Focused Feedback: Robotics Revamp

[Bionic_Flea]

Personally, I would like it if the bots retained at least a little -regen that could then be augmented by taking the MM's personal attacks. 

 

So if the original -regen was Bots 500 (max), MM Zero and the new version is Bots Zero, MM personal (200x3 = 600 max), then we could try a newer version of Bots 200, MM (125x3 = 375) for a Max of 575 if all three attacks are taken.  But more likely is that only one or two attacks would be taken and get to 325 or 450.  Or something like that.

[Sovera]

The endurance (and APM) is a problem, but tests (not math) showed that using the attack powers along side the minions was a 20% DPS.

 

Personally I found that my time on a MM was mostly spent desperately trying to keep my minions alive through heals, buffs, debuffs, resummoning, reupgrading. Who has time to attack on top of that?

[WindDemon21]

1 hour ago, Biske said:

I'm not sure I follow that logic either.  If they do as you suggest and instead make the personal attacks super powerful while bringing the pet damage in line, the same people complaining about the regen will just complain that the personal attacks are now mandatory instead.

The personal attacks are supposed to be there to give you something to do in slow support sets mainly. The point being, many choose to skip the attacks, or are too busy to use them.

 

Essentially not only losing the - regen with bots now, but also forcing you to get them to retain some of the - regen also at a huge detriment to end cost and cast/ support time.

 

Again, prior bots did average 650% -regen. If they want to lower that a little given the increased damage fine, but please devs keep it in the bots.

[WindDemon21]

1 hour ago, Bionic_Flea said:

Personally, I would like it if the bots retained at least a little -regen that could then be augmented by taking the MM's personal attacks. 

 

So if the original -regen was Bots 500 (max), MM Zero and the new version is Bots Zero, MM personal (200x3 = 600 max), then we could try a newer version of Bots 200, MM (125x3 = 375) for a Max of 575 if all three attacks are taken.  But more likely is that only one or two attacks would be taken and get to 325 or 450.  Or something like that.

I thought that too but even then would still want the majority in the bots so the attacks feel less NEEDED in comparison but still help. Even something like 50% per minnion, 75%lt, 100% assault, and 50% per attack.

 

That way it's 400% from the bots, and up to 150% from the attacks. Still less than current, or beta max, but let's be honest again that nobody really will use the aoe against a st usually, and lots still skip at least one of the st attacks.

[gameboy1234]

11 hours ago, Booper said:

Correct. It was a 30% chance for 500% regeneration debuff on two of its attacks.

 

Right, so the next question is what would be the expected uptime?  Would the -Regen last long enough that we'd expect it to double up?  Or would the -Regen wear off before the next 30% chance happened? Are we expecting to see -1500% regen or would it not trigger often enough and we'd only see an average of (for example) -250% regen?

 

I'm just trying to figure out whether a constant -600% is worth three attacks, and if it's equivalent to what is on live right now.  I guess I should grab a power analyzer and try it myself.

 

 

[Armaros]

i would like to know why Robots are the one singled out as having something inherantly already existing in the set moved over to personal attacks, while the personal attacks having no actual synergy with Robots themselves

 

Its literally the same powers but with -regen tacked on while removing something that has thematically been part of the assault bot and all other 'plasma' blast attacks. Necro summons specters that scale with attack slotting, Mercs get bonus damage and Ninjas get crit, while Robots get -regen that only really matters if you stack all three powers.

 

The worst part of the change is that its boring and even unthematic.

Edited October 15, 2022 by Armaros

[Booper]

1 hour ago, gameboy1234 said:

 

Right, so the next question is what would be the expected uptime?  Would the -Regen last long enough that we'd expect it to double up?  Or would the -Regen wear off before the next 30% chance happened? Are we expecting to see -1500% regen or would it not trigger often enough and we'd only see an average of (for example) -250% regen?

 

I'm just trying to figure out whether a constant -600% is worth three attacks, and if it's equivalent to what is on live right now.  I guess I should grab a power analyzer and try it myself.

 

 

This depends on many factors. For the MM on Beta, it is straight forward as you have complete agency over the debuff(s) and where it goes and when it goes. For the assault bot, you'd having to analyze a random process. We know the recharge and cast time of the power, but there is no guarantee as to when the Assault Bot will choose to use the power(s). There is also the variable on how many powers the Assault Bot has that it may cycle through. By default the Assault Bot has one -regen power and two attacks. This scenario will likely result in the highest -regen uptime as the other attack has a longer cooldown and you are less likely to run out of endurance. If you equip your Robots, the Assault Bot will gain two more attacks (one of which is another -regen power) and, if you upgrade your Robots, the Assault Bot will gain another two attacks. This results in numerous variables, all impacting when a -regen attack may happen and if it happens. This is very difficult math to solve if you're hoping to calculate an expected uptime.

