PDA

View Full Version : HDex Parses



Rten
11-20-2014, 06:35 PM
Ok here are 2 sets of parses and the supporting details for me with JUST HDex being varied between the parses. The only unintended consequence was I could only get sham haste when I ran these parses.

Atk 5356
Str 1556/457
Dex1406/368
Weapons: Hand of Destroyer 130/18 and Royal Thex Katar 130/18
Metal PS
OH Proc’d at 25%
No chanter haste buff, sham haste, rest raid buffs (no bard)
Drunken Monkey, Second Spire, IoT, Zan Fi, Eagle’s balance
Total Haste 225%
Total time 36,848 seconds
/tell Rten Rten -vs- Combat Dummy Ena: -- DMG: 1110454928 -- DPS: 30136 -- Scaled: 30136 -- Punch: 1018877879 -- Strike: 80513616 -- DirDmg: 11063433 -- Non-crit rate: 78.2% -- crit rate: 21.8% -- Attempts: 308955 -- Hits: 308955 -- Accuracy: 100% -- Avg Hit: 3594 -- Max hit: 12487 -- DMG to PC: 0

17,610 procs or 1 proc every 2.09 seconds
2,645 Brell’s Sacred Aloe Balm
11,183 Eagle’s Strike III
1,539 Fists of Fury
2,243 Infusion of Thunder Effect

24,688 Flurries
40.2 Flurries per minute
300,300 hits (punch and strike no proc hit count)
8.150 hits per second average

220,836 76.9% punches (normal)
66,226 23.1% punches (critical)

12,067 91.2% Strike (normal)
1,171 8.8% Strike (critical)

2,677 Punch Average Damage (normal)
5,761 Strike Average Damage (normal)

6,458 Punch Average Damage (critical)
9,387 Strike Average Damage (critical)

1,559 Punch Min Damage (normal)
1,911 Strike Min Damage (normal)

2,524 Punch Min Damage (critical)
2,850 Strike Min Damage (critical)

3,254 Punch Max Damage (normal)
7,594 Strike Max Damage (normal)

8,435 Punch Max Damage (critical)
12,487 Strike Max Damage (critical)



Parse number 2 which had as the only change the addition of HDex (well and AC/HPs but they should not matter at all).

Atk 5357
Str 1556/457
Dex1462/427
Weapons: Hand of Destroyer 130/18 and Royal Thex Katar 130/18
Metal PS
OH Proc’d at 25%
No chanter haste buff, sham haste, rest raid buffs (no bard)
Drunken Monkey, Second Spire, IoT, Zan Fi, Eagle’s balance
Total Haste 225%
Total time 26,075

/tell Rten Rten -vs- Combat Dummy Ena: -- DMG: 791376582 -- DPS: 30350 -- Scaled: 30350 -- Punch: 726034872 -- Strike: 57388417 -- DirDmg: 7953293 -- Non-crit rate: 77.8% -- crit rate: 22.2% -- Attempts: 218912 -- Hits: 218912 -- Accuracy: 100% -- Avg Hit: 3615 -- Max hit: 12487 -- DMG to PC: 0

12,549 procs or 1 every 2.08 Seconds
1,812 Brell’s Sacred Aloe Balm
8.039 Eagle’s Strike III
1,109 Fists of Fury
1,589 Infusion of Thunder Effect

17,634 Flurries
40.6 Flurries per minute
213,375 hits (punch and strike no proc hit count)
8.183 hits per second average

155,582 76.5% punches (normal)
47,708 23.5% punches (critical)

8,644 91.4% Strike (normal)
811 8.6% Strike (critical)

2,679 Punch Average Damage (normal)
5,671 Strike Average Damage (normal)

6,478 Punch Average Damage (critical)
9,353 Strike Average Damage (critical)

1,559 Punch Min Damage (normal)
1,911 Strike Min Damage (normal)

2,524 Punch Min Damage (critical)
2,850 Strike Min Damage (critical)

3,254 Punch Max Damage (normal)
7,594 Strike Max Damage (normal)

