|
|||||||
| Notices |
 |
| Thread Tools |
02-16-2006, 11:03 AM
|
#1 |
|
General
Join Date: Nov 2004
Posts: 415
|
Mathematical model of ranger damage before and after proposed proc changesSYNOPSIS:There are changes on the beta server that greatly reduce the procing oportunities available to rangers. While ranger damage output could certainly use some tuning, I believe that these changes will have a far more damaging effect than intended. I constructed a spreadsheet to model the (rather complex) damage output done by rangers in the current system and the proposed new system being tested. This post is a discussion of my findings and methodology.The spreadsheet containing my data model is here:http://home.comcast.net/~shader_eq2/ranger_damage_model.xlsMy model shows that rangers after the changes will be doing
EDIT 2006-02-16:Changed the following on the spreadsheet:+ Fixed problem where double and triple fire hits were always using the first hit damage, instead of the progressively larger values. (Noted by Calaglin/Pinski.)+ Fixed a proc rate typo. I had had Gleaming Strike set to a nominal (examine-window) proc rate of 50%, rather than 5%.+ Added a new tab, "Beta Changes - Feb 16" based on new information. It assumes that Quick Shot is capable of producing all of the other proc types, but double and triple shot arts don't proc Quick Shot or anything else on the 2nd or 3rd hit. (Noted by Calaglin/Pinski.)Recomputed the table shown below with the new changes. My spreadsheet is based around the idea of calculating the "expected damage" of any given attack or combat art. That is, for every possible combination of some damage event (proc, regular hit, or quick strike), the damage that a proc would do is multiplied by the probability of it happening. Then, all of those weighted damages are summed up to determine the expected damage of the attack as a whole.For instance, right now the third hit from our Triple Fire line of attacks can trigger a Quick Shot (an extra attack caused by our offensive stance), and that Quick Shot can trigger a poison. In order for that particular damage event to happen, all three triple fire shots have to hit (if one misses, the next ones don't have any chance), the Quick Shot has to proc, and the poison has to proc. So the expected damage of just that one part - poison proc from a Quick Shot on the third hit of a Triple Fire - is the damage it would do multiplied by the probability that all those prerequisits happened.It's not a perfect model. It makes some assumptions, and it's based off of the data I used. (Which was from my own examine windows.) The model assumes you're using poison, offensive stance, an imbued bow, and have 1 other type of proc. (For my data, the extra proc was the paladin buff Call of Glory.)Here are the numbers I came up with. (But if you want, download the spreadsheet and plug in your own.) All data below is what I use as a level 47 ranger with a legendary bow and using legendary DD poisons.  . + Stop and use Sharp Shot Message Edited by Bayler_xev on 02-16-200609:33 PM
__________________
Lokimor Human Coercer / Sage Venekor |
|
|
|
02-16-2006, 02:48 PM
|
#2 |
|
Loremaster
Join Date: Jan 2005
Posts: 474
|
These results are stunning. No pun intended. |
|
|
|
|
02-16-2006, 03:45 PM
|
#3 |
|
Loremaster
Join Date: Feb 2005
Posts: 308
|
Hmmm this data seems weird to me for triple fire. Even without the proccing it would sitll do around 1k+ damage. Just from reading what the spell does. I think you are putting the data on beta a bit low.
__________________
|
|
|
|
|
02-16-2006, 04:12 PM
|
#4 |
|
Join Date: Nov 2004
Posts: 56
|
Wow...All I can say is wow...thats REALLY [expletive haxx0red by Raijinn], and kinda insulting to the devs; as if they don't know any of this. Wow...I'm actually kinda interested in it, I've never seen anyone be really THIS [expletive haxx0red by Raijinn] over a game.
|
|
|
|
|
02-16-2006, 04:37 PM
|
#5 |
|
Join Date: Apr 2005
Posts: 1,332
|
This sort of model is entirely worthless.Until you actually play test it, your data is nothing but blind assumptions.
