KeyboardNinja
-
Posts
5,156 -
Joined
Content Type
Forums
Blogs
Events
Articles
Downloads
Gallery
Posts posted by KeyboardNinja
-
-
Oh WTH, I'll reply…
I haven't entirely disappeared. Just… mostly. Playing Guild Wars 2 and enjoying it a lot. Raided casually in 4.0 and (re)cleared everything but Brontes. Being even more casual now, waiting for the devs to confirm that they're not making real content anymore.
To the broader question on tank theory crafting in general, it's pretty boring since 4.0 due to the fact that defense is forced and the only "choice" is between shield and absorb. That choice has very little impact, numerically, and is broadly dictated by class mechanics anyway. Here's your rule of thumb:
- Make the numbers bigger (this always trumps any sort of min-maxing)
- Drop your defense as low as you can (in favor of shield/absorb)
- Start from balanced shield/absorb levels (equal amounts of both)
- If shadow/assassin, trade shield for absorb by about 10-20%
- If vanguard/powertech, trade absorb for shield by about 10-20%
- If guardian/juggernaut, reroll and go to step 1 (optional min-maxing step: if vanguard/pt, do the same)
The most meaningful question here is really how much endurance do we want. Health pools seem pretty darn generous in 5.0, and I haven't really seen anything that was so spiky that TTK was relevant. With that said, healers are so over-tuned that it literally doesn't matter. HPS basically falls from the heavens now, which means that optimizing to reduce HPS (which is what mitigation stacking does) is kinda pointless. Deeper health pools increase healer margins and reaction times at the cost of higher ongoing output requirements, but since HPS is easymode, the output requirements are a meaningless cost.
Additionally, this game basically doesn't have tanking in a form which can meaningfully impact the success of a group. The most successful groups have always been those with godlike healers, very very good DPS, and terrible tanks. Watch some old world first clears with a critical eye if you don't believe me. Very rarely you'll see apex groups that have legitimately good tanking (Zorz did in a few groups), but it's more of a coincidence than anything else. The healers carry most of the burden, followed by the DPS. The inverse is also true: if you have godlike tanks and mediocre healers, your group will never progress successfully. So basically, tank execution has a relatively minimal impact on group success, and tank gearing has even less.
Stat the way you feel like statting. Bant's numbers are optimal. The difference between optimal and "off the wall crazy" is almost non-existent from a numerical standpoint, and literally non-existent from a group success standpoint.
Now imagine how richer community would be should KBN use something people actually use, and not Mathematica
Actually if you had been around back in the day, you would know that I used to use a more conventional spreadsheet. However, there isn't a spreadsheet tool in existence (even excel) which can do the kind of multivariate optimization necessary to pin down shadow/assassin statting (pre-4.0), due to the interactions with Kinetic Bulwark. There are very few spreadsheets which can do simpler things like Guardian/Juggernaut (which still required at a minimum trivariate optimization). Mathematica was the best tool for the job, and you'll notice how basically all of the serious theory crafting in SWTOR's history has been done with a tool like it (Bant used Python, iirc; dipstik used matlab).
Anyway, every single tanking methodic used by different theorycrafters in past was flawed in it's own way (some -- in many ways at the same time), KBN's was probably best at what it was. Issue with it was that it solved problem nobody ever should care about. If your goal is minimizing DTPS period you are crippling yourself, plain and simple.Now that healers are overpowered in terms of HPS output, I agree with you. DtPS is meaningless in the 4.0 and 5.0 metas. Which is part of why it's boring. But previously, DtPS optimization was extremely important, simply because healer throughput margins were much lower.
I definitely don't agree with you that stacking shield is optimal in any objective sense of the word, but it's not going to make a huge difference either way. If your healers like it better, go for it! Most of what they "like" about it will be psychosomatic, but comfort is real and shouldn't be ignored.
- Make the numbers bigger (this always trumps any sort of min-maxing)
-
Honestly, I don't know why are you commenting so much in regards to our guild KBN while it's clear you have no idea what you are speaking about, your role in this community is clear and you were appreciated for it, progression or discussion on the topic however isn't.
