Average DPS for different jobs in T9?

This is a curiosity post more than anything.
How much DPS should I be grating in T9 as a BLM? I was told I had 310 my first clear, but that doesn't feel comfortable

Quote:

---

Originally Posted by Skeith-Adeline

This is a curiosity post more than anything.
How much DPS should I be grating in T9 as a BLM? I was told I had 310 my first clear, but that doesn't feel comfortable

---

Merged or unmerged ?

BLM in my raid's doing 350 with Laevateinn, merged

Our blm did 330+ with full BiS and unmerged :)

(please do not merge fights, waiting times are a part of the fight too!)

shld state total damage dealt nxt time XD

How can you tell your DPS numbers?!

I'm sorry, what is merged vs unmerged?

Quote:

---

Originally Posted by Skeith-Adeline

I'm sorry, what is merged vs unmerged?

---

It's when an encounter is split into multiple parses when there is any significant downtime in dps actions done that exceeds your program's time limit. What happens is while no one is attacking (say during Megaflare in T9) the last encounter is considered concluded and a new parse begins when you begin attacking again. What merged means is you take these multiple split up encounters and "merge" them together for an average. Unmerged means your program does not reset during down time, and tracks the entire fight start to finish including downtime. Merged encounters tend to show much higher numbers than unmerged ones, and many feel they are an inaccurate way to read dps number.

unmerged counts no dps time, such as phase transitions where no mob can be attacked (like titans jumps, or Nael's meteor streams). Merged cuts that time out, then combines the "dps only" pieces.

and I always thought "merged" time wasn't DPS but eDPS?

So, which is better to go off of?

Quote:

---

Originally Posted by Skeith-Adeline

So, which is better to go off of?

---

It's better to go by unmerged because depending on your timeout for combat end then merged can cause quite a bit of difference simply due to people using settings. Some may have a 6 second timeout other might use a 12 second timeout. Factor in multiple times during the fight where you have downtime it makes a big impact on the numbers if you are comparing to others. Merged can only be used effectively if everyone is using the same settings or if you're just comparing your own runs on the fight to previous runs.

For fights with long downtime just change the combat timeout to like 30 seconds or even a minute, and you avoid random breaks in combat while the fight is still happening.

Well, as always, it depends on what you want to measure. If you want to look great, compare your unmerged dps to the merged dps of others! XD For me, both are interesting. They give information on different things.

Merged is more equivalent to an instantaneous dps (well, in fact a dps curve is much more informative but OK). If at equivalent gear, your merged dps is lower than someone else, it may point out a problem in your rotation, or slacking between GCDs.

Unmerged is quite equivalent to damage done. The difference between both tells you your downtime. If this difference is much higher than for someone else, it means you could do more damage by only working on your gameplay for this particular fight (to increase active time).

The boss have 700 000 HP, the adds, i'm not sure... hum lets say there is 2 dragons, and 6 golem, it's what, 150-200k ?

Lets say 1 000 000 in total maximum (to count a little the downphases)

Knowing this, with 4 DPS, and 2 tanks (2 tanks does like the dps of one DPS if the warrior is good)
1 000 000/5 = 200 000, that's what they all need to do, kinda.
You have 13min, which is 780secondes.
200 000 / 780 = 256
Doesn't seems to be a lot, but I guess, it's more then 1 000 000.

But if you have more 280+ should be fine, I guess.

I know a good bard can do 350
Blm is kinda same as bard I think, around 350
A Monk/dragoon is much more, like 380 ? 400 ?

But OFC, if you ask average, is less then that.
I guess a decent Bard/BLM do 300-320 (and that's not bad actually, it's more then what you need for T9)

I am rocking 350-370 unmerged. If your 310 was unmerged that's still pretty decent and would probably get higher once you are more comfortable with the fight :)

As far as parser settings go: A 30 second encounter timeout will give you correct unmerged parses for T9. Keep that in mind while you're shuffling beads around on your abacus to come up with DPS numbers :cool:

Quote:

---

Originally Posted by Artiste

-snip doing the math-

---

That 280+ number is probably a decent baseline for DPS. Our group's first T9 kill a few months ago was SMN288/BRD282/DRG279/MNK257/WAR183/PLD108/SCH53 and still had over a minute and a half before enrage.

Hmm, I'll make sure to ask whether it was merged or unmerged and try to improve on it. There was times during the final phase in which I shuddered focusing on thunder (I always miss it) and end up using a Blizzard 3 with 1200mp left. Oh, I did die once towards the end of the battle, but I don't think it would have caused a huge DPS drop that late int he fight. (first divebom set, second divebomb)

Quote:

---

Originally Posted by Argentt

As far as parser settings go: A 30 second encounter timeout will give you correct unmerged parses for T9. Keep that in mind while you're shuffling beads around on your abacus to come up with DPS numbers :cool:

---

I can't emphasize more on this and this leads to the below

Quote:

---

Originally Posted by Artiste

I know a good bard can do 350
Blm is kinda same as bard I think, around 350
A Monk/dragoon is much more, like 380 ? 400 ?

