Werner

[ QUOTE ]
[ QUOTE ]

I let them beat on me for about 10 minutes, and it came up 5.47% for Slash, and 10.71% for Slice. Ah, the Broad Sword attacks have a little extra accuracy, which would account for Slash, or just the small sample size.


[/ QUOTE ]

Do Cims always lead with slash and follow with slice? That would somewhat explain your numbers, depending on what your def debuff res is like.

[/ QUOTE ]
I don't know, and it went away on my second test run. I'm chalking it up to a lucky streak and/or small sample size. I forgot to look at how many times they'd actually used Slice, which would have been relevant.

[ QUOTE ]
I'm saying MIDs is averaging the damage wrong. This is what its doing:

FinalDamage = BaseDamage + (TotalDotDamage * 0.8)

That is incorrect, because the dot isn't 80% chance for all of it, it's 80% per tick. It could stop after 3 ticks, or 2, or 1. In other words, it's giving you inflated damage numbers. As mentioned above, the proper way to do it is:

FinalDamage = BaseDamage + (TotalDotDamage * (0.8^NumTicks))

[/ QUOTE ]
If each DoT has an 80% chance, and the first failure cancels the rest, wouldn't it look like this?

BaseDamage + DotDamage*0.8 + DotDamage*0.8^2 + ... + DotDamage*0.8^NumTicks

That's more than your:

BaseDamage + NumTicks * DotDamage * 0.8^NumTicks

And less than Mids':

BaseDamage + NumTicks * DotDamage * 0.8

Also, is that how all Fire DoTs work? I want to make sure I'm doing this right, and I haven't fiddled much with Fire in the past.

[ QUOTE ]
I think his point was given the same defensive powerset (ie: sd to sd, da to da). Otherwise, I have some faceplanting tankers to introduce him to that aren't quite built for the 54 boss farms my scrappers run.

[/ QUOTE ]
I was chasing around a Fire tank who seemed to think that Rise of the Phoenix should be part of his offensive arsenal, used at every opportunity. I swear he must have had it slotted for recharge. Oh, and he had no knockback protection. I bought him a knockback protection IO it was so bad (this was back when they were much cheaper), but he didn't slot it. Good thing the rest of us were able to handle the aggro.

I'm sure there are equally bad stories about scrappers, and I've spent time around very good tanks as well, so I'm not picking on tanks. Just THAT tank.

[ QUOTE ]
Question, I currently have the PvP defense IO sitting in my base and I'm thinking about using it for this character. But I'm not positive yet. How powerful could this build be if done right?

[/ QUOTE ]
Probably very powerful, assuming you can pull off a build that does everything – soft capped defense, perma Dull Pain, Aid Self, Tough, and so on. You won't be setting any DPS records, but it'll be good enough to take down almost any target, just more slowly and safely than some of the DPS builds.

[ QUOTE ]
...contention held by another friend that..

"No Scrapper is ever better at surviving than any Tanker".

[/ QUOTE ]
Hahahahaaha!

Tankers are much better at surviving in general, but there's definitely an overlap area when you're comparing scrankers with DPS tanks.

(edit: By a similar token, some blasters are more survivable than some scrappers.)

Widows look awesome, but they have lower hit points than scrappers, right? That would seem to make them more susceptible to bad luck. On the positive side, when luck goes their way, look out.

[ QUOTE ]
What I really wanna see is what you can do w/Dark armor. Even w/capped S/L defenses, Fire has a lot of holes to cover up, and you can only go so far w/just pools & IOs.

