The Results are in...

I worry about attack chains in actual play because I design my characters to do one thing, do it well, and do it repeatedly. I go with a minimum number of attacks, tune the recharge so that they're ready when needed in a good attack chain, and spend the rest of the build trying to stay alive. So do I actually use attack chains in regular play? Absolutely. If I hit my best attacks as soon at they're recharged, I AM running my attack chain.

It's not for everyone, of course. Most people like more attacks and get bored hitting the same three keys over and over again for hours on end. I appear to be immune to that particular type of boredom.

---

"That's because Werner can't do maths." - BunnyAnomaly
"Four hours in, and I was no longer making mistakes, no longer detoggling. I was a machine." - Werner
Videos of Other Stupid Scrapper Tricks

[ QUOTE ]
I decided to do the math on Night Widows using the same numbers just to see how they compare. They score 220 DPS, placing them just below 10 target DM and a nice alternative for those who wants a more damage Claws toon. No Focus or Shockwave though, which are kinda set defining to me. They also have limited options to increase their damage. No -res IOs and no Shield Defense. Cheap Assault though. Here's the downside: 5.61 EPS.

[/ QUOTE ]

What's the chain you used?

---

Oh, I'm not saying I don't build with attack chains in mind and I also try to pick the best DPA attacks - just that the order that they're used gets changed based on in-game scenarios. For example, I frequently take at least one AoE simply because if I'm surrounded it's better DPA than any of the single-target attacks. Dark Obliteration is about even in DPA with the old Energy Transfer on the 5th target hit, and it's not hard to get 5 things in the radius of it. With 3 targets it's better than anything else Brutes get*, and near the top with 2.

Single target chains are all well and good, but they don't come into play very often in a meaningful situation (essentially only when trying to solo AVs and/or GMs). I just like the exercise in theory, and it gives me ideas for building - whether I stand toe-to-toe with a mob and continually use the chain or not.


Edited for footnote:
* - On thinking about it more, Ball Lightning is better (32.23 per target vs 31.60), but Gloom (55.64) beats out Mu Lightning (41.70) enough that you're better off with Ghost Widow than Scirocco as a patron. All numbers are base unenhanced at level 50 with 0 Fury.

---

[ QUOTE ]
I've actually got a more concrete method, but I'm a bit busy atm. I'll post it when I get 30 minutes to go over it.

[/ QUOTE ]

Are you still busy?

---

Blue
American Steele: 50 BS/Inv
Nightfall: 50 DDD
Sable Slayer: 50 DM/Rgn
Fortune's Shadow: 50 Dark/Psi
WinterStrike: 47 Ice/Dev
Quantum Well: 43 Inv/EM
Twilit Destiny: 43 MA/DA
Red
Shadowslip: 50 DDC
Final Rest: 50 MA/Rgn
Abyssal Frost: 50 Ice/Dark
Golden Ember: 50 SM/FA

[ QUOTE ]
[ QUOTE ]
I decided to do the math on Night Widows using the same numbers just to see how they compare. They score 220 DPS, placing them just below 10 target DM and a nice alternative for those who wants a more damage Claws toon. No Focus or Shockwave though, which are kinda set defining to me. They also have limited options to increase their damage. No -res IOs and no Shield Defense. Cheap Assault though. Here's the downside: 5.61 EPS.

[/ QUOTE ]

What's the chain you used?

[/ QUOTE ]

It's uh... long. FU > Slash > Lunge > Strike, repeat and move Slash one attack to the right each repetition (FU > Lunge > Slash... next one) until the last where Swipe replaces it and then restart.

---

I had all my chains built and began working on the new DPS calculations when I realized I had made a rather serious flaw in chain building logic.

I had slotted every attack 1acc/1end-red/1rec-red/3dam.

However, none of us would do that if dropping a dam for a recharge would tighten up our chains.

So... I'm starting all over tomorrow on the chains and adjusting all 14 of them to tighten them up.

This means that when I am doing the DPS calculations, instead of the easy .9908 (using lvl 50 basic IOs) across the board for damage, I'm going to have to keep track of the slotting for each attack.

