5.2 Trinket Discussion

[Deleted]

Btw excuse my ignorance, but when are stats snap shotted? when you start fireball cast or when you finish? if its when you start casting, then you could easily waste 25%(1sec) of the effect just from that. That is if trinket procs while your casting witch seems likely to happen.

[mrgreenthump]

Btw excuse my ignorance, but when are stats snap shotted? when you start fireball cast or when you finish? if its when you start casting, then you could easily waste 25%(1sec) of the effect just from that.

in fireballs case the stats are accounted when it hits as far as I know and now that I think of it, its that for every spell. Though thats not "snap shotting", Snap shotting is when a dot uses those values it has on apply for the whole duration rather than changing them mid duration when you gain or loose stats.

[elpadrino293]

Btw excuse my ignorance, but when are stats snap shotted? when you start fireball cast or when you finish? if its when you start casting, then you could easily waste 25%(1sec) of the effect just from that. That is if trinket procs while your casting witch seems likely to happen.

when the spell is finished casting...

[Deleted]

after some quick scribbling in my notebook... the chances of this trinket proccing when we need it to proc, especially in this scenario and even to simply make it worthwhile, are far far less than simply RNGing 3 crits in a row...

you'd be losing 1-2 ticks of the ignite pool easily, and if you tried to get 3 pyroblasts and use combustion in a 4 second window, you absolutely need to be GCD capped as well. and thats only if it proc'd right before you initially started the 3x pyroblast!

I'm perfectly happy to accept that, aside from on-pull it's going to be a bitch to try and line-up. I can't remember Nathyiel's maths on it but yeah the rough ICD based off PPM model doesn't line up amazingly with Combust; but hey RNG is RNG.

As mentioned, scenario is on-pull, so you have 2 "windows"; much like now. The difference is during these windows if you have HU+Pyro! it's 2 Pyros!, AT (reset to 4s window) 2 Pyros! (or one), Combust.

Just to clarify: The reason why I'm arguing about "on-pull" so much is that, even though the trinket is 0.5PPM, it's taken off time since last chance to proc on pull you are likely to cast more spells; ergo higher chance to proc. PPM model works similarly to ICDs but just slightly different. If there has been enough time since the last proc on-pull; it's likely that it'll proc. Saying otherwise is like saying 'OMG, X Trinket has only 15% chance per damaging spell land to proc, in theory could never proc'; yeah, in theory it's true, but in practice it's still most likely to proc on pull.

[elpadrino293]

I'm not in a "fantasy" world; I never maintained my maths was correct, and in my earlier post accepted could easily be wrong. Feel free to re-do it with the correct maths; still fairly sure you'll find the numbers ridiculous.

But, hey, let's address the issues you raised.

1. True; but on-pull, given enough time since last proc it's likely to proc.
2. True.
3. You're basing this off what? The fact it's RPPM only? Not how it works.
4. I acknowledged that the benefit is diminished as gear increases.

im not sure about #1 here and i could really use some clarification... everything ive read about RPPM says that "TimeSinceLastChanceToProc" which is capped at 10seconds... 3 seconds into the fight my last chance to proc was like 1 second ago. so my 1st attack, and 1st attack only, would be 10x more likely to proc than my 2nd? or 3rd assuming 1sec since last chance to proc? and any subsequent attack throughout the course of the fight at 1sec since last chance to proc? but if my average time since last chance to proc (direct damage, living bomb dot, pyroblast dot, ignite dot, possible flamestrike/combustion dot) is like 0.5sec then on pull is 20x more likely to proc?

[Deleted]

im not sure about #1 here and i could really use some clarification... everything ive read about RPPM says that "TimeSinceLastChanceToProc" which is capped at 10seconds... 3 seconds into the fight my last chance to proc was like 1 second ago. so my 1st attack would be 10x more likely to proc than my 2nd? or 3rd? and any subsequent attack throughout the course of the fight at 1sec since last chance to proc?

My bad about being vague; no it's not like that.

Basically on each spell hit, it has X% chance to proc, so the more spells you hit within a given time window, the more chance it has to proc. Typically, you are casting most at the start of the fight, so has the highest chance to proc.
Put it this way: given an extreme lucky streak; and given I understand RPPM correctly, I think it's possible to constantly get the proc over the course of the fight. I think the maths on this particular trinket was done earlier in the thread, it's ~1.8% chance to proc per Fireball cast, but as Fireball lands ~every 1.9s, that's why.

Kinda weird though; would seems that RPPM benefits DoT more as they are given more chances to proc it. Oh well.

[elpadrino293]

My bad about being vague; no it's not like that.

