There are two shadow priests in my 25s guild, and we're trying to find the discrepancies between her DPS and mine (often in the 1K - 1.5K range). I've done some analysis about it, and the guildie has already made some big improvements, but I wanted to get a second opinion. As an added bonus, any general review of either of our playing is also much appreciated.

Some Yoggy attempts and Ony: World of Logs - Real Time Raid Analysis
Keepers & El Generalisimo: World of Logs - Real Time Raid Analysis

me = Juzefon
other priest = Onihime

These are the only two recent logs that I have in which we are both around, although I realize that the linked fights aren't the greatest fights to compare DPS. When I looked through the logs, I looked at places like Mimi P1 and Ony.

My guildie knows that survivability is the #1 goal, so while that's certainly something to get from these logs, it is already a priority. I also learned that I take, on average, 0.4 seconds less between each of my casts than Oni does (after running casttimes for MFs and MBs in Mimi P1), despite our both having similar gear, similar lag, and Quartz. I think that time discrepancy is pretty big, and could account for the DPS difference, in which case the answer is likely "better execution" or somesuch. Am I thinking along the right lines, or am I missing something else?

I just looked at one fight (Thorim kill, p2), but it's what I normally see on DPS differences between SP's.

First, there's what you expect. You have better uptime on all your dots (Oni's 73% on VT is quite bad), more Mind Blasts, and evidence that you death on the move while Oni does not. There's definite room for improvement there.

Then, there's the big one. You had 137 Mind Flays to Oni's 95. That's 80k damage. I have tried to help other SP's DPS and this is usually what holds them back. They just flat don't Mind Flay enough. I don't know what Oni was doing all that time (That's 14 full length mind flays, or almost 40 seconds), but it wasn't casting more Mind Blasts, more Dots, or more Deaths. WAY too much idle time.

Most likely he was casting DoTs and kept clipping ticks--that's the only explanation I can come up with for fewer MFs and lower DoT uptime.

I need help! I seriously have no clue what I am doing wrong when it comes to raiding. I use Event Herizon and DoTimer to time my spells. My rotation is the normal rotation: VT, DP, MB, MF x2, SW.P and then VE. Rinse and Repeat! My computer runs just fine and there is hardly lag. I use quartz just in case!

I know I will probably NEVER be top ten in a guild full of burst dps, but I should be above 4200-4500 dps.

(btw...I have merlins robe and my triumph gloves but I chose to use the t-8 for the set bonus and the gloves I have on for haste)

My dps was alright until I changed out a belt with haste on it for what I have. It wasn't great, but it has went DOWN hill.

My WoL: World of Logs - Real Time Raid Analysis

We were doing 25 ToC and 25 ToGC that night. My recount always shows anywhere from 3500-5000 dps. The logs they use obviously shows me higher, but I know I am doing something wrong!

Can anyone help?

It's kind of hard to really get the numers right from your log. But let's take your longest heroic NRB fight (tyr 6) and your Lord Jaraxxus normal kill, since the other fights are kind of gimnik fights and are hard to compare. I will take your numbers and compare them to my numers.

Starting with Jaraxxus:
My log: World of Logs - Real Time Raid Analysis
Your log: World of Logs - Real Time Raid Analysis

- you hardly keep full dots on adds running
- I have 147 mindflay ticks, you have 94 despite my fight was shorter, same with other spells (but not such a huge difference)
- your shadowfiend did 20k dmg, my did 60k dmg, use it early if possible and you can use it twice. Also make sure to cast shadow crawl whenever it's up, it's a decent dps buff for your pet


NRB:
My log: World of Logs - Real Time Raid Analysis (sloppy kill and I got charged in the end :/)
Your log: World of Logs - Real Time Raid Analysis

