The Jedi way is to serve...

The math is pretty simple. Splitting the points gives more damage per Vital Shot.

Today I landed a Blade Storm crit for 16,638 on Dr. Lorrick in Lost Island. No screenshot, unfortunately, but it's in the log. Merciless Slash could probably hit even harder...

There are a few things you are overlooking. 1) Ataru procs automatically on every Blade Rush, and this is independent of the ICD on the random proc (sometimes it will proc twice on a single Blade Rush, once automatically, and once randomly). 2) The Blade Rush buff, which I endeavor keep up at all times, increases the random proc chance to 50% per attack. The parse was 5:27.

I agree with most of your advice, but there is nothing magical about 265 Surge - diminishing returns are gradual and continuous - and Accuracy Rating affects only one attack (Flurry of Bolts) or two (if you use Quick Shot, which is debatable) which make up a very small fraction of your damage. For my Scrapper, with 390 Surge and 0 Accuracy Rating, adding another point of Surge still increases my total damage output by almost twice as much adding a point of Accuracy Rating would. Of course, Surge Rating isn't a very good stat due to the DR, but Accuracy Rating is even worse; improving only 1 or 2 low-damage attacks that together make up 10% or less of your total damage is itself a very severe form of diminishing return.

Not trying to be snarky, but a 150 DPS swing from a single relic seems implausible. How many tests did you do? What is the standard deviation of the two DPS values you measured? The DPS of the proc relics is pretty easy to measure/estimate. Also, the proc relics benefit from your critical chance and critical multiplier, but not from weapon or bonus damage. It appears that some are force and some are tech, although the item itself doesn't indicate this. Based on the proc chance, ICD and my average rate of attacks, the Rakata (168 tooltip) internal damage relic adds about 30 DPS for me. The corresponding Campaign relic would add about 33 DPS. With my current gear (almost full BH, with one Campaign hilt), this is about +2.2 or 2.5%, respectively. The static +Power PvP relics are also pretty easy to analyze. The Recruit (+93) adds about +1.9%. The BM (+103) adds about +2.1%, the WH (+113) adds about 2.3%, and the +123 (not sure if in-game - I don't PvP) adds about 2.5%. These numbers are based on the exact damage formulas for each ability, and the relative damage done by each ability in my parses on the training dummy. Using the same method, the stats of the Matrix Cube add about 2.0%. The clicky is a bit more complicated, since you can control it. Worst case is to use it on cooldown, which means +310 Power (+6.3% with my gear) 1/6 of the time, or +1.1% on average for a very long fight. It is hard to see how selective use could more than double its relative benefit, although depending on the length of the fight, the uptime could be more than 1/6 (e.g. in a fight that lasts 2m20s, it is up 28.6% of the time, for a 1.8% overall damage increase; in a 5m fight, it is up 20% of the time, for a 1.3% overall damage increase, etc). Situationally, a large increase in damage for a short period of time may be more valuable than a constant increase, even if the constant increase is larger over the whole fight (which is why clickies have been banished from PvP...) In summary, the difference between the various relics for overall average DPS appears practically negligible, and an 11.5% increase in DPS by changing from one to another seems hard to credit.

When I ran the numbers, it was about 2%. Let me check again. Using a training dummy parse where I did 536,892 total damage, Ataru Form made up 116,326. Of the Atarus, 72/212 (34%) were crits. My crit bonus is 79.2%. So the Ataru damage without crits would have been about 116326/(1+0.34*0.792) = 91669. If the bug were fixed, Ataru would have a critical bonus of 109.2%, and Ataru would have done 91669*(1+0.34*1.092) = 125666 damage. The difference is (125666-116326) = 9340 damage out of my total of 536892, so percentage-wise: 1.74%. Yes, the curve I've posted was assuming the bug was fixed, and not fixing it lowers the value of Critical Rating slightly. But the optimal value is still right around 150 (~30% Melee Critical Chance with buffs) for me. Not any big change.

Yep, and even with 390 Surge Rating, Surge is still better than Accuracy for my Scrapper, since Accuracy only improves one ability that I use on bosses (the weakest - Flurry of Bolts). I am getting about 1350 sustained on the training dummy and bosses that behave one, with respectable (Rakata/BH) but far from optimal gear. I use Flyby a bit, but not on cooldown, which is probably something I should work on.

I don't think you've understood the graph, and I never said or implied any such thing. What the graph actually shows is that moving 250 points between Power and Critical Rating in either direction results in a 1% difference in total healing. I don't think most people would consider that "a drastic loss". Nevertheless, relatively high levels (>300) of critical rating are optimal, as the graph shows. What you said was that you use Diagnostic Scan to determine your stat priorities, and that 3% critical chance makes no difference to your healing output: There is nothing "simulated" about the graph. It is based on the exact amount that each ability heals, according to player stats and the actual data used by the game itself. Combat logs from real flashpoint and operations runs (rather than non-quantitative anecdotal claims) are to weight each ability's importance, as I explained earlier in the thread. Thanks for the compliment, and nothing personal, this is just math. Your original post suggested that a careful calculation was wrong based on a superficial and demonstrably incorrect argument, so I explained how it was incorrect.