 

This is why testing changes is important. Instead of understanding advanced probability concepts, we can simply take a Power Analyzer and test against a target. If you monitor your combat logs (always show time stamps with seconds) and/or video record your tests, you can see what is the total regeneration debuff on a target over time. If you do multiple tests, you build up a sample size that can be used to analyze the distribution of performance.

 

There are other factors at play to consider as well. The first thing is...how much -regeneration do you need? Once a target is at 100%, there is no need for additional debuffs. In such a case, the only real applicability is AVs.

 

Against a +0 con Level 50 AV, they will have 85% resistance to debuffs. So the 600% that the MM can do on Beta will result in 90% debuff in 4 seconds. But that is their cap. Once you hit that mark, the AV will be regenerating their health once every 150 seconds (typically).

 

On Live, the Assault Bot will be able to achieve a higher debuff, but it will typically take much longer to reach the same 600% the MM can achieve. But once it gets two debuffs on the target it will win out. For simplicity, let's assume the Assault Bot uses its -regen power once every 6 seconds. Let's not include the other -regen power because it has a 29s cycle time and also the once every 6 seconds feels overly generous already. With an attempt every 6 seconds we can do some math on how long on average it would take to get 2 successes with a 30% chance to success rate. In this situation, you can expect it to take 7 attempts to achieve 2 successes (2 divided by 0.3 = 6.67). Assuming the 1st attempt is at time 0s, then the 7th attempt would be at 6x6s = 36s. I'd say we shave that down to 30s to account for the other debuff power getting one attempt inside of that window.

 

I hope that helps. I strongly recommend testing over assuming my math is representative. My numbers suggest you will get better performance from the Assault Bot over the long run, but it would take roughly 30s to reach that better performance in a best case scenario (where your Bot never misses).

[Armaros]

your calculation fails the moment you go beyond +0 con AVs and you need more -regen% to stop their regen, which the new system cannot hope to reach.

 

This is especially important considering the new wave of hard mode content pitting you against 54+ enemies.

Edited October 15, 2022 by Armaros

[Astralock]

27 minutes ago, Armaros said:

your calculation fails the moment you go beyond +0 con AVs and you need more -regen% to stop their regen, which the new system cannot hope to reach.

 

This is especially important considering the new wave of hard mode content pitting you against 54+ enemies.

 

 

 

[gameboy1234]

19 minutes ago, Booper said:

I strongly recommend testing over assuming my math is representative.

 

I did some quick testing this morning on Live.  Eochai was up in Atlas Park so I used him as a test with my Bots/Time and I grabbed a Power Analyzer from P2W.

 

Short summary was that -Regen had about a 75% uptime over four minutes of testing, but most of that was one -Regen only, and I never saw the powers triple stacked.  Unless one is trying to solo GMs, I think 4 minutes is a pretty long fight, most things with a team will drop before then.

 

So I had some trouble with the bots getting wiped out early, so I actually started testing about 3 minutes into my stopwatch test when I got everything stabilized.  I tried to use no debuffs from Time on Eochia so it wouldn't interfere with testing.  I did use Farsight and Chrono Shift on the bots to help keep them upright.  Of course I also used Temporal Mending, the Leadership pool and Equip and Upgrade Robot.

 

From 2 min 48 second until 7 minutes exactly (when the assault bot died again) here's what I got: 1 minute 5 seconds no debuffs, about 35 seconds double stacked (one -Regen was -2.5 hp/sec, two of them were -5 hp/sec), and the remainder of the time 2 minutes 30 seconds roughly Eochai had only one debuff on him.  That's slightly more than a 50% uptime with just one buff.  At no point did I see a triple debuff (-5 hp/sec from Plasma Burn plus the second power, -2.5 hp/sec from I think it was called Double Plasma Burst)

 

Because my timing started in progress, Eochai already had one -Regen on him, which might skew the numbers to look like -Regen has a higher uptime than it otherwise might.

 

[arcane]

1 hour ago, Armaros said:

Necro summons specters that scale with attack slotting, Mercs get bonus damage and Ninjas get crit, while Robots get -regen that only really matters if you stack all three powers.

I cannot speak definitively for the devs but my observation is that you can’t compare the different revamps because they don’t have the same goal. The revamps for other MM sets were likely done with the intention of boosting underperforming sets. Under no stretch of the imagination could I consider Robotics underperforming.

[Booper]

1 hour ago, gameboy1234 said:

 

I did some quick testing this morning on Live.  Eochai was up in Atlas Park so I used him as a test with my Bots/Time and I grabbed a Power Analyzer from P2W.