Partially because I've always been generally annoyed at NGE (and FF before the rename) on a large number of issues, and now that the endgame is dead I don't care enough not to say things. Not annoyed at you personally in any way, but NGE. The other part is probably because NGE is one of the few guilds still here. Who else are we going to talk about?
It's also worth noting that, whether you value my contribution to the conversation or not, I certainly have every right to discuss progression. Count up the number of players remaining in the game who cleared everything at level. One would imagine that group would be the most qualified to speak to whatever progression-related topics even matter anymore, and I am among that group.
As I said, there really shouldn't be any drama here. There's nothing to get heated about because there's nothing left at all. We can squabble about this-or-that from days gone by, but I suspect that's all of us just missing the days when we had something to get riled up about. Rydarus was out of line posting the video, you were out of line poking fun at his server, and I'm out of line expressing my general displeasure with NGE. None of it matters though.
In any case, the video Rydarus linked was of Kanre's perspective. The one you linked is the second video that I had been thinking of, which was the only one I had ever seen prior to Rydarus's link. I think the sage perspective is better anyway, since you can see more of what's going on.
The rash of youtube takedowns on SWTOR videos is absurd. Basically the only ones that survived were videos where the game music was either muted or replaced with some other soundtrack.
-
It was there before, it's been around for a year or two. Video is from 2.0.
The video in question, for those who didn't hit the link before it was pulled, is Kanre's viewpoint on one of NGE's first kills of NiM Bestia 16 man (pre-nerf). I don't remember if it was their absolute first kill, but it was an early one. The video is notable for demonstrating what was, to my knowledge, a completely unique strategy for killing the boss. Namely, Kanre (on his shadow tank) spent the entire fight kiting around a sizable percentage of the monsters which spawned (if I recall, he ended the fight with five of them), aided by chained Transcendence from their several combat sentinels.
Interestingly, not only is Kanre's video gone, but another perspective (which I had seen a year or two ago) is also gone from the search results. I simply cannot find any videos of pre-3.0 NGE Bestia kills, despite the fact that I know I watched them previously.
In any case, there's no point in getting worked up. This game doesn't have progression anymore. There's literally nothing legitimate on the PvE scene to generate drama.
-
85% sure you wore empty shells, not no gear at all
I had a main hand and a left side. No empty shells because the shells weren't empty and were on my sorc.
Also, every time NGE acts elitist I just remember this bit of magic.https://www.youtube.com/watch?v=v0TJSZI2CTc
#420blazeitforumallthedramas
#4.0regressionsrsbsns
Nah, there's no drama here. I'm just teasing. Even without bolster, the current SMs would be doable naked. They're basically exactly as hard as leveling.
-
an unequipped slot gives no bolster, but good try.
Maybe they changed something, but the last time I did a naked run on my sage (because of accidentally leaving my legacy gear on my sorc) I simply didn't notice because my stats were higher than my then-Revanite gear would have provided.
-
bro SM doesnt even have mechanics anymore, my group went through all of them naked with a single mainhand equipped ignoring all the main mechanics, its just a faceroll. anyway above i meant more for HM. For nim its a must it goes without saying.
You actually have better stats when naked due to bolster. Come back to me when you've done it with low-level greens (or whatever the trick is these days to "reverse exploit" bolster). ;-)
Regarding mechanics, I really don't mind teaching people. It's always more helpful when they have an idea of what the mechanics are and how they work, but I want them to have an open mind regarding strat since I often want to do things differently than the standard guides. Teaching people HM Revan is a decent example of this, since the only strat which seems to be uniform between all groups is the mapping of aberrations to clock positions. So, I don't want or need people to have memorized a second floor path, since that path is quite likely different from the one that we will be using, but I would certainly appreciate it if they're aware of push/pull timings, what to do with blades, and why it is that we call for a DPS stop.