---

Try 30s timeout and see the numbers for yourself. The amount of downtime in T9 will make MNKs cry. 380-400DPS on T9 for a MNK? Maybe try again with BiS FCoB gear.

Quote:

---

Originally Posted by OneWingedSora

I am rocking 350-370 unmerged. If your 310 was unmerged that's still pretty decent and would probably get higher once you are more comfortable with the fight :)

---

What is your group DPS on those attempts?

Quote:

---

Originally Posted by HaroldSaxon

What is your group DPS on those attempts?

---

Dunno but other BLM was 320, BRD was 317, DRG was 270 and other BRD 240.

I know what you are trying to imply, the lower the groups DPS as a whole the better the DPS for me because of the downtime before Megaflare and yeah that's true. That's why T9 should never be used for DPS measurement unless you are parsing phase by phase.

Quote:

---

Originally Posted by Ooshima

Try 30s timeout and see the numbers for yourself. The amount of downtime in T9 will make MNKs cry. 380-400DPS on T9 for a MNK? Maybe try again with BiS FCoB gear.

---

Quote:

---

Originally Posted by OneWingedSora

That's why T9 should never be used for DPS measurement unless you are parsing phase by phase.

---

Which makes the most sense.

If multiple people want to parse then simply roll with default settings from the beginning. (Granted different parsers starting with the same settings). Once one person tries to tweak their parser to feel better about their DPS, it throws other ones off unless everyone does the same.

There is no benefit in trying to factor in downtime other then to try and create the illusion, your DPS isn't as bad as it may be. It just artificially lowers everyones dps number to account for a time when, you're DOING NO DAMAGE because NOTHING IS TARGETABLE, (by shortining the distance between party members), which yes is very silly. Saying someone did over 60 DPS more (on a merged parse), vs someone doing only 20 -30 DPS more (unmerged equivalent), influences the way it sounds so that one isn't as bad as the other, despite them both being equal.

Not to mention the general public as a whole parses phase by phase. Good luck bragging about your 280+ DPS, not many will sit and do the math for you even if you say it's unmerged. You're just gonna wind up looking bad. Parsing phase by phase also helps you target your weaker areas easier in a fight where as unmerged does not.

Unmerged parsing also gives you a fake number for what your really doing in damage over time. The actual damage you do in game is higher on average then what a parser factoring in downtime will show you. (in terms of the DPS number, by forcibly injecting time). You wouldn't even be able to accurately reflect your own numbers from a training dummy to gauge yourself correctly, because there is no forced downtime on a dummy to factor in to your parse.

Quote:

---

Originally Posted by Havenchild

Snip

---

Then you may have to understand most top FC players parse merged. Merged does not artificially lower your encounter DPS but it is unmerged that artificially increases your DPS.

This is because phase by phase benefits jobs that have high initial burst. For example, BRDs. It is pointless to parse unmerged and go say, hey wtf MNK why your DPS suck you are like more than 50 DPS less than me because some jobs just take time to build up. Setting a downtime of 30s will produce a single encounter parse that judges everyone's overall performance over the full duration of the battle.

Quote:

---

Originally Posted by Ooshima

Then you may have to understand most top FC players parse merged. Merged does not artificially lower your encounter DPS but it is unmerged that artificially increases your DPS.

This is because phase by phase benefits jobs that have high initial burst. For example, BRDs. It is pointless to parse unmerged and go say, hey wtf MNK why your DPS suck you are like more than 50 DPS less than me because some jobs just take time to build up. Setting a downtime of 30s will produce a single encounter parse that judges everyone's overall performance over the full duration of the battle.

---

You bring up a good point about MNKs for example but at the same time a merged parse and an unmerged parse wouldn't matter regardless in that arguement because the DPS they do per each phase would be the same as well the difference is you wouldn't be able to see it. Anyone who knows how Monks work, would understand that the loss of Greased Lightning is huge to their DPS.

If someone was gonna criticize the class based on that, they shouldn't be handling parsers to begin with. If the monk in question knew he was doing worse in a particular phase, he could attempt to adjust his rotation or save cds. It's way more effective to know where your weak spots are to improve your overall DPS. What use does a blanket number have if it can't tell you where your loss is happening accurately.

You can also judge the overall performance of people in a merged parse by simply wiping your parser every run and just clicking an ALL tab.

Quote:

---

Originally Posted by Havenchild

Snip

---

Similarly, any end game and top flight raiders who worth their salt should know where they would suffer a loss of DPS. For example, MNKs don't need a parser to tell them that downtimes and dropping GL3 would screw them, and a BLM who knows their shits would know how to control their UI/AF so that they don't start the next part in a crappy state and suffer a loss in DPS. These people knows where they would drop their DPS, but would want to know how much.