Basically on each spell hit, it has X% chance to proc, so the more spells you hit within a given time window, the more chance it has to proc. Typically, you are casting most at the start of the fight, so has the highest chance to proc.
Put it this way: given an extreme lucky streak; and given I understand RPPM correctly, I think it's possible to constantly get the proc over the course of the fight. I think the maths on this particular trinket was done earlier in the thread, it's ~1.8% chance to proc per Fireball cast, but as Fireball lands ~every 1.9s, that's why.

Kinda weird though; would seems that RPPM benefits DoT more as they are given more chances to proc it. Oh well.

"Here are even more nitty gritty details, if you’re interested:

It can proc from any damage/healing event. It keeps track of the last time it had a chance to proc for that enchant.

It calculates the difference in time since the last chance to proc. It uses that time to determine the chance for that event to trigger a proc.

For example, if you have 22% Haste, it was 1.4sec since the last chance to proc, and you’ve got Windsong, then the chance to proc is 2(ppm) * 1.22(haste) * 1.4(time since last chance) / 60 (sec per min) = 5.693%.

The ‘time since the last chance to proc’ is capped at 10sec, so that your first attack of a fight isn’t a guaranteed proc."

source

im not sure about #1 here and i could really use some clarification... everything ive read about RPPM says that "TimeSinceLastChanceToProc" which is capped at 10seconds... 3 seconds into the fight my last chance to proc was like 1 second ago. so my 1st attack, and 1st attack only, would be 10x more likely to proc than my 2nd? or 3rd assuming 1sec since last chance to proc? and any subsequent attack throughout the course of the fight at 1sec since last chance to proc? but if my average time since last chance to proc (direct damage, living bomb dot, pyroblast dot, ignite dot, possible flamestrike/combustion dot) is like 0.5sec then on pull is 20x more likely to proc?

[Deleted]

"Here are even more nitty gritty details, if you’re interested:

It can proc from any damage/healing event. It keeps track of the last time it had a chance to proc for that enchant.

It calculates the difference in time since the last chance to proc. It uses that time to determine the chance for that event to trigger a proc.

For example, if you have 22% Haste, it was 1.4sec since the last chance to proc, and you’ve got Windsong, then the chance to proc is 2(ppm) * 1.22(haste) * 1.4(time since last chance) / 60 (sec per min) = 5.693%.

The ‘time since the last chance to proc’ is capped at 10sec, so that your first attack of a fight isn’t a guaranteed proc."

source

im not sure about #1 here and i could really use some clarification... everything ive read about RPPM says that "TimeSinceLastChanceToProc" which is capped at 10seconds... 3 seconds into the fight my last chance to proc was like 1 second ago. so my 1st attack, and 1st attack only, would be 10x more likely to proc than my 2nd? or 3rd assuming 1sec since last chance to proc? and any subsequent attack throughout the course of the fight at 1sec since last chance to proc? but if my average time since last chance to proc (direct damage, living bomb dot, pyroblast dot, ignite dot, possible flamestrike/combustion dot) is like 0.5sec then on pull is 20x more likely to proc?

Ok, so if we assume Fire goes for 5K Haste cap in 5.2 (towards end; thinking. Not impossible to think), I think that, with Raid Haste; you would be at ~20% Raid Haste.

So calculations for Lei Shen ON PULL, w/ Hardcast Pyro is:

(0.5*1.2*10)/60 = 0.1

Which is a 10% chance to proc. That, in all honesty, isn't a bad proc chance. Obviously then diminishes, but still. Oh well.

[Polarthief]

in fireballs case the stats are accounted when it hits as far as I know and now that I think of it, its that for every spell. Though thats not "snap shotting", Snap shotting is when a dot uses those values it has on apply for the whole duration rather than changing them mid duration when you gain or loose stats.

When it's finished casting, not when it hits.

However, some spells, like Scorch, Fire Blast, and Arcane Blast hit at the same time they're finished casting.

---------- Post added 2013-02-19 at 06:58 AM ----------

Ok, so if we assume Fire goes for 5K Haste cap in 5.2 (towards end; thinking. Not impossible to think), I think that, with Raid Haste; you would be at ~20% Raid Haste.

So calculations for Lei Shen ON PULL, w/ Hardcast Pyro is:

(0.5*1.2*10)/60 = 0.1

Which is a 10% chance to proc. That, in all honesty, isn't a bad proc chance. Obviously then diminishes, but still. Oh well.

Is the 5k Haste cap our 5032 rating (with 5% raid buff of course) for our 4th tick of Combustion (Glyphed)? Just curious what the significance of 5k is :3

[elpadrino293]

Ok, so if we assume Fire goes for 5K Haste cap in 5.2 (towards end; thinking. Not impossible to think), I think that, with Raid Haste; you would be at ~20% Raid Haste.