-
Lethality is fine in raids. It has some advantages (namely its mid-range nature and strong mobility), but it doesn't bring much to the table other than sustained single-target damage. You have relatively long setup times and need to get a mostly complete rotation in order to do your damage. This makes the class inapplicable to some hard checks, like the orbs in Calphayus. Additionally, the AoE is… lacking. Lethality has more AoE damage than most people believe (most people can't use the DoT spread because of how fluky Carbine Burst can be), but it's still a lot less than most other classes.
These pronounced weaknesses would be fine if there were an off spec in the same class that you could swap to and rectify things, but there… isn't. Concealment solves the setup time problem, but the burst is relatively underwhelming (compared to what it has been in the past) and the sustained damage is abysmal. Furthermore, Concealment is one of the most melee-constrained specs in the game, with nearly everything in the core rotation being confined to 4 meters. To top it all off, Concealment's AoE is somehow even worse than Lethality's.
Other classes are more versatile and generally capable of doing Lethality's job better than Lethality can. With that said, you can make it work. On the positive side, operatives are one of two classes that bring the tech damage debuff (which is absurdly helpful for most of the meta classes right now), and Lethality does the most single-target damage of all four disciplines therein.
-
IIRC my WPM is 98 in typing tests. KBN's like 180. Chew on that.
Not quite. :-) It's 150-180 depending on what I'm typing and only if I'm typing on a real keyboard (Cherry MX switches ftw) and also only if I'm actually trying to type fast. On this precise keyboard (my macbook), I cap out around 130, and I usually sit closer to a comfortable 115 just because it's less effort.
Incidentally, Ry, your typing test may be selling you short. Most tests are designed for people who sit in the 50-90 WPM range (e.g. by showing a sentence at a time). A lot of people actually type quite a bit faster than that, but the test itself doesn't allow for higher speeds due to the way it interacts with the mechanics of typing so fast. A better test is something that literally just shows you a paragraph or two of text and asks you to copy it. The reason being that extremely fast typists have a tendency to plan their motions more than a sentence in advance (I generally think in terms of phrase "blocks"). Seeing the whole paragraph allows you to do that, while a more conventional test does not.
So google around for a real test and try again. You might find you're faster than you think you are.
-
Also, I think a part of optimal composition is basically making do with the resources that are available. The pool of competent raiders is finite, ESPECIALLY in 4.0 SWTOR. Often times, in my case it's simply not possible to get another class simply because I CAN'T PLAY another DPS class and do similar or better damage that I do on my guardian, and we really don't have that many people in the game anymore.
On a side note, I think that guardians are probably the best class to facetank brontes with a taunt, it's not quite hydraulics but unremitting can be used to block the fire and forget multiple times, saber ward with blade turning mitigates an entire set of shots, and focused defense heals you to full with one fire and forget. Your DPS is gonna be shmeh, but it means that you're basically freeing up a ton of uptime for your fellow dps.
I think in an ideal world, we can have ideal full fotm comps, but ideal doesn't really easily exist in a lot of cases. Better to just have a comp with players that wear it like a glove than a perfect comp that people are still trying to figure out.
You're just trying to head off the day when I ask you to swap to your sentinel… ;-)
-
I abstain and use my ballot to write in the margins about how important it is that they release more than 5 bosses per year. I might also spend a fair bit of time grandstanding about how shoddy the current endgame is and how the game as a whole has been mismanaged into the ground. There might be some abbreviated "story is not content," too.
And I would also be reminded that the developers are human and actually very nice people who want to do the right thing for the game. Over the years, I've met a reasonable percentage of the devs, and never came away from the encounter with an actively poor opinion of anyone. Doesn't mean I agree with their decisions. If anything, it makes me more disappointed to know that completely reasonable human beings made these decisions.
-
The need for aoe dps is overrated to be honest. I suggested an optimal comp in this thread (You can look back through) and so did KBN. He elaborated a bit more on his reasoning than I did. KBN also said something extremely relevant to the comps than most people are suggesting; "For the most part, this sort of optimization just makes easy fights easier (e.g. the Jugg solo deleting a trenchcutter pack)."
