Jump to content


Photo

Which damage meter?


  • Please log in to reply
34 replies to this topic

#21 CasT

CasT

    Piston Honda

  • Members
  • 204 posts

Posted 30 March 2007 - 11:30 AM

I often hear people lament when confronted with their (so-so) dps after a raid that 'those meters are inaccurate'. So I made it my mission to have the best dmg meter out there to catch out the slackers ;-)

So what is the accepted 'gold-standard' dps meter? I use Recap atm, but is there something better/more accurate?


It's very obvious to me that damagemeters all in all are slightly inaccurate. This is based on one single AV game with recap/ damagemeter and SW running all three gave different results and none where the same as the score board of the game.

Closest was SW stats.


But do remember that there is much more to this game than beeing to notch DPS. I firmly belive the pushing to much attention to the Damage meter will end up in ppl pulling aggro and forgetting their shackel/ sheep WE just to be no.1 healer/dpser


To really say who is slacking and who is not you need more to look at what abilities they use and how they react to the dynamics of the game. At least thats how we do it in PVP. Put a warlock in WSG midfield and he will easily win damage done. Try to compare him to a hunter who's guarding in the flag. and only will fight the occasional rogue/ druid who'll get by.

Summary: I value more the ability to act/react than Gear and 1337-DPS
Do not matter how much you play, you will never get the carrot.

#22 Quasar

Quasar

    Don Flamenco

  • Members
  • 445 posts

Posted 30 March 2007 - 03:50 PM

If you're looking to catch slackers, it doesn't really matter what damage meter you use. Either way, the players with the lowest damage/healing done are the slackers.


Not necessarily, but in a general sense yes.

Personally I use Recap, having tried SW Stats and DamageMeters. I like Recap's UI by far, and I like how it breaks down damage information with all the little details (averages, crits, % breakdowns by spells, etc).
JUICE! Aww I'm sorry. Did... did anyone want some juice?

#23 Pane

Pane

    Piston Honda

  • Members
  • 161 posts

Posted 30 March 2007 - 09:06 PM

But do remember that there is much more to this game than beeing to notch DPS. I firmly belive the pushing to much attention to the Damage meter will end up in ppl pulling aggro and forgetting their shackel/ sheep WE just to be no.1 healer/dpser


Well obviously. But at the same time, with smaller raids, people have to pull their weight. You can't drag along a 300 dps rogue anymore, or something like that. Pulling agro or forgetting crowd control are obvious no-goes, and infinitely more visible.

#24 CasT

CasT

    Piston Honda

  • Members
  • 204 posts

Posted 02 April 2007 - 12:26 PM

Well obviously. But at the same time, with smaller raids, people have to pull their weight. You can't drag along a 300 dps rogue anymore, or something like that. Pulling agro or forgetting crowd control are obvious no-goes, and infinitely more visible.



If you ar hunting 300 dps rogues they can not blame it on a damagemeter, well maybe if the poster is ranged and player is melee, but syncronization changed all that and the in accuracy is more down to the %.

But do keep in mind that healing is now not going to be usefull since both Prayer of Mending and Lifebloom counts as the targets heal and not the caster.
Do not matter how much you play, you will never get the carrot.

#25 Christmas

Christmas

    Piston Honda

  • Members
  • 112 posts

Posted 05 April 2007 - 01:48 PM

I use Recap for infight things and WoW Web Stats for post-raid analysis. Lossendil's work on the WWS server is absolutely fantastic.

#26 Durza

Durza

    Von Kaiser

  • Members
  • 44 posts

Posted 05 April 2007 - 01:55 PM

Have to agree with Christmas, absolutely in love with WWS and the new beta server since last week. Great job Lossendil!

#27 Elsia

Elsia

    Banned

  • Banned
  • 311 posts

Posted 11 April 2007 - 10:12 AM

I know this is probably too hard if not in many cases impossible, but having a damage meter that actually amortizes utility would be neat. What I mean by that is that if one class provides a damage buff to another class, the damage increase is attributed to the buffing class, not the damage dealing class.

For example some buff gives you 3% extra damage. That 3% should be attributed to the buffer, not the damage dealer.

This way one could see more clearly synergies and effects of group setup and overall raid composition.

I can see that it's hard to give actual utility for percentage chance increases, as it's chance, so one can't just factor out 3% crit increase trivially though doing so would be very good ballpark. Also for some abilities I'm not sure if it's detectable who procced a certain buff (say shadow weaving or expose weakness when multiple shadow priests or surv hunters are present).

Anyway, this would give a clearer picture of class and raid composition contribution to raid DPS than individual DPS as we have it now... seeing both would be very helpful I think.

#28 aya

aya

    Von Kaiser

  • Members
  • 67 posts

Posted 11 April 2007 - 10:22 AM

For example some buff gives you 3% extra damage. That 3% should be attributed to the buffer, not the damage dealer.

This way one could see more clearly synergies and effects of group setup and overall raid composition.


While I can see occasional use of such tables every now and then, I think it's called synenergy because it takes two to dance tango. A part of a given rogus DPS might be derived from a warriors battleshout, that doesn't mean the battleshout alone would do the DPS even if the rogue wasn't there.

Taking the idea further, I think it would be interesting to see statistics which only showed DPS from buffs!
"Windfury did this and this much for our group in this fight",
"This and this much was gained from keeping TSA on", and so on.

#29 Elsia

Elsia

    Banned

  • Banned
  • 311 posts

Posted 11 April 2007 - 10:37 AM

Well that's pretty much what I'm saying.

I think it'd be very valuable to learn where the DPS levels come from, because as you say it takes two to dance. Current damage meters show only one.