- again looking at your mindflay numbers: 88 mindflay ticks for you in ~400 seconds. That's about 1 tick every 4.5second. In contrast, I have 243 ticks in a 480 second long fight. => Every 2 second a tick
- you don't seem to keep dots on all the snobolds and the boss in p1 running, that's the best thing you can do to do damage in phase 1
- you don't keep dots on the other worm running in phase 2. Even if you focus down one worm, you should keep dots running on Dreadscale as well



Due to a rather big gear discrepancy between us, it's kind of hard to say how much of my dps advantage comes from my gear and how much is because of my play style. But it seems that your problem is the same that most other shadows have, which lack dps: Just not enough casts, especially a really low number of mindflay ticks. That the most important reason for your poor dps, because your idle time is just way to high. I don't know if you clip your mindflay, but in practise it isn't worth it for me. You should also use a channeling-macro for mindflay to not clip it by a accidentally.

Another big advantage for shadows is the way we work on fights with adds/more than 1 boss. Use our strenght and keep dots running on all targets, if it's worth it (most of the time it is worth it!).



A small note to your gear: If you put your 2 blue gems into slots, which give you a spellpower bonus, you'll end up slightly better (and you should switch your 20 haste socket from your belt).

Thank you for the suggestion about my blue gems!

I do realize that my dot uptime isn't really what I want it to be. In the majority of those tries with ToGC 25, we were trying two different strategies (neither of them worked). LOL! I did have a question. You are just spamming mind flay after refreshing right?

I have tried that (I do have the no channel MF macro). I have read so much on how there are spell rotations that should be followed and theories on clipping MF and so many things. I have tried switching my own spell rotation around to try those others and none of them work.

Use the addons Quartz and QuartzLatencyMF2. Spend maybe a couple hours practicing clipping it. Unless your ping is abnormally high, once you get the hang of it, you should see dps increases in the triple digits by properly clipping MF2 to increase dot uptime and MB usage.

I do use Quartz. I always have. So clipping is still relevant to being a spriest? I have read in some places how MF is random in ticks now. I have noticed it myself. So if it's random...is it even visable to do clipping?

Clipping is viable, but it is a very minor DPS increase unless you can do so flawlessly, and with our new set bonus, it becomes even less important.

I don't think the damage of MF is random. There's been no blue indication of such a chance, and I've never seen others talk about it. There is a delay from when the server notes that the damage happens to when it triggers on screen (not counting normal server-client lag), but it still happens at regular intervals during the cast.

Mind Flay ticks are most certainly not random. If you aren't using a cast bar addon for it already, I highly recommend downloading this one. It can be somewhat inconsistent depending on your internet stability, latency, or server stability, but it's still not random. I believe I saw a post that estimated that clipping isn't worth it unless your latency is stable around ~70ms, which I'm looking for at the moment to reference directly.

This should be obvious, how ever. You can not simply rely on a mod, or statement to determine whether you should do something.
This is indeed a very true and accurate statement by C4Tuna. Does the fact that the statement itself implies a dps increase? no. A lot of what is in that statement is "judgement call". You need to get very comfortable with your casting/timing of Mind flay before you start clipping it. I'll give you an example.

WoW Meters Online - Combatlog Replay

This is a parse of my guilds lady deathwhisper kill this past tuesday. The duration of fight was 7min and 50sec.. or 470 seconds. I had 233 "Ticks" of Mind flay for the duration of that fight. If my math is correct, my mind flay was performing a periodic damage every 2.01 sec. In my opinion that is about as spot on for up-time as can be asked. I also never...read: not once, allowed for a full duration of mind flay (meaning, at no point did I allow the cast of mind flay to perform its complete duration and expire from the bar), more than once per sequence. This was not something that happened overnight, or from a handful of spare-time dummy drills. This was the result of spending sometimes hours per day on the target dummy timing my mind flay, learning to get quicker and more accurate about recasting, or casting a new spell in the "sweet spot" when it hits the red latency line. I admit that at first it caused me to lose some dps at first as I was not comfortable with it during actual raids. Over time though, it is something that will become more natural. There are still times where I flat out clip the 3rd tick of mind flay, but I can you the only time that happens is if I NEED to queue up at least 2 more spells (most often this is VT/MB, or VT/DP).