There are no aoe dps requirements that the worst aoe dps can't meet. And no, you don't need to be extremely good, just average to meet them. My suggested comp focused on making the select few highly difficult encounters easier and I personally think that's what the aim should be. Making moderate or easy encounters easier isn't really a priority, especially in a perfect player scenario.
I think basically all of the encounters are pretty easy, with the possible exception of Brontes. With that said, you're right that your comp is better tuned for the fights that are on the harder end of the scale than mine is. Mine should do just fine on those fights, but a tweak or two would improve things (e.g. not running a lower DPS class, Jugg, primarily because of AoE and cheese).
Regarding AoE checks… There are four AoE DPS checks in the game at present, and only one of which approaches any modicum of seriousness. Draxus is the most severe. Or rather, it would be if people didn't five DPS the fight, which is probably a valid reason why the game doesn't have massive AoE checks. The other three in order of severity are: Warlord Kephess, Bestia and Dash'roode. Basically all DPS have enough AoE to beat Kephess, though having more AoE simply makes the healing proportionally easier. Bestia is an almost identical situation (though if you really, really lack AoE, you might find yourself leaving the final add up). Dash'roode is even more trivial since the sandstorm isn't a serious concern anymore. Sustained AoE in rotation makes the DPS more straightforward since most of your group can just tunnel.
Anyway, if there were actually a need for a second tank on Draxus, I think you'd see people being more aggressive about AoE comps (as in fact, people were more aggressive about AoE comps when NiM DF was current). Beyond that, AoE is mostly a matter of streamlining already farmable content.
My philosophy on Brontes and Styrak is basically this: either you're going to beat them, or you aren't. Composition isn't going to make as much of a difference as you would think. This isn't a pre-buff HM Revan situation (or sniper meta NiM Styrak), where the right composition could delete a major mechanic from the fight. Aside from some specs which are just not viable, you can mostly bring whatever. Class stacking isn't going to overcome a bad mid-burn strategy on Brontes or low DPS on Styrak. Rerolling your tank from jugg to powertech isn't going to make them any cleaner on swaps, any sharper at agro or any more precise on cooldowns.
-
Double Jugg Tanks
Double Operative Healers (double merc comes in a close second)
Four Operative DPS
No AoE. Mediocre (or at best, average) single-target DPS. Poor survivability. Very low utility. Limited raid buffs/debuffs. Generally sub-optimal tanks. Healers completely incapable of handling burst.
-
What action queue bug ?
Following any cast equal to or greater than the GCD in duration, there is a constant (it appears to be around 100-110ms) gap before the game will allow another action to take place, even when subsequent actions have been queued. This has been brought down somewhat recently (it seems), but it's still present. Instants are also affected, but not to the same degree.
What this means in practice is that you cannot have an ideal APM on any class, and classes with higher cast-reliance are proportionally worse. What makes this even more severe is the fact that alacrity doesn't affect the gap between actions, meaning that running higher alacrity actually makes the problem linearly worse (a higher percentage of your time is wasted). Mercs are the most cast-heavy healers by far, and they are also more reliant on alacrity due to the supercharge mechanic. In practice, the issue is very, very noticeable when healing a group that is progressing on high damage content.
-
Sin/PT tanks (sin/jugg is also acceptable since HO isn't as important as it once was)
Sorc/Op heals (merc is alright, but the action queue bug hits them hard and sorc/op is simply stronger now)
Mara/Merc/Sniper/Jugg DPS
The Jugg DPS is probably going to be contentious, but here's the reasoning. First, Vengeance has the second highest sustained AoE DPS in the game right now (barely behind Madness) while maintaining completely acceptable single-target DPS and decent burst. Marauders, Mercs and Snipers are all lacking somewhat in sustainable AoE, so this fills a niche which the trio is weaker in without substantially hindering what the group is strong in. Second, Juggs guarantee the armor debuff, allowing greater spec flexibility from the other three (all of whom can provide the armor debuff, but it isn't necessarily always optimal). Finally, the Saber Reflect cheese is REAL on some bosses with tight DPS checks (FB&S and EC Kephess come to mind). This is an advantage strong enough to merit special consideration.