8,435 Punch Max Damage (critical)
12,487 Strike Max Damage (critical)


The min hit and max hits for normal and critical are exactly the same but the average hit with HDex is higher as is the number of critical hits (0.4% higher). The difference is fairly trivial and even though these are relatively long parses I cannot rule out that it is within the statistical variance of the hits that there may actually not really be a true coding based benefit. But on the practical level, if HDex effects crit success rates, then it *should* also modify critical chances for our special attacks so that flying kick should also have a benefit from it. The thing I do feel comfortable in absolutely declaring is that HDex does NOT affect your proc rate in any parse-able way. Also this test has not in any way tested to see if HDex has an impact on our tanking capabilities. But as of this test result, I have to conclude that HStr which adds to straight to our min/max of all hits (normal and critical) has a larger impact on our DPS. I say this because where our DPS matters is in general raid mobs or named kills for the casual player. And the higher the mobs AC the lower our average hit falls, so anything which has the effect of pushing up our average hit and of course (min/max) is in our best interest. With this in mind, I have to also conclude that for straight DPS, going for significantly more cleave has an even more pronounced affect (see my posts on cleave stacking). As a follow up to this parse, when I get a 2nd raid aug will again parse the effect of adding a raid aug into our DPS. Historically, an item showing cleave IX +42 will be adding +42 to your min/max hits of all kinds, which is the equivalent of 420 HStr…

Phreaky
11-21-2014, 09:02 AM
Thanks for this Rten. Did you happen to see a difference between the parses with Thunderkick procs? I've been told that hDex helps with things like headshot for other classes, I wondered if that helps us with Thunderkick. Tanks have worked out that hDex indeed affects block/parry/riposte skill checks so it should help us defensively in the same way I'd assume.

I'd be curious to see what a hStr parse would do as I have not gone that route at all.

My understanding in the past was that hStr doesn't help our dps in an appreciable way and certainly nowhere near as much as the cleave augs will (as you mentioned). The raw AC on augs on the other hand does help in a meaningful way so I've gone the AC/dex route for those reasons.

The augs haven't kept up with the gear increases over the years so to me it feels like the only stats on them that actually impact gameplay in a measurable way is ac, hDex/hAgi and mod3s (specifically spell dmg and healing amount) for casters.

Rten
11-21-2014, 02:01 PM
The following are opinions for the most part... so a disclaimer! My parse showed there was no correlation between procs and HDex I am therefor drawing the unproven conclusion that the proc for thunderkick would also be unaffected. But that is complete conjecture on my part. I suspect that back when HDex modified hit damage and crit damage it may have also impacted what may have happened in the past with regard to procs.

When pretty much all of us even the casual gamer have 100k hps and 8K AC, and raiders are routinely 10%+ of those numbers, I just do not think there is a compelling case for any specific augment stat. There are 19 augment slots. Putting some statistics to it:

HPs max per aug is ~250 that would max it as 4750 HPS (or 4-5%)
AC max per aug is ~40 that would max it as 760 AC (or 9-10%)

HStat of your choice max per aug is ~8 that would translate to 152...
I think the average player probably runs ~350 to any HStat, so this in theory makes this the biggest change % wise (30%?) but also not very much of an impact. In my parse 60HDex translated for me into a 214DPS increase or 0.007% increase. If HDex scales linearly that would be a maximum of a 2% increase with complete focus on the HStat.

TDS seems to be a direct counter to the mudflation we have been seeing in our capabilities. When trash mobs have 3-8 Million HPS, even if I can sustain 35K dps, the increase of 214DPS is so negligible as to be sad. And I am not sure what you all are seeing, but even with a respectable buffed AC of ~10.4K, I still cannot molo without using 2hb, and even that is very dependent on mend and defensive discs.

I went trolling the interwebs looking for any relatively recent HStr parse and I found old stuff (2010-2011) but that is far enough back to be unreliable. And everything current in posts is pretty much straight speculation. Some of the people speculating are pretty smart, but with the impacts being so small, and the /rng being so twitchy, I fall back to if it is not a long parse, it is just conjecture.

