Arcanaville

I just posted a build for fun in the I19 discussion thread as an example of inherent fitness leveraging. Since I don't often post builds, I wouldn't want to cheat the number-crunching community of a chance to tear apart one of my min/max attempts. So I'm reposting here, so the I19 thread doesn't become a scrapper-dominated bloodfest.

My question to you all is: what's wrong with this build? How could it be improved? Is it really something that benefits from inherent fitness, or is there a way to replicate it that I haven't found. And here's a real conundrum: there's actually a redundant set bonus in there, something over the rule of 5. Oddly, whenever I try to shuffle things around to get rid of that redundant bonus, I tend to make the build *worse* in some way. Very curious.


Repost follows:

Just for giggles, I tried to see what I could come up with in terms of an inherent fitness rebuild of my MA/SR scrapper. This is just for fun, and I'm not saying the results are going to be typical. On the other hand, I don't think this is necessarily the best anyone's going to be able to do either under inherent fitness. This build takes every MA/SR power except thunder kick, challenge, and elude, burns a power choice on travel (as opposed to just Ninja Running everywhere) and still ends up pretty tough. More importantly, its not trivial to replicate without inherent fitness due to the power choice requirements: losing three power choices means losing fighting, or losing MA attacks. You could eliminate EC, DT, and maybe Super Jump and then rebuild as a more recharge-heavy build (you'd need more recharge to make consistently full attack chains and also to boost single target to compensate for the lack of the crit-enhanced combo of EC->DT). But I think this build is a product of inherent fitness in large part.

Reading off the totals:

>48% defense to melee/ranged/AoE
~16.5% defense to s/l/f/c/e/n
~14.5% defense to psi
19.9% resistance smash/lethal
3% resistance f/c/t/p
6.8% resistance energy
9.9% resistance negative

388% regen (31.5 h/s at level 50)
145.5% health (1948 at level 50) [with accolades]
190% recovery (3.49eps)
end use: 1.12eps (with FF, FS, Ev, CJ, Tough, and Weave all running)

Not bad. It has some PvPIOs in it (in Focused Fighting and Focused Senses of all places, but also a couple in Tough) but that's mostly extravagance for the +3% procs and a few more +health and +dmg bonuses. They could be replaced with LotG without seriously altering the build (without the Glad +3% I would shift the slot to Weave for some extra defense insurance, although I would still be above 45% to all positions either way - honestly, that's the one I would use merits to buy above all others if I was going to implement this build; the others are entirely optional).

Interestingly, the strategy of this build is not to steal slots from fitness, but rather to steal slots from elsewhere to leverage the additional power choices. Which I think will happen as often as, if not more often than, people pull slots from stamina or health themselves. It still means that extra pool choice and four more power choices are still very meaningful.

I'm curious to know if anyone can recreate this build more or less without inherent fitness. I couldn't, but then again I'm not the worlds most experienced Mids jockey.

Also, I rarely post builds, so now's everyone's chance to rip my min/max skillz a new one by posting a far superior build that does more or less the same things. Hit me with your best shot, people.




(Note: the build uses version 1.8 with the inherent fitness database hacks if you want to load it)

Hero Plan by Mids' Hero Designer 1.803
http://www.cohplanner.com/

Click this DataLink to open the build!

Violet Rumble: Level 50 Natural Scrapper
Primary Power Set: Martial Arts
Secondary Power Set: Super Reflexes
Power Pool: Leaping
Power Pool: Medicine
Power Pool: Fighting
Ancillary Pool: Body Mastery

