So are we SURE Curtain of Fire is a 45% chance to proc?

[ArchangelLBC]

So I decided what the hell, I'll play around with Assault a bit on the dummy, see if I can't get some decent numbers and get non-sucky enough at the rotation. One thing I noticed IMMEDIATELY was how, well seemingly reliable Charged Bolts was at resetting the cooldown on High Impact Bolt. Not one time in over an hour of parsing did I ever have HiB finish it's cooldown entirely. Most of the time it never got below 6 seconds, and the way it worked out I was almost never using Full Auto to do it (often, FA would come off cooldown just as a CB reset the cooldown anyway so I'd go ahead and fire off FA).

 

Now, I'm quite aware that bias will make you "notice" something that isn't really there (like when you had a little bit of crit and then claim to really notice the difference even though, really the actual difference wasn't all that big), but still given how often I do completely finish the cooldown on FA, while having spammed GR the whole time, I have to seriously wonder if the talent isn't bugged.

 

To that end, I'm highly considering going onto the dummy and just spamming grav round for an hour and trying to get a large enough sample size to see if the proc % is what it claims to be, because right now, compared to CB resetting HiB for assault with the same %, it just seems way too ridiculously streaky.

[klham]

I've been noticing this for a long time, but it became really noticeable once Charged Barrel was changed to require only a single Grav Round to apply the full armor debuff. I thought this change would liberate us from the obligation to spam Grav Round. Alas, no.

 

I haven't tested it systematically with a training dummy and timer, but even after the 6 second rate limit I've often had to spam GR repeatedly to reset Full Auto. Many times it doesn't proc before the 15 second cd. So it almost feels like there's a hole in the Gunnery rotation that gets filled with (what else?) Grav Round.

[Helmholtzz]

I agree with you. It is not that rare when I use FA without a curtain of fire that means FA came off cool down as I did not get the proc. If you need parse data I can help you out with that. Just let me know.

[Helmholtzz]

So I could not wait and tried it out.

 

Test method:

1. Use grav round till I get a curtain of fire proc.

2. After curtain of fire proc I use full auto (2.9 seconds) Then I use 2 hammer shots (2 GCD = 3 seconds) then grav round channel (1.5 seconds). So total time passed = 7.4 seconds. Which ensures that every single grav round counts towards the chance of proccing curtain of fire.

3. I only used full auto when curtain for fire was available. So number of full auto activated denotes the number of curtain of fire procs.

 

Here is the log.

 

Results:

Grav rounds: 176

Full auto ticks: 255; Full auto activated = number of curtain of fire procs = 255/3 = 85;

 

Percentile = 100*85/176 = 48.29%.

 

Thats the results I found. More parse data will make it more robust.

[salaciousc]

I think that problem is caused by the high alacrity values of the Assault Specialist.

Alacrity shortens the duration of single attacks, but it does not affect not the reset timer for HIB.

 

With alacrity =0 exactly 4 attacks a 1,5 seconds fit into the the 6-second lockout period.

 

Alacrity shortens each of those 4 attacks around 5%-6%, as a result resetting HIB fails.

 

 

The Red Eclipse - Scaeva, Salacia, Lowblow

[ArchangelLBC]

So I could not wait and tried it out.

 

Test method:

1. Use grav round till I get a curtain of fire proc.

2. After curtain of fire proc I use full auto (2.9 seconds) Then I use 2 hammer shots (2 GCD = 3 seconds) then grav round channel (1.5 seconds). So total time passed = 7.4 seconds. Which ensures that every single grav round counts towards the chance of proccing curtain of fire.

3. I only used full auto when curtain for fire was available. So number of full auto activated denotes the number of curtain of fire procs.

 

Here is the log.

 

Results:

Grav rounds: 176

Full auto ticks: 255; Full auto activated = number of curtain of fire procs = 255/3 = 85;

 

Percentile = 100*85/176 = 48.29%.

 

Thats the results I found. More parse data will make it more robust.

 

That was the exact test method I was going to use. So far looks to be about 45%. I'll do it myself using the same exact method and we'll see what we see.