[/ QUOTE ]
What you can do with Dark Armor:

- Soft Cap
- Great Resists
- Great Heal
- Sustainable Endurance (in most situations)

<font class="small">Code:[/color]<hr /><pre>| Copy &amp; Paste this data into Mids' Hero Designer to view the build |
|-------------------------------------------------------------------|
|MxDz;1404;703;1406;HEX;|
|78DAA5935B4F134114C767BB2D4BE98D52DA722D37A19496B 544A35C14A3A206A50|
|911E5496DD632B4156C9BDD46415FFC001A0DA03EF9E6F523 18BF89FA05BC7E0353|
|CF65DAA8AF6EE0FF9B3D33FF9973E674733B4B7E211E9C125 AF0ECB6E538F9B5826|
|DD56AD2F6E4AC62B9200C2144B419CBE7E4B694E625AB6E55 ACFE5674496ECA8A23|
|CD25CBDECA9FB66F57EDF072A5246D59A99BCD817FB55ADD3 6CF97EB15E938FCB22|
|2AD5AB9520CA89962A90E6FA1E6D486B49D52B9163E572B17 CC33D58DDD7CCE72EA|
|D2DEED818C52F0FF0853A3A7E1119B8019E12A324A0477997 18BD0B64548A14B239|
|72EB21ACD1C25B41F214CC7018362CA50CB44C3ADDD216BC7 5D827F8710BC47E8B9|
|4F48C3725D6DABF3B641DEB6F3AF6DDB609D87D709CF24599 F40C8D05C5487C175B|
|4731D1D5C4717D7D1CD7574731D197079D546DE4580A18B18 C47C1CF3F83EBA28D9|
|4F84CECF84C42F8149A41AF088E9E3EC8A832BC095BA02EC8 AB12BC6AE695810A2C|
|B6878420B14EA9B656F5F978E1B0E74EB181D88131231C21E D8C2AA3FE11F2E5CD7|
|F585D0F39590F84618FECEF849780CAE08BB44649E4F31201 65575452FC22029469|
|619171857E98EC7D6198B74D5EDE08A2B577C8FD21EDB271C 3A603C653C234C3C27|
|EC83AB5765DDCBDD98E06E4C703792DC8D247723C9DD30C1D 5CF6769FDEA5E87153|
|503B3E18A06E9E60F2032A4CE189AA56353738C79C2D409C6 49C6089D9105D7A8EA|
|F8E831DE5B87D8B83A775CFD0A328A6E989B5473932A978C6 218E6D26A2EFDCF5C0|
|4E64C756FE64BCAE0F02BC66BC63BC2CC1BC65BC21AF43CCB 3569D934253CC31870|
|B7BE53F8C3FA57FE8C78C9BEEA6E7D6E4263DF90BBF5A5FED 7F3DEDFDCC77BB903F|
|40ACA2A88B68EA36B28D7F1F5068E2C1CDD4429E06BE383BF 996B6816139A439947|
|594079888B0C1F880FC58F124009A284503A51222831943E9 417288DDF0628F3F9|
|-------------------------------------------------------------------|</pre><hr />

Empty slot is for the PvP +resistance IO. Second Steadfast Protection unique is a PvP +defense IO.

[ QUOTE ]
The developers don't want Tankers "taking down" anyone.

[/ QUOTE ]
Fortunately for tankers that want to take down some bad guys, it doesn't matter what the developers want. It only matters what the sets are capable of achieving.

[ QUOTE ]
I think the Defender forums have generally the right mindset to make good use of it if I could: they would probably accept reasonable calculations if they could be demonstrated to reflect in-game reality, and conversely they would tend to suppress the urge to use the numbers to "prove" particular sets "not worth playing" which is always a danger of analysis like that.

[/ QUOTE ]
Yeah. It's good to avoid the trap of drawing too broad a conclusion from some calculations.

As a “random” example, take DPS. No matter how accurately calculated, it isn't very representative of many or most in-game situations. Pylons? Yeah. AVs? Mostly. Running around on a team doing missions? Nope. That's AoE and front-loaded damage. It won't matter if your Dual Blader does twice the DPS as my Broadsword if it takes 20 seconds before your Blinding Feints stack up and catch up to my built up initial chain. It won't matter if my Broadsword does a lot of burst damage to one or two targets if someone's Spines scrapper is just tearing the whole crowd apart. And I think the scrapper forum overall, despite a fascination with DPS and more recently AVs, does realize these things. Nobody is saying “Spines Sux” because of its comparatively poor DPS.