So calculations for Lei Shen ON PULL, w/ Hardcast Pyro is:

(0.5*1.2*10)/60 = 0.1

Which is a 10% chance to proc. That, in all honesty, isn't a bad proc chance. Obviously then diminishes, but still. Oh well.

then quickly diminishes to less than 1% when your time since last chance to proc is < 1sec
4% of random uptime just isn't good enough.
perhaps if it was On Use, and it had a 6 second duration, and had a 90sec cooldown, and then i would use it.

[Deleted]

then quickly diminishes to less than 1% when your time since last chance to proc is < 1sec
4% of random uptime just isn't good enough.
perhaps if it was On Use, and it had a 6 second duration, and had a 90sec cooldown, and then i would use it.

On-Use, less base Int, wouldn't mind keeping same base 4s duration (can AT), 90s CD, fine. Oh well.

[Polarthief]

then quickly diminishes to less than 1% when your time since last chance to proc is < 1sec
4% of random uptime just isn't good enough.
perhaps if it was On Use, and it had a 6 second duration, and had a 90sec cooldown, and then i would use it.

Because a meta gem is going to become on-use

I don't even know how in the hell they would code that. How would you drag the gem to your bars? If it enchants the helmet with an on-use... again, coding issues.

Regardless, I really don't know if I want to bother with this thing. 10s isn't all that long/great, and I'm not sure if it's worth the loss of the Burning Meta.

[Deleted]

Because a meta gem is going to become on-use

I don't even know how in the hell they would code that. How would you drag the gem to your bars? If it enchants the helmet with an on-use... again, coding issues.

Regardless, I really don't know if I want to bother with this thing. 10s isn't all that long/great, and I'm not sure if it's worth the loss of the Burning Meta.

We're talking about the trinket; this is a trinket discussion thread...

[ArchitechSK]

Ok, so if we assume Fire goes for 5K Haste cap in 5.2 (towards end; thinking. Not impossible to think), I think that, with Raid Haste; you would be at ~20% Raid Haste.

So calculations for Lei Shen ON PULL, w/ Hardcast Pyro is:

(0.5*1.2*10)/60 = 0.1

Which is a 10% chance to proc. That, in all honesty, isn't a bad proc chance. Obviously then diminishes, but still. Oh well.

Initial math points to the likely haste softcap to aim for is the 11150 breakpoint for fire (Gcd capped under hero+meta proc)

[mrgreenthump]

Initial math points to the likely haste softcap to aim for is the 11150 breakpoint for fire (Gcd capped under hero+meta proc)

Doubtful that we will ever get there. We will just stack as much crit as we can, then look at closest haste cap and reforge according and put rest to mastery..

[ArchitechSK]

Doubtful that we will ever get there. We will just stack as much crit as we can, then look at closest haste cap and reforge according and put rest to mastery..

I think you misunderstood me - and while this will all be clarified in the coming days by the new release of simcraft. As the current stay weight stay the same, haste will be better then crit for fire. the 11150 haste mark is a point where for certain the stay weights will significantly change, so I am doubtful that my predictions remain true past that point. This change is predicated by the fact that the rppm System scales with haste while only one trinket scales with crit (and this trinket scales better with haste them with crit)

Sorry for typos and odd English - was typed from my phone

[Deleted]

I think you misunderstood me - and while this will all be clarified in the coming days by the new release of simcraft. As the current stay weight stay the same, haste will be better then crit for fire. the 11150 haste mark is a point where for certain the stay weights will significantly change, so I am doubtful that my predictions remain true past that point. This change is predicated by the fact that the rppm System scales with haste while only one trinket scales with crit (and this trinket scales better with haste them with crit)

Sorry for typos and odd English - was typed from my phone

RPPM isn't some "new" thing; it's been in the game since 5.0

Haste won't have enough of an effect on RPPM to make it viable for Fire, or indeed any other class/spec combo to stack PURELY for trinket gains.

[mrgreenthump]

RPPM isn't some "new" thing; it's been in the game since 5.0

Haste won't have enough of an effect on RPPM to make it viable for Fire, or indeed any other class/spec combo to stack PURELY for trinket gains.

Not to mention crit is going to start scaling better and better, cause of hot streak still proccing from 2 crits in a row, given IB lessens it some bit

[ArchitechSK]

Not to mention crit is going to start scaling better and better, cause of hot streak still proccing from 2 crits in a row, given IB lessens it some bit

This is incorrect. Because you have to give up one point of haste to gain one point of crit, the number of hot streak procs you expect in a given time period from two crates in a row increases linearly with with increasing crit at the cost of not increasing haste.

[Deleted]

This is incorrect. Because you have to give up one point of haste to gain one point of crit, the number of hot streak procs you expect in a given time period from two crates in a row increases linearly with with increasing crit at the cost of not increasing haste.

Crit will scale continually until Fire reaches CM cap. Considering CM cap is now ~77% crit fully buffed; we'll probably never reach it xD