I will also admit that I still mess up on fights, or get caught up in panic situations.. this is part of the process to becoming a better Shadow priest in the end.

Just looking to get some advice/input on analyzing logs in the 3.3 era where haste affects VT/DP. Prior to 3.3 it was easy to check DoT uptimes from one priest to another just by looking at number of ticks and comparing that against the fight duration. Do we just take the parser calculated uptimes (such as from World of Logs) as gospel now?

Here's an ICC-25 alt run where an applicant and myself were the two shadow priests. I'm looking specifically at the Festergut fight where the kill was just past the enrage. Both of the priests were completely new to this fight. Running Simcraft on both priests gives an estimated 7749 dps for Calistella and 7729 dps for Teona using 10000 iterations and a fight length of 300 seconds on the high latency setting. One thing that was noted on the kill was that she prepotted with Wild Magic and then used a second one in the fight while I just popped a single Potion of Speed. The Simcraft numbers assume that Potion of Speed is used.

The main things I'm seeing with Teona's performance is much fewer Mind Flay ticks and 7 fewer Mind Blasts on the kill, which accounts for 240k less damage done, out of a 400k damage differential (it appears to be less on WoL but it's also counting Blighted Spores as damage done which shouldn't be the case). But the VT/DP stats are harder to interpret and give feedback on. For VT it's 114 ticks for myself vs. 116 ticks for her, and her uptime is higher on VT as calculated by WoL. For DP it's 125 ticks for myself vs. 120 ticks for her, and again she has the higher uptime on WoL. From the number of Improved DP procs it looks like she must be clipping DP as she cast 2 more DP but she has fewer ticks overall.

Although Teona has higher haste on Armory we both are using trinkets that affect haste: she is using the [Fetish of Volatile Power], while I have both [Scale of Fates] and [Embrace of the Spider]. Optimally the trinkets would be used so that both VT/DP expire and are recast at least once while under the effects of the trinkets - and earlier in this forum it was discussed how clipping the last tick or so of VT/DP could be beneficial if a haste buff was about to expire. From using the "Buffs gained" and "Buffs cast" graphs it's possible to see the uptimes of DoTs and of trinkets, but not so easy to pinpoint when exactly something was recast if it was clipped. See Teona's DP in the last minute and a half - the uptime bar is solid and while the haste trinket was used during that period, it's impossible to tell if DP was recast when the trinket was up without using the Log Browser. Am I reading too much into VT/DP uptimes and optimal casts, or is there something here worthy of feedback?

I recall it being theorycrafted on shadowpriest.com that the latency cutoff for wanting to clip Mind Flay for just Mind Blast was ~70ms. Obviously that threshold is going to change if you're clipping for more than one spell, but that becomes more complicated to calculate.

First off, it was not a latency cutoff. It was a calculation of how soon after the second you would have to be able to consistently clip for it to be a DPS increase.

Second, that theorycraft was faulty, because it assumed a delay after MF2, but that MF3 was always perfect. That's obviously false: It's no harder to clip a MF at MF2 than to clip it at MF3, it's the exact same operation. The obvious exception being if you use nochanneling, in which case you never clip but always incur a time loss equal to your latency with every cast. If you redo the math accounting for that, you'll find that MF2 clipping is almost always a dps gain on paper.


That said, the DPS gains from MF2 clipping are extremely small. Even in an absolute unrealistic perfect clipping scenario you will be lucky to reach low triple digit dps from clipping at MF2. A shadow priest that ignores the concept of MF2 clipping altogether may very well out damage a similarly geared MF2 clipping priest, if he's only slightly more aware of his surroundings.

