Elitist Jerks
Register
Blogs
Forums


Go Back   Elitist Jerks » Class Mechanics

 
 
LinkBack Thread Tools
Old 11/18/07, 12:17 PM   #1151
songster
Chief Passenger
 
Gnome Rogue
 
Earthen Ring (EU)
Originally Posted by Saegon View Post
That could be due to cold blood.
Don't be an idiot. Cold blood applies to one attack every 3 minutes.

Great Britain Offline
Old 11/18/07, 4:53 PM   #1152
Tiridor
Glass Joe
 
Human Mage
 
Feathermoon
Fixed the primary bug and a few other ones I hadn't found; remaining TODOs are properly testing to make sure that damage is physical, abilities with damage multipliers, and multi-word target names. As it is you can copy this thing, compile it, and use it to get a very precise (including armor reduction!) estimate of your personal damage from hemo and a ballpark estimate of raid damage from hemo (including armor reduction, but again not differentiating between spells and physical).

It's going to remain available as source only, but hopefully soon I'll have it in a stage where I can make it an executable - hell, maybe even with a GUI.

Please, for the love of all that is holy, if you have the java SDK download this thing and test it; I only have one 45-minute log right now and that isn't a lot to test against.

// Hemo damage parser. Takes an argument of a filename and reads standard-format combat logs to estimate contributed damage from hemo.
// Programmed by Tirith from Feathermoon.

import java.io.BufferedReader;
import java.io.FileReader;
import java.io.IOException;
import java.io.InputStreamReader;
import java.util.StringTokenizer;



public class HemoParser {

	
	public static String fileName = "C:\\Program Files\\World of Warcraft\\Logs\\WoWCombatLog.txt";
	/**
	 * @param args
	 */
	public static void main(String[] args) {
		String name = "";
		int mHemo = 0;
		int range = 0;
		
		BufferedReader stdin = new BufferedReader( new InputStreamReader( System.in ) );
        
		if (args.length == 1) {
        	fileName = args[0];
        }
        
        
		System.out.println("How wide is your damage range?");
		try {
			range = Integer.valueOf(stdin.readLine());
		}
		catch(Exception e) {
			System.err.println("Use a number you idiot " + e.toString());
			
		}
		try {   
			mHemo = maxHemo();			
		}
        catch(Exception e) {
        	System.err.println("File read error. I bet it's your fault. " + e.toString());
        }
        try { 
			System.out.println(hemoDamage(mHemo, range));			
		}
        catch(Exception e) {
        	System.err.println("File read error. I bet it's your fault. " + e.toString());
        }
	}

	static String hemoDamage(int maxHemo, int range) throws IOException {
		int armorless = (maxHemo-range/2); // Yes, this means that sometimes the hemo mod will behave as if armor is less than zero; it also means that sometimes hemo will behave as if armor is greater than it is. It should balance, statistically speaking. Any better ideas?
		double armorFactor = 0;
		double damage = 0;
		double yourDamage = 0;
		int charges = 0;
		int wasteCharge = 0;
		int crit = 0;
		String line = "";
		String tar = "";
		StringTokenizer st;
		BufferedReader in = new BufferedReader (new FileReader (fileName));
		while (in.ready()) {
			line = in.readLine();
			if (line.length() > 40 && (line.substring(20,40).contains("Your Hemorrhage hits") || line.substring(20,40).contains("Your Hemorrhage crit"))){ //on a hemo
				wasteCharge+= charges;
				charges = 10;
				crit = 1;//these next lines hunt down how hard your hit or crit was to estimate armor
				st = new StringTokenizer(line);				
				while(st.hasMoreTokens()) {
					String tok = st.nextToken();
					boolean done = false;
					if (tok.equalsIgnoreCase("crits"))
						crit = 2;
					for (int j = 0; j + 1 < tok.length(); j++) {
						if (tok.charAt(j) <= 48 || tok.charAt(j) >= 57) {
							break;
						}
						if (j + 2 == tok.length())	{
							armorFactor = (1.0)*(Integer.valueOf(tok.substring(0, tok.length()-1))/crit)/armorless; // this hemo's damage (adjusted for crit) divided by the maximum hemo.
							done = true;
							break;
						}	
						if (done)
							break;
					}
				}
				st = new StringTokenizer(line.substring(39 + crit)); 
				if (st.hasMoreTokens())
					tar = st.nextToken(); //TODO: add multi-word targets	

				
			}
			else if (charges > 0) { //on a non-hemo line with charges remaining
				st = new StringTokenizer(line.substring(20));
				boolean onTarget = false;
				while (st.hasMoreTokens()) {
					if (st.nextToken().contains(tar))
						onTarget = true;
				}
				st = new StringTokenizer(line.substring(20));
				if (physical(line) && onTarget) { //Test for physical damage on the hemo target
					boolean you = false;
					while(st.hasMoreTokens()) {
						String str = st.nextToken();
						if (str.equalsIgnoreCase("you")) { //others' benefit from hemo is more interesting than yours, since yours already shows up as your damage.
							you = true;
						}
						if (str.contains(("hit"))){
							if (you) {
								yourDamage +=36*armorFactor;
								charges--;
								break;
							}
							else {
								damage += 36*armorFactor;
								charges--;
								break;
							}
							