While this will add a great deal of time to the final work, I feel that this is the correct way of going about it.

Especially when considering that with level 50 basic IOs, the loss of 15.76% damage enhancement means a gain of 40.92% recharge reduction enhancement when changing from 1/1/1/3 to 1/1/2/2 slotting.

Followup, Blinding Feint, Buildup, Soul Drain and Rage will all have 3 rec-reds.

---

Be well, people of CoH.

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
I decided to do the math on Night Widows using the same numbers just to see how they compare. They score 220 DPS, placing them just below 10 target DM and a nice alternative for those who wants a more damage Claws toon. No Focus or Shockwave though, which are kinda set defining to me. They also have limited options to increase their damage. No -res IOs and no Shield Defense. Cheap Assault though. Here's the downside: 5.61 EPS.

[/ QUOTE ]

What's the chain you used?

[/ QUOTE ]

It's uh... long. FU > Slash > Lunge > Strike, repeat and move Slash one attack to the right each repetition (FU > Lunge > Slash... next one) until the last where Swipe replaces it and then restart.

[/ QUOTE ]

So it is FU, Slash, Lung, Strike, FU, Lung, Slash, Strike, FU, Lung, Strike, Slash, FU, Lung, Strike, Swipe, repeat.

How do you get Lung to recharge in time for the second bolded instance?

---

I knew I should have written it down. It might have been FU, Lunge, Slash, Strike, FU, Lunge, Strike, Slash, FU, Lunge, Strike, Swipe. It was something like that, anyway. Did quite a few and don't remember which one it was that won. It might have been the one I said and I missed the Lunge - Lunge thingie. If that's the case the DPS will be a few notches lower, not much though.

---

[ QUOTE ]
What methodology can I use besides flat out trial and error to get the best chains possible using only single target attacks (and cones of <90degrees) for all the melee primaries with each power only getting 42.4?

I'm not asking for the chains themselves, but rather a methodology that could then be used across the board for all attack sets.

[/ QUOTE ]

...I don't know, Bill. I always assumed it was an np-complete problem.

---

[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
[ QUOTE ]
I decided to do the math on Night Widows using the same numbers just to see how they compare. They score 220 DPS, placing them just below 10 target DM and a nice alternative for those who wants a more damage Claws toon. No Focus or Shockwave though, which are kinda set defining to me. They also have limited options to increase their damage. No -res IOs and no Shield Defense. Cheap Assault though. Here's the downside: 5.61 EPS.

[/ QUOTE ]

What's the chain you used?

[/ QUOTE ]

It's uh... long. FU > Slash > Lunge > Strike, repeat and move Slash one attack to the right each repetition (FU > Lunge > Slash... next one) until the last where Swipe replaces it and then restart.

[/ QUOTE ]

So it is FU, Slash, Lung, Strike, FU, Lung, Slash, Strike, FU, Lung, Strike, Slash, FU, Lung, Strike, Swipe, repeat.

How do you get Lung to recharge in time for the second bolded instance?

[/ QUOTE ]

That chain can be maintained if you toss a swipe in after the strike as long as slash recharges in 4.488 that was the highest one I came up with. Your attack chain where slash is moved down is indeed higher, I wouldn't have thougth of that, nice one.

Followup > Lunge > Strike > Swipe > Slash
.86/2.1/1.3/.985/1.98 in 6.072 = 1.19

FU > Lunge > Slash > Strike > FU > Lunge > Strike > Slash > FU > Lunge > Strike > Swipe
.86/2.1/1.98/1.3/.86/2.1/1.3/1.98/.86/2.1/1.3/.985 in 14.52 = 1.22

---

[ QUOTE ]
How do you get Lung to recharge in time for the second bolded instance?

[/ QUOTE ]

Take a deep breath...

---

[ QUOTE ]
[ QUOTE ]
How do you get Lung to recharge in time for the second bolded instance?

[/ QUOTE ]

Take a deep breath...

[/ QUOTE ]

Thank you. I've been waiting for someone to say that.