|
|
|
|
|
02-16-2006, 04:38 PM
|
#6 |
|
Loremaster
Join Date: Dec 2004
Posts: 473
|
This is almost exactly what soloing is like for an averagely-geared assassin at any level. And believe me it's not a whole lot easier with above-average gear either unless you're fighting green solo cons. |
|
|
|
|
02-16-2006, 04:39 PM
|
#7 |
|
General
Join Date: Feb 2005
Posts: 573
|
Mathematical models are nice to play with, but yours doesnt seem to reflect real number that are happening in practise. maybe your model is missing some parameter or calculating wrong or maybe some stuff simply isnt doing what we suppose it to do in theory. My model shows that rangers after the changes will be doing 45% of their current damage. This indicates that ranger would be going to lose 55% of their damage. Assuming that even after the change ranger will still do some proc damage, just to a lesser amount, lets say 1/5 of what they are procing now, this would mean that currently 70% of ranger damage comes alone from poison and procs if your mathematical model is correct. Which is not the case. Your model is flawed. |
|
|
|
|
02-16-2006, 06:33 PM
|
#8 |
|
Loremaster
Join Date: Nov 2004
Posts: 40
|
>...this would mean that currently 70% of ranger damage comes alone from poison and procs if your >mathematical model is correct. Which is not the case.Ehm.... it simply .... *IS* the case.Well you might discuss if it is 50% or 70%. But its within that area. Of course, if you have Master1 on all your combat arts and do not use a legendary high DD - low DoT poison it might be lower.Greetings,tom1301
|
|
|
|
|
02-16-2006, 06:52 PM
|
#9 |
|
Loremaster
Join Date: Nov 2004
Posts: 294
|
Mind posting some logs or something? I sincerely doubt that 50% of your damage comes from procs. Also, as BG (?) said, in the beta, Rangers were doing 4000 DPS. With this, it's going to shorten it to where it should be.
__________________
 If they ever come up with a swashbuckling school, I think one of the courses should be laughing, then jumping off something.- Jack Handey |
|
|
|
|
02-16-2006, 07:02 PM
|
#10 |
|
Loremaster
Join Date: Dec 2004
Posts: 618
|
I'd really be surprised if the changes affect DPS to that extend. I would imagine they have been testing it. Rangers are still in their Teir 1 model, athough I don't know who to model a Tier 1 after.
|
|
|
|
|
02-16-2006, 07:20 PM
|
#11 |
|
Loremaster
Join Date: Nov 2004
Posts: 134
|
Im Raid leader of my guild, and i parse 99% of them. so far i can assure you that ranger dps comes in a 50% from proccs, 35% of poison and the rest of self buff and other proccs like weapon proccs, shield, etc. Im at work and cant post it but im talking about raids, godking, Goaa, Coaa etc... i will try to find some data later on. |
|
|
|
|
02-16-2006, 07:33 PM
|
#12 |
|
Loremaster
Join Date: Nov 2004
Posts: 152
|
I have played a ranger from day one and that sounds about right Edit: Also, there have been times when it has just been too late or too early and I forget to use poisons(I know, stupid me) but it doesnt make all that big of a difference(maybe take time and a quarter more to kill something), and I buy the best poisons on the market for my level range. Message Edited by DarkLegacy2005 on 02-16-200608:35 AM
__________________
|
|
|
|
|
02-16-2006, 07:37 PM
|
#13 |
|
Join Date: Apr 2005
Posts: 1,332
|
But of course, you're not losing all the procs, so it's not going to cut your DPS in half. You're only going to lose SOME of the procs, so it's still a much smaller nerf then you're making it out to be.
|
|
|
|
|
02-16-2006, 07:42 PM
|
#14 |
|
Loremaster
Join Date: Jun 2005
Posts: 173
|
You have no ideah what you are talking about. Play a ranger then come talk about how this effects your class. |
|
|
|
|
02-16-2006, 07:50 PM
|
#15 |
|
Loremaster
Join Date: Nov 2004
Posts: 152
|
I do play a ranger, and I agree with magus. You all are screaming the sky is falling when its just raining on our heads for once. Relax and enjoy the game. If the game gets harder, all the better. It isnt going to be game breaking for us, only a small percentage(like 10 - 25% at the most) which would put us in line with the other T1 classes. And if it drops you below T1, my question would be if you are playing it right...? I have 3 other buddies who play a warlock, assasin, and wizzy. I out class them in DPS by a long shot. Might be time to realize we are broken. Its not just them either. When grouped with others of those classes I still outclass them in DPS.