						} 
						else if (str.equalsIgnoreCase("crits")){
							if (you) {
								yourDamage +=72*armorFactor;
								charges--;
								break;
							}
							else {
								damage += 72*armorFactor;
								charges--;
								break;
							}
						}
					}
				}				
			}
			
		}
			
		return ("Others' damage from hemo: "+ (int)damage +"\nYour damage from hemo: " +(int)yourDamage+"\nWasted Charges: "+wasteCharge);
	}
	
	static boolean physical(String line) { //TODO: make this an actual test for whether the attack is physical >.>
		boolean phys = true;
		return phys;
	}
	
	static int maxHemo () throws IOException {
		int maxHemo = 0;
		String line = "";
		StringTokenizer st;
		BufferedReader in = new BufferedReader (new FileReader (fileName));			
		
		while (in.ready()) {
			line = in.readLine();
			if (line.length() > 40 && line.substring(20,40).compareTo("Your Hemorrhage hits") == 0){
				st = new StringTokenizer(line);
				for (int i = st.countTokens(); i > 0; i--) {
					String tok = st.nextToken();
					boolean done = false;
					for (int j = 0; j + 1 < tok.length(); j++) {
						if (tok.charAt(j) <= 48 || tok.charAt(j) >= 57) {
							break;
						}
						if (j + 2 == tok.length())	{
							if (Integer.valueOf(tok.substring(0, tok.length() -1)) > maxHemo) {
								maxHemo = Integer.valueOf(tok.substring(0, tok.length() -1));
								done = true;				
							}
						}	
					if (done)
						break;
					}
				}
			}
		}	
		return maxHemo;
	}

}

Last edited by Tiridor : 11/18/07 at 9:21 PM.

Offline
Old 11/19/07, 11:42 AM   #1153
Left
Don Flamenco
 
Left's Avatar
 
Draenei Paladin
 
Darkspear
Originally Posted by songster View Post
...

In practice this means the only raiding builds with SF are dagger builds with the extra crit from improved backstab (i.e. Mutilate builds).
/derail

The crit from Improved Backstab doesn't apply to Mutilate. Mutilate builds have a slightly lower base crit than Combat builds due to the lack of Vitality (but usually higher total crit to due crit-oriented gear selection).

The advantage of Seal Fate with Mutilate is that either of the two attacks (main hand or off hand) can crit, effectively increasing your crit rate for the purposes of Seal Fate by a significant amount. Thus, your chances for an additional combo point are precisely the chance that either hand will crit:

(crit %)*2 - ((crit %)(crit %))

For example, with a 25% crit rate, a Mutilate rogue has a ~43% chance of one or the other of his Mutilate attacks critting, so a ~43% chance for an additional combo point.

/end derail

Last edited by Left : 11/19/07 at 11:46 AM. Reason: Clarification

Offline
Old 11/19/07, 12:16 PM   #1154
Killars
Von Kaiser
 
Killars's Avatar
 
Gnome Rogue
 
Alterac Mountains
Well did a full BT/Hyjal clear while hemo and even though no WWS was put up I think Hemo is definitely on par or close to it, thats for sure. While being near what combat is give or take for personal damage, it definitely more than makes up for it in raid damage. I truely think im sold and would go as far to say rogues no longer have 1 single spec, but like druids, have multi specs for raids and you'd want at least 1-2 hemo rogues per raid.

To go along with a previously asked question/response, yes the Ashtongue trinket is definitely not even close to the same usefulness as it was with combat. I've been using WSC/MoTB with Executioner on my offhand and I seem to get a pretty good rotation of armor pen that typically doesn't overstack, plus the hit rating from both trinkets is nice. As mentioned I do have Excutioner on my offhand (1.4 speed) and I was debating if MH or OH gets more attacks off esp considering that I use Hemo now. I figured OH just as a guess since it's always breaking first, but I seem to think I may be wrong, someone let me know.

I don't know if I'm really late or completely being obvious here, but the thread got a bit derailed and I never got a chance to put in my last thoughts after all my questions. O and someone direct me to a place where I can find a new main raiding rogue for my guild, one of our three just went casual. I'm an officer so just send em my way =P


Offline
Old 11/19/07, 1:42 PM   #1155
Anked
Von Kaiser
 
Gnome Rogue
 
<Kin>
Azuremyst
Originally Posted by PSGarak View Post
You don't actually need to explicitly watch the enemy's debuff bar. All you need is a mod that says HEMO UP--HEMO DOWN, or "time since last hemo: 0:01.45" or beeps on a hemo application or something, by parsing the combat log. And as left is saying, it's just an offset. With enough it and expertise its an entirely deterministic half-second or one-second offset, too. The second rogue is losing maybe half a second of slice and dice over the course of the fight, which altogether adds up to probably not even one autoattack.

