SnD usage as combat question

[Warstar]

Since combat damage is all about CD usage and getting as many uses of KS/AR&Blades as possible and with restless blades giving us CDR I have a question about potentially letting SnD fall off for the sake of lower CDs. Since refreshing SnD doesn't effect restless blades and only stuff like eviscerate does would it be a good idea to let SnD drop for a few seconds to use a 5 point finisher on evis and then the ruthlessness bonus CP on SnD which in turn can proc an additional ruthlessness CP sometimes giving a "free" (cp wise) SnD but sometimes not.

From what I've read in the past the standard rule of thumb was refresh SnD as soon as it wears off with whatever CP you have be it 1 or 5 but I also read this before they re-added ruthlessness in.

So what about in a situation where you have 5 CP and SnD wears off? Do you evis for the CDR? Or SnD for a long SnD?
What about if you have 4 CP and SnD wears off? Do you let SnD slide for a few seconds, generate 1 more CP and evis to reduce an extra 10 seconds off your major CDs? Or do you just refresh SnD right there (and potentially not even generate a point of ruthlessness)?

With our current style of game play with the AoC involved and high levels of haste cooldowns are being used so much and that's where the vast majority if our DPS is coming from so it seems like potentially getting there CD up faster would be worth a few seconds here and there of SnD down time. And getting semi free SnD usage seems to be nice as well.

I'm probably just looking at the short end of this and it "feels" faster and potentially mathed out over a whole fight it's a dps loss and I'm screwing up. Don't think I've seen anywhere how the addition of ruthlessness effected our usage of SnD and how it potentially changes its use.

[Squirl]

No. SnD is our single highest damaging move. Always keep it up with as many CPs as you have; 5 is best since it's the most energy efficient.

[Warstar]

No. SnD is our single highest damaging move. Always keep it up with as many CPs as you have; 5 is best since it's the most energy efficient.

In one of the CD usage threads someone had mentioned that even delaying 5-10 seconds is a dps loss. In a situation of 5 cp, snd is falling off and you have to choose between a 5 CP evis with a 1 CP SnD vs. a 5 CP SnD and then building back up to 5 points for an evis that it just feels like delaying the usage of more CDs.

So in a similar situation where SnD doesn't fall off or potentially falls off for a fraction of a second. 5 CP, 1 second on SnD. Do you evis and then take SnD with the 1 point of ruthlessness getting your CDs up faster? Or do you 5 point SnD and wait 10 more seconds delaying your CD because its more energy efficient?

[Squirl]

5 CP SnD. You're thinking on too short of a time scale. Doing a 1CP SnD means you're going to have to drop more energy into it for the same uptime, which means less energy for builders that will go into other finishers, which means you're getting less RB over an encounter.

The other thread you referenced was likely referring to holding CDs for various damage boosts (such as Red Insight, trinket procs, etc.), which is almost never worth it. Using SnD over Evisc does not delay your CDs.

[Sesshou]

So Warstar to be clear, you're suggesting never using anything but 1cp SnD because you want to put your cps into damage finishers? That is just going to waste a huge amount of energy and combo points keeping SnD up like that which is less energy and cps on evis and thus still delaying your cooldowns.

You would need 3 1cp SnDs to last what a single 5pt SnD lasts. 3 1pt SnDs is on average 60 energy, costs 2.4cp. 1 5pt SnD is 0 energy and costs 4 cp. 60 energy is 1 and 1/5 of a SS and an SS is 1.2cp so 60 energy is 1.44 cp. So 1.2 casts of SS versus .16cp. Discounting the cp from SS since it was included, that is damage from SS plus 1.2 chances for poison and MG and CP procs from the MG. If you aren't in deep insight then you also get BG advancement. I'd bet that all beats .16cp.

Edit: Though I would say it is situationally useful. If I got a 10 second trinket proc and my cooldowns were around 10 seconds left, I would definitely do a 5cp evis and 1cp snd to squeeze cooldowns into a gigantic agi proc.

