Hello mando fellows !

I have posted this thread in the mercenary forum about a problem I/we have with Barrage (Curtain of Fire) and Prototype Particle Accelerator (Ionic Accelerator) procs.

I post here to know if you have some problems to proc CoF, or if this is only for mercenaries ?

Both procs behave the same, for both classes. Assault/Pyro has predictable procs, while Gunnery/Arsenal is random (and highly annoying).

Assault/Pyrotech procs appear to be additive, whereas the Gunnery procs seem to be a flat % every ability use. Id like to see Gunnery get the additive approach as well so that the CoF procs are much more reliable. If you get unlucky your DPS and ammo management can suffer.

Assault/Pyrotech procs appear to be additive, whereas the Gunnery procs seem to be a flat % every ability use. Id like to see Gunnery get the additive approach as well so that the CoF procs are much more reliable. If you get unlucky your DPS and ammo management can suffer.

I'm not sure what you mean by this when you say the Assault proc is additive.

Assault/Pyrotech procs appear to be additive, whereas the Gunnery procs seem to be a flat % every ability use. Id like to see Gunnery get the additive approach as well so that the CoF procs are much more reliable. If you get unlucky your DPS and ammo management can suffer.

Seconded.

I'm not sure what you mean by this when you say the Assault proc is additive.

I assumed that he meant that if there's a a 33% chance for something to happen, if it doesn't proc on the first, the chance for the second time would be 66%, etc.

I assumed that he meant that if there's a a 33% chance for something to happen, if it doesn't proc on the first, the chance for the second time would be 66%, etc.

Perhaps. Has anyone done a test with the proc from this? It would be a pretty weird but spectacular fail of programming for BW to have done this and screw up the tool tip so badly.

I assumed that he meant that if there's a a 33% chance for something to happen, if it doesn't proc on the first, the chance for the second time would be 66%, etc.

that's wrong.

When measuring the procs, we can't add them because the events are independent on chances. Think, if heads is 50% likely, then two flips should guarantee a heads at 100%.... but also a tails.... which is clearly not guaranteed because I get two heads in a row a lot.

Take Flame barrage - It's a 30% proc, with no cooldown. If you use it twice, there's a 49% chance that nothing happened, and a 51% chance something did happen. But, that 51 can be divided into three subcategories:

• 9% - both attacks proc'd for 30% of 30%
  
• 21% - The first attack proc'd but the second one did not; 30% of 70%
  
• 21% - the second attack proc'd but the first one did not; 70% of 30%
  

Taking Pyro's Ionic Accelerator crap, let's just say it was two bolts in a row to parallel the above example - It's a 45% proc with a cooldown (so the math is slightly different) There's a 55% of 55% chance, or 30.25% chance that nothing happens. The other 69.75% chance to happen refers to the proc happening, and can almost be divided into three subcategories:

• 24.75% - the first bolt failed but the second bolt proc'd
  
• 24.75% - the first bolt proc'd and the second bolt did not (because it didn't roll)
  
• 20.25% - the first bolt proc'd and the second bolt proc'd (isn't possible? it didn't roll)
  

2 and 3, however, are the same (thus add them), so my revised list is:

• 45% - first bolt procs
  
• If that fails, then there's a 24.75% that you get it on the second bolt.
  
• If that fails, then the remaining 30.35% chance means you failed
  

probability is simply binomial expansion:

• a + b = 1 {either a or b has to happen - the sums of their probabilities must be 1}
  
• (a + b)^2 = 1 {raise both sides to the power of x, where x is the number of times you use an ability/roll dice/flip coins/whatever}
  
• a^2 + 2ab + b^2 = 1{expand the binomial. This shows a^2 means they both proc'd, b^2 means they both didn't proc, 2ab refers to either first one procing and the second one not, and then vice versa}
  

that's wrong.

 

When measuring the procs, we can't add them because the events are independent on chances. Think, if heads is 50% likely, then two flips should guarantee a heads at 100%.... but also a tails.... which is clearly not guaranteed because I get two heads in a row a lot.