edit for kellhus's reply: Because the hemo combat cycles are so deterministic, one rogue holding back until the other's hemo is down results in the two rogues working in offset lockstep as a natural consequence. The comparison to mutilates is a valid point in that, unlike a GCD-limited class, being constantly energy-starved means waiting half a second (or even six seconds) is valid strategy without hampering dps.
Actually it would be easier then that. You could just set up you're own custom macros to set the refresh on your hemo to what you want it to be. In fact, you could work out your whole cycle in a matter of no more then 2 macros. If there are only 2 rogues in the raid w/ hemo I wouldn't think it would be that difficult to offset their rotations, even if that means losing 1.5 sec of dps time for the second rogue on the rotation.

LOL, you know what would be even funnier... to use whisper cast and you could actually play 2 hemo rogues as one including offsetting your times.

Offline
Old 11/19/07, 2:37 PM   #1156
Captain Winky
Von Kaiser
 
Gnome Rogue
 
Shadow Council
Originally Posted by Killars View Post
I've been using WSC/MoTB with Executioner on my offhand and I seem to get a pretty good rotation of armor pen that typically doesn't overstack, plus the hit rating from both trinkets is nice. As mentioned I do have Excutioner on my offhand (1.4 speed) and I was debating if MH or OH gets more attacks off esp considering that I use Hemo now. I figured OH just as a guess since it's always breaking first, but I seem to think I may be wrong, someone let me know.
If you wanted it to proc more, wouldn't MH be the way to go no matter what your OH speed is? I was under the impression that normalized PPM meant that a 1.4 offhand would only get half the proc chance per swing of a 2.8 mainhand. Add in that every Hemo gets the 2.8 proc chance, and if you're rocking swords every OH swing that procs a MH swing gives another 2.8 proc chance, and you come out with a massively larger proc rate from the MH.

Let's say it's normalized at 1 PPM. You get 21.4 mainhand swings per minute, and 42.8 offhand swings. That means a 2.8 mainhand has ~4.7% chance per swing to proc. A 1.4 offhand would get ~2.35% chance per swing. Add in another 17.1 MH swings per minute from Hemo (ignoring marginal energy gains/losses from finishers), ~1 swing per minute from MH Sword Spec, and ~2 swings per minute from OH Sword Spec. That means you actually get ~41.5 swings per minute, meaning your MH will proc 1.95 times per minute, compared to 1 per minute from the OH. This would change if you use Shiv a lot, but I've never seen that in a consistent ability rotation, so I'll ignore the possibility. Correct me if I'm wrong on this, because it's the assumption I've been running on

Offline
Old 11/19/07, 2:44 PM   #1157
Brandontu
Glass Joe
 
Human Rogue
 
Wildhammer
Originally Posted by Captain Winky View Post
If you wanted it to proc more, wouldn't MH be the way to go no matter what your OH speed is? I was under the impression that normalized PPM meant that a 1.4 offhand would only get half the proc chance per swing of a 2.8 mainhand. Add in that every Hemo gets the 2.8 proc chance, and if you're rocking swords every OH swing that procs a MH swing gives another 2.8 proc chance, and you come out with a massively larger proc rate from the MH.

Let's say it's normalized at 1 PPM. You get 21.4 mainhand swings per minute, and 42.8 offhand swings. That means a 2.8 mainhand has ~4.7% chance per swing to proc. A 1.4 offhand would get ~2.35% chance per swing. Add in another 17.1 MH swings per minute from Hemo (ignoring marginal energy gains/losses from finishers), ~1 swing per minute from MH Sword Spec, and ~2 swings per minute from OH Sword Spec. That means you actually get ~41.5 swings per minute, meaning your MH will proc 1.95 times per minute, compared to 1 per minute from the OH. This would change if you use Shiv a lot, but I've never seen that in a consistent ability rotation, so I'll ignore the possibility. Correct me if I'm wrong on this, because it's the assumption I've been running on
You're right. Killars doesn't have a clue how PPMs work if he thinks that his OH is going to be procing the enchant as much as his MH (because of instant attacks) unless he uses shiv as his main attack.... which I wouldn't put past him.

Offline
Old 11/19/07, 3:05 PM   #1158
Hanos
Back in my day...
 
Hanos's Avatar
 
Human Rogue
 
Sen'jin
Originally Posted by Brandontu View Post
You're right. Killars doesn't have a clue how PPMs work if he thinks that his OH is going to be procing the enchant as much as his MH (because of instant attacks) unless he uses shiv as his main attack.... which I wouldn't put past him.
Right now the mechanics of Executioner are still being tested and it isn't clear if it is PPM or % based.

Last edited by Hanos : 11/19/07 at 3:19 PM.

