When do powers start to recharge? - "In the weeds" about single target attack chains

[EnjoyTheJourney]

A very good single target attack chain for any given set of powers will fully realize the single target damage potential of those powers, keep endurance management in a good place, be relatively easy to execute, and use slots and other resources in an efficient way to achieve these objectives. 

 

Fully realizing single target damage potential ...

 

Much about this is straightforward. One issue is straightforward in theory, but potentially a bit challenging to sort out in practice. 

 

In theory, it helps to have a seamless transition between powers used in the single target attack chain. One power is used, the next one is ready to go and used immediately, and so on ... with no gaps between the availability of powers to be used (ie: no "dead space" in the attack chain). This is relevant because time spent waiting for the next attack to recharge is time spent doing no damage*.

 

* A possible exception is that sometimes a short wait will allow a key power to fully recharge (ie: a power that boosts damage) so it can be used just ahead of a barrage of heavier damage attack powers. 

 

In practice, figuring out how to minimize "dead space" in a single target attack chain can be challenging. This is true in part because of uncertainty concerning when powers start to recharge. Do they start to recharge immediately after you press a button to use them? Do they start to recharge when they activate, but before they finish their animation? Do they start to recharge after their animation fully completes, but before arcanatime finishes up? Do they only start to recharge after arcanatime casting time finishes up? Are there different answers to these questions for different powers and, if there are differences, is there an observable patterns to the differences? 

 

Thank you in advance for any insight provided.

Edited May 19, 2022 by EnjoyTheJourney

[EnjoyTheJourney]

There seems to be three possible reasons why there's no response to this question.

 

1. The answer is so obvious and simple that people are embarrassed to tell me. 

2. Nobody who has read this thread is quite sure of the answer, and so "no response" seems like the best response.

3. A number of people know the answer and are trying their best to respond. But, their hands are being held by their sides, away from their keyboard, by a Very Mysterious Force. 

 

Regarding possible reason #1 above, feel free to embarrass me with your understanding of when the recharge countdown starts after clicking a power to use it. Knowing this helps noticeably for figuring out how to slot single target attack chains and there are probably plenty of people who don't know. 

 

Regarding possible reason #3 above, get voice recognition going so you can continue to post and to ask for help getting rid of that Very Mysterious Force holding your arms by your sides. 

[Linea]

On 5/19/2022 at 5:44 PM, EnjoyTheJourney said:

Do they only start to recharge after arcanatime casting time finishes up? Are there different answers to these questions for different powers and, if there are differences, is there an observable patterns to the differences? 

Assume the answer is Arcanatime.  Due to backend changes, Arcanatime is not always accurate now, but is the best estimate you are going to get.

 

Some powers vary due to internal activation issues.  See: BuildUp.  It's actual duration is only ~9.3 seconds.  I wouldn't be at all surprised to see these same issues have minor effects on recharge in some rare cases as well.

 

Edited May 25, 2022 by Linea

[MoonSheep]

i like to optimise my attack chain by thinking about which order to bash out my powers, especially on my mind/fire dom which has very fast animations. i haven’t put much more thought into it though as i think it’s such a minimal gain that it’s not worth thinking about

 

in my opinion people would get more out their characters by learning the best time to use their powers, mob positioning for survival and which targets to attack to make best progress 

[Bill Z Bubba]

3 hours ago, Linea said:

Assume the answer is Arcanatime.

 

This.

 

Here's how I do it.

Generally speaking, I want a chain of no more than 4 attacks. Ever. This isn't always possible with freak powers like seismic smash and the like. But for claws, I want an attack chain of followup, focus, slash, repeat with no pauses.

 

Followup's base recharge is 12 seconds. It has a given cast time of .83 seconds and an arcanatime of 1.056.

Focus has base recharge of 6.4 seconds, cast 1.17 and arcanatime of 1.32.

Slash has base recharge of 4.8 seconds, cast 1.33 and arcanatime of 1.584.