 

I think that problem is caused by the high alacrity values of the Assault Specialist.

Alacrity shortens the duration of single attacks, but it does not affect not the reset timer for HIB.

 

With alacrity =0 exactly 4 attacks a 1,5 seconds fit into the the 6-second lockout period.

 

Alacrity shortens each of those 4 attacks around 5%-6%, as a result resetting HIB fails.

 

 

The Red Eclipse - Scaeva, Salacia, Lowblow

 

I think you have this backwards, but once you correct for that I think you're right. It's Gunnery which has the higher alacrity since the stance gives 3% and most of us take the 2% alacrity talent in Assault, or 1% alacrity from assault plus a point in critical responder in CM. Also I was complaining about HiB being really easy to reset while FA seems hard to get the proc on. But I think that yes indeed, the fact that we essentially need the full cast of full auto plus two extra GCDs before we can actually reset it again, but even then it is just ridiculously streaky. I don't know that I've had a parse recently where FA didn't finish it's cooldown, spamming grav round with 5 stacks of CB. That never happens with Charged Bursts and HiB (though the second ability to reset doesn't hurt either).

 

That last bit may be the other portion. I wonder if we could have Demo Round proc CoF with a 70% chance.

[kennethdale]

That was the exact test method I was going to use. So far looks to be about 45%. I'll do it myself using the same exact method and we'll see what we see.

 

I think you have this backwards, but once you correct for that I think you're right. It's Gunnery which has the higher alacrity since the stance gives 3% and most of us take the 2% alacrity talent in Assault, or 1% alacrity from assault plus a point in critical responder in CM. Also I was complaining about HiB being really easy to reset while FA seems hard to get the proc on. But I think that yes indeed, the fact that we essentially need the full cast of full auto plus two extra GCDs before we can actually reset it again, but even then it is just ridiculously streaky. I don't know that I've had a parse recently where FA didn't finish it's cooldown, spamming grav round with 5 stacks of CB. That never happens with Charged Bursts and HiB (though the second ability to reset doesn't hurt either).

 

That last bit may be the other portion. I wonder if we could have Demo Round proc CoF with a 70% chance.

 

I'm still not so sure. Its not just perception because I have seen FA CD naturally whereas the only time HiB ever gets close is if I'm refreshing other things and miss a chance to proc. Even so I'm not sure I've ever had it go the full 15s whereas FA definitely has. I never use FA to proc IA (only use it after a IA proc to give me 4.5 seconds of pure regen) so its the 45% with a 6s lockout for both and FA definitely resets less often.

 

Perhaps if it seems that CoF is getting procced enough, maybe its IA thats bugged in our favor?

 

As it happens I was messing around on a dummy today in Pyro so lets take a look... (for reference, again, I do not use Unload/FA to reset the CD on RS/HiB)

 

Power Shot - 504 uses (504/2) 252 Activations

Rail Shot - 342 uses (342/2) 171 Activations

 

I start every rotation with an unprocced RS so subtract the total number of separate parses for the approximate number of times I used RS without a proc (once or twice I re-started my rotation after making a big mistake to simply waste time as I was waiting for CDs so I'll add 2 to the number of uses of non-proced) for 8 and take that away and you get

 

Rail Shot - 163 Activations from Proc

 

That means that I used Power shot 1.55 times per time I used Rail shot. That seems WAY higher than 45% chance. I'll take some time tomorrow and do a parse with just Power Shot and Rail shot (With 3 rapid shots after a RS to keep up a DoT) to get some more data but based on that very loose math it seems to be very skewed in the favor of Ionic Accelerator.

 

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Update

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

I did a 15 minute parse using just Rail Shot, Power Shot, and Rapid shots. I used 3-4 Rapid Shots after every PPA proc to ensure that I was only using Power Shot for procs.

 

Power Shots: 278/2 = 139

Rail Shots: 172/2 = 86

 

1.61 Power shots/Rail Shot.

 

So its still pretty accurate to my other parses but I can tell you that I never used more than 2 Power Shots to proc a PPA. Not once. That cannot be just 45%.

Edited October 20, 2013 by kennethdale