Well, kind of surprised at the first parse results, getting a weapon to verify my new theory based on my results.
I am at work right now, but I have a parse at home running where I am using a 26 delay 1hb. My calculated delay was a little over 3 seconds, I am betting that when I get the 26 delay 1hb parse done it will have a similar result. Since all of our logs only timestamp whole seconds, delays like 26 will end up reading as 3 seconds.
Here is 9+ hour parse (no buffs, non-weapon secondary, naked level 50 human monk with maxed skills): Test Fifty in 33237s, 449k @14sdps --- #1 Shekai 449k@14sdps (14dps in 33237s) Shekai -vs- Test Fifty: -- DMG: 449429 -- DPS: 14 -- Scaled: 14 -- Punch: 449429 -- % dmg as normal: 100% -- Non-crit rate: 100% -- Attempts: 14394 -- Hits: 9539 -- Missed: 4855 -- Accuracy: 66.3% -- Avg Hit: 47 -- Max hit: 100 -- DMG to PC: 0 1 attempt per sec: 3660 2 attempt per sec: 4557 3 attempt per sec: 540 "2x damage spike" at 50 (1597 hits) And the math: attack rounds = 1 attempt per sec + 2 attempt per sec + 3 attempt per sec delay = 10*time/(attack rounds - 1) = 10*33237/(3660+4557+540-1) = 37.959 damage = (2x damage spike - damage bonus)/2 = (50-8)/2 = 21 So, 21/38
Ok, so here are my 7 hour parses, I added the durations just to the right of the character's name. I forgot that my twink had a Haste Belt on, but it is on for both parses so its effect is negated. First, the bare fisted with 270 HtH skill with 24% Haste Belt: Next, the 26 dly 1HB with 24% Haste Belt: Well, my theory was in fact wrong, as was my thought that Laurb's "seconds per attempt" averaging thing was incorrect. Looking back at it now after having done these calculations I am betting that Laurb's haste must have been at around 66% during that parse. Adding up the Punch Rounds (i.e. the Attempt(s) per second numbers) we have 3509 + 4205 + 517 = 8231; with the duration of the HtH parse being 25201. If we take the duration and divide it by the number of rounds we get 25201/8231 = ~3.061717 seconds; so the Hasted Delay would be ~30.61717. Since Hasted Delay = Delay/(1+Haste), then Delay = Hasted Delay * (1+Haste); in this case ~30.61717 * 1.24 = ~37.96530; which does seem to support the idea of it being 38 delay. Adding up the Crush Rounds we have 4980 + 6271+785 = 12036; with the duration of 1HB parse being 25202. Here Duration divided by the number of rounds we get 25202/12036 = ~2.093885 seconds; making the Hasted Delay here to be ~20.93885. As such the actual Delay would be 20.93885 * 1.24 = 25.96417; with the actual delay being 26. For reference, I used a Flawed Defiant Brawlstick for this parse; and I am not anywhere near max on my twink's 1hb skill; thus the abysmal dps. While the 20 damage amount that was told to the then Monk Correspondent by the Devs of that time does appear to be correct as demonstrated earlier in this thread; it does seem that, as others have stated, the delay on unarmed HtH does in fact seem to be considerably higher than the 26 delay that was given out at that time. In fact upon looking at the list we have used since the change 38 delay is even higher than what the delay is supposed to be at level 5 (37 delay). Presently, I almost want to bet that the delay on our bare fists is 38 at all levels and skill; even if it is supposed to reduce. I also wonder if the Iksar are at 38 or 39 delay. It would really be nice if someone could look into this. I am not saying that anything even necessarily needs to be changed. I think that it would be reasonable for us to be able to see the formulas that are supposed to determine our bare handed damage and delay with relation to our skill, since the numbers for the Delay that we've been using since the revamp were apparently wrong. I am wondering if the Damage being at 21 might have something to do with the skill we are using being 270, when the skill listed in the table that was given to us has 260 for level 50. 270 might be the earliest it kicks up to 21. Just another reason for us to ask for the formulas; because it will really affect choosing when to start using weapons for people, especially on the TLP servers.
Xianzu, Here is the data on the short parse for fist only. Not sure why you think I was hasted. Putting these numbers into the formula I just learned from Raynard (Thanks Raynard): (10 x 1802 secs)/(209+237+29) = 18,020/475 = 37.936 Delay rounded to 38 delay agreeing with Raynards longer parse. Xianzu said, “It would really be nice if someone could look into this. I am not saying that anything even necessarily needs to be changed. I think that it would be reasonable for us to be able to see the formulas that are supposed to determine our bare handed damage and delay with relation to our skill, since the numbers for the Delay that we've been using since the revamp were apparently wrong.” I agree totally with this. One of two things happened. Either the chart provided in this link is incorrect http://www.monkly-business.net/forums/showpost.php?p=207148&postcount=32 or we have had a stealth nerf at some point.
I could do the testing, I'm curious myself. Here is what I have in mind, so weigh in now if you see a flaw or would like something added. 1. Verify ratio is tied solely to h2h skill: Find ratio at level 1, then level to 50 without using h2h, and verify ratio is unchanged. Note max h2h skill along the way to tie ratios to levels at the end. 2. Alternate skilling up with ratio parsing with steps no larger than every 5 levels. As Brogett pointed out, long parses are not necessary here, so I'm thinking 10 minutes or so should suffice. 3. Continue on to higher levels. I can trivially power level to 75, but I don't want to commit much beyond that. We'll see. If ratio is tied to level, I doubt I'll do the testing. It would be super tedious to run back and forth leveling and parsing. You can skill up on the dummies in the arena in test when they are active, and parse without skill ups when they are not. It's very convenient.
