[Ret] Trinket ratings estimation [Dragon Soul 4.3+]

[Butosai]

why would it be bugged? i dont get it. just insane amount of weapon procs

http://www.worldoflogs.com/reports/8...2&e=1121#Lymez
thats ~1/2 the dps

oh fuck 10 man wtb 5 minute kill time

[Deleted]

doesnt change the fact that the stars guy got double the amount of procs
i didnt use my 3rd round of cds. but overall i think you will be doing more dps in the 25 man version because of the 2nd guardian

[Reith]

That is why I just stand between the 1st and 2nd line. But yes I understand however I do think it could be a bugged parse time to dive into it since nothing else going on.

That is the best spot, right in the center half-way between those lines. I got 8k tentacle DPS on LFR Ultraxion as I posted earlier so yeah it's definitely the optimal spot. The fact that it's such a high percentage pretty much rules out the possible uncertainty of the position.

As for a possible bugged parse.. meh, maybe. I've noticed that my recount's different from our uploader's combatlog - sometimes extremely different, like for example I saw a discrepancy of 11,000 DPS on our Combat Rogue for a Madness kill between my recount and the combatlog that was uploaded, so I dunno. Is it possible to have multiple people upload their combatlogs for better accuracy?

BTW I am seriously shaking my fist at the holidays because we've only been able to kill heroic Morchok and the rest of the time had some people skip out on raids. Oh well.

[Butosai]

doesnt change the fact that the stars guy got double the amount of procs
i didnt use my 3rd round of cds. but overall i think you will be doing more dps in the 25 man version because of the 2nd guardian

personally i find the delay of cooldowns kills my dps

on my best attempts, im usually at 48-51k at the 4 minute mark

[Reith]

i didnt use my 3rd round of cds. but overall i think you will be doing more dps in the 25 man version because of the 2nd guardian

Completely agree with this. Heroism should also be used around 4:50-5:20 (depends how fast you kill it and/or when your raid leader decides when to use it, obviously) so it's actually ideal for it to last at least 5:30.

---------- Post added 2011-12-30 at 10:51 PM ----------

personally i find the delay of cooldowns kills my dps

You won't get a 4th set of cooldowns on Ultraxion anyway. Going by the earliest possible times you could do it: 0:02-2:02-:4:02-6:02 and by that time you're most certainly dead before you could possibly use the 4th. Delaying is no loss, especially if you're timing that third set with Heroism.

[Deleted]

yes i didnt watch the hp it was my bad. i wanted to use guardian +wings/zeal together.
As soon as 10man kills will be at ~4:30-4:40 10man dps will most probably be superior to 25man. Before that i think its quiet similar.

heroism should be used at the beginning with prot4set and prepot.

[Butosai]

Completely agree with this. Heroism should also be used around 4:50-5:20 (depends how fast you kill it and/or when your raid leader decides when to use it, obviously) so it's actually ideal for it to last at least 5:30.

---------- Post added 2011-12-30 at 10:51 PM ----------


You won't get a 4th set of cooldowns on Ultraxion anyway. Going by the earliest possible times you could do it: 0:02-2:02-:4:02-6:02 and by that time you're most certainly dead before you could possibly use the 4th. Delaying is no loss, especially if you're timing that third set with Heroism.

Well on heroic ultraxion you lust at the start for DPS and the healers phase down to not get lust then lust for healers at the end. The reason delaying cooldowns is a loss is because you are forced to wait 3 minutes inbetween your 2nd and 3rd set of cooldowns since like you said you wont get a 4th set and the third set lines up with your guardian. On 10 man with a 5 minute kill you pop cds at 0s, 2minutes, and 4 minutes with no delay, and they fade right before the boss dies when your dps is at its peak.

[Deleted]

yep true, next time ill be better prepared

[Bonkura]

You guys also assume you always have all raid buffs in 10 man raids, which is very hard to have. I've been doing DS without 5% crit all the time... sigh. :P

[Reith]

yes i didnt watch the hp it was my bad. i wanted to use guardian +wings/zeal together.
As soon as 10man kills will be at ~4:30-4:40 10man dps will most probably be superior to 25man. Before that i think its quiet similar.

heroism should be used at the beginning with prot4set and prepot.

Well yeah if your kill times are less than 5:20 you'd probably want it at the beginning... but there is ramp-up to consider for some specs, and you most certainly won't get full use of that potion like you would with a Heroism around 5 minutes. Just depends on the situation though, I guess.