 

Recharge equation: Time = BaseRecharge / (1+X) where X is the amount of recharge needed

Tweaked around for my purposes, X = (BaseRecharge - Time) / Time

The Time we're dealing with is the period between activations of followup which is the sum of the cast times for focus and slash.

 

If we ignore arcanatime, then followup has to recharge within 2.5 seconds. (12-2.5) / 2.5 = 3.8 or 380% total recharge in followup to pull it off. But that's not the reality of the situation. With arcanatime, it works out instead to  (12-2.904) / 2.904 = 3.13 or 313% total recharge in followup. Using pylons as the primary test bed, I know that my chain becomes pauseless at 313% and not 380%.

 

To get the other recharge values needed, you do the exact same but Time will change.

To calculate focus: (6.4-2.64) / 2.64 = 1.42

To calculate slash: (4.8-2.376) / 2.376 = 1.02

 

Total side note: because claws' secondary effect is that all attacks cost 20% less endurance and recharge 20% faster, Followup's recharge should be 9.6 instead of 12 seconds. I've no clue why this was never corrected.

 

Let's look at Dual Blades instead. There, I want the chain to be Blinding Feint, Ablating Strike, Sweeping Strike, Ablating Strike, repeat.

BF 12 sec recharge, again, and a 1.452 arcanatime-cast. 1.188+1.452+1.188 = 3.828 seconds between activations.

AS 6 recharge and a 1.188 a-cast. But for the requested chain, the time between activations would be 1.452! So we'd better see if that's possible first.

SS 11 recharge and 1.452 a-cast. Also 3.828 between activations.

(6-1.452) / 1.452 = 3.13  ...... well, ain't that weird? 313% recharge again. Spooky.

BF only needs (12-3.828) / 3.828 = 2.13 or 213% total recharge.

 

 

Edit: I know I've posted this before. I should save it this time so I don't have to do it again.

Edited May 25, 2022 by Bill Z Bubba

[Bill Z Bubba]

Oh yea... so why not BF, SS, AS, repeat for DB? Cuz you'd need 355% recharge in BF. Pretty sure that ain't possible without heavy outside buffs.

[EnjoyTheJourney]

4 hours ago, Linea said:

Assume the answer is Arcanatime.  Due to backend changes, Arcanatime is not always accurate now, but is the best estimate you are going to get.

 

Some powers vary due to internal activation issues.  See: BuildUp.  It's actual duration is only ~9.3 seconds.  I wouldn't be at all surprised to see these same issues have minor effects on recharge in some rare cases as well.

 

Activation time is one of the issues I'm wondering about. For a noticeable number of powers their activation time will occur 0.5 seconds to 1 second earlier than the finish of their animation time. For such powers noticeably less recharge would needed to maintain a continuous attack chain if powers start to recharge as soon as they take effect. 

 

When expanded across quite a few powers, this could be a noticeably source of difference between DPS for different builds. 

 

Some examples of differences between arcanatime and animation time before effect ...

 

Dominator earth assault, Seismic smash: arcanatime = 1.716 sec, animation time before effect = 0.833 sec (about 0.88 seconds difference)

Dominator earth assault, power up: arcanatime = 1.167 sec, animation time before effect = 0.5 sec (0.667 seconds difference)

Dominator earth assault, heavy mallet: arcanatime = 1.848 sec, animation time before effect = 0.8 sec (1.048 seconds difference)

Dominator psi assault, telekinetic thrust: arcanatime = 1.188 sec, animation time before effect = 0.533 sec (0.655 seconds difference)

Dominator psi assault, mind probe: arcanatime = 1.32 sec, animation time before effect = 0.567 sec (0.676 seconds difference)

 

... and so on

 

1 hour ago, Bill Z Bubba said:

 

This.

 

Here's how I do it.

Generally speaking, I want a chain of no more than 4 attacks. Ever. This isn't always possible with freak powers like seismic smash and the like. But for claws, I want an attack chain of followup, focus, slash, repeat with no pauses.

 

Followup's base recharge is 12 seconds. It has a given cast time of .83 seconds and an arcanatime of 1.056.