I suppose it depends what you mean by "significantly", but I'm not convinced of this. I am only about 50-75 DPS behind our raid's Watchman Sentinel on the dummy for 5+ minute parses; he has two BH Hilts while I only have one, and he has a couple marginally better mods (e.g. Might Armoring 26 in his Belt/Bracers, compared to my 25's in those slots) and probably a better relic. About 1700-1750 for him vs. 1625-1675 for me with stims and no Inspiration. Neither of us may be the best players in the game, but we are both pretty experienced at our specs. I figured the bug fix and truly equal gear would probably make up most of this difference. That bug-which-is-no-longer-a-bug is costing me about 2% of my potential damage, which is about half the difference right there.

It looks to me like yet another example of a spec getting screwed in PvE for the sake of PvP. Even on the training dummy, I think Combat is *at best* on par with Watchman, or more likely a little bit (couple percent) worse. If so, for any real boss fight where you can't just stand there and wail on the target, Watchman must be superior thanks to doing DoT damage while Combat is doing none. For PvP, burst damage is what generates whining, so we get to stay gimped in PvE as a result. So infuriating. When they nerf a spec, either directly, or by deciding not to fix a clear bug we have suffered with for MONTHS, why don't they show us these all-knowing, all-seeing metrics? Probably because they are all based on PvP.

Math is not a matter of opinion. You came into the thread saying, "I don't believe in math, I just use Diagnostic Scan to make all my stat decisions. Oh, and 3% Critical Chance doesn't make any difference." I tried to explain why that doesn't lead to the correct answer, and then you had a tantrum. Of course it's not random, there is a fully known formula for all abilities. And the formula dictates that critical rating is superior to power up to around 39-40% critical chance (buffed). If you had read the thread, you would see that I even posted a graph showing that. Here's an updated version of the graph, showing that currently, if I changed any of my 319 Critical Rating into Power, or any of 543 Power into Critical Rating, my overall healing would decrease: http://sdrv.ms/RAC0Rv

The BH/Campaign itemization for Sentinels is so bad, it defies description. You are going to need to shop for armoring/mods/enhancements on a slot by slot basis, usually getting one or two mods that are useless in order to get one that isn't. For the armorings, some of the Guardian DPS pieces are better than Sentinel. For the rest, you pretty much need to buy about half a dozen pairs of boots and rip the mod/enhancements out. Extremely expensive, slow, and annoying. If you have a lot of credits, you may also find someone to craft Deft Mod 26 and Adept Enhancement 26 with mats you buy, or if you are rolling in credits, you can just buy them off the GTN. Otherwise, you can look forward to getting boots, and more boots. You certainly want the Rakata 2-piece bonus. The 4-piece bonus helps a little, but is considered optional by most Sentinels I know. For the weapons, your choices are to get off-hands from Firebrand/Stormcaller Hard Mode and transfer the hilt to your main-hand, or to get a main-hand drop from Kephess Hard Mode. The main-hand is actually the less desirable drop, since all the hilts have high endurance and low strength...

A little preview of the support we can expect with Free-to-Play, I suppose...

No, your math is wrong. Kolto Pack's healing is not proportional to your bonus healing; it has a large base value which doesn't scale at all. That is true of every other healing ability too, except for Diagnostic Scan. So a 5% increase in bonus healing does not result in a 5% larger heal for any ability, except Diagnostic Scan. All the relevant formulas are known, even if you apparently don't know them yourself.

Wrong. Saber Storm was still buffing the *force* critical damage bonus of Ataru Form, but since Ataru Form was changed to a melee ability in 1.2, this buff did not actually increase damage. Customer Service eventually acknowledged that it was a bug.

Not ever, apparently: What a slap in the face.

Accuracy is all but useless as a Scrapper. It increases the expected damage of exactly one ability (Flurry of Bolts, the lowest damage ability). I am at 78% Critical Multiplier and 91% Accuracy (legacy bonus only), and even with severe DR, Surge is still almost twice as effective as Accuracy in terms of increasing my expected damage.

So your hope is to disprove mathematics then? It also has no critical bonus healing from talents (unlike the abilities that do most of our healing), and a very high critical chance bonus from talents. Are you claiming that a 3% increase in critical chance does not, in fact, increase your critical chance by 3%? It would take thousands of casts of establish that. How many thousands of casts did you tabulate at each value of critical rating? All you are really proving is that Power increases the tooltip and Critical Rating doesn't. If you ignore the increased rate of criticals, obviously Power will look better. You do realize that Diagnostic Scan is unique, in that it is ONLY based on your bonus healing, and that every other heal has a very large base value which is independent of your bonus healing, right? Obviously, if 3% of a critical heal's extra healing is more than about 21, it is certainly "worth it". My Kolto Pack lands for around 2.8K without a crit, and crits for about twice that. 3% of that difference is 84, and 84 >> 21.

