Ground Zero offensive proc activation rate drastically reduced since i28p2

[Maelwys]

As noted here by ZemX.

 

I've tested this on Live vs a Single Target (Pylon) with 3 damage procs in Ground Zero on my Scrapper. Whilst they ARE activating its very far short of the expected 90% rate.

 

Something appears to have gone wrong with proc activation chance in whatever tweaks were done to get Ground Zero to have different target limits for friendly and unfriendly targets. As CoDv2 isn't updated yet I can't determine the exact cause but I strongly suspect it's now calling a child effect that has a lower base recharge time for the offensive properties (including these Procs)

 

 

[Captain Powerhouse]

Just double checked and the internal executed power has a 90s recharge and 3s activation for a total 93s cycle. 

[Maelwys]

5 hours ago, Captain Powerhouse said:

Just double checked and the internal executed power has a 90s recharge and 3s activation for a total 93s cycle. 

Thanks for checking.

Unfortunately I'm not sure then why the current observed activation rate in Ground Zero on Live is drastically lower than what is being predicted by the PPM formula. (The observed rate seems to be roughly ~30%; but it ought to be 90%. And until recently the observed proc activation rate in Ground Zero was in line with that 90% prediction!)

Ground Zero has a 90s base recharge + 15ft radius + 360 degree arc + 3s cast time; so if it is slotted with 80.9% local recharge aspect (as in my Scrapper) then each 3.5PPM proc within it should be getting an activation rate that is capped at 90% (~114.5% before clamping). My Scrapper's Ground Zero currently has three such 3.5PPM procs in it; each of which should be getting a separate independent 90% chance of activation; meaning that the likelihood of me observing only one or fewer of these three procs firing should be 2.8%. However currently I'm consistently seeing that very situation (only one or no procs firing) happening over and over again in my testing; as shown in the above screenshots.
(Also there should be a 72.9% likelihood of all three procs triggering simultaneously; yet I only observed that happening once across a series of over 25+ tests this afternoon!)

So something still appears to be very wrong here... 😕

[Maelwys]

FWIW here's a combat log of twenty back-to-back Ground Zero activations, standing next to a RWZ Training Pylon with the above Scrapper (80.9% recharge, 3x 3ppm procs):

CombatLog-GZTest.txt

Breakdown:
0 procs = 5 occurrences
1 procs = 9 occurrences
2 procs = 6 occurrences
3 procs = 0 occurrences
Total = 20

(Admittedly a lowish sample rate, but enough to highlight the problem. That distribution appears to be more in line with a 30%-40% chance of activation than a 90% one!)
 

Edited Saturday at 10:02 PM by Maelwys

[Troo]

Is there an extra accuracy check that the procs trigger off of? Changing it from basically an autohit might have something to do with the lower proc results. (WAG)

[Maelwys]

26 minutes ago, Troo said:

Is there an extra accuracy check that the procs trigger off of? Changing it from basically an autohit might have something to do with the lower proc results. (WAG)

IIRC Procs that are in "autohit" powers require their own Tohit check; otherwise they just inherit the Hit Check of the parent power.

Ground Zero still requires a hit check; and all my rolls were capped at 95% (and successfully hit!). I guess it's possible that a new internal power execution/redirect has been set up in such as way that the child power isn't set to inherit the parent power's accuracy or boosts properly... but it certainly appears to be inheriting the damage enhancement OK.


 

Edit: Nope it doesn't appear to be accuracy related. 

Just retested quickly as a Level 50+1 /Rad Scrapper versus a pack of level 23 Sky Raiders + Freakshow in Terra Volta.
The vast level difference should have made any accuracy discrepancies meaningless:

That's 11 hits, so 33 possible Proc Activations... and only 12 kicked in (~36% activation rate)



Second check versus another low-level group in the same zone (level 25 Freakshow + Lost):

That's another 11 hits, so 33 possible Proc Activations... and only 11 kicked in (~33% activation rate)

So the observed activation rate seems pretty consistent here and also in line with my earlier pylon testing.
(e.g. my Scrapper's GZ proc likelihood is currently sitting at about 30-40%, whenever it should be capped at 90%!)
 