Focus has base recharge of 6.4 seconds, cast 1.17 and arcanatime of 1.32.

Slash has base recharge of 4.8 seconds, cast 1.33 and arcanatime of 1.584.

 

Recharge equation: Time = BaseRecharge / (1+X) where X is the amount of recharge needed

Tweaked around for my purposes, X = (BaseRecharge - Time) / Time

The Time we're dealing with is the period between activations of followup which is the sum of the cast times for focus and slash.

 

If we ignore arcanatime, then followup has to recharge within 2.5 seconds. (12-2.5) / 2.5 = 3.8 or 380% total recharge in followup to pull it off. But that's not the reality of the situation. With arcanatime, it works out instead to  (12-2.904) / 2.904 = 3.13 or 313% total recharge in followup. Using pylons as the primary test bed, I know that my chain becomes pauseless at 313% and not 380%.

 

To get the other recharge values needed, you do the exact same but Time will change.

To calculate focus: (6.4-2.64) / 2.64 = 1.42

To calculate slash: (4.8-2.376) / 2.376 = 1.02

 

Total side note: because claws' secondary effect is that all attacks cost 20% less endurance and recharge 20% faster, Followup's recharge should be 9.6 instead of 12 seconds. I've no clue why this was never corrected.

 

Let's look at Dual Blades instead. There, I want the chain to be Blinding Feint, Ablating Strike, Sweeping Strike, Ablating Strike, repeat.

BF 12 sec recharge, again, and a 1.452 arcanatime-cast. 1.188+1.452+1.188 = 3.828 seconds between activations.

AS 6 recharge and a 1.188 a-cast. But for the requested chain, the time between activations would be 1.452! So we'd better see if that's possible first.

SS 11 recharge and 1.452 a-cast. Also 3.828 between activations.

(6-1.452) / 1.452 = 3.13  ...... well, ain't that weird? 313% recharge again. Spooky.

BF only needs (12-3.828) / 3.828 = 2.13 or 213% total recharge.

 

 

Edit: I know I've posted this before. I should save it this time so I don't have to do it again.

 

Thank you for this. It's very helpful and I've incorporated these ideas into a spreadsheet that will help with building and streamlining attack chains. 

[biostem]

I mean, neutrino bolt for defenders, (haven't checked on corruptors), has like only a 2 second recharge and animates pretty quickly.  If you wanted to max out its recharge, you could get that down to about 1 second.  I'm not sure on its exact animation time, but it's pretty quick.  Permitting 2 attacks would seem like a better option, IMO...

[Doomguide2005]

Neutrino Bolt -> Defender version

• Cast time 1.0 sec, Arcanatime 1.188 sec
• Root time 1.0 sec
• Animation time 1.0 sec
• Animation time before effect 0.567 sec
• Recharge time 1.5 sec

Corruptor values the same per CoD v2.0

[Bill Z Bubba]

38 minutes ago, EnjoyTheJourney said:

For a noticeable number of powers their activation time will occur 0.5 seconds to 1 second earlier than the finish of their animation time.

 

Don't get me started. Animation time SHOULD be equal to root time which SHOULD be equal to arcanatime.

[Bill Z Bubba]

27 minutes ago, biostem said:

I mean, neutrino bolt for defenders, (haven't checked on corruptors), has like only a 2 second recharge and animates pretty quickly.  If you wanted to max out its recharge, you could get that down to about 1 second.  I'm not sure on its exact animation time, but it's pretty quick.  Permitting 2 attacks would seem like a better option, IMO...

 

My swipe-only/invul scrapper knows all about the attempt to get a single attack perma.... it is not and can not be as recharge doesn't start until the animation/cast/rooted time completes.

[Doomguide2005]

Lol

Brawl is faster than Neutrino Bolt except recharges in 2 seconds rather than 1.5

[Bill Z Bubba]

9 minutes ago, Doomguide2005 said:

Neutrino Bolt -> Defender version

• Cast time 1.0 sec, Arcanatime 1.188 sec
• Root time 1.0 sec
• Animation time 1.0 sec
• Animation time before effect 0.567 sec
• Recharge time 1.5 sec

Corruptor values the same per CoD v2.0

 

Even at the cap, (+300?) you're left with a pause of .375 seconds. The horror.... the horror... the horror.

