Cut to the Chase

(Much of this article is reiterated from other posts I've made on the subject, but there's some new material on cycles with low-CP Eviscerates.)

Cut to the Chase is the talent that seems to be intended to crown Assassination as a viable PvE DPS spec alongside combat swords in Wrath of the Lich King. If you don't have to refresh Slice and Dice, then you can devote more of your generous combo point supply towards finishers, increasing your DPS. In theory, it's a great idea.

The problem lies in the proposed implementation: Slice is refreshed to its 5 CP duration only on Eviscerate and Envenom crits. To understand why this isn't a great implementation, you must first understand combo point cycles. A cycle iteration begins right after performing Slice and Dice, and consists of those actions which you may perform within the duration of Slice and Dice, and terminates upon performing the next Slice and Dice.

Mutilate cycles can be generalized thus: Mutilate twice, perform secondary finisher, Mutilate twice, perform Slice and Dice, repeat. Sometimes there are variations, like Mutilating three times for the secondary finisher when procs don't go your way, then Mutilating only once so that you can still refresh Slice before it drops.

However, in general, assuming two Mutilates per finisher and two finishers per cycle is reasonably accurate. Two Mutilates guarantees at least 4 CP, and almost always yields 5. Four Mutilates in total cost 240 energy, and two finishers of at least 4 CP will cost at most 10 energy after Relentless Strikes (assuming your secondary finisher is Rupture, as it is for most Mutilate rogues currently). The total cost of the cycle therefore ranges from 240-250. A 4 CP Slice gives you 26.1 seconds of Slice, so you have just enough Slice uptime in this cycle.

Now let's look at Cut to the Chase. Recall that a cycle iteration starts immediately after performing Slice and Dice. Two Mutilates guaranteed us at least 4 CP, and a crit Eviscerate/Envenom would have given us a 5 CP Slice, so we know we have at least 26.1 seconds of Slice time. This is enough for the same four Mutilates and two finishers. Let's say that we Mutilate twice and perform Eviscerate as our secondary finisher. If our Eviscerate crits, it has the same effect as performing Slice and Dice, therefore our current cycle iteration ends and we return to the start of this paragraph. If our Eviscerate doesn't crit, we now have roughly half the Slice buff remaining -- enough time to Mutilate twice and perform another finisher. If we use the next finisher to Eviscerate again, it's possible that it crits and refreshes Slice. However, it's also possible that it doesn't crit and Slice drops. This is a catastrophic event and we must avoid it. Thus, if our first Eviscerate doesn't crit, then we have to use our next two Mutilates to perform Slice and start the next iteration.

So what's the actual gain? Well, every time we crit an Eviscerate, we get to perform another Eviscerate. In statistics, this is called a geometric series; given a chance to crit C, and the idea that if a crit happens, we get to Eviscerate again, we can calculate the average number of Eviscerates performed per cycle as 1/(1-C). Obviously, as C increases in this formula, the number of Eviscerates performed grows pretty rapidly. For C < 0.5, the growth is kind of slow, and at C = 0.5, you're performing two Eviscerates on average per cycle. For C > 0.5, you're rapidly gaining in Eviscerates per cycle. However, crit levels that high are pretty unrealistic even fully-buffed in raid situations.

Thus, with Cut to the Chase, a typical cycle goes from 4+s/4+e (read as Slice with at least 4 CP, followed by Eviscerate with at least 4 CP, repeat) to 4+s/4+e…, which is simply my notation for saying "repeat 4+e until Cut to the Chase doesn't proc." In 4+s/4+e, 50% of your cycles are spent performing Eviscerate. In 4+s/4+e…, with C = 50%, you are performing two Eviscerates per cycle, or 66.67% of your cycles. This is a gain of roughly 33.33% in Eviscerate DPS. If you manage to reach C = 60%, you are performing 2.5 Eviscerates per cycle, or 71.43% of your cycles, for a gain of 42.86% in Eviscerate DPS. And naturally, when C = 100%, you no longer need to refresh Slice, so 100% of your cycles are spent performing Eviscerate, a gain of 100% Eviscerate DPS over 4+s/4+e.