Technically it's 50% harder. A MF3 gives you 3 ticks, one third of which are subject to latency. A MF2 gives you 2 ticks, half of which are subject to latency. Or put another way, if there was some magical MF30 spell, you could cast that and it would be like casting 10 MF3s in a row with zero latency after the first nine casts. That's clearly better than 10 copies of MF3.

That said, I suspect the math is still correct that a MF2 clip is generally a DPS gain.

Alritey I'm curious >_>

I don't feel my DPS is quite where I want it.

Soooooo here are a couple of links.
World of Logs - Real Time Raid Analysis Rotface wipes
World of Logs - Real Time Raid Analysis Marrowgar kill/Deathwhisper Wipes...
World of Logs - Real Time Raid Analysis Deathwhisper/Saurfang
World of Logs - Real Time Raid Analysis ToC 25... where I thought I did very well

K some notes on the ToC: I was on orbs on Anub, I died in the first two fights... the first death was due to a tank not picking a worm back up and the second was due to healers not picking up some heals. Lemme know what I can do improve my DPS guys!

Also... as a side note... I'm running simcraft and it's returning 0s for all of my scaling factors... can anyone fix this?

I am not a shadow expert, but no one else has commented, so here goes. You are over the hit cap, and almost every piece of gear you have has spirit. It also looks like your dots are falling off, at least on that fight. It took me a lot of practice to keep my dots up, but it made a huge difference. Also, I did 5k DPS on Marrowgarr 10 man, with lesser gear, and my percentages were way different than yours? Specifically, Mind Flay is the highest percentage of your damage. Finally, are you getting two casts of Shadow Fiend on a 6 minute fight?

My log:

World of Logs - Real Time Raid Analysis

My damage by spell:

World of Logs - Real Time Raid Analysis

Yours:

World of Logs - Real Time Raid Analysis

Sorry if I am rambling, my brain is fried. I'll try to come back and look at this tomorrow.

You might want to hit your troll racial more than twice in a 2 hr period. Maybe even macro it to VT. Also, your raid appears to be missing a scorch debuff. Drop some hit, drop some spirit pieces, and you should definitely start to see some improvement. I also think you went a bit overboard with the reckless gems. On some of the fights your dot times and mind blast cooldown seem really good, like Jaraxxus, but on others there's some room for improvement. You do seem to SW:D a lot more than necessary.

Honestly, considering your gear, I'd say your dps is not that bad.

Edit: Just noticed you did have an affliction lock for scorch in ICC, although not in ToC.

OK: Some updates.

While those logs are not that old, I guess I should have mentioned some of the changes I've made since I did those raids.

A: I've downloaded PowerAuras which is so incredibly helpful. I've got it letting me know when my Berserk is off cooldown. I was using it for VT and DP... but that just doesn't seem helpful as the only times they're not on the boss are when I'm bursting down adds, CCed or doing an excessive amount of movement in which case I cast DP anyway.

B: I use my berserk CD at the first re-cast of Vampiric Touch, so that it benefits from the haste. I also re-cast devouring plague early for the same reason. When my berserk is active I do not mind blast as the cast time seems hardly beneficial over a Mind Flay.

C: I use Shadow Fiend at 70 percent mana, so that it's able to be cast twice in almost every fight but the mana return is not wasted.

These are all conscious changes I've made since these logs were registered. I'll have to get some more recent ones for you to break down. Thank you for the efforts so far and the advice has not been ignored.

In terms of my gear, I've already got the hit cap situation under control. As soon as I've got my second tier ten piece, and thusly breaking my tier nine four-piece bonus, I've got the 264 crafted pants to replace my tier nine ones.

I will post an updated log and armory once all of these things are in effect.

Super high dot times and mind blast hit counts are pretty rare these days and for a reason: as your gear gets better clipping mind flay in favor of mind blast can barely be called an dps upgrade and high dot uptimes can mean you are waiting to cast a dot instead of channeling a new mind flay which is also in most cases a dps loss. So one should not blindly look at 98% dot uptimes or high mind blast hits.
Our last 10man for reference