Hero Profile:
Level 1: Storm Kick -- Mako-Acc/Dmg(A), Mako-Dmg/EndRdx(3), Mako-Dmg/Rchg(3), Mako-Acc/EndRdx/Rchg(5), Mako-Acc/Dmg/EndRdx/Rchg(5), Mako-Dam%(7)
Level 1: Focused Fighting -- SW-Def(A), SW-Def/EndRdx(19), SW-Def/EndRdx/Rchg(21), SW-Def/Rchg(21), SW-ResDam/Re TP(45)
Level 2: Focused Senses -- SW-Def(A), SW-Def/EndRdx(25), SW-Def/EndRdx/Rchg(25), SW-Def/Rchg(27)
Level 4: Cobra Strike -- T'Death-Acc/Dmg(A), T'Death-Dmg/EndRdx(7), T'Death-Dmg/Rchg(9), T'Death-Acc/Dmg/EndRdx(9), T'Death-Dmg/EndRdx/Rchg(11), T'Death-Dam%(11)
Level 6: Agile -- S'dpty-Def(A), S'dpty-Def/Rchg(33), S'dpty-Def/EndRdx/Rchg(33)
Level 8: Focus Chi -- GSFC-Rchg/EndRdx(A), GSFC-ToHit/Rchg(13), GSFC-ToHit/Rchg/EndRdx(13)
Level 10: Practiced Brawler -- RechRdx-I(A)
Level 12: Crane Kick -- Mako-Acc/Dmg(A), Mako-Dmg/Rchg(15), Mako-Acc/Dmg/EndRdx/Rchg(15), Mako-Dam%(17), P'ngS'Fest-Acc/Dmg(17), P'ngS'Fest-Dmg/EndRdx(19)
Level 14: Combat Jumping -- GftotA-Def(A), GftotA-Def/Rchg(37), LkGmblr-Rchg+(39)
Level 16: Dodge -- GftotA-Def(A), GftotA-Def/EndRdx(23), LkGmblr-Rchg+(23)
Level 18: Crippling Axe Kick -- Mako-Acc/Dmg(A), Mako-Acc/Dmg/EndRdx/Rchg(39), Mako-Dmg/Rchg(39), Mako-Dam%(40), P'ngS'Fest-Acc/Dmg(40), P'ngS'Fest-Dmg/EndRdx(40)
Level 20: Super Jump -- Jump-I(A)
Level 22: Quickness -- Run-I(A)
Level 24: Aid Other -- Numna-Heal(A), Numna-Heal/Rchg(48)
Level 26: Dragon's Tail -- Sciroc-Acc/Dmg(A), Sciroc-Dmg/EndRdx(27), Sciroc-Dmg/Rchg(29), Sciroc-Acc/Dmg/EndRdx(29), Sciroc-Dam%(31), Armgdn-Dam%(31)
Level 28: Lucky -- LkGmblr-Def(A), LkGmblr-Def/Rchg(31), LkGmblr-Rchg+(33)
Level 30: Aid Self -- Numna-Heal(A), Numna-Heal/Rchg(43), Numna-Heal/EndRdx/Rchg(43)
Level 32: Eagles Claw -- Mako-Acc/Dmg(A), Mako-Acc/Dmg/EndRdx/Rchg(34), Mako-Dmg/Rchg(34), Mako-Dam%(34), P'ngS'Fest-Acc/Dmg(37), P'ngS'Fest-Dmg/EndRdx(37)
Level 35: Evasion -- LkGmblr-Def/EndRdx(A), LkGmblr-Def(36), LkGmblr-Def/EndRdx/Rchg(36), LkGmblr-Rchg+(36)
Level 38: Boxing -- Acc-I(A)
Level 41: Tough -- S'fstPrt-ResDam/EndRdx(A), GA-ResDam(42), S'fstPrt-ResDam/Def+(42), GA-3defTpProc(42)
Level 44: Weave -- LkGmblr-Def(A), LkGmblr-Def/EndRdx(45), LkGmblr-Rchg+(48)
Level 47: Focused Accuracy -- EndRdx-I(A)
Level 49: Physical Perfection -- Numna-Heal(A), Numna-Heal/EndRdx(50), Numna-Heal/Rchg(50)
------------
Level 1: Brawl -- Empty(A)
Level 1: Sprint -- Empty(A)
Level 2: Rest -- Empty(A)
Level 1: Critical Hit
Level 4: Ninja Run
Level 1: Swift -- Run-I(A)
Level 1: Hurdle -- Jump-I(A)
Level 1: Health -- Numna-Heal(A), RgnTis-Regen+(43), Numna-Heal/Rchg(46), Mrcl-Rcvry+(46), Mrcl-Heal(48), Numna-Regen/Rcvry+(50)
Level 1: Stamina -- P'Shift-EndMod(A), P'Shift-EndMod/Rchg(45), P'Shift-End%(46)[/QUOTE]