Now wait a second. Let's reread the talent. Your Eviscerate and Envenom critical strikes will refresh your Slice and Dice duration to its 5 combo point maximum. It says nothing about how many combo points you need to spend on Eviscerate. What if we tried using multiple low-CP Eviscerates to increase our opportunities to proc Cut to the Chase? Using a single Mutilate guarantees at least 2 CP and at most 4 CP, and a single Mutilate plus a 2-4 CP Eviscerate costs 75-85 energy. Two of these would cost 150-170 energy. Two Mutilates and a 4-5 CP Slice cost at most 125 energy and generate at least 26.1 seconds of Slice, just enough to manage two 2-4 CP Eviscerates (assuming Focused Attacks).

So we now start an iteration of this cycle, 4+s/2+e/2+e…. If the first Eviscerate crits, then our current cycle iteration ends and we return to the start of the paragraph. Otherwise, we perform our second Eviscerate, and if it crits, then our current cycle iteration ends and we return to the start of the paragraph. Otherwise, we have to Mutilate twice and perform Slice, again ending the current cycle iteration.

How many Eviscerates are performed? Well, let's think about it. At the start of an iteration, we perform one Eviscerate. We perform a second Eviscerate in the same iteration only if the first Eviscerate does not crit. Therefore each iteration consists of 1 + (1-C), or 2-C Eviscerates. Now, the average number of crits experienced per iteration equals C times the number of Eviscerates performed, or C*(2-C). Again, for the sake of explicitness, the number of Eviscerates per iteration is 2-C, and the chance per iteration to proc another iteration is C*(2-C). Thus, we can calculate the number of Eviscerates performed as (2-C)/(1-C*(2-C)).

Thus, with Cut to the Chase, we're going from 4+s/2+e/2+e -- where, of every 285 energy (averaged), you are spending about 160 towards Eviscerates, or 56.14% -- to 4+s/2+e…/2+e…. With C = 50%, you perform 6 Eviscerates per cycle, for a total of about 480 energy out of every 605, or 79.34%, an increase of about 41.32%. With C = 60%, you perform 8.75 Eviscerates per cycle, for a total of about 700 energy out of every 825, or 84.85%, an increase of about 51.14%.

So using two Eviscerates per cycle, you do gain slightly more benefit from Cut to the Chase than you do when using only a single Eviscerate per cycle. However, in either case it's only at best a 50% increase in your Eviscerate DPS, and only at unrealistic crit levels. And in all cases, you still have to use Slice and Dice at least periodically. As a result, you can't use any finishers but Slice and Eviscerate/Envenom.

I don't really think Blizzard intended this talent to further limit the selection of finishers by Assassination rogues, considering that Assassination is the finisher tree, and given the existence of Blood Spatter. However, as long as crit rates of 50% and above are unrealistic, this talent can't really be designed to depend on critting finishers.

In either case, any less than 100% guaranteed chance to keep Slice up without having to perform Slice and Dice after full investment in Cut to the Chase would be a disappointment. I can see what was intended with this talent, and I like the direction it's going. However, the way cycles are actually used in practice, finisher variety just won't increase unless taking 5/5 Cut to the Chase guarantees 100% ability to sustain Slice and Dice without actually having to perform Slice and Dice. Even 90% chance just means that eventually we'll get screwed.

As I proposed in my feedback thread on the beta forums, this would be a better way to implement Cut to the Chase to enable it to perform its intended purpose:

Cut to the Chase
Rank 0/5
Require 45 points in Assassination talents
When you perform Eviscerate or Envenom, you have a 20/40/60/80/100% chance to refresh your Slice and Dice effect as if you had performed Slice and Dice with the same number of combo points.

Now I'm not necessarily certain as to the DPS increase provided by this proposed version of the talent, since it's tough to say without access to level 80 gear. However, I can say for sure that this version of the talent would enable the full adoption of an alternating Rupture/Eviscerate or Rupture/Envenom cycle rather than having to use Slice/Eviscerate…. You wouldn't be able to use low-CP Eviscerates in the same way with this version of Cut to the Chase, but you also wouldn't need them since there would be 100% chance to proc the talent on your Eviscerates.

If Cut to the Chase stays the way it is, no matter how you game it it's just not a worthwhile expenditure of 5 points, at least not when compared against the possibility of a 45/21/5 build. However, it's a neat idea and I'd really like to see Blizzard carry it through to viability.