There are two major problems with this setup. First, raid debuffs are split and suboptimally allocated. The Merc/Sniper pair are lacking a tech debuff, and their ranged debuff is wasted on the Marauder/Jugg, who are missing a force debuff. This isn't the worst thing in the world, but it does reduce the damage dealt across the board. One could argue that a PT or Operative would be superior in the Jugg slot, but I truly believe that they both simply bring less to the table (the PT has lower DPS than the Jugg, while both the Op and the PT have less AoE and less cheese). The second problem is that it's an even split of melee vs ranged, which isn't treated kindly by most of the current content in the game. 3 ranged, 1 melee is generally a superior setup given the choice.
To that end, if you are worried about the ranged vs melee issue and (to a lesser extent) the debuff issue, you could swap the Jugg with a Sorc DPS, or you could bite the bullet and take a PT or Op (with the latter coming at the cost of very weak group-wide sustained AoE).
You can flip things around quite a bit though without hurting yourself much. Comparatively few fights are tuned tightly enough that you need to deeply care about composition. For the most part, this sort of optimization just makes easy fights easier (e.g. the Jugg solo deleting a trenchcutter pack).
-
KeyboardNinja. I did the test about two months ago after reading something about it in the Sorc/Sage forum. DoTs applied under alacrity affects are always affected by these alacrity affects as long as they don't fall off.
The game resets the exact same debuff when you reapply it if it's still on. Test it on a Sorc. it works as I said sooner. I don,t play these class right now, And my theorycrafting for AP isn,t ready but I'm retty sure an alacrty adrenal would net a DPS gain due to 1 extra RB tick for everytime you have 6s between your RS which is quite often. I just need to math out how much time you need for it to win over the damage lost from power adrenal during your high CD's.
It is absolutely a DPS gain for Lightning since it is reseted way less often than in AP. For Madness, it would be a DPS gain if you can ALWAYS resets them at the 13.5s mark (they would last 15s after Polarity Shift being factored.. maybe closer to 13.5s if you add an alacrity adrenal).
For Pyro and IO, it's a clear DPS loss. For Virulance, it doesn't matter if you use the 3-cull rotation. For Lethality, I think it is a DPS loss, but I can't be sure. For Hatred, I don't know.
This is basically insane. :-S It means that alacrity is not working quite the way that it is supposed to be, and the standard sorc dps rotations (for both specs) need to be adjusted accordingly. It's not much of a difference for Lightning (since you're just moving Polarity Shift up one GCD in your opener), but it absolutely affects Madness to a significant degree. It also makes downtime quite punishing for both specs.
-
Wait, Soa has a melee attack?
Yep! I noticed it in my logs years ago, back when Soa was actually relevant (sort of). If memory serves, it is actually called "Melee". It only happens on the third floor, and only when he's on his knees, and only when you're in melee range (which isn't common since you're mostly kiting him to the next pylon at that exact point).
-
As a general rule, every spec that does not have an armor debuff benefits hugely from having one... all the other ones are more minor since most of those classes provide the other ones they need...
Basically when making a group, make sure to have an armor debuff and the rest will take care of itself
This isn't strictly true. Lethality operatives do not provide an armor debuff, but their damage increase from having it is comparatively quite minimal. Pyro powertechs are in a similar boat. There is actually a methodical way of balancing availability of buffs with damage ratios to compute an optimal damage boost (the most a group can get turns out to be right around 14% overall).
In general, it doesn't matter all that much though. Other class factors matter a lot more than debuffs. And, as you imply, the armor debuff is the most important one for most fights.
-
I may not understand well, but i think you're wrong.
The dot mechanism seems to be an 'independant' debuff on the target. So the alacrity changes on the caster do not affect the dots when they are on the target (increase or decraese).