---

QR
I'm currently fiddling around with tankers(mmm fire blast) but can't find a proper recharge formula.
Werner posted a chart with the =ROUND.UP(100*Recharge/Req.Rech;0)/100-1
but it's not working for me the numbers are way off when i check in mids and sometimes there negatives.
What am i doing wrong?

---

[ QUOTE ]
I had all my chains built and began working on the new DPS calculations when I realized I had made a rather serious flaw in chain building logic.

I had slotted every attack 1acc/1end-red/1rec-red/3dam.

However, none of us would do that if dropping a dam for a recharge would tighten up our chains.

So... I'm starting all over tomorrow on the chains and adjusting all 14 of them to tighten them up.

This means that when I am doing the DPS calculations, instead of the easy .9908 (using lvl 50 basic IOs) across the board for damage, I'm going to have to keep track of the slotting for each attack.

While this will add a great deal of time to the final work, I feel that this is the correct way of going about it.

Especially when considering that with level 50 basic IOs, the loss of 15.76% damage enhancement means a gain of 40.92% recharge reduction enhancement when changing from 1/1/1/3 to 1/1/2/2 slotting.

Followup, Blinding Feint, Buildup, Soul Drain and Rage will all have 3 rec-reds.

[/ QUOTE ]

If you are going with such high recharge, I would assume someone has at least Frankenslotted the chain?

Let me suggest this as a base line fore melee:

5 Focused Smite set and a Smashing Haymaker Dam/End/Rec. Every single recipe is between 5000-25000 and cheaper than buying standard IO recipes from the university. None use rare salvage.

Acc 61.54, Dam 95.97, End 60.59, Recharge 78.90

To customize you can either swap to a Smashing Haymaker Dam/Rech to tweak Recharge up to 83.03 or even regular Recharge IO to tweak recharge to 93.77 and damage down to 86.30. That's for Follow Up or other attacks where high recharge is better than damage.

---

Moonlighter

50s include MA/SD, MA/SR, DP/Elec, Claw/Inv, Kat/Dark, Kat/Fire, Spine/Regen, Dark/SD

First Arc: Tequila Sunrise, #168563

[ QUOTE ]
QR
I'm currently fiddling around with tankers(mmm fire blast) but can't find a proper recharge formula.
Werner posted a chart with the =ROUND.UP(100*Recharge/Req.Rech;0)/100-1
but it's not working for me the numbers are way off when i check in mids and sometimes there negatives.
What am i doing wrong?

[/ QUOTE ]
The recharge formula the game uses is:

Final Recharge Time = Base Recharge Time / (1 + Recharge Buffs)

Now, for some math, using R(f) for final recharge, R(b) for base recharge, and Buffs for recharge buffs:

R(f) * (1 + Buff) = R(b)

R(f) - (R(f) * Buff)) = R(b)

R(f) * Buff = R(b) - R(f)

And, substituting things back in with the full name:

Recharge Buff (needed) = (Base Recharge Time - Final Recharge Time) / Final Recharge Time


So, for example, if you needed to know the recharge needed to get Bone Smasher (base of 8 seconds recharge) to a 2.9 second recharge (so you can squeeze it around an Energy Transfer) it would be:

(8 - 2.9) / 2.9, or 1.75862...

That's +176% recharge, counting slotting and any global recharge buffs such as set bonuses, specific power buffs, or Hasten. As per the original post, the chains were just set with +250% recharge in each power, so the recharge time on each would be (base recharge / 3.5).

---

[ QUOTE ]
[ QUOTE ]
I had all my chains built and began working on the new DPS calculations when I realized I had made a rather serious flaw in chain building logic.

I had slotted every attack 1acc/1end-red/1rec-red/3dam.

However, none of us would do that if dropping a dam for a recharge would tighten up our chains.

So... I'm starting all over tomorrow on the chains and adjusting all 14 of them to tighten them up.

This means that when I am doing the DPS calculations, instead of the easy .9908 (using lvl 50 basic IOs) across the board for damage, I'm going to have to keep track of the slotting for each attack.

While this will add a great deal of time to the final work, I feel that this is the correct way of going about it.