__________________
|
|
|
|
|
02-16-2006, 07:56 PM
|
#16 |
|
Loremaster
Join Date: Jun 2005
Posts: 173
|
If you don't realise that this will effect your dps by @ 50% as a ranger then I question your ability to play your ranger. I parse dailey and see first hand how critical of a nerf this is. To say anything else is simply lack of knowledge in your class. |
|
|
|
|
02-16-2006, 08:05 PM
|
#17 |
|
Loremaster
Join Date: Nov 2004
Posts: 294
|
Ok, Raids I can see about 50% coming from procs. He's talking about procs though and then using a soloing situation to prove his point with theoretical numbers.
__________________
If they ever come up with a swashbuckling school, I think one of the courses should be laughing, then jumping off something.- Jack Handey |
|
|
|
|
02-16-2006, 08:11 PM
|
#18 |
|
General
Join Date: Nov 2004
Posts: 415
|
There's certainly value in applying a "common sense test" to any model. But you're trying to refute my model, constructed from actual game data and mechanics, with guesswork. Your 70% and 1/5 figures are out of nowhere.
__________________
Lokimor Human Coercer / Sage Venekor |
|
|
|
|
02-16-2006, 08:12 PM
|
#19 |
|
General
Join Date: Nov 2004
Posts: 415
|
I think you're missing the point. My model doesn't assume that all procs will be gone. It calculates exactly how often different proc situations will occur, and how much damage - over all - can be expected from them.
__________________
Lokimor Human Coercer / Sage Venekor |
|
|
|
|
02-16-2006, 08:25 PM
|
#20 |
|
Loremaster
Join Date: Feb 2005
Posts: 1,039
|
The average DPS that comes out of procs for my ranger (level 59 toon, with adept 3's and some master 1's) is 39% damage done with poison, 19% damage done with archers frenzy, and 4% with my bow proc - Compiled these stats with acouple of fights in a clops group using advanced combat tracker.His data isent entirely wrong with what he has come up with, however his data isent entirely right. I tend to average around that DPS.
__________________
RIP Oakleaf 1949-2006 |
|
|
|
|
02-16-2006, 08:33 PM
|
#21 |
|
General
Join Date: Nov 2004
Posts: 415
|
Well, the DPS numbers I listed above are a theoretical max, based on the data from the examine windows. I don't claim that that's what a parser shows for DPS. I only used it as a way to compare before and after effects in a way that reflects how often a given art or attack can be used.It's worth mentioning that the examine data wasn't completely in line with my logs. My logs usually showed higher average damages for everything - CA hits, procs, autoattack hits - than what the examine window showed. But they were always pretty close, and seemed proportional. So again, I present these numbers for comparison with each other - not for comparison against real-world parses.