When applying a dot (on a target that has not the debuff on it) under alacrity bonus (polarity shift, adrenal, commando class buff), the debuff duration on the target is reduced, and the tickrate increase in proportion.
When refreshing this dot (I mean applying the dot on a target that already has it active), the debuff timer restarts from 0. So as long the debuff does not fade on the target, the "alacrity effect" from the buffed one remains.
If you pop polarity shift before refreshing a "normal" dot on a target, the effect is... nul. The duration of the "active dot" will be set to 0, the alacrity will not have any effect on the duration and the tickrate of the dot.
What you point with rotation issues is true, tho. Any dot cannot be used this way : to be effective, the dot HAS to be re-applied before it fades. Depending the spec used, that should be very tricky to find proper rotations that reaapply the dot just before it fades (minimum loss) without delaying anything important. Some other spec have automatic-dot-re-applied-side-effect.
And concerning the dot spreading, I never did any test. It could function... or not. It depends on how the spread was coded.
If this is the case, then IO, Pyro and Virulence would all suffer from very noticeable rotation issues any time they're in a fight where the merc raid buff is used at the wrong time. This would be particularly noticeable in the case of IO, since DoTs are applied at the very start of the opener. All it would take is for the alacrity buff to be delayed from the fight start by half a second for the DoTs to end up clipped later in the rotation.
I'm not saying that's definitively not the case. 4.0 introduced a lot of bugs in the combat math, so I guess I wouldn't be too surprised if this were an issue. I would just be surprised if no one noticed it before now.
-
Defense would only be bad if you were fighting a boss that didn't do any melee/ranged attacks. Currently there aren't any bosses like that, so having more Defense stat will always increase your mean mitigation.
Soa does a single melee attack if you get close to him on the third floor, but its coefficient has been bugged for four years and does almost zero damage. He is sort of infamous as a purely force/tech boss.
If you run a zero-swap strategy on Firebrand and Stormcaller, the Stormcaller tank will never take any melee/ranged damage (though they will take kinetic/energy damage from the barrage while kiting spires). The Sunder tank in Cartel Warlords will take no melee damage, and the only ranged damage they will receive will be from Horic cleaves or from Horic/Sano random single-target attacks (and sane Sunder kite patterns are well out of Horic's singlet-target range). Brontes has astonishingly little melee/ranged damage (only sources I can think of are p1 tentacles, kephess, droids and Fire and Forget which shouldn't be on the tanks anyway).
Broadly speaking though, you're absolutely right. :-) I'm just picking nits because it's fun. Trading away some mitigation for no mitigation is never the right choice, even if the some mitigation is marginal.
-
Thank you for your detailed answer! I used to play a Monk but apparently I should've paid more attention to the mechanics.
:-) Snapshotting is really important for Uliana's Monks, and was previously somewhat important for Static Charge (irrelevant now that SC has been nerfed). It's also extremely important for Arcane Orb DMO Wizards (snapshotting the AO cast at 5 arcane dynamo stacks). Short-duration effects in Diablo are actually so potent that a tiny, tiny change to how snapshotting interacts with the Bane of the Stricken in 2.4 actually resulted in a dramatic DPS reduction for many classes on high-GR elite packs.
/offtopic
Since posing the question I've tried 2) on the dummy. I used Polarity Shift and reapplied the DoTs manually - they'd always last the shorter, alacrity-reduced duration. When I applied the DoTs, then used PS and reapplied them, they still had the longer duration. Didn't try it for Lightning and auto-refresh but it seems like the general consensus among Lightning Sorcs is to use PS before Affliction.So what does that mean? Does the DoT slow down but its duration stays the same? That'd make it quite terrible to use DoTs during PS. Is the text wrong? Or did they change the mechanics?
How did you do your testing? In other words, exactly how did you determine the DoT duration? Absent a hilarious bug, you should notice a proportional reduction in your DoT duration (and corresponding increase in tick rate) when using Polarity Shift after DoT application. In fact, you would actually get a slightly more pronounced effect than you would see if you applied your DoTs under Polarity Shift, since 100% of the alacrity boost would be happening within the effect duration.