Especially when considering that with level 50 basic IOs, the loss of 15.76% damage enhancement means a gain of 40.92% recharge reduction enhancement when changing from 1/1/1/3 to 1/1/2/2 slotting.

Followup, Blinding Feint, Buildup, Soul Drain and Rage will all have 3 rec-reds.

[/ QUOTE ]

If you are going with such high recharge, I would assume someone has at least Frankenslotted the chain?

Let me suggest this as a base line fore melee:

5 Focused Smite set and a Smashing Haymaker Dam/End/Rec. Every single recipe is between 5000-25000 and cheaper than buying standard IO recipes from the university. None use rare salvage.

Acc 61.54, Dam 95.97, End 60.59, Recharge 78.90

To customize you can either swap to a Smashing Haymaker Dam/Rech to tweak Recharge up to 83.03 or even regular Recharge IO to tweak recharge to 93.77 and damage down to 86.30. That's for Follow Up or other attacks where high recharge is better than damage.

[/ QUOTE ]

Negative. SetIOs are not allowed in Take 2 of this little exercise.

And round 2 is posted!

---

Be well, people of CoH.

That works better, thanks.
Hmmm inc, fb, gfs, fb should work then for fiery melee and sm, fb, mg, sm, fb, sl for dm.
<3 fire blast.

---

[ QUOTE ]
[ QUOTE ]
QR
I'm currently fiddling around with tankers(mmm fire blast) but can't find a proper recharge formula.
Werner posted a chart with the =ROUND.UP(100*Recharge/Req.Rech;0)/100-1
but it's not working for me the numbers are way off when i check in mids and sometimes there negatives.
What am i doing wrong?

[/ QUOTE ]
The recharge formula the game uses is:

Final Recharge Time = Base Recharge Time / (1 + Recharge Buffs)

Now, for some math, using R(f) for final recharge, R(b) for base recharge, and Buffs for recharge buffs:

R(f) * (1 + Buff) = R(b)

R(f) - (R(f) * Buff)) = R(b)

R(f) * Buff = R(b) - R(f)

And, substituting things back in with the full name:

Recharge Buff (needed) = (Base Recharge Time - Final Recharge Time) / Final Recharge Time


[/ QUOTE ]
Continuing to simplify that equation:

Recharge Buff (needed) = Base Recharge Time / Final Recharge Time – 1

And if you want that in a whole percentage, you need to round up, because if you need 95.2% recharge, 95% won't do.

= roundup(Base Recharge Time / Final Recharge Time, 2 decimal positions) – 1

I can't remember why I did the *100 and /100 to get the 2 decimal positions rather than just specifying 2 decimal positions in the formula. I obviously knew that I could specify decimal positions, as I specified 0 explicitly. Perhaps I ran into a calculation error in Open Office. Perhaps I just didn't trust other people's approaches to rounding off decimal places after seeing a reporting package at work use binary floating point as its internal format for these calculations, which DOES lead to incorrect decimal results.

In any case, the formula I posted works just fine for me. I don't know why it wouldn't be working for you.

---

"That's because Werner can't do maths." - BunnyAnomaly
"Four hours in, and I was no longer making mistakes, no longer detoggling. I was a machine." - Werner
Videos of Other Stupid Scrapper Tricks

[ QUOTE ]
[ QUOTE ]
I've actually got a more concrete method, but I'm a bit busy atm. I'll post it when I get 30 minutes to go over it.

[/ QUOTE ]

Are you still busy?

[/ QUOTE ]

I'm still a bit busy (getting my Summer plans fleshed out and dealing with my Saturday D&D game), though I've got the time to actually give a basic breakdown of what I hope to eventually turn into "Umbral's Guide to Attack Chains" (it's still in the alpha phase).

The basic idea I operate with is that there are 3 types of attack chains. 2 of them are short and sweet and the other is long and reasonably convoluted. They're all effective, and, depending on your build, will sometimes different types of attack strings will be required at different times for different powersets and amounts of +recharge. Oftentimes, one type may be optimal at a certain threshold of +recharge and significantly worse at another.