[shadowboy]

Sesshou: Maintaining 1-point SnD will ultimately use FEWER combo points, not more. a 1-point SnD offers more time-per-cp than a 5-pt one.

SnD is only more energy efficient. SnD grants less uptime-per-cp at higher cp finishers (a 5-point SnD is only thrice as long as a 1-pt SnD). Looking at the energy cost of SnD is meaningless. You need to look at the energy cost of all the builders, too. (otherwise a 5-point SnD has infinite damage-per-energy but it's not, it cost you energy to generate those 5 cps).

A 1-point SnD grants you 12 seconds of SnD @ a cost of 59.1 energy, or 4.9 seconds of SnD uptime per energy spent.
A 5-point SnD grants you 36 seconds of SnD @ a cost of 195.8 energy, or 5.4 seconds of SnD uptime per energy spent.

A 5-point is only 10% more efficient than a 1-point with the T16 4-set. Without the set bonus it's 5.1/5.7.
Keep in mind, these factors are ignoring the effect of ruthlessness (just quick math),

[Oggiva]

Sesshou's math looks solid to me. You shouldn't factor in the cost of the sinister strikes used to generate the combo points as part of the Slice and Dice cost. Sinister Strike is something you are doing anyway, so there will be a constant stream of combo points. The question boils down to whether it's best to use SnD on one or five combo points. Using SnD on one combo point gives you more uptime per combo point spent, but the energy cost of doing that makes it not worth it in the general case.

[shadowboy]

Sesshou's math looks solid to me. You shouldn't factor in the cost of the sinister strikes used to generate the combo points as part of the Slice and Dice cost. Sinister Strike is something you are doing anyway, so there will be a constant stream of combo points. The question boils down to whether it's best to use SnD on one or five combo points. Using SnD on one combo point gives you more uptime per combo point spent, but the energy cost of doing that makes it not worth it in the general case.

You most certainly DO include the cost of the sinister strikes because those combo points generated for the SnD are combo points NOT going towards restless blades which is the whole point of the OP.

Besides, you can't look at finishers in a vacuum. You need to consider the entire finisher cycle. The logic is just as flawed as the whole CT vs Evis discussion on EJ for exactly the same reason.

[Oggiva]

If you were using Expose Armor to generate combo points I could agree. If you make the argument that the cost of the Sinister Strikes is needed then the damage and advancement of Bandits Guile surely have to be a part of the calculation as well?

[Kael]

It's getting to the point of too complex for napkin math, if you want to compare the cost of energy across finishers vs. CP generators and BG state advancement... which is why we like modeling and sims so much. In the end, the 5cp SnD cycle has proven time and again to be superior.

I do think that in isolated situations, using a 1cp SnD after a 5cp offensive in order to push CDs into an active trinket proc would be worthwhile, as you say, Sessh.

Edit: if, by some fluke, there were issues in the APL or coding the last time we went to check, it is, presumably, possible that there could be a different answer. If you want to change the APL to use 1 CP finishers only, it should give a quick answer.

[Warstar]

Mostly I just wondered if anyone had really figured out optimal SnD usage (even if it's the same as always) since ruthlessness had been re-added to the game. Trying to push CDs rather than 5 cp SnDs just feels like those CDs are coming up faster although it's probably a messed up perception and not actually timing them.

[shadowboy]

I think general convention is optimum usage is when there is less than 2 sec on your current SnD, regardless of number of combo points. Excessive SnD clipping is far worse than an undersized one. You lose no SnD effectiveness with a <2sec clip. If you clip at 3 seconds, you wasted 2 sec of the previous SnD. If you have 5 cp and 4 sec left on SnD and can pool for 2-3 sec without capping energy, then it's optimum to do that.

[Sesshou]

Sesshou: Maintaining 1-point SnD will ultimately use FEWER combo points, not more. a 1-point SnD offers more time-per-cp than a 5-pt one.