Offline
Old 11/19/07, 4:43 PM   #1159
Anked
Von Kaiser
 
Gnome Rogue
 
<Kin>
Azuremyst
As most love to say, "look at the spread sheet" whenever anyone asks a gear/spec question; I decided to do just that. I've done configurations in regards to 3 different builds using my current gear including the upgrades I covered on page 40 or so. My findings are somewhat interesting, so much so I thought I would share them with you guys as they may change how you think about gearing/gemming for a hemo build.

Here are the unbuffed talent builds:
19/42/0 Combat Swords (focusing on Hit)
Stats: 24.86% Crit, 22.12% Hit, 1656 AP
DPS: 769.56

31/0/30 Assassionation/Hemo (focusing on Agi/Crit)
Stats: 26.75% Crit, 13.00% Hit, 1859 AP
DPS: 846.66

11/27/23 Combat/Hemo (focusing on Agi/Crit)
Stats: 26.75% Crit, 18.00% Hit, 1690 AP
DPS: 818.95

11/27/23 Combat/Hemo (focusing on Hit)
Stats: 24.61% Crit, 22.12% Hit, 1646 AP
DPS: 818.91

Agi/Crit focus gems:
8x [Delicate Living Ruby]
2x [Smooth Dawnstone]
4x [Shifting Nightseye]
Socket bonus: 7 crit, 7 agi, 3 hit

Hit focus gems:
3x [Glinting Noble Topaz]
2x [Delicate Living Ruby]
7x [Rigid Dawnstone]
2x [Shifting Nightseye]
Socket bonus: 7 agi

If the spreadsheet is accurate, then stacking hit to max is now irrelevant post 11% when looking at hemo. Granted I haven't tested all of these in game, but from what I have seen going from 14/42/0 Combat Swords to 11/27/23 Combat/Hemo post 2.3 this feels right. I actually noticed a boost in my personal DPS, I know this is due to my currently miserable hit rating and how pivotal that is to a combat sword build.

I think I'm going to respec to 31/0/30 then next time I log in. You get a lot more utility, and even with my current gear (upgrades pending) it's a dps gain on the spreadsheet. As far as utility goes, I would have CB and Prep at the small cost of BF. It looks like 31/0/30 would be a very viable PvE and PvP spec to boot.

Has anyone done any in game testing already to prove/disprove this observation?

Offline
Old 11/19/07, 5:06 PM   #1160
• Aldriana
Mike Tyson
 
Night Elf Rogue
 
Doomhammer
Originally Posted by Anked View Post
I think I'm going to respec to 31/0/30 then next time I log in. You get a lot more utility, and even with my current gear (upgrades pending) it's a dps gain on the spreadsheet. As far as utility goes, I would have CB and Prep at the small cost of BF. It looks like 31/0/30 would be a very viable PvE and PvP spec to boot.

Has anyone done any in game testing already to prove/disprove this observation?
I recall seeing something go by that indicated that the high performance of 30/0/31 was due to a bug in the spreadsheet, and that the real behavior wasn't expected to be as good. My instinct is to say that, in practice, this build should get outperformed fairly badly by the more traditional combat and combat hemo builds; however, it is hard to say for certain as Cold Blood is a hard ability to model from a theoretical perspective.

Offline
Old 11/19/07, 5:09 PM   #1161
Ariashley
Von Kaiser
 
Blood Elf Paladin
 
Dunemaul
For some reason - this posted twice...

Offline
Old 11/19/07, 5:11 PM   #1162
Ariashley
Von Kaiser
 
Blood Elf Paladin
 
Dunemaul
Got a chance to do some live testing of Hemo uptime against Void Reaver over the weekend. The lack of improved eviscerate in that fight was probably not my favorite feature of 11/27/23

WoW Web Stats

I used an Excel spreadsheet to calculate an average length of hemo charges (based on 5163 physical attacks against VR over 504.959 seconds = 10.22 attacks per second = 0.978034 sec average uptime on hemo charges).

Edit: I filtered a previously untouched combat log for that fight (the log I had been using, for whatever reason, had filtered all the "fades from" messages) to determine hemo uptime of hemo. The actual uptime during the fight was 106.985 seconds for a total of 21.18%. Average time that hemo was up was 0.922 sec, the fastest consumed (likely during a bloodlust) 0.282 sec, and the longest to be consumed was 3.156 sec. The standard deviation of uptime was 0.421 sec. I'm not a statistician so I'm pretty sure I would mess up a calculation of a confidence interval about uptime being around 21.18% for one hemo rogue in 25-man raid. Not to mention that 116 hemorrhage debuffs is not a decent statistical sample.

I have a mod (Dotimers) that has a bar that turns blow when the charges on Hemo run out. Took me a little while to notice that it was working properly (and the bar would disappear when the charges were used, not just based on the 15 sec timer).

It was clear that for this particular fight, my damage and DPS were FAR inferior (about 30% lower or more) to the last time we killed Void Reaver. I've been working on comparing it to the last Void Reaver killing that I was in (where I was #1 on damage overall). There was an awful lot different about the 2 fights. As I filter out that noise and get more data from other fights, I'll be able to comment on how 11/27/23 compares with 19/42/0 in a variety of different situations. The coming week will probably not be a good data gathering week - given that it's Thanksgiving week and I don't expect we will be on our normal raid schedule.

