Is there a way to capture the combat log?

I want to test something as far as DPS output … but I would need to review the entire combat log over a long period (say, 5mins).

I’m sure this is possible, I just don’t know how. Advice, tips, suggestions? :slight_smile:

edit: And I don’t mean an addon. Not Warcraft logs. I mean officially, in-game. Just … like, what I’d like is a way for the GAME to output a file to my WoW directory. Is that possible?

maybe try warcraftlogs

2 Likes

I would just use a combat logging add-on like “Details!”- that should give you everything you need.

1 Like

Turn on advanced combat logging in settings, “system” settings, type “/combatlog” and upload your log to Warcraftlogs to analyze it. Hope it helps.

7 Likes

aside from addons there is no way to review it. warcraft logs will take a log of the data that wow provides.

I mean technically you could open the log file and read through it if you want your head to explode and/or need help falling asleep :joy:

5 Likes

Not a surprise that I said “NOT Warcraftlogs” and all but 1 of the response is “use Warcraftlogs.” :stuck_out_tongue:

@Jonashai: yes, it may be dull and redefine boring, but it’s just for a couple of experiments, not a full time job. lol

The point is that I’m trying to figure out why the different DPS meters have such huge differences from each other. They all parse the combat log, add it up, and divide by time, yes? So how can they be SO different??

Like maybe some piece of data gets dropped here or there… maybe one of them is 2-4% different somehow/someway.

But in my Raid last night, the difference between two of the meters was more than THIRTY PERCENT. I don’t want the canned responses from these boards like “that meter sukx”, I want to know WHY.

I haven’t done a LOT of research on this, but the little I’ve done all ends on ZERO actual data, and all anecdotal evidence. Well, that’s not evidence. Every comment is “the old meter is inaccurate, use the new one.” None of them explain WHY. There is no reason or explanation, we’re all just supposed to accept it.

I want to know WHY. I want some kind of proof that the tool people are basing every DPS decision in this game is working correctly.

I already know the first-line response to this: you think the RWF-type people don’t already do this? I don’t know. Again, that’s the point. If they DO have this kind of research, I can’t find it anywhere. It’s not posted, it’s not available. Why not?

I just want to know how DPS meters work. Why/how some of them could potentially be so inaccurate if they are just doing the simple process of “add it all up, divide by time.” And the only way to verify this is to do it myself: I’ll run a sim and add it all up and divide by time… and see what the meters say in comparison.

Combat Logs may be written to a file using the /combatlog command. The resulting file may be found at Logs\WoWCombatLog.txt.

But it’s going to be a boatload of text for a 5 minute parse and it’s not going to be easily readable. I killed 4 mobs on my rogue and got 250 lines of combatlog data and it looks like this