[ QUOTE ]
the scrapper forums will test almost anything (which reminds me, I still have to work on that all-SO Regen AV build).

[/ QUOTE ]
Way off topic, but likewise. I've backed off of my even level SOs personal requirement and have a bunch of 52 and 53 SOs now. Took a bunch of runs on Werner against Marauder. No luck so far. I'm thinking Sergei will have a better chance since there are a couple easy AVs that are susceptible to negative energy damage, but I haven't found a build I'm happy with. And I just keep doing other things.

Random numbers are themselves remarkably patterned at times. They'd actually be less random if patterns were avoided. Combine that with the human trait of finding patterns, and it's just something you're going to see a lot, even if, as I suspect, the random number generator in the game is a perfectly decent one.

[ QUOTE ]
P.S. - I've been playing Tankers for longer then Sarrate and Werner .... combined.

[/ QUOTE ]
No you haven't. I've been playing tankers (and blasters and scrappers) since beta. Forum registration date isn't the same thing as when people started playing. That was a big assumption to make.

(edit: Wait, even by forum registration date, we'd have significantly more time when combined. Do you mean hours played? That's possible, but on what are you basing that assumption and just stating it as fact? I have over 5000 hours in this game, though I'll admit that only a small portion of that is on tankers.)

Also completely irrelevant. How long we've been playing has nothing to do with whether we're calculating DPS correctly or not. On the other hand, you weren't the one that brought up how long we've been doing this, so I can't blame you for that particular train of thought. You were just responding to someone else.

Back on topic and now talking to the OP, OK, so you want to go with Invulnerability as the primary, and then either Super Strength, Fire Melee or Energy Melee as a secondary? That certainly makes the question easier.

First big question - what is your budget? Top end DPS typically happens at very high levels of recharge, and very high levels of recharge are very expensive. Even if I show, say, that Super Strength does significantly better DPS, it won't matter to you if it only does so at +287% recharge, and you don't have or don't want to spend the influence to get there.

[ QUOTE ]
Not to make all that hard work seem worthless, but you can see mob Hit rolls on your combat tab now.

&gt;.&gt;
&lt;.&lt;

[/ QUOTE ]
&lt;face palm&gt;

Right. I, uh... ummmm...

Look! A baby wolf!

&lt;runs away&gt;

[ QUOTE ]
This build has plenty of recharge to run SP - Smite - SL - SP - Smite - MG. I don't recall the exact totals needed but the highest requirement was MG at about 120 ish? %. I was running that chain with no gaps before I even had the purple set, so no worries. Maybe Werner or another numbers guy could chime in with the with the actual numbers needed to ease your worries.

[/ QUOTE ]
+128% recharge in Midnight Grasp, +90% recharge in Smite, +49% recharge in Siphon Life

For Invuln, I'd probably be double-stacking Divine Avalanche. You can hit the lethal soft cap easier than that, but I'd want to hit the Melee soft cap as well. Then ideally, soft cap FCEN, and you only have a ranged and AoE smashing hole, which shouldn't be a problem with still decent defenses and very good resists.

So I'd suggest my usual defensive chain: Divine Avalanche -&gt; Gambler's Cut -&gt; Golden Dragonfly -&gt; Gambler's Cut -&gt; Divine Avalanche -&gt; Gambler's Cut -&gt; Soaring Dragon -&gt; Gambler's Cut. That only takes +90% recharge in Gambler's Cut, which is to say that you don't need Hasten at all, let alone perma Hasten. You'll probably want Hasten anyway to get perma Dull Pain, but you won't need it for your attack chain.

And yes, perma Hasten can still be done, most easily on Super Reflexes. It's just uber expensive, and dominates the build. I'm about 2 seconds away from perma on my Dark Melee/Super Reflexes. I could have been there, but there were other things I wanted more than those last two seconds.

[ QUOTE ]
[ QUOTE ]
And completing formulas often does not make you an expert. It makes you practiced.