Last edited by Ariashley : 11/20/07 at 3:20 AM. Reason: Better Information Obtained

Offline
Old 11/19/07, 5:11 PM   #1163
Left
Don Flamenco
 
Left's Avatar
 
Draenei Paladin
 
Darkspear
Originally Posted by Anked View Post
As most love to say, "look at the spread sheet" whenever anyone asks a gear/spec question; I decided to do just that.

...

I think I'm going to respec to 31/0/30 then next time I log in. You get a lot more utility, and even with my current gear (upgrades pending) it's a dps gain on the spreadsheet. As far as utility goes, I would have CB and Prep at the small cost of BF. It looks like 31/0/30 would be a very viable PvE and PvP spec to boot.

Has anyone done any in game testing already to prove/disprove this observation?
There has been a significant amount of discussion about the difficulty of modeling Seal Fate and the fact that a 31/0/30 build has never really competed before. However, only one person (that I've seen) has tried it. They found it hard to play (hard to settle into a cycle for), and thus experienced a DPS loss when compared to combat.

Interestingly, the spreadsheet shows a 1s/5r "cycle", but the problem with Seal Fate is that the procs are random enough that cycles aren't really possible. I would find it almost better to try for a 3-5s/5r/3-5e type cycle, where you basically dump your additional combo points into an eviscerate at the end of each cycle prior to renewing SnD, based on how the crits are proccing. It strikes me as being very hard to play, although it would be very fun to try as well. I say go for it and tell us what you think.

Offline
Old 11/19/07, 5:15 PM   #1164
Hanos
Back in my day...
 
Hanos's Avatar
 
Human Rogue
 
Sen'jin
Originally Posted by Anked View Post
I think I'm going to respec to 31/0/30 then next time I log in. You get a lot more utility, and even with my current gear (upgrades pending) it's a dps gain on the spreadsheet. As far as utility goes, I would have CB and Prep at the small cost of BF. It looks like 31/0/30 would be a very viable PvE and PvP spec to boot.

Has anyone done any in game testing already to prove/disprove this observation?
For the 100th time, 31/0/30 in the spreadsheet is grossly misleading and the DPS is very likely wrong. Seal Fate is not properly modeled and the energy bar is not sufficiently large to make it function properly. The spec is far too dependent on crit'ing or non-crit'ing, and even with the Ashtongue Trinket it is generally considered to be not viable for raiding (Mutilate is more viable due to the fact the each mutilate has 2 chances to crit), with something absurd like 50% crit it might become more realistic, but that isn't a gear level you have access to.

To address some of your other comments:
-Yes Hit is more important in Combat and less important in Hemo (especially if you go 4/5 Sword Spec)
-Agi is essentially equivalent to hit in Hemo
-Crit is still over priced and you should never get for it

Personally I have been running with my Pre-2.3 308 Hit Rating with Precision and WE, and popping +20 Hit food for bosses, and I have liked the results so far, I would say my personal DPS is very similar to Combat with my total DPS contribution being a reasonable amount higher.

Offline
Old 11/19/07, 5:25 PM   #1165
ekval
Piston Honda
 
Night Elf Rogue
 
Stormreaver (EU)
Originally Posted by Left View Post
There has been a significant amount of discussion about the difficulty of modeling Seal Fate and the fact that a 31/0/30 build has never really competed before. However, only one person (that I've seen) has tried it. They found it hard to play (hard to settle into a cycle for), and thus experienced a DPS loss when compared to combat.

Interestingly, the spreadsheet shows a 1s/5r "cycle", but the problem with Seal Fate is that the procs are random enough that cycles aren't really possible. I would find it almost better to try for a 3-5s/5r/3-5e type cycle, where you basically dump your additional combo points into an eviscerate at the end of each cycle prior to renewing SnD, based on how the crits are proccing. It strikes me as being very hard to play, although it would be very fun to try as well. I say go for it and tell us what you think.
5s/5r cant keep SnD up atleast. That spec has potential with specific gear, most of I think it needs some decent cycle built for it. Decent cycle or lack of good 3rd PvE finisher probably fails this spec.

Edit: huge missrate is kinda problematic also, when it starts to affect yellow damage its quite bad, Not sure if some expertise from gear would solve that thought.

Its nice spec to play with, 15% movement speed, 2,5-3,0k hemos etc.

Last edited by ekval : 11/19/07 at 5:59 PM.

Offline
Old 11/19/07, 5:25 PM   #1166
Anked
Von Kaiser
 
Gnome Rogue
 
<Kin>
Azuremyst
Originally Posted by Hanos View Post
For the 100th time, 31/0/30 in the spreadsheet is grossly misleading and the DPS is very likely wrong. Seal Fate is not properly modeled and the energy bar is not sufficiently large to make it function properly. The spec is far too dependent on crit'ing or non-crit'ing, and even with the Ashtongue Trinket it is generally considered to be not viable for raiding (Mutilate is more viable due to the fact the each mutilate has 2 chances to crit), with something absurd like 50% crit it might become more realistic, but that isn't a gear level you have access to.