__________________
Lokimor Human Coercer / Sage Venekor |
|
|
|
|
02-16-2006, 08:40 PM
|
#22 |
|
Loremaster
Join Date: Nov 2004
Posts: 152
|
This is bull that is gonna affect you by 50%. Are you completely wack? Thats as if they said 'poison is now cosmetic,' which is far from the case. Its making it so that the procs are based off the first strike of a CA. Not to mention its painfully obvious we as rangers needed a fix. Even moorgaurd pointed out rangers doing 4000 dps... DPS. Thats nuts. I have never, EVER, seen my poison account for more then 40% of my damage (and I buy the BEST poisons for my tier), and I parse ALL the time. So even if this cut this to 1/3, which I doubt it does, that means a drop of ~30%, but lets be pessimistic and say 35%. Even that sounds off beat. Example: If I have 10 CA's that all have 3 hits involved in them and are 3 second cast times, I effectively have one strike every second, right? This is not the case with Rangers, but we are being pessimistic about this, remember? This would pose the greatest drop theoretically. I have noticed that the CA's do on average a little more then the tick of a poison, or DD, depending on what you want. We are going to choose DD poison for this, hence if each hit did 100 damage, the DD on the poison would be roughly 80 or 90, but lets just say 90. Remember this is on average they are better... I know some are an exception. Lets pretend I have a 25% to hit on my poisons, so saying I have good luck and it goes 40%. That means on 12 hits, my poisons proc. Now it dots every 6 seconds for 1/4 the damage, but lets continue being pessimistic and say it dots every 4 seconds for 1/3 the damage. Only one poison can be on at any given time, so for simplicity's sake, it strikes an additional 1/3 normal poison damage every 4 seconds. That means: My damage = 30 Strikes * 100 damage due to CA's = 3000 12 Strikes * 90 damage due to poison = 1080 8 Stirkes * 30 damage due to dot = 240 Total = 4320 Where in the hell do I lose 50%? Poison isnt even equal to 50%. But you know what, we are being pessimistic, lets change this model so that poison = 50% of my damage. That means I need 1320 to = 3000, which is roughly a 3x jump. So now: My damage = 30 strikes * 100 (CA) = 3000 12 strikes * 270 (poison) = 3240 8 strikes * 90 (dot) = 720 Total = 6960 Now poison accounts for about 57% of my damage. Perfect pessimistic model, lets use it. Now lets say I only proc on the first hit of each CA. That means 40% of 10 hits, so 4 procs. The dot amount stays constant (isnt affected due to the fact that its over time and we are pretending the first hit lands a poison for simplicity's sake). My NEW damage = 30 strikes * 100 (CA) = 3000 4 strikes * 270 (poison) = 1080 8 strikes * 90 (dot) = 720 Total = 4800 Or rougly 70% of what it used to be. ON AN ENTIRELY PESSIMISTIC VIEWPOINT. Realistically, it should proc once every 6 seconds, do less damage then my average CA's, and dot for 1/4 of that damage. My damage = 30 Strikes * 100 damage due to CA's = 3000 12 Strikes * 90 damage due to poison = 1080 5 Stirkes * 23 damage due to dot = 115 Total = 4320 My NEW damage = 30 strikes * 100 (CA) = 3000 4 strikes * 90 (poison) = 360 5 strikes * 23 (dot) = 115 Total = 3475 Or roughly 80% of my original. So somewhere between 20 and 30% is what you can expect, but this is even still less accurate as I am assuming you are hitting with CA's that all have multiple strikes in them, not counting auto attack, and not counting other CA's you might be doing. All those should drop the % even more, probably enough to put it near the 20% range, which is a perfectly acceptable 'nerf' considering 4000 DPS on some rangers and the fact that I almost always do near 1.5x the other DPS classes I have grouped with. 'The sky is falling, the sky is falling!'
__________________
|
|
|
|
|
02-16-2006, 08:44 PM
|
#23 |
|
Loremaster
Join Date: Nov 2004
Posts: 294
|
If you think that this will effect your DPS by 50% as a ranger, then I question your ability in math.First, during Autoattack it's going to be the same amount of procs as before this "nerf".Second, during CAs, I'm guessing (speculating, hypothesizing) that the chance to proc per CA will be the % listed, such as 20% chance to Proc with poisons per CA. Assuming that you guys proc poison off of 4 of every 5 CAs (which you probably don't), that's a 75% (You're hitting them 3/4 less) nerf to your poison proc damage.So let's say that 75% of your procs are from CAs, the other 25% from autoattacks.For simplicity, let's say that you proc 4 times in a fight, 3 (75%) from CAs and 1 (25%) from autoattack now. With a 75% nerf to the CA procs. That will bring you down to 0-1 proc from CAs (25% of 3) using real numbers. This would be a 50%-75% nerf to your proc damage. Which translates to 25%-37% nerf to overall damage. Not the 50% nerf that everyone is worried about. Granted, these are rough estimates and numbers. Nobody will be sure until it's actually tested.