[Bonkura]

If you have a fire mage you wouldn't only have your healers crying if you pop BL at beginning of the fight.

Prepot, prot4set and AoK obviously won't work together either.

Which also makes me realize how AoK can fuck you over during this fight in 10 man since you'd sometimes have to wait 30-45 secs for those 5 stacks to pop first CD's. Worst case scenario it will be CD's at 00:45->02:45->04:45. There's no way it will work with a proper BL timing and might even solve the mystery (YES I KNOW ITS FARFETCHED LOL) of this Taiwanese dude's decision to go with VB.

Personally if we're on the topic of doing prot4set and prepot I'd say VB wins hands down because it always procs within 10 seconds of the fight. Then you should be able to time another proc when it's BL time.

[Butosai]

If you have a fire mage you wouldn't only have your healers crying if you pop BL at beginning of the fight.

Prepot, prot4set and AoK obviously won't work together either.

Which also makes me realize how AoK can fuck you over during this fight in 10 man since you'd sometimes have to wait 30-45 secs for those 5 stacks to pop first CD's. Worst case scenario it will be CD's at 00:45->02:45->04:45. There's no way it will work with a proper BL timing and might even solve the mystery (YES I KNOW ITS FARFETCHED LOL) of this Taiwanese dude's decision to go with VB.

Personally if we're on the topic of doing prot4set and prepot I'd say VB wins hands down because it always procs within 10 seconds of the fight. Then you should be able to time another proc when it's BL time.

you use preguardian then you have 25 seconds to get off 3 rotations of 3 holy power. At the end of your third rotation guardian will have about 20 seconds left and you pop your cds there for the last 20 seconds. AoK should have 5 stacks. Lust and t11 guardian are both 45 seconds

[Bonkura]

you use preguardian then you have 25 seconds to get off 3 rotations of 3 holy power. At the end of your third rotation guardian will have about 20 seconds left and you pop your cds there for the last 20 seconds. AoK should have 5 stacks. Lust and t11 guardian are both 45 seconds

Sounds like you've tried it out a lot and I'm not gonna question it. Just because I will have to answer negative opinions on using BL on pull, what would be the good reasons for it despite benefiting your Ret's DPS?

[Reith]

you use preguardian then you have 25 seconds to get off 3 rotations of 3 holy power. At the end of your third rotation guardian will have about 20 seconds left and you pop your cds there for the last 20 seconds. AoK should have 5 stacks. Lust and t11 guardian are both 45 seconds

I thought lust was 40 seconds?.. but yeah, even in the best-case scenario, you won't have your potion active during AW/Zealotry for more than ~13 seconds. Heroism at the beginning is not ideal especially since AoK can be a bit sporadic sometimes. If anything, in a ~4:40 situation, I'd wager that pre-protguarding at the beginning isn't worth it and I would phase myself out just like the healers so I could have heroism popped for my third set of cooldowns in which I'd be using Guardian and the perfectly-timed Golemblood Potion.

Like I said though, it really depends how fast you kill it.

[Requital]

why would it be bugged? i dont get it. just insane amount of weapon procs

http://www.worldoflogs.com/reports/8...2&e=1121#Lymez
thats ~1/2 the dps

Ok if you compare some of the top parses you will see why I think it's registering as a bugged parse.

http://raidbots.com/comparebot/4efe8...d000228#damage

On a comparison breakdown the compare bot is only calculating 38K DPS if you go through all his numbers the compare bot isn't missing anything but the numbers he ha are all lower than both the other 2 parses I compared them with. So something is not calculating properly, I'm not sure what it is but there is definitely something wrong.

It might not have the pet merged, I'm still breaking it down.

[Butosai]

Sounds like you've tried it out a lot and I'm not gonna question it. Just because I will have to answer negative opinions on using BL on pull, what would be the good reasons for it despite benefiting your Ret's DPS?

its not for benefiting ret dps at all, infact i may go down with the healers next week to try to have it at the end. Its to benefit most of the classes that have little ramp up time and they can sync all their cds with procs like tailoring and make use of prepot, generally speaking a lust at the start of any encounter yeilds the highest raid dps

---------- Post added 2011-12-31 at 05:45 AM ----------

Ok if you compare some of the top parses you will see why I think it's registering as a bugged parse.

http://raidbots.com/comparebot/4efe8...d000228#damage

On a comparison breakdown the compare bot is only calculating 38K DPS if you go through all his numbers the compare bot isn't missing anything but the numbers he ha are all lower than both the other 2 parses I compared them with. So something is not calculating properly, I'm not sure what it is but there is definitely something wrong.