[biostem]

2 minutes ago, Bill Z Bubba said:

My swipe-only/invul scrapper knows all about the attempt to get a single attack perma.... it is not and can not be as recharge doesn't start until the animation/cast/rooted time completes.

I mean, even if a power had less than a second recharge time, you'd still have to wait for the animation to play out, so what we're really talking about, in practical terms, is "as close to perma as possible"...

[Doomguide2005]

9 minutes ago, Bill Z Bubba said:

 

Even at the cap, (+300?) you're left with a pause of .375 seconds. The horror.... the horror... the horror.

The cap would be +400 (500% total).  

 

Acck, arrg brain melt down.  YES +300 (400%) total for a recharge time of 1/5th base time

Edited May 25, 2022 by Doomguide2005
Woops

[Bill Z Bubba]

4 minutes ago, Doomguide2005 said:

The cap would be +400 (500% total).  

 

Still .3 second pause of NOT KILLING!!!

[Bill Z Bubba]

6 minutes ago, Doomguide2005 said:

Acck, arrg brain melt down.  YES +300 (400%) total for a recharge time of 1/5th base time

 

YAY! My memory actually worked for once!

[EnjoyTheJourney]

If the powers used in a single target attack chain averaged 1.584 seconds of arcanatime per cast and have an average of 0.375 seconds of "lost time" between them, then that would be about a 20% DPA loss*. Under these circumstances 250 damage becomes 200, 500 damage becomes 400, and so on. If arcanatimes are shorter than 1.584 seconds then even more DPA would be lost (with an average of 0.375 seconds of lost time between attacks). 

 

* This is being a bit loose with the numbers. Still, the presence of one "extra" attack at the beginning or end of the attack sequence, compared to the number of "lost time" incidents helps to reduce what would be fairly close to a 25% DPA loss to about a 20% DPA loss (depending on the number of attacks). 

 

Ergo, repeatedly losing small amounts of time can add up to a non-negligible DPA drop. 

Edited May 25, 2022 by EnjoyTheJourney

[Mezmera]

I like to get all of my characters down to 4 of the best attacks in their assaults that I'll use in a sequence and typically the aoe attacks mixed in where advantageous to use unless it's my tank where I want to soak up aggro with pbaoe attacks so those get moved into the first 4 rotation.  

 

For instance on my Mind/Nrg dom my sequence can look like this: 

#1 key: Good fast lead in attack either Bone Smasher or Dominate (go-to control)

#2 key: Power Burst; My best follow up power

#3 key: Total Focus; The attack I look to lean on the most with great damage and leads into other things like the Energy Mechanic

#4 key: Snipe; Second follow up after TF and Power Burst

 

My recharge is usually good enough that I can use #3 and then go with 2, 4 and 1 by the time 3 recharges to start it all over again.  #5 usually is where I slot a good aoe attack, after that I'll be ready to click other areas in the trays or binds.  

 

But yes you need to consider animation times of most things on top of those in your main chain because things that you'll want to use like Build Up, Soul Drain or Sleet eat into the fluidity of your chain and mostly those powers are too beneficial to not use.  

 

 

Edited May 25, 2022 by Mezmera

[Hjarki]

3 hours ago, Bill Z Bubba said:

Recharge equation: Time = BaseRecharge / (1+X) where X is the amount of recharge needed

Tweaked around for my purposes, X = (BaseRecharge - Time) / Time

The Time we're dealing with is the period between activations of followup which is the sum of the cast times for focus and slash.

This is the same way I do it - and I'm pretty sure it's slightly wrong both from my observations in game and my general understanding of how servers tend to be implemented.

 

CoH appears to run on a 0.132 sec tick. That is, everything which requires server communication can happen with no more granular of time than 0.132s.