If anyone feels the urge to parse HStr, the easy way to do a variance is to remove HStr augs and to turn off HStr tribute, then to re-equip it all, turn on the trib and parse away. That should give most of us the ability to hit ~50-80 HStr swing for the tests. And FYI even at 7 hours in, my parse was still wobbling by ~5dps, so that to have concrete you probably need to do 12 hours but 7-9 is reasonably accurate.

sojuu
11-21-2014, 04:48 PM
Did you take into acount hdex is % based? After +300 hdex returns should be so minimal they dont matter same for hagi.

Phreaky
11-21-2014, 08:33 PM
I disagree on the AC front. I'm in raid gear and my augments make up 19% of the AC from worn gear. That's a a LOT from augments! That does effect our ability to tank/molo.

MasterRin
11-22-2014, 12:55 AM
Last I checked, pre-TDS going full cleave was only adding like 2k dps. The TDS set lookin' much better?

brogett
11-22-2014, 03:12 PM
Nice work. I think my own parses (many years back) I concluded somewhere between 100 and 150 hdex for 1% crit. Your parses are in line with that too.

HDex *does* also modify certain procs, but possibly not all. Eg assassination has an hdex component. It's possible it also applies to some specific monk things too so it's worth testing.

The question of whether +crit or +attack (str) is most significant is a hard one, especially as parse dummies all have rubbish AC it seems. Arena test ones at least allow for "/say increase mitigation", but it takes a lot (100?) to make them equivalent to normal mobs.

Kaliaila
11-22-2014, 10:49 PM
Considering how very small all of the bonuses from Heroics are as well as the fact that they tend to only come at steps of 10 and 25, is it really realistic to declare that there is no parse-able increase to proc rate when there is only a 59 point change in HDex, and only a 52 point change in useable Dex?

Additionally, if all procs which occurred are bundled up into a single group of procs doesn't that also poison the well for proc rate data? Since only one offensive and one defensive proc can occur at any single point in time, with someone procing once every 2 seconds it is highly likely that any increase in proc rate would be blocked by two items trying to proc at the same time. We would really need to have the parse set up so that there was only 1 proc possible to happen in both the baseline and experimental parses.

Mris
11-23-2014, 02:35 AM
Nice work. I think my own parses (many years back) I concluded somewhere between 100 and 150 hdex for 1% crit. Your parses are in line with that too.


That lines up with what I parsed, as well, back when HDex had a per-hit damage addition to melee, so very long ago... If I feel highly motivated at some point here, I will do a longer parse with a larger margin. (I think I can manage around 150 HDex from augs and 20ish from trib without changing anything else but HP/AC.)


HDex *does* also modify certain procs, but possibly not all. Eg assassination has an hdex component. It's possible it also applies to some specific monk things too so it's worth testing.

With reference to Headshot (mentioned earlier as a specific example) HDex is specifically designed to help ranged (archery and throwing) damage, so it's possible that Headshot got an HDex figure in there somewhere to keep in play with that. That aside, I think weapon procs, at the very least, were mentioned by the devs as being increased by both normal Dex and HDex separately, at intervals of 35 for regular and 25 for heroic.


The question of whether +crit or +attack (str) is most significant is a hard one, especially as parse dummies all have rubbish AC it seems.

Further compounding the issue is the higher AC on raid mobs as opposed to group mobs. More attack might be pointless in group content, but push HStr over the edge in a raid environment. (Not saying it does, that's just theoretical.) That would make monks similar to necros, who, as I understand, prefer different Specialize casting skills depending on their raider or grouper preference. (Because they use different types of spells for different environments.) I'm no nec expert, so feel free to correct me if I'm wrong on that.

HStr also has the +damage component. (Not that I'm pushing HStr, as it's not a favorite of mine, but gotta be fair.) 1 damage for every 10 HStr is what I recall, and that was tested by several of us as accurate.


Arena test ones at least allow for "/say increase mitigation", but it takes a lot (100?) to make them equivalent to normal mobs.