9/1 12:08:49.025 COMBAT_LOG_VERSION,19,ADVANCED_LOG_ENABLED,0,BUILD_VERSION,9.1.0,PROJECT_ID,1 9/1 12:08:49.025 ZONE_CHANGE,2364,"The Maw",0 9/1 12:08:49.025 MAP_CHANGE,1648,"The Maw",5733.330078,2972.919922,8666.669922,4527.080078 9/1 12:09:01.292 SPELL_CAST_START,P,0x511,0x1,0000000000000000,nil,0x80000000,0x80000000,315584,"Instant Poison",0x1 9/1 12:09:02.799 SPELL_AURA_APPLIED,P,0x511,0x1,P,0x511,0x1,315584,"Instant Poison",0x1,BUFF 9/1 12:09:02.799 SPELL_CAST_SUCCESS,P,0x511,0x1,0000000000000000,nil,0x80000000,0x80000000,315584,"Instant Poison",0x1,0000000000000000,0000000000000000,0,0,0,0,0,0,-1,0,0,0,0.00,0.00,1648,0.0000,0 9/1 12:09:03.050 SPELL_CAST_START,P,0x511,0x1,0000000000000000,nil,0x80000000,0x80000000,3408,"Crippling Poison",0x1 9/1 12:09:04.482 SPELL_AURA_APPLIED,P,0x511,0x1,P,0x511,0x1,3408,"Crippling Poison",0x1,BUFF 9/1 12:09:04.482 SPELL_CAST_SUCCESS,P,0x511,0x1,0000000000000000,nil,0x80000000,0x80000000,3408,"Crippling Poison",0x1,0000000000000000,0000000000000000,0,0,0,0,0,0,-1,0,0,0,0.00,0.00,1648,0.0000,0 9/1 12:09:06.372 SPELL_AURA_REMOVED,P,0x511,0x1,P,0x511,0x1,1784,"Stealth",0x1,BUFF 9/1 12:09:06.372 SPELL_ENERGIZE,P,0x511,0x1,P,0x511,0x1,8676,"Ambush",0x1,0000000000000000,0000000000000000,0,0,0,0,0,0,-1,0,0,0,0.00,0.00,1648,0.0000,0,2.0000,0.0000,4,0 9/1 12:09:06.381 SWING_DAMAGE,Creature-0-4224-2364-7652-165978-00002FA55F,"Scavenging Mawrat",0x10a48,0x0,P,0x511,0x1,0000000000000000,0000000000000000,0,0,0,0,0,0,-1,0,0,0,0.00,0.00,1648,0.0000,0,66,76,-1,1,0,0,0,nil,nil,nil 9/1 12:09:06.381 SPELL_DAMAGE,P,0x511,0x1,Creature-0-4224-2364-7652-165978-00002FA55F,"Scavenging Mawrat",0x10a48,0x0,8676,"Ambush",0x1,0000000000000000,0000000000000000,0,0,0,0,0,0,-1,0,0,0,0.00,0.00,1648,0.0000,0,563,400,-1,1,0,0,0,1,nil,nil 9/1 12:09:06.381 SPELL_CAST_SUCCESS,P,0x511,0x1,Creature-0-4224-2364-7652-165978-00002FA55F,"Scavenging Mawrat",0x10a48,0x0,8676,"Ambush",0x1,0000000000000000,0000000000000000,0,0,0,0,0,0,-1,0,0,0,0.00,0.00,1648,0.0000,0 9/1 12:09:06.381 SWING_DAMAGE,Creature-0-4224-2364-7652-165978-00002FA551,"Scavenging Mawrat",0xa48,0x0,P,0x511,0x1,0000000000000000,0000000000000000,0,0,0,0,0,0,-1,0,0,0,0.00,0.00,1648,0.0000,0,56,64,-1,1,0,0,0,nil,nil,nil 9/1 12:09:06.381 SWING_DAMAGE,Creature-0-4224-2364-7652-165978-00002FA546,"Scavenging Mawrat",0xa48,0x0,P,0x511,0x1,0000000000000000,0000000000000000,0,0,0,0,0,0,-1,0,0,0,0.00,0.00,1648,0.0000,0,75,86,-1,1,0,0,0,nil,nil,nil 9/1 12:09:06.443 SPELL_CAST_FAILED,P,0x511,0x1,0000000000000000,nil,0x80000000,0x80000000,185763,"Pistol Shot",0x1,"Not yet recovered" 9/1 12:09:06.676

3 Likes

You’re certainly welcome to reinvent a wheel that exactly fits the cart your pushing, but anything you did would more-or-less be identical to what you’ll find there.

2 Likes

I’m admittedly coming from a place of complete ignorance, but that’s what this post is about: for me to learn.

I guess I’m not completely understanding how Warcraftlogs (the Raiding app) works.

But don’t forget the prevailing question here: how can one of the meters be SO inaccurate?

