
 
Troll Mage
 
Mal'Ganis
I'm 90% positive combatlog timestamps are local, and have always been. The order of events will vary from player to player, as well as display different values. This is the core reason why you should never ever merge logs -- all it does is end up with double curse of doom ticks during the same minute and things like such.

<Eej> YOU"RE GONNA PULL
<Eej> IF YOU SQUEEZE OFF ANOTHER ARCANE BLAST
<Spectear> You've obviously never played with Manly.
<Spectear> That's hardly a reason to stop DPS.
Very Manly Staff

Canada Offline
 

Go Back   Elitist Jerks » Class Mechanics

Thread Tools

Similar Threads
Thread Thread Starter Forum Replies Last Post
[Mage]: Ignites working correctly? Cryic Class Mechanics 21 05/27/07 10:08 AM
Back to WOW - Mage Help drlogico Public Discussion 4 03/01/07 1:52 PM
Does Threat Change Dynamically On Ignites? DarthB Public Discussion 31 10/12/06 1:05 AM
How do rolling ignites work? Copernicus Public Discussion 204 09/29/06 6:13 AM