[/ QUOTE ]

Sure, and they have a lot more practice than you do. Werner's also tested his theories with a high degree of accuracy, something you haven't done. In professional circles, being able to recognize expertise is a very good skill to have. I'm not trying to drive you further into uber-defensive mode - I'm genuinely glad the topic got brought up because I *have* been curious about single target tanker DPS for quite a while after all the scrapper work. That said, might want to give Werner and Sarrate a tad more credit - they've been doing this a lot longer than you have.

[/ QUOTE ]
Eesh. *chuckle* Thanks for the praise, but I don't want to be oversold, and don't guarantee that I have more practice and time doing this. I haven't done any tanker calculations, and I'm not sure I want to take the effort of entering all the data for all the tanker secondaries, trying out a whole bunch of different attack chains, figuring out what real world slotting would look like for the best of them, and giving a reasonable comparison. I already have a lot of unfinished scrapper projects, so I'm not sure I need to add a tanker project to the mix. Heck, we still can't agree on top DPS primary in the scrapper world because the answer is usually “it depends”. I'm suspect that will hold true here as well.

[ QUOTE ]
DPS can only be approximated accurately. Various variables (which the Scrapper boards rarely acknowledge I might add) effect the actual end number after a particular "run" has been done.

These can include hit/miss ratio (especially on an AV where there is more of a base minimum chance to miss), insp gobbling (global activation means you're not doing damage (or as much damage in the case of auras)), alternate power activation (i.e. Build Up, Consume, T9 secondary power, etc ...), server lag (especially in AE missions now), missing a beat on your Q because you're going crosseyed slamming buttons for 10 mins straight and so on.

As some of these are totally random, you can only approximate DPS at best. That make sense in a real world context?

P.S. - Can you miss a pylon? If not, then you can more accurately approximate raw DPS, but that still has no real world place except maybe as a "perfect score" sort of scenario.

[/ QUOTE ]
You can miss a pylon. I strive for a 95% chance to hit +4s, and I include the 5% chance of missing in my DPS formulas. A long time ago, I compared everyone at 100% accuracy and just knew that my numbers were high, but this was making sets that require a number of attacks to hit in a row (like Dual Blades) appear better than they really were. That has long since been corrected. My DPS calculations will be wrong for those rare occasions that the enemy gets my chance to hit lower than 95%, but that is a rare occasion, even among AVs.

For inspirations, we typically don't use inspirations when we're doing “challenges” like AV soloing. Similarly, though a very different subject, when I calculate damage mitigation and survivability, I ignore inspiration use. So yes, inspirations will absolutely change things in the real world, and differently for different builds. But I ignore them in the calculations because I ignore them in game on any challenge of interest.

When I go full bore, I account for the Build Up cycle, the Consume cycle, the tier 9 powers, the Hasten cycle and so on. These accountings may not be 100% accurate, but they're a lot more accurate than ignoring them and just saying “too many variables to account for” and flying by the seat of your pants (not that that's what you do).

Anything but heavy server lag shouldn't be an issue if you queue your attacks up.

Missing a beat and screwing up your chain every now and again makes a difference, yes. But how much? 1%? Also, it's not the kind of thing that would invalidate a DPS comparison between two different chains. I suppose you could say that you're more likely to screw up complicated chain X than simple chain Y, and therefore reduce chain X's DPS by 1%, but I'd really rather leave the human factor out of it in this case.

As far as hitting other powers, let's take my Katana/Regen scrapper as an example. Regeneration is a very click-heavy secondary. And I don't, in fact, take those clicks into account when calculating DPS. So against really tough targets, like most AVs, I'll see slightly lower DPS than I calculate and see against easier targets that don't require the clicks, like pylons. But it's not enough to actually work into the numbers. It's just something to keep in the back of my mind. I just know that 200 DPS from Katana/Super Reflexes is marginally superior to 200 DPS from Katana/Regen when the character is pushed the the wall. On the other hand, maybe that Katana/Super Reflexes picks up Aid Self, and when pushed to the wall, has to spam it. That's a much longer activation, so might affect DPS more rather than less. If your point is that these factors exist, I agree. If your point is that they belong in a DPS calculation, and that if they aren't there then the DPS calculation isn't very good, then I would disagree.