 

Take Flame barrage - It's a 30% proc, with no cooldown. If you use it twice, there's a 49% chance that nothing happened, and a 51% chance something did happen. But, that 51 can be divided into three subcategories:

• 9% - both attacks proc'd for 30% of 30%
  
• 21% - The first attack proc'd but the second one did not; 30% of 70%
  
• 21% - the second attack proc'd but the first one did not; 70% of 30%
  

 

Taking Pyro's Ionic Accelerator crap, let's just say it was two bolts in a row to parallel the above example - It's a 45% proc with a cooldown (so the math is slightly different) There's a 55% of 55% chance, or 30.25% chance that nothing happens. The other 69.75% chance to happen refers to the proc happening, and can almost be divided into three subcategories:

• 24.75% - the first bolt failed but the second bolt proc'd
  
• 24.75% - the first bolt proc'd and the second bolt did not (because it didn't roll)
  
• 20.25% - the first bolt proc'd and the second bolt proc'd (isn't possible? it didn't roll)
  

2 and 3, however, are the same (thus add them), so my revised list is:

• 45% - first bolt procs
  
• If that fails, then there's a 24.75% that you get it on the second bolt.
  
• If that fails, then the remaining 30.35% chance means you failed
  

 

 

 

probability is simply binomial expansion:

 

• a + b = 1 {either a or b has to happen - the sums of their probabilities must be 1}
  
• (a + b)^2 = 1 {raise both sides to the power of x, where x is the number of times you use an ability/roll dice/flip coins/whatever}
  
• a^2 + 2ab + b^2 = 1{expand the binomial. This shows a^2 means they both proc'd, b^2 means they both didn't proc, 2ab refers to either first one procing and the second one not, and then vice versa}
  

If the first bolt fails to proc shouldn't the second bolt then still have a 45% of proccing? It's a dependent event in that the second bolt cannot proc if the first one does, however 45% of the times that the first bolt fails to proc the second bolt will still proc.

What you mean I think is that if you were to fire a pair of charged bolts 100 times (always with IA possible to proc) then, on average (key words there), 45 of those attempts would proc on the first attempt. Of the remaining 55 attempts 24.75 of them (or 45% of the remaining 55 events) would proc on the second bolt. And the other 30.25 times you would have gotten no proc in those two attempts.

I think that's what you meant all along, just your post made it seem like the second bolt only has a 24.75% chance to proc if the first bolt fails, which can be a tad confusing. It has a 45% chance to proc, but only during the 55% of attempts where the first bolt didn't proc in the first place. It's a subtle difference, and again I think you meant exactly that. Just wanted to make sure I, and everyone else, understand what's happening.

Now, here's my question: How is this manifestly different from the CoF proc which also has a 6 second ICD, and also has a 45% chance to proc on CD?

Honestly what I think is happening is that in reality Full Auto is proccing a lot more of our IA HiBs than we really think it is. The 75% chance for FA to do this essentially smooths out the times when otherwise the proc could also get really streaky. Really tempted to try it out on a dummy and not use FA at all and see if IA is still as well behaved as we all think it is.

The issue right now seems to be that, allegedly, the proc of IA "is" additive.

i.e. if the 1st CB doesn't proc (55% chance), the second one has a 90% (45+45) chance of procing, thus making it infinitely more reliable than the CoF proc.

This is from someone testing it, not statistically significant but....

Pyro procs:

12 times on the first power shot

21 times on the second power shot

Arsenal procs:

14 times on the first tracer missile

11 times on the second tracer missile

4 times on the third tracer missile

3 times on the fourth tracer missile

once on the fifth tracer missile

If this is actually the case then, as someone said, BW screwed up the coding spectacularly. "If" it is a bug and not a feature , it should be fixed, which would be a nerf to assault.

Or.... under the "it's not a bug, it's a feature" mindset, they could do the same thing to the CoF proc.

Or.... under the "it's not a bug, it's a feature" mindset, they could do the same thing to the CoF proc.

Yes please. This alone would pretty much solve all problems with Gunnery DPS, and I wouldn't have to bring my gunslinger anymore.