 

That's why you get arcanatime. You send a message to the server saying "I want to activate this power". On the next tick, it recognizes the message and starts counting one per tick until it's counted the right number of ticks - at which point the power activates. Because you can never hit the tick exactly, there's always a 'dead tick' between powers.

 

So if your power takes 1.0 sec to activate, that's means the server has to count at least roundup(1.0 / 0.132) = 8 ticks before it will let the power activate. Add a 'dead tick' between power activations that occurs because you can never precisely hit the exact moment when a tick occurs and you've got your arcanatime.

 

However, this would also apply to durations and recharges. So let's say you have a power with a 1.0 sec activation and a 10.0 sec recharge. You'd see something like:

0 tick: Before you activate power

1 tick: The first tick after you activate power and we start counting

8 tick: You finish the count and start your recharge count

9..83 ticks: Counting down the recharge

84 tick: Recharge up and power is ready if you want to use it (same spot as our 0 tick above)

 

This entire sequence takes 11.088 sec.

 

The sequence where we just use arcanatime for activation (but not recharge) would be 11.188 sec (a tenth of a second longer). If we're waiting to activate our power, it would be 11.22 sec. Essentially, we're double-counting that 'dead tick' by assuming recharge/duration are continuous while activation is discrete.

 

The reason I think it works this way is that I've noticed that you don't need to fully close rotations to make them work. If you miss by a little bit, you're still activating powers as quickly as they can possibly activate. Moreover, it makes sense that they're using the same tick-based timekeeping for everything (not just activation).

 

Note: Some powers activate midway through the animation. I assume this activation is also impacted by arcanatime but has no impact on when the recharge starts counting, only when the duration starts counting.

Edited May 25, 2022 by Hjarki

[EnjoyTheJourney]

13 hours ago, Hjarki said:

This is the same way I do it - and I'm pretty sure it's slightly wrong both from my observations in game and my general understanding of how servers tend to be implemented.

 

CoH appears to run on a 0.132 sec tick. That is, everything which requires server communication can happen with no more granular of time than 0.132s.

 

That's why you get arcanatime. You send a message to the server saying "I want to activate this power". On the next tick, it recognizes the message and starts counting one per tick until it's counted the right number of ticks - at which point the power activates. Because you can never hit the tick exactly, there's always a 'dead tick' between powers.

 

So if your power takes 1.0 sec to activate, that's means the server has to count at least roundup(1.0 / 0.132) = 8 ticks before it will let the power activate. Add a 'dead tick' between power activations that occurs because you can never precisely hit the exact moment when a tick occurs and you've got your arcanatime.

 

However, this would also apply to durations and recharges. So let's say you have a power with a 1.0 sec activation and a 10.0 sec recharge. You'd see something like:

0 tick: Before you activate power

1 tick: The first tick after you activate power and we start counting

8 tick: You finish the count and start your recharge count

9..83 ticks: Counting down the recharge

84 tick: Recharge up and power is ready if you want to use it (same spot as our 0 tick above)

 

This entire sequence takes 11.088 sec.

 

The sequence where we just use arcanatime for activation (but not recharge) would be 11.188 sec (a tenth of a second longer). If we're waiting to activate our power, it would be 11.22 sec. Essentially, we're double-counting that 'dead tick' by assuming recharge/duration are continuous while activation is discrete.

 

The reason I think it works this way is that I've noticed that you don't need to fully close rotations to make them work. If you miss by a little bit, you're still activating powers as quickly as they can possibly activate. Moreover, it makes sense that they're using the same tick-based timekeeping for everything (not just activation).

 

Note: Some powers activate midway through the animation. I assume this activation is also impacted by arcanatime but has no impact on when the recharge starts counting, only when the duration starts counting.

 

Essentially, all times given for duration and recharge for powers are estimates because of how ticks work. This explanation seems very, very plausible. Thank you for posting this. 

 

It would seem that arcanatime is the best available basis for estimating recharge times and animation time before effect is the most plausible beginning point for estimating duration.