I'll take a bit of a different tack, since the OP's problem sounds like survivability. By Level 30 you have a number of useful defensive abilities, in addition to Rebuke (which you should should just before any pull you think might be problematic): 1) Disable droid: take the biggest droid you can out of the fight whenever possible 2) Pacify: this is an extremely powerful debuff against landscape mobs (strongs and elites) 3) Saber Ward: another ability that allows you to take almost no damage for a moderate period of time, and the cooldown is not that long 4) Transcendence: yet another damage reduction ability; in most cases, Zen will be preferable, but this does have the ability to increase your survivability slightly. 5) Force Camo: yet another significant damage reduction with a relatively short cooldown In addition, Force Sweep and Blade Storm can stun mobs. Later (Level 40) you get Awe, which does the same. Useful for reducing damage at the start of a fight, when fighting a pack. You also have Pommel Strike by Level 30, which will often 1-shot a stunned normal mob if it crits, immediately reducing your incoming damage. When you get Opportune Strike (Level 38), that will allow you to one-shot, or almost kill, another normal mob immediately after Force Leap. Once you have that, Force Leap -> Opportune Strike -> Force Sweep or Blade Storm -> Pommel Strike should take 2 normals out of every pack almost immediately. Basically, use your damage reduction early, when the largest number of mobs are still alive. Always kill mobs from weakest to strongest, using stuns where possible, and save Pacify for the final, toughest mob if necessary. I used Kira exclusively from Coruscant onward, and didn't really have many survival issues. Kira will generally solo one mob at a time, without dying unless she decides to go after an elite (and in that case she'll take some hit points off it before dying while you deal with everything else).

While taking cover doesn't use a GCD, there is still some delay before you can use a cover ability like Sabotage Charge. I suppose if a particular fight allows you to stay in melee range, and in position, and in cover, it might be worthwhile. The Stormcaller/Firebrand fight in Denova is about the only one I can think of that might fight that description.

This combination is very close (few points) to mathematically optimal for Campaign/BH gear: 303 Defense Rating (+43 stim) = 26.4% Defense Chance 545 Shield Rating (+20% Kinetic Ward) = 64.9% Shield Chance 563 Absorb Rating = 59.9% Shield Absorption Stat weights for this build (equal means optimal): +1 Defense Rating = 2.22 Armor +1 Shield Rating = 2.10 Armor +1 Absorb Rating = 2.27 Armor Changing any mod, enhancement or augment would increase my average damage taken (squishiness). I may have to switch an augment when I get the Campaign shield, but it will be a very small net change to defensive stats.

I don't think K.O. is very useful for flashpoint/operation content. Shifty-Eyed gives you a 2% chance to avoid ALL damage (melee/ranged/force/tech) and I would take that over K.O. for any kind of challenging content. It is easy to show that (1/2) Mortal Wound + (1/1) Open Wound is superior to (2/2) Mortal Wound. Your suggested opening of Shoot First -> Back Blast seems inefficient. Better to wait for the DoT from Flechette Round to tick off, then reapply again with Back Blast. I find it easier to think of our "rotation" in reference to Blaster Whip. After every Blaster Whip, you have 3 GCDs to do other stuff before Blaster Whip is available again. You will always want to Sucker Punch during one of these 3 GCDs if you have 2 stacks of Upper Hand, and also if you have 1 stack and Pugnacity is already up. It shouldn't be the first unless you have 2 stacks of TUH, because if you don't proc Round 2, you will be without the Upper Hand buff for anything you do afterward. And you shouldn't wait until the last, because if you *do* proc Round 2, you normally would want to Sucker Punch again, before you Blaster Whip. Back Blast (+Flechette) is your big attack, and you will want to use it after every second Blaster Whip (12s c/d). Vital Shot is a decent damage per energy, and with the suggested traiting, you need to reapply it every third Blaster Whip (18s c/d). Finally, there is Pugnacity, which should be maintained at all times. Thinking of the three GCDs after Blaster Whip as "slots" to fill, this seems like a reasonable slot-by-slot priority system: Slot 0: Blaster Whip (every 6 seconds) Slot 1 Priorities: Disappearing Act + Flechette + Shoot First (if possible) Flechette + Back Blast (unless DoT from Shoot First is still active) Sucker Punch (if 2 stacks of TUH) Flurry of Bolts (or perhaps Quick Shot if you are nearly full on Energy) Slot 2 Priorities: Pugnacity (if not active) Sucker Punch (if 2 stacks of TUH) Fly-by (fills Slots 2 and 3) Sucker Punch Slot 3 Priorities: Vital Shot (if not active) Sucker Punch Flurry of Bolts The opener is then just Flechette + Shoot First, before starting into this steady-state 6-second priority list (i.e. with Blaster Whip)

Did you try having a tank taunt their target after, or even while, the tank kills his own?

This. In my experience, levelling 4 toons to 50 so far, tank companions become useless around Level 40 or so. They can't hold aggro except during the six seconds after their taunt, and unless you do nothing but heal them, they can't survive holding aggro even for that long against anything where you'd like to have a tank.