To address some of your other comments:
-Yes Hit is more important in Combat and less important in Hemo (especially if you go 4/5 Sword Spec)
-Agi is essentially equivalent to hit in Hemo
-Crit is still over priced and you should never get for it

Personally I have been running with my Pre-2.3 308 Hit Rating with Precision and WE, and popping +20 Hit food for bosses, and I have liked the results so far, I would say my personal DPS is very similar to Combat with my total DPS contribution being a reasonable amount higher.
Noted on the bit about SF and the spreadsheet. I think I'll give it a test anyway and see how it feels. I've done a couple hydros runs as 14/42/0 and 11/27/23, so it would be easy to notice a difference to 31/0/30.

Anyway, considering the other bit about agi vs hit for combat/hemo... Have you tried plugging in agi gems rather then hit on the spreadsheet? If my observations are correct you should see some nice dps boosts. I'm curious how it would change things for one that has access to gear above my own ability.

The real kicker is I'm not seeing that combat swords is better in the least then a hemo hybrid build for personal dps. Is there another glitch in the spreadsheet I don't know about?

Offline
Old 11/19/07, 6:03 PM   #1167
Hanos
Back in my day...
 
Hanos's Avatar
 
Human Rogue
 
Sen'jin
Originally Posted by Anked View Post
Noted on the bit about SF and the spreadsheet. I think I'll give it a test anyway and see how it feels. I've done a couple hydros runs as 14/42/0 and 11/27/23, so it would be easy to notice a difference to 31/0/30.

Anyway, considering the other bit about agi vs hit for combat/hemo... Have you tried plugging in agi gems rather then hit on the spreadsheet? If my observations are correct you should see some nice dps boosts. I'm curious how it would change things for one that has access to gear above my own ability.

Hydross is a horrible test mob for many reasons: bleed immune, variable BF damage depending on when you pop it, time on target based on transitions etc, poison immune in the nature phase.

What were you doing raiding 14/42/0? Just couldn't decide what to spend those last 5 points on?

As far as gemming, I normally work with what I label an "ideal" version of the spreadsheet which contains my future gear plans, so that I know where I want to end up, and I kind of wing it in the interm. In this scenario I lose 0.74 DPS for every Glinting Pyrestone I swap out for a Delicate Crimson Spinel, and gain 0.74 DPS for every Glinting Pyrestone I replace with a Ridgid Lionseye, however this is a set that includes about as much -armor as you can get, which puts a higher value on white damage, so the numbers will shift some, however taking armor penetration from gear completely out of the equation, I still see a 0.55 DPS loss going from 5 Hit/5 Agi to 10 Agi. So no you will not want to stack Agi at the expense of hit at least with BT level gear.

However that assumes every buff and debuff up at all times for hit to be better then agility, as you lose buffs/debuffs the margin gets closer.

The real kicker is I'm not seeing that combat swords is better in the least then a hemo hybrid build for personal dps. Is there another glitch in the spreadsheet I don't know about?
Do you have "Include Hemo Debuff DPS Estimate?" on the Talents tab turned on? Or Hemorrhage under the Boss Buffs in the Gear_Buffs Tab?

Offline
Old 11/19/07, 6:07 PM   #1168
Abaxial
Piston Honda
 
Abaxial
Gnome Rogue
 
No WoW Account
Last week I tried out 0/31/30 using a 3s/5s/5r cycle and dps wise I was impressed that it was able to keep up with my numbers for combat. Note, I forgot to extend my combat log after formatting the weekend prior so dps time will seem off but dps and damage done should not be effected by this. I ran out when Teron was at 27% and had not used my second AR.

0/31/30 Talent Calculator - World of Warcraft
Loading...

I appear to not have any wws of combat where I had similar buffs and did not get shadows early(most recent one seems to have expired). But under similar conditions I do stay within the 1900-1950 area as combat swords.


As a side note, from some short testing with a shaman I have noticed that at least earthshock and lightning bolt do in fact remove a charge of hemo and gain no benefit. I don't tend to have a vast amount of time to find someone from different classes to test this individually but figured someone else here might have that spare time.


Offline
Old 11/19/07, 6:44 PM   #1169
Anked
Von Kaiser
 
Gnome Rogue
 
<Kin>
Azuremyst
Originally Posted by Hanos View Post
Hydross is a horrible test mob for many reasons: bleed immune, variable BF damage depending on when you pop it, time on target based on transitions etc, poison immune in the nature phase.

What were you doing raiding 14/42/0? Just couldn't decide what to spend those last 5 points on?