Edited Saturday at 10:27 PM by Maelwys

[Captain Powerhouse]

@Maelwys I just ran a test, and keeping in mind I'm working locally on a newer branch it seems to work as expected. Against 10 foes, I hit 9. This was on a scrapper.

 

Edit: I tested with Touch of Lady Gray and although I still saw a few more misses than I would expect, it still is RNG on 10% floor and that can go any way. I dont see any reason why proc rate of any power would be affected here. 

 

[Maelwys]

10 hours ago, Captain Powerhouse said:

@Maelwys I just ran a test, and keeping in mind I'm working locally on a newer branch it seems to work as expected. Against 10 foes, I hit 9. This was on a scrapper.

 

Edit: I tested with Touch of Lady Gray and although I still saw a few more misses than I would expect, it still is RNG on 10% floor and that can go any way. I dont see any reason why proc rate of any power would be affected here. 

 

Thanks - 

I picked up some unslotters and retested this morning with a single Touch of Lady Grey and different combinations of local recharge aspect slotted in Ground Zero:


This was seeing proc activations roughly ~90% of the time on both Live and Brainstorm.


This was seeing activations roughly ~70% of the time on both Live and Brainstorm.


This was seeing activations roughly ~40% of the time on both Live and Brainstorm.


This was seeing activations roughly ~30% of the time on both Live and Brainstorm.



In theory Ground Zero ought to be able to take up to 134% local recharge aspect slotting in it before seeing any difference, yet versus a pack of 10-11 foes I'm very consistently seeing 7+ procs firing whenever it has 0% recharge slotted in it and 3 or fewer procs firing whenever it has 99% recharge slotted in it. 





If you're unable to see the same behaviour happening internally and there've been no related changes then something very weird is going on! 🤯

(I guess worst case whatever is causing activation rate to work properly on that newer internal branch will get pushed to Live at some point...) 🤞

 

Edited 15 hours ago by Maelwys

[Captain Powerhouse]

I should had added recharge enhancements to my test... 

 

Anyways just calculating:

Ground Zero is 90s

15ft radius

3s cast time

 

That is a base 57.7% PPM multiplier with no recharge 

 

Doing this in a bit of a rush as I'm taking a very short break from some chores, but this Is what I am coming up with:

 

 

This seems to match your results. Are your expectations based off 8ft radius calculation by any chance?

[Maelwys]

3 hours ago, Captain Powerhouse said:

This seems to match your results. Are your expectations based off 8ft radius calculation by any chance?

My expectations are going off a 15ft radius.
The damage procs I've been using are each 3.5PPM and (as shown by your table!) in theory each of them should be getting their activation rate capped at 90%. 

FWIW I normally use a copy of MacSkull's Google Sheet for any quick PPM calculations (mainly because it lets me see at a glance exactly how far over the 90% cap I'm going + therefore exactly how much leeway I have to increase my local recharge aspect enhancement!) plus CoDv2 to find the base values to plug into it.
Currently CoDv2 is still using the pre i28p2 patch figures for Ground Zero; of 15ft radius; 360 degrees, 3s activation, 90s base recharge. These values predict that for 3.5PPM Procs like Touch of Lady Grey, my Scrapper's Ground Zero with 80.90% recharge should be sitting comfortably over the 90% cap (~114.5.% before getting clamped to 90%!)



If I increase the "Recharge slotting" value it shows that GZ ought to be able to take up to ~134% recharge before 3.5PPM procs will dip under the 90% activation rate cap.

These expectations do certainly all appear to tie in with your table.
 

3 hours ago, Captain Powerhouse said:

Doing this in a bit of a rush as I'm taking a very short break from some chores, but this Is what I am coming up with:

The values in this table look right. However unfortunately they're not reflective of my current combat log results on either Live or Brainstorm.

My Scrapper typically runs with 80.9% recharge in their Ground Zero; but I'm currently only seeing each 3.5PPM damage proc trigger roughly 30-40% of the time (rather than the predicted 90%!) regardless of whether I'm activating it on a single target (like the pylon) or multiple targets (like the large spawns in Terra Volta and Cimerora).