Just a short update. I've started a new batch of tests, and while they aren't complete yet one thing that seems to be suggested by the data is that, as you might expect but this isn't necessarily obvious, recharge may be bound to the animation clock. Meaning: its possible you may only be able to reduce the recharge of a power in 0.33s increments (approximately). Plus or minus about 5 milliseconds - and that plus or minus seems to be gaussian jitter, not random error - timing measurements cluster around those staircases. Also: the server seems to be capable of making one or two frame errors, and then "catching up" on the next attack. This ability to lag one power than make up the time on the next power is part of why ArcanaTime seems to work: individual powers don't always consistently clock exactly according to AT predictions, but sequences of powers tend to follow the predictions more closely than any one power does, because offsets in one power are evened out in the very next power very consistently.

For the numerically inclined, this is what I mean:

So, swipe->slash-swipe.

First, I tested with this chain: swipe->slash->swipe->strike. Why? Primarily for this reason: I can only tell when an attack starts in demorecords, not when one ends. So I like to bookend an attack chain with a dummy attack, because that is the attack that will tell me how long the final attack in the chain took, which can provide valuable information.

Here's the summary information. I ran this chain fifteen times. I timed the attacks as SwipeA, Slash, SwipeB, Strike. The average execution time of each attack was:

SwipeA: 1089ms
Slash: 1629ms
SwipeB: 1082.5ms

That's at +9.6% recharge (a +3 TO). Here are the same numbers at +19.2% recharge (2 +3 TOs):

SwipeA: 1087ms
Slash: 1489ms
SwipeB: 1087ms

It might be interesting to note that Slash is now timing out actually significantly *less* than its calculated ArcanaTime (1584ms). But that's actually something I observed when testing ArcanaTime in the first place. Due to how attacks "line up" with the animation and server clocks, it can happen that an attack will *look* a little longer because it "slops" over a clock, but then when that happens usually the next attack in the chain "picks up" that time. In other words, its an aliasing issue: attack A edges a little into the next clock, which makes it seem longer but also the following attack seem shorter. And that's what we're seeing here. Note the total duration of the three attacks in the second run: 3663ms. That's *very* close to the calculated ArcanaTime of the chain of 3696ms. This clock aliasing tends to average out. You're not really seeing Swipe run slower than ArcanaTime predicts, you're really seeing SwipeA stealing some time from Slash, making it look longer and Slash look shorter.

It also means my theory takes a bullet. 9.6% recharge is not the best you can do; at higher recharge the chain speeds up, which suggests something else is going on. But this also seems inconsistent with my cycle time measurements for single attacks.

Sleepy again. Will think about it tomorrow. There's probably a simple explanation for the discrepancy, but I'm not going to find it tonight. If someone else figures it out, by all means post. In fact, I already have a hunch what it might be, but I'd rather figure it out tomorrow.

There's something else I want to address, because apparently we're overdue and its been a while since this has been hashed out. Umbral (and others) have often made the claim that because fights rarely last longer than 30 seconds, only the 30 second calculations are relevant (to balance). The 180 second calculations are basically meaningless, because almost no fights last that long. 1800 second calculations are even more meaningless, because they have no connection with reality.

This is false. But rather than talk about the math, I'll illustrate with an example first because I think it will make more sense. I'm going to use a farming example. I have no interest in discussing the merits or detriments of farming: I'm only using the example because I think it will make the most sense to the most people in an intuitive manner.

Suppose I decide I want to figure out how fast I can earn XP (or Inf, or whatever). So I make a farm mission. It doesn't matter what kind for our purposes. And when I test myself in it, I discover that I can sustain up to 0x4 and still kill every spawn without risk to myself, with enough health so that by the time I move on to the next spawn I'm basically back to full health, which means this is sustainable all the way to the end of the mission. But if I set to 0x5, while I survive, I end up with sufficiently low health that if I don't rest before engaging the next spawn, I end up with even less health and eventually I have to rest or die. And when I factor in rest time, I actually earn less XP per hour than if I just stayed at 0x4. And when I increase to 0x6 for giggles, the first spawn just kills me.