The first is simply a straight, non-convoluted attack string like the non-gap Broadsword string mentioned before (HS>Dis>Hack>Slash). The design is simple and elegant (get your 4 highest DPA attacks and string them in a row), though, as is shown by adding gaps, can sometimes be improved. Of course, the adding gaps portion is actually reasonably easy to analyze. I'll get to that later.

The second type is the "alternating" attack string. Dark Melee does this with Mg>Smite>Siphon>Smite. The entire concept is that you're alternating between a fast recharging attack (that you've gotten to the recharge below the animation time of the other attacks) and 2 slow recharging attacks. Katana also does this with GC>GD>GC>SD, though uses a slightly different mentality because, rather than building around the slow recharging attack, it builds around its fastest attack.

The last "convoluted" type is generally most useful for partially optimized sets and is often the the hardest part to factor in. The basic design method here is in having 2 attacks that must be used as often as possible (rather than just 1) but have too dissimilar of recharge times. Oftentimes, the attack string will actually be the 2 or more "sub-attack strings" with a single power switched in with the others. A non-optimized Claws build I designed an attack string for used FU>Foc>Slash>Swipe>Foc>FU>Slash& gt;Focus>Swipe>Slash. Gotta use Focus and FU as soon as their recharge and use the other attacks only when necessary.

As to the adding-a-gap option (which is only recommended when you've got attacks with highly disparate DPA), all comes down to math. A gap will be more beneficial than actually using an attack in the last slot of the attack string is lower DPA than the DPA of the first attack with an activation time that is longer by duration of the gap. That's a bit complicated, so here's an example.

Take an arbitrary imaginary attack string. The first attack has a DPA of 100 and the middle attacks don't matter. The final attack has a DPA of 50 but is going to extend animation time of the attack string .5 seconds longer than the animation time + recharge time of the first attack. So, here's the question: is the DPA of the "last attack + first attack" sub-attack string higher than the DPA of the "gap + first attack" attack string? For some sets that have very long attack strings (Broadsword), the effect of an extra 5 seconds on the big, high DPA attack is minor. For others that have fast animations (Claws, Katana), the effect is going to be significant. In the example, the animation time of the first attack would have to be less than 1 second in order for the gap to not be the better option.

Your best bet for creating attack strings is going to be figuring out what pivotal attacks you're going to build it around (HS and Disembowel; FU and Focus; etc.). In general, they're simply the highest DPA attacks you've got available. Everything else is filler. Then, figure out priority, including which filler attacks are more important. After that, it's just an issue of following the priority list until a cycle emerges and analyzing the use of filler attacks immediately before use of highest priority attacks to determine whether the presence of a gap would be better.

For any detractors, this is still pretty alpha for a set of protocols. I'm stilling working on getting all of the wording done for the processes I already follow innately. If you've got any questions, ask away. It'll help me chisel this into something a bit more comprehensible.

---

I knew there was a good reason to put this thread on my Favorites list. heh.

---

Y'know, I never bothered to further simplify just because I usually end up having the recharge time needed stored in memory in the calculator (having just calculated out either the addition of the animation times between activations) and hitting / MR was easy enough that I never thought about making it any simpler once I isolated the variable.

---

I love this thread.

14 pages about Brute/scrapper balance issues that only come up in builds achievable by 1% of the player base.


Really?

****** head to one side trying to get a better angle*

Really????

---

When something good happens to me, I can never enjoy it....
I am always too busy looking for the inevitable punchline...

BEHOLD THE POWER OF CHEESE!

Go join the same argument in part 2 where the builds are easily achievable by 100% of the player population that can hit level 50.

---

Be well, people of CoH.

[ QUOTE ]
Go join the same argument in part 2 where the builds are easily achievable by 100% of the player population that can hit level 50.

[/ QUOTE ]

Yes, Part 2 is a hoot lol

---

I'm curious what this would look like with the higher Crit chances taken into account for Scrappers. After all, the targets that the increase applies to are the ones where single target DPS is going to matter more. I think that, coupled with the usage of procs (which will give an advantage to the shorter-animating attacks) are the main reasons my numbers are so wildly different from Billz'.

---