As far as gemming, I normally work with what I label an "ideal" version of the spreadsheet which contains my future gear plans, so that I know where I want to end up, and I kind of wing it in the interm. In this scenario I lose 0.74 DPS for every Glinting Pyrestone I swap out for a Delicate Crimson Spinel, and gain 0.74 DPS for every Glinting Pyrestone I replace with a Ridgid Lionseye, however this is a set that includes about as much -armor as you can get, which puts a higher value on white damage, so the numbers will shift some, however taking armor penetration from gear completely out of the equation, I still see a 0.55 DPS loss going from 5 Hit/5 Agi to 10 Agi. So no you will not want to stack Agi at the expense of hit at least with BT level gear.

However that assumes every buff and debuff up at all times for hit to be better then agility, as you lose buffs/debuffs the margin gets closer.



Do you have "Include Hemo Debuff DPS Estimate?" on the Talents tab turned on? Or Hemorrhage under the Boss Buffs in the Gear_Buffs Tab?
Sorry that was a typo, should have been 19/42/0... i've been messing around with a lot of specs today.

Anyway, I did have the include hemo debuff estimate turned on but not the hemorrhage under the boss buffs. However, I had these same setting for everything i compared today. Even after correcting this 11/27/23 is superior to 19/42/0 according to the spreadsheet. I even went back and decked out the char with full T6 and BT gear and gems with both Warglaives. 11/27/23 still comes out on top! And that's with hit gear/gems.

If that's true no one should be noticing a decline in their dps once they get the playstyle down, and that would mean 11/27/23 is the new top raiding dps spec. Somehow I find that hard to believe, and I'm thinking the spreadsheet isn't taking everything into account.

Offline
Old 11/19/07, 6:58 PM   #1170
Hanos
Back in my day...
 
Hanos's Avatar
 
Human Rogue
 
Sen'jin
Originally Posted by Anked View Post
Sorry that was a typo, should have been 19/42/0... i've been messing around with a lot of specs today.

Anyway, I did have the include hemo debuff estimate turned on but not the hemorrhage under the boss buffs. However, I had these same setting for everything i compared today. Even after correcting this 11/27/23 is superior to 19/42/0 according to the spreadsheet. I even went back and decked out the char with full T6 and BT gear and gems with both Warglaives. 11/27/23 still comes out on top! And that's with hit gear/gems.

If that's true no one should be noticing a decline in their dps once they get the playstyle down, and that would mean 11/27/23 is the new top raiding dps spec. Somehow I find that hard to believe, and I'm thinking the spreadsheet isn't taking everything into account.
Hemo is going to scale better at the high end due to the better modifiers, the question is where is the point of inflection (hence the thread). From what Killars and I have seen with a reasonable amount of BT Gear we are coming about break even with Hemo, ahead on some fights and behind on others due to mechanics, My guess is you will do as well or better if you are in Hyjal BT, and might take a personal DPS hit in T4-5 Content. Gemming for hit is going to be better at the high end due to White Damage, especially if you stack armor penetration, which I have already seen compared to the non-hemo rogues in my guild.

Don't assume the spreadsheet is 100% accurate, however based on what we have seen so far it appears to be very close to accurate.

Offline
Old 11/19/07, 7:02 PM   #1171
Balmong
Glass Joe
 
Night Elf Rogue
 
Silver Hand
Originally Posted by Anked View Post
I think I'm going to respec to 31/0/30 then next time I log in. You get a lot more utility, and even with my current gear (upgrades pending) it's a dps gain on the spreadsheet. As far as utility goes, I would have CB and Prep at the small cost of BF. It looks like 31/0/30 would be a very viable PvE and PvP spec to boot.

Has anyone done any in game testing already to prove/disprove this observation?
I have done 1 SSC run through Leo with 31/0/30. Playing the actual build was a difficult shift in style. I finally set myself on 5s/4-5r.

WWS - Loading...

We don't usualy run 5 rogues in the raid, but a few people had RL priorities. Somatra is our 11/27/23, Leiji and Sintaro run 41/20/0 and Fraseir (our least raid experienced rogue) is running a 20/41/0 (or some variant of it). So compare and contrast if you want to.

I am not really convinced about this build yet, so build your own opinion on WWS and trying it out yourself.

Edit* for stupid errors

Offline
Old 11/19/07, 8:27 PM   #1172
Leoki
Glass Joe
 
Night Elf Rogue
 
Darksorrow (EU)
Since live i have been testing some hemo builds. I dont have any data but want to share what i found.

With 11/20/30 (i dont have swords since i was mutilate) i fall slightly behind my normal position with the rogues on teron but only slightly. I think it is alot more suitable for a sword rogue to take the 11/2x/2x builds.

So lastnight i decided to try 30/0/31. A build which i had no faith in to do well. However i didnt lag behind in dps as i thought i would. It was an illidan fight so it isnt the best test, but i actually beat the other rogues and was beating them for the whole fight. We had two combat sword rogues one 11/27/23 and me with an enhancement shaman.