The reason why individualized reattribution would be neat is because you'd learn more about group composition. So shammy X was in group 2 and shammy Y was in group 3. Damage contributions were different and you can look at the group makeup to learn a whole lot from it.

I think about this like I think about recap, an analytical tool to understand the situation better. For a hunter at least cast bars are more important to gauge and improve your DPS performance than a damage meter I found. But I know that damage meters have a pride factor, but that doesn't have much to do with understanding the situation of the fight.

#30 kaib

kaib

    Don Flamenco

  • Members
  • 349 posts

Posted 11 April 2007 - 10:43 AM

SWstats is just a huge memory leak. The functions are awesome, but whenever using it, my memory used goes from 35 to about 40-42 and increases more and more with that 'saving old data' function. I like it better then good old damagemeters, but as most instances are already eating my fps alive these days, I can't really justify using it.

#31 Dirich

Dirich

    Piston Honda

  • Members
  • 123 posts

Posted 11 April 2007 - 11:05 AM

If you're looking to catch slackers, it doesn't really matter what damage meter you use. Either way, the players with the lowest damage/healing done are the slackers.


That's the more noobish thing it could possibly be said about damage/healing done.
Ok, maybe dps-ers with CC don't have to fall too much in the ladder cuz of the time "lost" on CCs, but damage output need to be considered togheter with the class max possible dps. I mean, i am shadowpriest and i usually do great in 5 man teams (usually 2nd in damage), but i presume in 10/25 man raid shadowpriest dps is not going to be one of the best in standard fights (with multiple non-aoeable mobs i think we can do great btw).
Also i would like to let u notice that if any "backup healer" switch to dps to healing mode on just a boss he will surely lose A LOT of positions in the DPS' ladder.

The worst part of ur sentence is the one concerning the healing done... i've raided up to AQ40 pre tbc, always as healing priest. Most of the time i was on top of the healing meter, but that did not mean anything. Considering hots canceling each other, overheal, main targets (if u r not random healing), etc. healing meter is totaly worthless (except for really evident slackers... but if u r not random healing than u'll spot them anyway cuz their targets will die really often if no one helps.
Considering an "healing ladder" to find who's good and who's not will only lead to wasting mana.

P.S.
There will always be a "last in damage/healing" cuz the number of dps-ers/healers is finite. Last don't need to be equal to slacker.

#32 aya

aya

    Von Kaiser

  • Members
  • 67 posts

Posted 11 April 2007 - 11:18 AM

Perhaps Renaldo worded it out a bit wrong (or simply forgot to add "pontial" on front of "slackers"), but I think you're somewhat missing his point. I think he's saying that the difference between various damagemeters (or combat loggers, if you will) is minimal for most of the users.

Ofcourse in many cases something important might not be presented fully in the statistic, and for those cases it's important for you to understand the encounter and the role each person is playing there. Like you said it's quite common for someone to be simply channeling and cancelling heals to quickly react to burst in that particular fight. However I hardly see a problem asking "Hey, McHealington, howcome you're always so low on effective healing" and then hearing "Well, I need to keep on channeling highrank Gheal here because occasionally the boss bursts tank down really low, and those are the REALLY important moments on this fight, so having me cover them is vital".

It's not "noobish" to read data from a fight. It's noobish to be ignorant slave to it perhaps, but surely we can respect eachother enough to not suspect such things of eachother here?

#33 Valoran

Valoran

    Piston Honda

  • Members
  • 248 posts

Posted 11 April 2007 - 11:26 AM

Ignoring the value of tracking statistics because they may be misinterprated is the real problem with how people view damage tracking addons. The "winners" of the damage meters are the people who have used the data gathered on a boss encounter to better understand what seperates the people at the top from the people at the bottom and how to improve either their own or the raids output, be it through group synergy, consumables, or innate class characteristics.

#34 TheOnly

TheOnly

    Don Flamenco

  • Members
  • 302 posts

Posted 16 April 2007 - 05:29 PM

My combat log settings are at 2000 yards (I think DamageMeters did this, if I set it to 200 it gets put back to 2000). But all my meters still fail to get data from about 2/3 of the players I encounter. Its been like this since WoW 2.0. I can be standing on top of a mage and near a target, and get absolutely zero data bout them, yet still get all the data from the tank. I get data from almost all druids and priests, and almost never any at all from mages, hunters, and rogues.

Has anyone else seen this?

This happens with both SWStats and DamageMeters (I have not tried recap).
SWStats is nice, but does not show fight-by-fight dps in real-time which is critical for assessing my own performance on the fly. Well it can for raid-wide dps but who cares, I need per person dps per combat interval.

#35 Northerner

Northerner

    Great Tiger

  • Members
  • 825 posts

Posted 16 April 2007 - 11:03 PM

I've thought about trying to work out a marginal value system for players (or "contribution" or "attributable damage" or whatever you like to call it) but the logic seems to be fairly complex. This obviously wouldn't be for the catching slackers type of situation but more for the "what happened when we added a shaman to group 2?" type of scenario. The reverse of the process (What was Rogue_003's un-modded dps?) would however be extremely useful for comparative evaluation over the present system of "well, hunter_001 had a spriest and a fury warrior so sure he's ahead of hunter_002 who is in group five spamming misdirection".

I like post-fight analysis tools and the more information the better. Even a barebones framework that posted buffs with uptime, non-dps activities and so on would be quite nice.

EDIT: Obviously this goes beyond the information available from the combatlog alone, which is a large part of why I abandoned the concept early on. Another issue that seems to crop up more for me lately though is lines of information that are just plain MIA from the combatlog. It seems that there is some throttling going on which causes some rare dataloss and while that's no huge issue statistically over a long fight, it plays absolute havoc with conditional logic for events.




0 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users