__________________
If they ever come up with a swashbuckling school, I think one of the courses should be laughing, then jumping off something.- Jack Handey |
|
|
|
|
02-16-2006, 08:49 PM
|
#24 |
|
Join Date: Nov 2004
Posts: 433
|
Ive seen 50-60% of our rangers dps comes from parses during raids. On average the rangers are dpsing in the range of 1.3-1.8k. Now currently on beta procs of any kind are hardly firing at all, in some cases once or twice ever 10 fights. To be generous and only say that the reduction will not completely remove the proc dps a ranger has but will lower it to about 10% of their current proc dps (currently on test its much lower than that). This drops a rangers average dps to 650 - ~800 now this is (in raid) below our average wizard dps which is 800-1.2k.I guess we will have to see what happens when this hits live, but i can really see rangers being made t2 dps for a good while (probably a few months) before they are fixed and put back into t1
Message Edited by perano on 02-16-200610:55 AM
__________________
- Igus 70 Dirge, Second Dawn Best Quote Ever: "BREAKING: New York subways smell funny" |
|
|
|
|
02-16-2006, 08:50 PM
|
#25 |
|
Loremaster
Join Date: Nov 2004
Posts: 152
|
Can you edit this and word it better? I cant understand it.
__________________
|
|
|
|
|
02-16-2006, 08:51 PM
|
#26 |
|
Join Date: Nov 2004
Posts: 433
|
OMG this is not a small nerf its a massive nerf to all classes that rely on procs
__________________
- Igus 70 Dirge, Second Dawn Best Quote Ever: "BREAKING: New York subways smell funny" |
|
|
|
|
02-16-2006, 08:53 PM
|
#27 |
|
Loremaster
Join Date: Nov 2004
Posts: 152
|
I'd have to disagree. I feel like I am one of the few people left that has any faith in the devs.
__________________
|
|
|
|
|
02-16-2006, 08:55 PM
|
#28 |
|
General
Join Date: Nov 2004
Posts: 3,040
|
Your spreedshot is incorrect. Your triple fire data is screwed up on the beta assumptions. You're using the damage for the first hit for the 2nd and 3rd hits, rather than the 2nd and 3rd hit damage. So you're missing out on 420 damage or so. Plus your DPS total is not applying recovery times, which is where you're going to hit your auto-attacks, in between each of your CAs.Also quickshot still had the chance to proc poison on Wednesday(yesterday), so you're missing damage on the beta assumptions there.All it seems they are changing is having the 2nd+ hit on a CA doesn't proc anymore, and CAs have their own system for determing proc rate rather than using weapon delay which is a cheap system. So, fix your spreadsheet with this new correct data, then try to show your mathematical model of ranger damage before and after.
__________________
Calaglin, Former Illusionist/Guild Leader of Dissolution on Nektulos Calaglin, Former Illusionist/Guild Leader of Confirmed on Unrest |
|
|
|
|
02-16-2006, 08:57 PM
|
#29 |
|
General
Join Date: Nov 2004
Posts: 3,040
|
I think it's not small, but it's also not massive. They're not changing that much, in fact if people had the correct information before posting bull[expletive haxx0red by Raijinn] like the OP did. Ranger DPS is NOT dropping by 45%, in reality it's dropping by more like 10-15%, it isn't that large.
__________________
Calaglin, Former Illusionist/Guild Leader of Dissolution on Nektulos Calaglin, Former Illusionist/Guild Leader of Confirmed on Unrest |
|
|
|
|
02-16-2006, 08:58 PM
|
#30 |
|
Join Date: May 2005
Posts: 117
|
Not being able to see the macros makes it hard to comment on the accuracy of the model, but it does seem flawed. How is autoattack being effected? Autoattack should remain constant through the models, should it not?
__________________
Mykel - 70 Wizard Shadospawn - 70 Brigand |
|
|
|