No. I don’t care to do the work required to investigate this myself as I’m sure it’ll take me quite a bit of time but if you believe this post from reddit then it could explain it

There’s been a few responses just saying recount is terrible and one who touches on the idea of the difference.
Recount tracks someones DPS for as long as they are active in the fight, where as Details tracks DPS for the duration of the fight.
For example, I was running MOTHERLODE yesterday and had to go AFK for a couple of minutes, I came back and joined in a trash pack just for the last 5 or so seconds, according to Details my DPS was about 15K, on Recount my DPS would have been much higher.
Putting some numbers to it:
If there is a 1 minute fight but you died 45 seconds in and you do 1.2M damage, on Details this will show as 20K DPS, on Recount it will be 26,667 DPS.

https://www.reddit.com/r/wow/comments/icm38g/is_recount_more_unreliable_than_details_dps_meter/

The meters aren’t all doing the exact same thing they are doing something similar but differently.

What is the difference in their active time in the fight?

https://i.postimg.cc/Dy7twWkL/image.png

If you want an answer that you can accept then you’ll have to learn to read code and read the code for each addon :smiley:

The question you should ask yourself is “what is the meter showing me” then you can determine if it is inaccurate, you’re making an assumption that it’s

“add it all up, divide by time.”

which may not be the case

1 Like

Great response, thank you so much!

Unfortunately for me, the example I’m looking at was a Raid boss (Fatescribe).

So there is no downtime, and I didn’t die.

Both meters started at the same time (the pull), the both ended at the same time (we 1-shot the boss), and I didn’t die.

So they both should have tracked the EXACT same data, over the EXACT same amount of time. And yet… 30% difference.

That just… can’t be.

I know recount has settings to add pet damage to player’s amount. I also think some damage meters try and calculate how much your particular group buff adds to dps. I’d check into the damage meter’s preferences also damage meters seem to differ in what hey consider ‘your’ dps vs. a proc.

https://memegenerator.net/img/instances/75647162.jpg

1 Like

I’m a Warrior- no pets.

Why would it try and “calculate” anything? If it’s not damage output, it shouldn’t track it.

Same with procs. If I didn’t hit something, if there isn’t damage… how/why would it count it as damage?

I have a Jaithys that procs ALLLLL the time. But it’s a Strength proc, there’s no damage to track. I hit HARDER when I proc, which then equates to more damage… but that is what all meters should be tracking, not the proc.

The point I was making is that the meters don’t don’t all use the same algorithm as it parses the damage. That would be my opinion as to why the wide variance, also check that whatever meter your testing is current to the latest patch.

What “algorithm?” Why would there be ANY algorithm??

Add up all the damage, divide by time. That’s the ONLY thing that should be happening here.

Maybe it’s just semantics, but I wouldn’t call that an algorithm, that’s just basic math.

The only way you’re really going to know the true why is if you look at the code :man_shrugging: not trying to be a jerk but that’s really the only option unless you can get the devs of the addons to say how exactly their numbers are determined or someone you trust that can read code to figure it out.
Most people just go with whatever they are told is “the best” as far as meters are concerned. I started with Recount back in the day, then Skada became “the best” and now it’s Details. I really don’t like Details and would rather use Skada but Details surpasses the other addons in customization so I’m stuck with it :[

1 Like

Some have already provided responses I would probably also respond with. However, and maybe you covered this already. But which logs are you comparing that are 30% off? Are they the same addon? Different addons? Which addons, details, recount, skada, another one?

There are definitely different variables to how these addons work and count your DPS. One variable I remember reading about a while ago was the distance the player logging was to the person you want to track. For example, a ranged player that never stands in melee logging the fight could potentially have different numbers than a melee person logging their own fight.

But if I wanted to get to the nitty gritty details, I would review my logs with my addon and compare it to the logs I’m questioning (details has a detailed breakdown of how many times your abilities hit/crit/etc. and how many casts vs hits … also shows aura uptime).

But if you don’t have access to the other person’s logs, there is also the in game combat log, “my actions” that outputs everything you do (but doesn’t show buffs), but it will show each ability you used and what it did to the target.

Also, regarding WCL, often times they won’t align with Details or other damage meters, since the developers at WCL sometimes update the output to reflect their own agenda (for example, removing damage to adds considered padding). Not sure if they are doing anything like that in Shadowlands (don’t think so), but it’s a possibility.