Now i am not saying this build will beat the others or that i think it should normaly. But i am now intrested to find out if it will work and am going to give it a chance on our next teron encounter and will try to get some data. My crit chance isnt exceptionaly high but i did use the BT rep trinket. With the cycles it was hard to not waste any energy or uptime but nothing i am not used to with playing mutilate. if you get too meny combo points just wait and let your energy fill up while the last 6-8 secs of S+D run out and then spam again and you dont waste any uptime.

p.s. I use syphon mainhand and boundless offhand

Last edited by Leoki : 11/19/07 at 8:39 PM.

Offline
Old 11/19/07, 9:59 PM   #1173
Anked
Von Kaiser
 
Gnome Rogue
 
<Kin>
Azuremyst
Originally Posted by Hanos View Post
Hemo is going to scale better at the high end due to the better modifiers, the question is where is the point of inflection (hence the thread). From what Killars and I have seen with a reasonable amount of BT Gear we are coming about break even with Hemo, ahead on some fights and behind on others due to mechanics, My guess is you will do as well or better if you are in Hyjal BT, and might take a personal DPS hit in T4-5 Content. Gemming for hit is going to be better at the high end due to White Damage, especially if you stack armor penetration, which I have already seen compared to the non-hemo rogues in my guild.

Don't assume the spreadsheet is 100% accurate, however based on what we have seen so far it appears to be very close to accurate.
If my observation of how the spreadsheet says things should be, even at my current gear, 11/27/23 is better then 19/42/0. According to my gearing choices from my current gear, to my gear plus the upgrade we covered on page 40, and then all the way up to full T6 filling the rest out with BT gear w/ both Warglaives. No matter what 11/27/23 always came out 1.5%+ better.

I agree no spreadsheet can rule exactly how things will go, but as you stated it has been proven to be a good baseline... well besides the whole 31/0/30 thing =) So, unless there is some other unforseen issue with the spreadsheet it would seem that starting as early as Kara (perhaps sooner, as hemo isn't dependent on hit) 11/27/23 Combat Hemo > 19/42/0 Combat Swords.

Offline
Old 11/20/07, 2:34 AM   #1174
ekval
Piston Honda
 
Night Elf Rogue
 
Stormreaver (EU)
Edit: double.

Last edited by ekval : 11/20/07 at 2:40 AM.

Offline
Old 11/20/07, 2:39 AM   #1175
ekval
Piston Honda
 
Night Elf Rogue
 
Stormreaver (EU)
Originally Posted by Leoki View Post
So lastnight i decided to try 30/0/31. A build which i had no faith in to do well. However i didnt lag behind in dps as i thought i would. It was an illidan fight so it isnt the best test, but i actually beat the other rogues and was beating them for the whole fight. We had two combat sword rogues one 11/27/23 and me with an enhancement shaman.

Now i am not saying this build will beat the others or that i think it should normaly. But i am now intrested to find out if it will work and am going to give it a chance on our next teron encounter and will try to get some data. My crit chance isnt exceptionaly high but i did use the BT rep trinket. With the cycles it was hard to not waste any energy or uptime but nothing i am not used to with playing mutilate. if you get too meny combo points just wait and let your energy fill up while the last 6-8 secs of S+D run out and then spam again and you dont waste any uptime.

p.s. I use syphon mainhand and boundless offhand
Please enlighten me did you outgear other rogues badly or how did you make 30/0/31 spec worth of it? I've tried similar spec in Illidan also and was blown away by combat sword rogues by about 40-80k (2-5%) only in Phase1+2 combined. I am probably missing some key things with that spec, what cycles/finishers did you use? The damage just felt kinda weak when I tried it, even with BT trinket.

Edit: Forgot that mine Illidan testing was prepatch in 2.2, could've things changed that much? That probably means I need to give this spec another go. Any discussion about Vigor vs Premed vs some other talent for last talent point for PvE? Premed didn't feel useful at all. Last time I also used [Rod of the Sun King]+[Messenger of Fate] combo, would [Syphon of the Nathrezim]+[Boundless Agony] be better when you can't combine Rod with combat potency? Don't think Rod/Syphon proc mean anything with this type spec due so high missrate so it's up to stats. On the other hand 3x/0/3x doesn't seem to benefit from armor penetration on Boundless Agony at all because of low hitrate and no DWSpec. Or which weapons would benefit this spec most?

Last edited by ekval : 11/20/07 at 3:03 AM.

Offline
 

Go Back   Elitist Jerks » Class Mechanics

Thread Tools

Similar Threads
Thread Thread Starter Forum Replies Last Post
[Rogue] Combat Daggers Question Mayor Class Mechanics 2 06/18/07 10:34 AM
Rogue - Dodge vs. Parry Talents, One Roll Combat Theory, Combat Sword Spec Questions tok3n Class Mechanics 30 04/12/07 2:15 PM
Pre-raid Hemo Rogue TBC gear Zoro Public Discussion 4 01/11/07 2:06 AM
Ambush-Hemo: Points in combat or Assassination Hass Public Discussion 6 09/13/06 8:42 PM
Replacing a combat rogue with a hemo rogue, result? darun Public Discussion 14 09/12/06 1:35 AM