If this weren't working, then any class with a rigid rotation timed around DoT applications and cooldowns (or GCDs) would be broken right now. Notably: IO, Pyro and Virulence.
As a sidebar, when playing Lightning, you should use Polarity Shift after applying Affliction in your opener. Outside your opener it's irrelevant, since Affliction is auto-refreshing anyway. Specifically, the standard Lightning opener:
(precast) Crushing Darkness > Affliction > Polarity Shift > Adrenal > Unlimited Power >
Thundering Blast > Lightning Strike > Recklessness > Lightning Flash > Chain Lightning > etc…
-
Just logged in to say that I'm amazed with you yet again, KBN. That's just brilliant
EDIT: Also, 1/40 ms is... err... Are you even human?
LOL. I suspect most gamers are capable of achieving that rate without even thinking too much about it. Here's a bit of timing intuition…
Imagine you're looking at a user interface element, like a button. You know that when you click the button, something is supposed to change state (like, the UI turns green). You click that button and it feels like the response was not quite instantaneous, but very close. In other words, you perceived enough of a delay that you wondered for a split second if the button click had even registered. That amount of time is 150 milliseconds (200 for most people, but as a gamer you have faster reflexes, so probably closer to 150). So scale off of that. One button click every 40ms is right around four clicks in the span of time it takes you to wonder whether or not a UI is frozen. As I said, I'm pretty sure most gamers are capable of achieving that input rate. :-)
Not necessarily cyclic, but you reminded me - if you can make some programmably-fed input, it's totally worth to make it pseudo-random with different density parameters so to eliminate weird frequency response that might happen otherwise.Indeed. If we could run automated testing of this sort (the devs can, so the engine is certainly capable), we would be able to provide much more statistically robust input.
Isn't that making latency strike at its full? I've always thought queued ability actually IS sent to server, all that is needed is to contact client twice and not just once.The queued ability is sent to the server, but only once it is actually necessary. The client does pre-send it, but only "action queue window" time prior to the GCD interval.
I mean, let's suppose my lag is 100 ms. If i set ability queue to 250ms and server contacts me 250ms before my GCD ends, finds an ability, executes it exactly at the end of GCD, during the remaining 250ms I find myself in a pretty bad place so I try to press some defensive but some other ability is already being executed. So in this case I probably have killed myself with the ability queue setting, thus I lower it to 100ms or something and voila!The server is actively syncing GCD boundaries with the client. It's not that the server asks the client "what ability do I do now?" Rather, the server actively keeps the client apprised of "I will need the next ability at this point in time", and the client contacts the server at some point prior to that time, as determined by the action queue window.
Also, you can most certainly kill yourself with the action queue window. A notable place where this happens is interrupts on casters, since the action queue window is used to define a period at the tail end of a cast where you're "locked in" and the cast is going to complete, regardless of what you do (another place where this happens is reflects, as on Ruugar).
Also, how do you resolve alacrity issues? Especially those coming from raid buffs etc.?I mean, don't you have to contact clients at the rate provided by alacrity hard cap rather than 0% alacrity?
Yep. This is taken into account with the server's active syncing. When the next boundary shifts, it notifies the client accordingly. If you have high enough latency, this can result in client-side "hitching" if the window moves dynamically from a period in the future to the past (from the perspective of the client).
-
1.) If you use an alacrity buff before using a DoT and then automatically refresh that DoT (Lightning, Affliction or Retractable Blade for AP), will it still benefit from the alacrity buff?
Yes, but only for the duration of the alacrity buff. Once the alacrity buff falls off, the DoT slows down again. This can be problematic for rotations that use multiple DoTs (e.g. IO, Pyro, Madness, etc), since one DoT has been accelerated (cumulatively) more than the other DoT, resulting in a more significant desync.
2.) If you use an alacrity buff before using a DoT and then manually refresh that DoT before it wears off, will it still benefit from the buff?Yes, but just as with above, once the alacrity buff is gone, the DoT slows down.