It might not have the pet merged, I'm still breaking it down.

looking at that reminded me we didnt have mark of the wild for the first minute of the fight /sad

[Adhal-Blackrock]

His tentacles did 7000 dps.

This is common with anyone ranking high, no matter the trinket. Fact is you won't rank high without the sword having amazing procs, so discounting that log for that reason is just idiotic. There are people below him that have had even better tentacle procs. However it could just be the fight lenth, 6 min fight gave him 7 procs from VB, which tend to be fairly easy to align with cds.

---------- Post added 2011-12-31 at 12:50 AM ----------

Brooch has a very high uptime I use mine all the time on certain encounters and it's maintaining 35-40% uptime and that alone beats everything else, I'm not surprised by his numbers at all.

exactly, like I said this is proving to be very interesting tier for trinkets. Not discounting anyone, but hell truth is the only thing that may matter is Gurthalak and EoU if you want to rank.

---------- Post added 2011-12-31 at 12:52 AM ----------

Do you guys know Chinese? Because I don't get shit of that log. Would like some translations to see what he's actually doing and with what gear. Sigh.

Anyway, VB is freaking awesome and I'm switching back to it after some terrible tries with AoK on Yorsahj HC and only 35K during the kill. It just keeps giving me a headache. Bear in mind that I only have the normal version so I'll have to use AoK:H eventually but I wouldn't say no to a BLF N/H version.

Edit: I did 44K on Yorsahj N the week before with worse gear and VB, so yeah. FML for shit logs.

If you hover over buffs in buffs gained it will give you it in english. where you can very clearly seem him getting titanic strength (EoU) and varothen's brooch

---------- Post added 2011-12-31 at 01:06 AM ----------

why would it be bugged? i dont get it. just insane amount of weapon procs

http://www.worldoflogs.com/reports/8...2&e=1121#Lymez
thats ~1/2 the dps

And looking at your log you had a large amount of TV crits, 37% on TV

Compare that to 21% TV by the Stars guy and thats a large amount of damage right there.

Chalk it up to RNG again. Though I will say its an impressive amount of damage either way, I'm not out to discount any logs either way since we all know RNG is a large factor, just to open discussion.

Hell Flash from Vodka is Ranking extremely well with BLF, though i do tend to discount him a bit since hes like 2 pieces away from full BiS already.

---------- Post added 2011-12-31 at 01:07 AM ----------

oh fuck 10 man wtb 5 minute kill time

this 2, all of those 25 man ranks would have been higher had it ended at a 5 min mark

[Deleted]

Ok if you compare some of the top parses you will see why I think it's registering as a bugged parse.

http://raidbots.com/comparebot/4efe8...d000228#damage

On a comparison breakdown the compare bot is only calculating 38K DPS if you go through all his numbers the compare bot isn't missing anything but the numbers he ha are all lower than both the other 2 parses I compared them with. So something is not calculating properly, I'm not sure what it is but there is definitely something wrong.

It might not have the pet merged, I'm still breaking it down.

if you look at the summary all three did ~14m.
but looking at the dps on top only the stars guy didnt get his tentacle dps added.
i guess comparebot has some problems with tw logs.

some people get lucky with crits others get lucky with weapon procs. in the end someone very lucky will be at the top

[imortals]

still love the Vessel of Acceleration (heroic) (mastery) use when u have enough inq time left, bloodlust is up and you use knight+ zealotry
massive amstery+inq let my hand of light hits harder than the crusader strike or templars verdict.
TV->CS->TV->CS->TV->CS->TV->CS->TV->CS->TV->CS->TV->CS->TV->CS->TV->CS->TV->yay dps

[pbad]

The problem with this whole line of reasoning is that it really doesn't prove anything. Single parses will always be filled with RNG, from crit, to DP, weapon procs, fading light debuffs, even the logging can vary a fair amount on H Ultrax because of all the phasing in and out of the twilight realm. If you've ever compared your own log to whomever is logging for your guild on Ultrax there is always a fair amount of discrepancy.

That's why we approach these problems mathematically like Kisko has done, or using spreadsheets/SimC, etc. If you feel something is being undervalued by a model or just modeled incorrectly then our best bet is to try to improve that model. Just linking a single parse and saying hey this guy was using x,y,or z trinket, maybe we all should look in to using it doesn't really help anyone.