My character can *sustain* 4, and can *survive* up to 5. Now my friend comes along and wants to try my mission. He discovers that just like me, he can sustain 4. But when he tries to increase to 0x5, the first spawn kills him. While I am 4/5, he is just 4/4 - he can sustain 4, and he can only survive 4. He can't survive 5. You could say I am more powerful than he is: we both can sustain the same amount, but I have the edge in the highest I can go without dying. So we tie on sustainability and I win on maximum survival, so I win. That's simple enough.

Now our friend wants to try *her* luck. She discovers that she can set this mission all the way up to 0x7 and not die. 0x8 kills her. She is much more survivable than either of us. But curiously, she can only sustain 0x3. Although 0x4 is no threat to her, at 0x4 she ends each fight with less and less health, and eventually has to rest. And she finds that she earns XP faster going continuously at 0x3 than stop and go at 0x4. So while I am a 4/5 and my other friend is 4/4, this friend is 3/7.

Now who's more powerful, she or I? She's obviously tougher. But I'm faster. When fighting spawns set to 4, they whittle her down faster than me. But when fighting spawns set to 7, I'm just plain dead. She can survive. How do you reconcile that?

Clearly, neither number tells the whole story. In terms of what individual threat can we survive she wins. In terms of what individual threat can we mow down continuously for rewards I win. The first number represents best case reward earning, and the second number represents toughest survivable threat. In terms of *game* balance, the first number is more important. But there are times the second number is more critical.

If I am playing a scrapper, I will tend to think the first number is more important. What matters most is what maximum threat I can simply play through. But if I am playing a tanker, I will tend to think the second number is more important. What's the worst case scenario I can tank for, especially in case of sudden emergencies.


Thinking that the 30 second number is "more important" than the 180 second number misses the point. Both numbers are important *always*. The first one tells you *what* you can kill. The second one tells you *how often* you can kill. The combination of the two represents a qualitative measure of the strength and sustainability of a character. Which one is more important, and to what degree, is somewhat situational. But usually, the shorter window number (30s in this case) represents more of a limit than a metric. In other words, you don't really want the character with the best 30s score. What you want is the character with the best 180s score of all the characters that meet a certain minimum 30s score. In other words, switching scales a bit, a 30/90 is better than a 10/50, but a 40/70 might be better than the both of them simply because you would rather sustain 40s and be able to take on the occasional 60 than be forced to sustain only 30s even though you can take on the rare 90. If anything higher than 60 is fine with you, then both 30/90 and 40/70 "qualify" but then 40 beats 30, and the fact that 90 beats 70 is less important to you.

Personally, I think this is how *most* people look at survivability *except* for tankers and other characters that are played like tankers by their player (like some brutes, masterminds, crazy blasters, etc). For those people, the exact opposite is probably true. They have a qualifying *sustainable* number, such that any lower would be too mind-numbingly slow, but anything above a critical value is fine, and then they want all out maximum burst survivability. So maybe they want at least 25, but anything above that is fine. So 30/150 and 50/90 both qualify, but then they would rather have the 30/150, because even though the 50/90 sustains a far higher level of activity, they really want the 150 and are willing to pay for it.


And that's why my mitigation spreadsheet *AND* my I7 calculations look at different time windows. Each looks at just one aspect of survivability, neither of which makes the other redundant.

And that is the relevance of 1800 second calculations. It presumes an average 30 minute mission, and looks for the best sustainable activity within it. Because the unit of measurement for reward earning is not the critter or the spawn, but the mission. Once the mission is cleared, you can't sustain any more activity even if you want to, because it will be empty. So any health you have at the end of the mission becomes essentially worthless. A player that can finish that hypothetical farm with full health has no advantage over the player that finishes the farm with half health. And that is why that calculation is relevant, even though no fight lasts that long. Because its not a measure of the best fight you can take on anyway. Its a measure of how many fights you can take on over a long period of time, a completely different limiting factor on combat.