 

Short summary was that -Regen had about a 75% uptime over four minutes of testing, but most of that was one -Regen only, and I never saw the powers triple stacked.  Unless one is trying to solo GMs, I think 4 minutes is a pretty long fight, most things with a team will drop before then.

 

So I had some trouble with the bots getting wiped out early, so I actually started testing about 3 minutes into my stopwatch test when I got everything stabilized.  I tried to use no debuffs from Time on Eochia so it wouldn't interfere with testing.  I did use Farsight and Chrono Shift on the bots to help keep them upright.  Of course I also used Temporal Mending, the Leadership pool and Equip and Upgrade Robot.

 

From 2 min 48 second until 7 minutes exactly (when the assault bot died again) here's what I got: 1 minute 5 seconds no debuffs, about 35 seconds double stacked (one -Regen was -2.5 hp/sec, two of them were -5 hp/sec), and the remainder of the time 2 minutes 30 seconds roughly Eochai had only one debuff on him.  That's slightly more than a 50% uptime with just one buff.  At no point did I see a triple debuff (-5 hp/sec from Plasma Burn plus the second power, -2.5 hp/sec from I think it was called Double Plasma Burst)

 

Because my timing started in progress, Eochai already had one -Regen on him, which might skew the numbers to look like -Regen has a higher uptime than it otherwise might.

 

Thank you. I love that you went through and provided this detailed breakdown of your Live test results. There are certainly some good tidbits here in how gameplay can impact performance. The death of the pet will certainly result in poorer performance. I am somewhat surprised that over the 4:12 test run that only only 35 seconds had two stacks (14%). I would have thought that would be better, but again, I'm sure pet deaths play a part in that.

 

 

2 hours ago, Armaros said:

your calculation fails the moment you go beyond +0 con AVs

Normally, this goes without saying whenever someone shows math on one thing and try to apply the same math to another thing where purple patch plays a part. However, this is not the case. Re-read what I was calculating and you'll see that other than the value of the regen debuff after resistance (600% became 90%), everything I calculated was simply a comparison of when does Assault Bot overtake the MM in debuff. Purple patch does not play a part in that. You still need 2 debuffs stacked from the Bot to overtake the MM, and when that happens would still be roughly after 7 attack cycles (on average).

 

Now, what is different when it comes to the purple patch is how much stronger the Bot becomes once it gets two or more debuffs on the target. That is certainly a valid point to consider, however that point was never stated as being a part of my calculations. I would highly recommend testing against a Level 54 AV on Live and compare it to a Level 54 AV on Beta to see what the performance truly is. Based on @gameboy1234 test results above, perhaps we can expect the in-game performance to perform slightly worse than initially thought. But I won't be making that statement as fact. Test it, and provide results.

[WindDemon21]

35 minutes ago, Booper said:

Test it, and provide results.

*sigh* AGAIN, we're not arguing the percentage, of -regen, vs the new damage of the bots etc. We ARE arguing, that it shouldn't be in the MM attacks and should remain in the bots...

[Booper]

4 minutes ago, WindDemon21 said:

*sigh* AGAIN, we're not arguing the percentage, of -regen, vs the new damage of the bots etc. We ARE arguing, that it shouldn't be in the MM attacks and should remain in the bots...

You might be arguing that, but go back and look specifically at what I quoted in my replies to gameboy/Armaros and how I responded. There was no discussion about who should have -regen attacks. The discussion was specifically in regards to how does Live compare to Beta.

[WindDemon21]

Just now, Booper said:

You might be arguing that, but go back and look specifically at what I quoted in my replies to gameboy/Armaros and how I responded. There was no discussion about who should have -regen attacks. The discussion was specifically in regards to how does Live compare to Beta.

Ah so then there may be  a fix to put it back in the bots!?!

[tiberius0476]

Comparing the -regen performance for live vs beta is a little disingenuous. Even if somehow the new beta result is equivalent against AVs compared to live, achieving that comparable result requires selecting (and then likely slotting) 3 additional attack powers that most builds weren't intending to use, instead of the powers we were already using to help the team. Whereas on live, it is happening in the background already, without slots or endurance cost or animation time. That power tax is also independent of purple patch, which I agree is a problem but a different problem\

 

EDIT: Typo 

 

 

Edited October 15, 2022 by tiberius0476

[Bionic_Flea]

Robotic Flea: Bots/Rad, level 50 (Nerve radial, barrier, reactive, support) but only used passives and barrier.  No personal MM attacks.  Used rad toggles to start, Perma AM and Lingering Radiation.  Occasionally used Nemesis Staff or Blackwand if bored.  Heal aura and resummon as needed.  This was against a RWZ Pylon.

 

Live: 3:13

Beta: 2:24

 

It's only one data point, but informative.  True that Rad brings it's own significant -regen, but lots of MM secondaries have -regen.