I recall someone saying that those ups (and downs) worked only for 5 before rolling over or capping, but I'm not certain if that's true. I've not done any testing on the test dummies to make sure that further testing can be tested accurately. Test Test Testy Test TEST! *head explodes*



The thing that really gets me in all this is, monks in general seem to push for HDex (or at the very least, seem evenly divided) while all I see from zerkers (or the vocal zerkers) is HStr. With HDex (theoretical) benefit being % based, and HStr (theoretical) benefit being per-hit, you'd think the class with the higher hits per second would prefer HStr, if either did. (Also not a zerk expert, this is just what I tend to see on the Live forums.)

Rten
11-23-2014, 03:31 PM
Interesting discussion, if the HDex proc rate is in 25+ increments, then it *may* have an impact, but in 7+ hours of parsing 60 was showing a negative impact of 0.1% to my procs, which I interpreted as /rng being the wonky but basically being the same. It is of course also possible that FK or thunder are more effective, I do not have my system setup for doing repeated button mashing for the testing. it would be great to see a 150 HDex parse to compare to.

Mris
11-24-2014, 07:26 PM
Did a couple auto-attack tests for several hours each, with a change in HDex of 160. End result, Melee critical hit rate went down after removing my HDex augs and HDex tribute. Procs didn't change in any way that can't be attributed to the RNG. I didn't have the comat dummy attacking me, so I have no data for defense. (For now.)

533 HDEX, 336 HSTR, 4801 ATTACK

Khoza -vs- Test One Hundred Three: -- DMG: 794106289 -- DPS: 19300 -- Scaled: 19300 -- Crush: 788333381 -- DirDmg: 5772908 -- Non-crit rate: 75.5% -- crit rate: 24.5% -- Attempts: 194715 -- Hits: 163440 -- Missed: 31275 -- Accuracy: 83.9% -- Avg Hit: 4858 -- Max hit: 10378 -- DMG to PC: 0

Time 41146
DD attempts 4456 Avg 9.23 seconds
Fists of Fury 838 Avg 49.10 seconds
Total Procs 5294 Avg 7.77 seconds

373 HDEX, 336 HSTR, 4801 ATTACK

Khoza -vs- Test One Hundred Three: -- DMG: 498962685 -- DPS: 18983 -- Scaled: 18983 -- Crush: 495287761 -- DirDmg: 3674924 -- Non-crit rate: 77.1% -- crit rate: 22.9% -- Attempts: 124085 -- Hits: 104207 -- Missed: 19878 -- Accuracy: 84% -- Avg Hit: 4788 -- Max hit: 10378 -- DMG to PC: 0

Time 26285
DD attempts 2844 Avg 9.24 seconds
Fists of Fury 540 Avg 48.68 seconds
Total Procs 3384 Avg 7.78 seconds

Rten
11-26-2014, 12:58 PM
Very interesting parses, the raw data combining our parses to show:

HDex 368 had 21.8% crit rate (Rten)
HDex 373 had 22.9% crit rate (Mris)
HDex 427 had 22.2% crit rate (Rten)
HDex 537 had 24.5% crit rate (Mris)

60HDex added 214 DPS (Rten)
160 HDex added 317DPS (Mris)

I do not have all of the new AA so I *think* that has an impact, and I am going guess that has an impact on the crit rate, but we are clearly seeing an increase in crits associated with HDex.

My parses had greater than 8.1 swings (hits) per second. With 10 Hstr adding 1 dmg per hit (to min and max) I would project that to say that 60 Hstr would add 6dmg. 6 dmg times 8.1 swings per second being 48.6 DPS added by HStr, where the increase in crits added 214DPS for 60 HDex. Mris was showing a hit rate of ~4 per second, which at half of my hit rate makes for some VERY interesting conjecture. More hits with more crits should scale up noticeably. Taking Mris' swing rate, an addition of 160HStr would have added 16Dmg per hit and at 4 Hits per second, would have assumedly added 64DPS. So based on these 2 samples, HDex adds in a range of 4.3 to 4.9 times the damage of HStr. I am not saying HStr is not important or beneficial (we are not factoring in ATK at all at this stage), but I think that with straight DPS in mind, that you have to weight HDex pretty heavily over HStr.