OK, I think I finally see what your formulas mean.

The formula for attack chain DPS assumes you're looking at the unslotted damage for the attack, and that you'll average 100% damage slotting. It doesn't adjust for Arcanatime explicity, but you could replace “Total Animation Time of Chain” with “Total Arcanatime of Chain” and it would do the trick. I'd definitely skip the * 2 part, though. It makes your formula completely inapplicable except as the first very rough approximation. Mind you, I make very similar very rough first approximations when I'm looking for a good DPS chain. In my case, I just add up the Mids' average unslotted damage and divide by the total Arcanatime. It's nothing even approximating what you'll see in game, but it's good enough to start looking for better chains for more detailed analysis. So if that's what you're doing here, then I'm on board.

The formula for animation time of the chain is even rougher, anticipating that your chain will be limited by its longest recharging attack, have +100% recharge, and have a fudge factor of 1.25 seconds for Arcanatime. To me, that's just too rough to be of much use, even as a first approximation. You might find the best chains at exactly +100% recharge, but you won't find the best chains overall unless they happen to run at that low a point.

I hit it from another angle. If there's a buff attack (Blinding Feint, for instance), I'll amost certainly want it. Then I'm looking at the highest damage per Arcanatime attacks, and trying to fit them all together. Bonus points if the attack is a defense debuff, since those can take an Achilles' Heel proc, which is gold for AV soloing. I try out some chains, calculate the required recharge for each attack, and so on. As candidates start sticking out, the calculations get more and more detailed. For instance, with Dual Blades with perma Hasten and about +200% global recharge, can we do more DPS with Blinding Feint -&gt; Ablating Strike -&gt; Sweeping Strike -&gt; Ablating Strike (doubling up on the Achilles' Heel) or with Blinding Feint -&gt; Sweeping Strike -&gt; Ablating Strike (tripling up the Blinding Feint buff on Sweeping Strike). Questions like that can take some pretty in-depth analysis.

[ QUOTE ]
This is not to mention the futility of having 5 attacks in an attack chain that at most should be fitting 3 attacks given your criteria (but we'll assume different secondary effects or end costs or something to make up the logic).

[/ QUOTE ]
Completely correct. In the example, exceptional circumstances to the contrary, we should just have two attacks in our attack chain. I was just throwing out numbers to demonstrate the apparent problem with your formulas.

The Luck of the Gambler global is like a set bonus, so you don't have to use Divine Avalanche to benefit. You can't slot an Achilles' Heel in Divine Avalanche (a mistake I keep making in build discussions). But if you're going for what I consider the best defensive chain (DA&gt;GC&gt;GD&gt;GC&gt;DA&gt;GC&gt;SD&gt;GC), and you stick one in Gambler's Cut, you'll get almost four chances to proc per ten seconds, which is enough that that I probably wouldn't bother putting one in any of the other attacks.

A double stack of Divine Avalanche with a Luck of the Gambler global recharge in it will give you 34.8% defense. If you have 10.2%+ melee defense in the rest of your build, I wouldn't slot more defense in Divine Avalanche. I'd make sure I had 10.2% melee defense from the rest of the build rather than stick any more defense in Divine Avalanche.

I'm going to agree with Umbral. It'll be better for your own survivability at this point to have more defense debuff resistance, and it will also better help you keep others around you alive when there's aggro bleed.

[ QUOTE ]
Those remarks simply meant that all the theory crafting and number crunching in the world ultimately cannot take the place of experience.

[/ QUOTE ]
I completely agree.

That said, calculating DPS can be done accurately. Over on the scrapper board, we often verify our calculated DPS by soloing a pylon. It's the "perfect" target for this, as if you have sufficient DPS to do it at all, you can easily calculate your actual in game DPS from the time it takes to solo one (DPS = 38343.75 / seconds + 127.8125). These calculations are in very high agreement with the theoretical DPS calculations we use.