 

Also, thanks to Booper, for mentioning setting the chat to clock with seconds.  Much easier to log precise times than trying to juggle a stopwatches start and stop.

Edited October 15, 2022 by Bionic_Flea

[tiberius0476]

1 minute ago, Bionic_Flea said:

Robotic Flea: Bots/Rad, level 50 (Nerve radial, barrier, reactive, support) but only used passives and barrier.  No personal MM attacks.  Used rad toggles to start, Perma AM and Lingering Radiation.  Occasionally used Nemesis Staff or Blackwand if bored.  Heal aura and resummon as needed.

 

Live: 3:13

Beta: 2:24

 

It's only one data point, but informative.  True that Rad brings it's own significant -regen, but lots of MM secondaries have -regen.

 

Also, thanks to Booper, for mentioning setting the chat to clock with seconds.  Much easier to log precise times than trying to juggle a stopwatches start and stop.

What were you fighting? As a bots/storm MM, the -regen changes do concern me, but I am glad to see most people reporting the increased base damage makes up for it (not including purple patch)

[Bionic_Flea]

23 minutes ago, tiberius0476 said:

What were you fighting?

Derp!  Edited original to add RWZ Pylon.

[ScarySai]

Another page, another wall of untested complaining about the -regen "loss".

 

Guys, its a straight buff no matter how you twist this. +0, to +4. Save it for something that's actually a nerf.

Edited October 16, 2022 by ScarySai

[Sovera]

2 hours ago, ScarySai said:

Another page, another wall of untested complaining about the -regen "loss".

 

Guys, its a straight buff no matter how you twist this. +0, to +4. Save it for something that's actually a nerf.

 

Easy to test to boot. All a matter of finding a +4 AV and wail on it on live and then on brainstorm, report results.

 

So far we've had tests on pylons that are not indicative (but brainstorm was faster than live) unless muddying the waters with math afterwards.

[Qualith]

4 hours ago, Sovera said:

 

Easy to test to boot. All a matter of finding a +4 AV and wail on it on live and then on brainstorm, report results.

 

So far we've had tests on pylons that are not indicative (but brainstorm was faster than live) unless muddying the waters with math afterwards.

It is easy, keep in mind mms do inflated damage on pylons. So pylon times aren't an accurate indication.

 

[Sovera]

25 minutes ago, Qualith said:

It is easy, keep in mind mms do inflated damage on pylons. So pylon times aren't an accurate indication.

 

 

Hence what I said of using a 54 pylon.

[tiberius0476]

7 hours ago, ScarySai said:

Another page, another wall of untested complaining about the -regen "loss".

 

Guys, its a straight buff no matter how you twist this. +0, to +4. Save it for something that's actually a nerf.

That's not quite what I'm trying to convey. Let me phrase this differently. It's not at all that I'm fixated on -regen values, or the well known hard mode/AV/purple patch situations where MMs often underperform.

 

With this patch, many Mastermind primary attack powers are being given new features to increase their appeal and flavor and usefulness, which is fantastic. Necromancers summon even more pets, ninja's give crit chances to pets, mercenaries straight increase pet damage. These are all awesome, and great ways to make these powers better. This approach is a great idea that I am happy to see the devs implementing.

 

But, the bots one falls short of this creativity, simply taking -regen that we previously had automatically on the assault bot, and moving it to powers that cost endurance, animation time, and powers/slots. 

 

That is the only change I am criticizing. The rest of the damage, animation and power changes to bots are awesome. The tests posted from beta prove this. Across the board they are better. Their animations are smooth and they pop off attacks fast enough that they even use all their mana sometimes. I don't even sweat the AOE changes to incendiary missiles because I think overall it'll be fine. Nobody has posted any data to say otherwise. Nobody is asking for those changes to be reverted or calling them a nerf.

 

I'm mostly just disappointed that robotics attacks got no new flavor or enticement, and in fact were given a feature we previously already had for free, and one that only matters in specific AV-type situations.

 

Was the assault bot -regen broken on new bots changes, so that it had to be moved? If that's the case, then this all makes sense.

 

Or would it be a reasonable option to revert the -regen changes back to assault bot free like live, if we could brainstorm a different feature we would consider giving the robotics attacks? (some sort of temporary stacking res/def bonuses for all pets, since robotics is a more defensive set to begin with? -maxhp instead of -regen, or anything that would be more helpful against more targets than just AVs, encouraging these attacks to be used more often? I'm just spitballing now)

 

Anyway, this was a feedback thread and that's my feedback.

[Communistpenguin]

So, if the devs are considering alternate models for the maintenance drone, I would like to suggest the sky raiders force field generator bot. Also, we cant heal the maintenance bot?