2.) a) If so, is it worth it for Madness to always keep the Polarity-Shifted DoTs up?Uh… no. You want to try to avoid applying DoTs during Polarity Shift, and instead use PS immediately after their application. The reason for this is two fold. First, as I mentioned above, you don't want to desync your DoT durations through asymmetric alacrity application (alliteratively). Second, and more importantly in Madness's case, you want to make sure that the GCD between your DoTs (which both Madness and Lethality preserve) remains exactly a GCD. If you apply during an alacrity window, and that alacrity window closes before you reapply, your DoTs will end less than a GCD apart from each other, meaning that you will have some downtime that might cause problems (e.g. with Force Leech).
Now, using PS after DoT application will speed up the DoT tick rate (because, remember, every calculation is relative to the current state of your character), but it will speed the DoTs up symmetrically and avoid generating gaps in the rotation.
The best approach with Madness (and with Lethality) is to use Polarity Shift immediately after the second DoT is applied.
3.) Does the game calculate the crit% for each DoT tick the moment it happens? Meaning, if I use a crit adrenal while a DoT is up, is that a dps increase?Yes. Don't use crit adrenals though as they are not as much of a DPS increase as the power adrenals.
From the game's perspective, each tick of a DoT is a bit like if you had activated a fresh ability at that exact instant. It recalculates based on your buffs and the target's debuffs. This is actually very easy to see on a dummy. Click off your class buffs. Apply Creeping Terror and observe a few ticks, then hit your class buffs. You'll see the DoT ticks go up in value (for maximal accuracy, remove your relics beforehand).
4.) If I spread my DoTs, do those DoTs have the same alacrity as the original DoT? What if the original DoT is unbuffed but I use an alacrity buff right before spreading?Nope. Basically the same as the answer to question 1.
All of these questions hint at a core mechanic that SWTOR's combat system does not have, which is to say, snapshotting. Snapshotting is a well-known technique in Diablo 3, as well as many other RPGs and MMOs. A common example from D3 would be a Monk applying Divine Palm (a debuff that explodes on damage) only during the "cold" proc of Convention of Elements ring (an item which rotates through four extremely strong damage buffs, each buff lasting 1 second). These palm applications would "snapshot" the buff state of the caster and hold onto that snapshot until detonated, even if the CoE buff had moved on from the snapshotted state.
In games with snapshotting (like D3), timing of delayed damage activation is what is important. SWTOR doesn't have snapshotting though, and every damage calculation happens relative to the fight state in the moment, and so timing of delayed damage application (i.e. when you see the flytext) is the important bit. This holds for every effect in the game, including DoTs, channels, buffs/debuffs, etc.
-
The good news is that everything is still very doable with the bug. (p2 NiM Council is the hardest, but still doable with a disciplined movement pattern) The bad news is the way this reflects on their attitude toward endgame PvE, which we unfortunately already knew. :-(
I've been playing Guild Wars 2, which is ironically an MMO which completely lacked any serious endgame PvE for almost 4 years of its life. It should tell you something that I feel more engaged in that game than in this one.
-
It's fine in NiM. Haven't tried it in any other mode since the patch.
Unsubbing
in Off-Topic
Posted
At one point, it was very fashionable for players ragequitting SWTOR to post their unsubscription rant in this subforum. Then unsubbing became more common (partially because of F2P), and I think also people got tired of reading the rants. Call me old fashioned, but I'm posting mine.
I have just canceled my subscription, and for the first time in over five years, I will not have an active SWTOR account. In the unsubscription form, you're asked to write some free-form text describing why you're leaving. This was mine:
Since I was at one point quite well known on these forums, and since I still have several pinned threads on the class forums, I thought it was appropriate that the last ("most recent") post in my profile would be this one. I did everything in the game. Twice. (thanks Jessie Sky! next time listen to what I say when you ask for my opinion on something) I had a lot of fun over the years, but it's clear that those years are past and not future.