AV soloing is also very close to an ideal DPS situation. You stand there and beat on an AV for ten minutes or whatever. There aren't that many variables. From a damage perspective, just about everything important can be accounted for.

So while I agree that accurate DPS calculations have little to do with in-game damage, and very little to do with in-game performance overall, particularly for tankers, they DO have a LOT to do with answering the question posed by the OP - what is the best tank for DPS and AV soloing.

[ QUOTE ]
Werner: I typo'd the original formula I posted. Reread it and apply it again to your example.

[/ QUOTE ]
OK, reapplying your edited formulas:

[ QUOTE ]
Formula for attack chain DPS: (Total Damage of Chain / Total Animation Time of Chain) * 2

*EDIT* (corrected typo) Formula for determining an attack chain: Total Animation Time of Chain + 1.25 = Longest Recharge in Chain / 2

[/ QUOTE ]
So again, my example is 5 attacks doing 200 damage each, activation time 2 seconds, recharging in 3 seconds.

Total Animation Time of Chain + 1.25 = Longest Recharge in Chain / 2
Total Animation Time of Chain + 1.25 = 3 / 2
Total Animation Time of Chain + 1.25 = 1.5
Total Animation Time of Chain = 1.5 – 1.25
Total Animation Time of Chain = 0.25

Attack Chain DPS = (Total Damage of Chain / Total Animation Time of Chain ) * 2
Attack Chain DPS = (5 * 200 / 0.25 ) * 2
Attack Chain DPS = (1000 / 0.25) * 2
Attack Chain DPS = 4000 * 2
Attack Chain DPS = 8000

OK, while debuffed below 0% defense for a few minutes against some Cimerorans, Herostats says they hit me 80% of the time with Slash and 66.67% with Slice. Probably not a long enough test to count as proof, but certainly indicative that everything is functioning as expected.

[ QUOTE ]
whats the maximum dps you can get out of DA considering frankenslotting and all of the procs you could slot in it ?

[/ QUOTE ]
Assuming you want some accuracy, maybe 2 +3 Nucleolus, Hecatomb damage, Hecatomb proc, Mako's Bite proc and Touch of Death proc for 175.7 average damage in 1.584 seconds, or 111 DPS. It'll do more with global damage and Build Up, of course. I wouldn't slot that way, though.

I'm pretty sure that what the math suggests is what happens in game. Anecdote isn't evidence, but it sure SEEMS like negative defense behaves as to-hit buffs. And it would make sense for it to behave that way. Since we have the Achilles' Heel proc, I know that resistance debuffs work that way.

Just to make sure there was no I15 change, I pulled up Herostats. Now, this is against a crowd of Cimeroran minions with 45.39% positional defense, so they can technically debuff my defense below the soft cap, but it looks like a single hit takes me to 44.99%, so they'd have to stack to get anything significantly over 5%, in theory.

I let them beat on me for about 10 minutes, and it came up 5.47% for Slash, and 10.71% for Slice. Ah, the Broad Sword attacks have a little extra accuracy, which would account for Slash, or just the small sample size. I have no idea why Slice would have twice the accuracy as Slash, though. My AoE defense is the same as my melee defense. Hmmm.

OK, I turned off Maneuvers, dropping my defense to 41.81%. I'm routinely getting debuffed now, but generally something between 0 and 2 at a time. I'm probably averaging about 41.5% defense. Just saw a very brief “cascading defense failure” down to about 39%, but it didn't last long. I again ran the test for about 10 minutes, and Slash was at 8.90% and Slice was at 9.09%, so about what I would expect.

So other than what was probably just an unlikely string of hits from Slice in the first test, everything looks about like I would expect.

That not to say that you can't see what appear to be anomalous results. There are a lot of different enemies doing a lot of different things. Maybe there were some to-hit buffs when you were paying attention to the hits, or maybe just a lucky streak.