Excellent plan and well thought out. I don't see anything to add. Good idea about leveling to 50 with no H2H. Cheers!
The only reason that I was thinking that you were somehow hasted is that 16.6 value you got from dividing the other other delays by their time and in turn multiplied the HtH value would be how you would determine haste. Because seconds between rounds * 10 * (1+Haste) = Delay; so Delay/(seconds between rounds) = 10 * (1+Haste); in your case 16.6/10 = 1+Haste, so Haste = .66 or 66%. That was the only way I could make sense of how your math worked. The issue here is mainly on the damage side. You should be able to verify that the delay stays the same but the damage amounts can possibly change with level due to the damage modifier and the damage caps that are tied to levels. But you can probably find those values and compensate somewhat. The other points sound reasonable.
Xianzu, I can see why my math only kinda works. I was using the total attempts number provided by the parser to try and calculate the delay when by definition I need to use the number of rounds. Raynard method works. I got close to the right delay but the my calculated delay would get less accurate as the number of double and triple attacks varied from parse to parse. Since I was ignorant of the formula to determine delay I simply used the attack rates for known delay weapons to get a constant and multiplied the unknown fist rate by the constant to determine delay. Close but not accurate. You were correct in questioning it. Cheers!
Given all this information, where does this leave whitened treant fists on TLP? With bare fists @ 21/38, we have a ratio of 1.809, with the treant fists @ 14/28 giving a ratio of 2.0 Mainhand, would the 1 second faster delay, and thus better application of damage bonus(plus the addition of a melee augment proc) overcome the ~.19 ratio difference? I know speed is king in mainhand, and 38 is abysmally slow, so would this be a viable item? Wu's Fist of Mastery is obviously worth using mainhand if you can get it, as well. Also, this information means the information we have on how the epic works is most likely completely wrong as well. All of this is just super interesting, and I look forward to seeing more testing come out
This is from Prathun, I asked for the correct charts. Thank you to the parsers who figured this out. Looking at the code, hand to hand delay is 38 for non-iksar and 39 for iksar, unless the character is wielding the Epic, in which case the delay is 28. (Small caveat, at level 53 or below, wielding the epic scales the H2H delay based on the skill of the character)
I am waiting for a response, with the correct hth charts hopefully. so let's wait till tomorrow and see.
I want to say that my lvl 50 monk had epic equipped when I did the parse and calculated 38. I am almost thinking that we will end up seeing 38/39 delay for all the skill levels. I had asked Dzarn about it, but he had said something about it likely taking a long time to parse the area in the code where this is located at, I am guessing that Prathun was actually around when they did the HtH revamp, so likely knows better where to look. I am starting to think that back when the monk correspondent was given those numbers that was the intention, but something cause issues with the delay side so they cut that out all together.
With this new found information, does this place Rogues as the higher tier melee DPS on TLP servers for both sustained and burst when going into Kunark/Velious/Luclin/PoP? Appreciate all the information being gathered on this.
I think that depends on a lot of factors. We need to know the ratio at 60, and we need confirmation of the epic effect on the ratio. Rogues get 15 damage pierces in kunark, stay at 15 in velious, and then get 16 damage in luclin. Monks fists wont scale at all from kunark-> luclin, although a few 1hand options will come about. In velious, they have gharn's rock of smashing at 16/19 for mainhand, which should be a nice upgrade over bare fists, and in luclin they get the fangs off VTC at 17/19. In all accounts, fists will most likely be the best choice for offhand. So monks get slighty better ratios in velious, while rogues remain at 15 dmg piercers, and in luclin at the same time that rogues go from 15 to 16 dmg piercers. Post luclin, monks start getting crazy fists weapons as rogues 1hp scale up to 17 and beyond, so they should match up a bit better at that point. At this point its really tough to say without having more data on monk fist ratio and the epic.
I'm in the dilemma where I'm trying to find the best pure melee dps for this era, essentially your saying i can't go wrong with either though if I'm melee focused? If I'm a monk and my overall dps never dips below 10% of a similar geared rogue, I'm happy with that. Also will always be in a group with a shaman and a bard, which benefits more?
They should be pretty similar, yes, but like i said it depends on how the fists look at 60(which we dont know yet) and how the epic effects them(which we also dont know yet) Rogues will most likely be the safer bet, given the information we know now. There are no unaccounted variables with them. Monks and rogues will most likely end up being quite similar, unless something goes really wonky with the fists from 50-60 and with epic. As for shaman/bard in group, both get a strong boost from the haste/stat buffs. Monks are more self sufficient outside of shaman/bard grouping, due to the epic providing spell haste, which will stack with an item. Assuming the 41% haste belt from posky, a monk epic is self buffed to 81% haste. A rogue with epic will only self buff to 40% haste. However, both will be haste capped in the group with shaman/bard, so this doesn't matter unless they're soloing. Also, increased dex from shaman may disproportionately buff rogues(all though not by a crazy amount) if monks end up using bare fists with epic, because monks will not have procs and rogues will. This wont be a huge advantage and will shrink as kunark and luclin hit when monks have access to weapons with procs they can use.