Um, did you somehow misread:

3 1pt SnDs is on average 60 energy, costs 2.4cp. 1 5pt SnD is 0 energy and costs 4 cp.

Because, yeah, I did say that maintaining SnD with 1cp SnDs takes less cp per duration.

And I'm with Oggiva, I don't think you should be considering how you got those combo points into this. The entire point is how many more cp could you put towards restless blades doing it one way over the other (which is .16 after considering the extra cps from the energy difference but you give up an awful lot for that .16 of a cp). How you got the cp doesn't matter.

And I don't think your math actually addresses the question. It was stated early on to be known that 5cp SnD was more energy efficient which is all I'm really getting from your math. The idea thrown out was that it might be better to ignore the energy efficiency in order to get more cp into RB.

@Kael
Yeah to get an actual estimated damage difference you would need that. The idea that was thrown here though was that you'd get more RB time off cooldowns which is why I posted that in the first place to show even if you ignore the damage from all the things you miss out on by doing it, you at the very best end up a tiny fraction of a combo point ahead of a 5cp snd.

[shadowboy]

Not considering the generators, as I said leads to faulty divide-by-zero math. It should be avoided. I refer you to the Evis vs CT discussion on EJ for exactly that reason. How you got the CP doesn't matter in terms of WHICH builders you use (whether you use EA, spam RvS, or use SS), since we're comparing the relative cost of undersized versus maximum cp SnD, but keep this in mind, if you consider the cost of SS, you CAN NOT consider the refund of RS simply because that energy is not used on SnD. It's used on the builders for the next finisher. (or alternatively, you could sum up all of your RS procs and average it out as part of your total energy regen, or distribute the RS energy among the combo builders--and yes, adding MG-CP to the builders is completely valid and optimal, too). You can't have your cake and eat it too. Either consider both the builders AND relentless strikes or NIETHER the builders NOR relentless strikes.

All of that is really beside the point. See my previous post on optimum usage on why we don't plan on either 1pt or 5pt SnD, or anything in between and just use whatever we have (trying to prefer 5pt). I'm not arguing that a 1pt SnD is better than a 5-pt. I'm arguing that trying for one or the other is less effective than just using whatever you have when <2sec is left on SnD, because as it turns out, a 5-pt SnD is not THAT much more effective than a 1-pt. Clipping SnD with 5 seconds left on the buff, however, is extremely ineffective, or letting it lapse for 3 seconds is extremely ineffective.

[Warstar]

So end all is stick with standard only trying to "push" CDs to refresh early potentially for trinket procs or burn phases or what not?

[Sesshou]

Except that your math completely avoids the point. The point was to see how many more cps you can use with RB and at what potential cost where as all you mathed out was the efficiency of snd by seconds per energy spent. The cost to generate cps themselves will be identical regardless of what you use them on. There is no division by zero... that is just simplifying the fraction. What is important is the opportunity cost of generating a cp and this is the same regardless of what you use the cp on. Since it is a common factor in both cases, you can remove it.

[Squirl]

So end all is stick with standard only trying to "push" CDs to refresh early potentially for trinket procs or burn phases or what not?

Yes, if by standard you mean maintain 100% SnD uptime and prefer 5CP while avoiding clipping.

Or sim it and find out for yourself.

[shadowboy]

Edit: Wasting words.

[Battlebeard]

So much discussion over a simple answer:

Keep Slice and Dice up 100% of the fight. No matter what. Period.

[InfinityRogue]

I think that insight is an incredibly important factor that you are not taking into account. I want to know the math behind what happens if you estimate the amount of combo point snd you are going to need to make sure that it falls off in no insight or green insight. By refreshing snd in a lower insight we give up the cost of having to spend combo points on non damaging abilities during yellow or red. It's actually very easy to do this in a fight if you understand the rotation and cooldowns you have available.