I'll try to test proc rate in some similar high-recharge, high-radius AoE powers (a Blaster Nuke ought to do...) later this evening just in case this oddness isn't just limited to Ground Zero - I know at one point during i28p2 Open Beta damage procs were dealing lower damage than they should across multiple powers so it's probably worth ruling that out. You've got me wondering now whether 3.5PPM procs are being mistakenly treated as 1PPM or 1.5PPM procs somewhere; since the predicted rates for that are just about matching my currently observed numbers... 🤔 🔍


- - - - - - - - - - - - - - - - - - - -

Edit: So I did a quick check of a different PBAoE with a similar predicted proc activation rate (my Fire Blaster's Inferno) and it appears to be functioning just fine.
Therefore whatever is happening here to skew proc activation rate; thankfully it appears to be limited to Ground Zero rather than a global thing.

Predicted 3.5PPM proc activation rate is ~121.5 (where Ground Zero's was ~114.5) before clamping; so it's being capped at 90%.


With 2x 3.5PPM Procs slotted:
+ First BOOM - 9 Foes hit. 18 proc opportunities, 15 proc activations (~83.3%)


+ Second BOOM - 9 Foes hit. 18 proc opportunities, 18 proc activations (100%)


So that's a predicted proc activation rate in Inferno of 90%, and an observed proc activation rate of ~83.3% and 100% - well within the margin for error.


If Ground Zero's proc rate was working properly then I'd be expecting to see a very similar number of activations to the above.
But instead, it's currently acting like the below (despite having the same predicted proc activation rate and the same number of slotted 3.5PPM damage procs!)

Predicted 3.5PPM proc activation rate is ~114.5 before clamping; so it's being capped at 90%.

With 2x 3.5PPM Procs slotted:
+ First FOOM - 10 Foes hit. 20 proc opportunities, 7 proc activations (35%)


+ Second FOOM - 9 Foes hit. 18 proc opportunities, 6 proc activations (~33.3%)


So that's a predicted proc activation rate in Ground Zero of 90%, but an observed proc activation rate of 35% and 33.3% - much lower than it ought to be.

I could understand if this was just me being "unlucky" a few times; but other tests vs multiple targets and prolonged testing vs a Single Target (see my pylon log above!) is all showing roughly the same observed 30-40% proc activation rate instead of the predicted 90% proc activation rate.

And it's not just me either - @ZemX noted here that the activation rate of their Knockdown Proc (which I think is 2.5PPM?) has plummeted from ~90% to ~25%...

 

Edited 7 hours ago by Maelwys

[ZemX]

3 hours ago, Maelwys said:

And it's not just me either - @ZemX noted here that the activation rate of their Knockdown Proc (which I think is 2.5PPM?) has plummeted from ~90% to ~25%...

 

Actually, it's 3.5PPM.  It's the Superior Avalanche Recharge/Knockdown proc.  Unless that has changed, it should be 3.5PPM 0.67 knockback.  Prior to the update, this sucker would put EVERYBODY on their butts.   Now, I know it got a reduction from 22.5ft radius to 15ft radius for Tankers in this page and I am seeing that.  It does not cover the same area.  But in the area it DOES cover, and now that I've had more time with it in game, I am seeing at best half and frequently less than half of the enemies getting knocked down.  So say a quarter to maybe half of enemies get knocked down in a big spawn (I use this whenever it's up as my opening shot).  I have not done all the data logging you have.  I have other 3.5PPM procs in GZ too, but the Sup. Avalanche is just really easy to see when it's not working. 

 

Pretty sure this isn't just perception.  It was so reliable before the patch that something has to have changed now.

Edited 5 hours ago by ZemX

[Maelwys]

35 minutes ago, ZemX said:

Pretty sure this isn't just perception.  It was so reliable before the patch that something has to have changed now.

 

Agreed!

 

Also FWIW, Uberguy has just updated CoDv2. Looks like the new redirect that handles the healing component is indeed set to "Player Teammate (Alive)" rather than the original "Ally (Alive)", which explains why it's no longer healing any nearby friendly targets that you aren't teamed with.

 

Unfortunately I still can't see anything obvious in either the main power or the offensive redirect to explain the observed proc activation rate reduction.