I AM reassessing my value to HDex, my leaning right now is to clearly favor HDex over HStr (which is a reversal from my pre parse logic). My weighting formula will now be something like:

1AC = 50HPS
1HStr = 3AC
1HDex = 9AC
1HSta = 15HPS = 0.3AC
1HAgi = 3AC*
1Corruption = 25HPS = 0.5AC*

* indicates gut based valuations where I cannot defend it in a parse or anything but "I think"
I would appreciate any thoughts you all have of the value of Corruption resists on raids now days.


I have put formula's into an excel spreadsheet where I can then "value" different augs to give me a score which helps me rank which ones I want to use.

I love that the community can have these kinds of conversations with the goal and efforts being that we can all have the raw data to try and make what we think of as informed decisions.

Mris
11-26-2014, 06:35 PM
The difference in our swing speed is because I was using a 2 hand blunt. Oddly, when I test copied, it didn't copy my one-handed weapons, so I went with the 2 hander. (Pretty sure it has something to do with which bags they were in, as about half my bags disappear when I test copy.)

Rten
11-26-2014, 07:39 PM
That also helps explain your lower DPS too, though it is very interesting how consistent other aspects of our parses were (like proc rate).

Phreaky
12-01-2014, 05:17 PM
Did you guys notice any difference in Thunderfoot procs in these parses?

BTW in going hDex you'll also realize gains in AC as many +hDex augs are great AC augs as well while that's rarely the case with hStr.

Rten
12-02-2014, 02:39 PM
Hey Phreaky, my parses did not contain any FK at all, I was a straight melee setup with no special attacks (I do not have a setup that will autofire FK). And as I posted above, I value AC, HStr, HDex, HAgi, HPS, Corrupt resists and Endurance when trying to decide on augs.

Sokon
12-11-2014, 09:11 AM
HStat of your choice max per aug is ~8 that would translate to 152...
I think the average player probably runs ~350 to any HStat, so this in theory makes this the biggest change % wise (30%?) but also not very much of an impact. In my parse 60HDex translated for me into a 214DPS increase or 0.007% increase. If HDex scales linearly that would be a maximum of a 2% increase with complete focus on the HStat.



Two decimal places off when converting fraction to percent. It's actually a 0.7% increase.

It's too bad we can't swap in regular dex to see if hdex and regular higher dex change our crit chance by exactly the same amount. Or wait... there are some potions that up dex... but how can I not be capped, hrmm maybe armor with no stats?

I think this would be the method to determine what if anything hdex is doing. Is it just because our dex is higher, whether it is hdex or dex, or is it because hdex is effecting crit rate.

Mris
12-11-2014, 07:32 PM
Two decimal places off when converting fraction to percent. It's actually a 0.7% increase.

It's too bad we can't swap in regular dex to see if hdex and regular higher dex change our crit chance by exactly the same amount. Or wait... there are some potions that up dex... but how can I not be capped, hrmm maybe armor with no stats?

I think this would be the method to determine what if anything hdex is doing. Is it just because our dex is higher, whether it is hdex or dex, or is it because hdex is effecting crit rate.

Well, if you do a testcopy, you could reset AAs and buy relevant DPS AA EXCEPT the Planar Dexterity AAs. That should add about 200 Dex once bought, enough to tell if it's base Dex adding to crit rate.

And for HDex doing anything... well, it's upping our dex cap, and raising your base Dex by the same amount. Some people argue that this isn't true, but it's easy to test. Check the Stats tab of the inventory window. Use a low stat like Wis for monks. Mine shows:
737 / 825 + 182
Which means 737 base, 825 total, and 182 Heroic. If I check the total stats, however, I see 919. If we used a capped stat, it might be harder to tell, but since my total is higher than my base, the Heroic must be adding to the total, not just the cap.

(BTW, I'm not trying to indicate you were saying otherwise. I'm just getting it in there before someone brings it up, because that seems to happen a lot.)