So we are still back to a culled (downtime included) parse because we want to know what is on the table. We want to see the whole picture, as a fight itself. Because you want to see how each person, or even each job, can bring to the table throughout the whole of the fight. Let us use an extreme example - say maybe FCoB will have a turn where there are short 3 phases at first, maybe lasting 1mins each and the final phase just drags until the boss is down. However, the rDPS demanded from this fight is so high like how T8 started to be, your grp decided to go solo tank with 5DPS. Now, which DPS job shall your off tank change to?

If you judge by unculled (downtime not factored) then high burst DPS like BRD would likely to have some inflated number to fool your judgement if you believe it blindly. However, we know that jobs like MNK would build up DPS overtime but if you see the fight as unculled (downtime not factored) then the first 3 phases would penalize MNK already. So is bringing in MNK better or BRD better? You need to see the MERGED parse to be able to tell which job brings the most to the table for the entire length of the fight.

Inflated numbers from the unculled (downtime not factored) also gives a wrong feeling for fights with downtimes, and at the same time, a hard enrage. T9 for example. You hit the hard enrage for "some reason" and then you look at the parser and say, everyone seems to be doing 300DPS and why the hell are we hitting hard enrage? That's because the 300DPS marker is a culled (downtime included) marker for the whole T9 duration with downtime factored. Which means that you need to push harder than 300 during phases where you can engage the enemy, and don't be too happy when you see 300DPS for an unculled (downtime not factored).

It is also like, hey I am doing 350++ DPS during phase 1 part 1 but why are we getting 5 or even 6 meteor instead of 4? That's because your downtime timer is at a default 6 or below and doesn't take into account of the downtime during the dives.

There is a reason why people at the top and crunching the numbers stick by culled (downtime included) parses.

Quote:

---

Originally Posted by Ooshima

snip

---

You basically just proved my exact point of why a merged parser is better then unmerged for all the exact same reasons I stated.

Quote:

---

Originally Posted by Ooshima

There is a reason why people at the top and crunching the numbers stick by merged parses.

---

You said it yourself, now you're arguing with nobody.

Ooshima do you realize that you are completely contradicting what you are saying? Merged parses do not include downtime( Merged Parse is simply each phase merged together), yet you are arguing that they do, I think you are confused. Un merged is one single parse with all downtime, Merged is not. There is really no difference in the two, other than downtime. You can easily take the Damage done in a Merged parser add the downtime that was not there and get an UN merged all downtime included dps. Either way the numbers are the same. All that matters is when comparing numbers both parties that are comparing should know whether or not downtime is being included or not.

okay I am indeed confused with the terms you guys used, cos I am used to call it culled and unculled. :(

I am saying that the culled (single parse with no parts, downtime included) should be used.

It was in response to those who said factoring downtime is pointless like this post

Quote:

---

Originally Posted by Havenchild

There is no benefit in trying to factor in downtime other then to try and create the illusion, your DPS isn't as bad as it may be. It just artificially lowers everyones dps number to account for a time when, you're DOING NO DAMAGE because NOTHING IS TARGETABLE, (by shortining the distance between party members), which yes is very silly. Saying someone did over 60 DPS more (on a merged parse), vs someone doing only 20 -30 DPS more (unmerged equivalent), influences the way it sounds so that one isn't as bad as the other, despite them both being equal.

Not to mention the general public as a whole parses phase by phase. Good luck bragging about your 280+ DPS, not many will sit and do the math for you even if you say it's unmerged. You're just gonna wind up looking bad. Parsing phase by phase also helps you target your weaker areas easier in a fight where as unmerged does not.

Unmerged parsing also gives you a fake number for what your really doing in damage over time. The actual damage you do in game is higher on average then what a parser factoring in downtime will show you. (in terms of the DPS number, by forcibly injecting time). You wouldn't even be able to accurately reflect your own numbers from a training dummy to gauge yourself correctly, because there is no forced downtime on a dummy to factor in to your parse.

---

Well like i stated, it doesnt matter. Both parses are correct. I personally parse with a megred parse. I like to see phase by phase dps, has nothing to do with inflating number. Cause for me to figure out what dps I did with downtime is very simple. Add downtime to parse, take damage done and divide it by the number of seconds with the new added downtime. Boom now i have a parse that includes downtime. There is no need to agrue over who is parsing correct and who is not. Just simply know whether the person you compare with is using downtime or not and adjust so you cam compare correct numbers. There is only one wrong way to parse, and thats not parsing at all.

Quote:

---

Originally Posted by Hakmatic

Well like i stated, it doesnt matter. Both parses are correct. I personally parse with a megred parse. I like to see phase by phase dps, has nothing to do with inflating number. Cause for me to figure out what dps I did with downtime is very simple. Add downtime to parse, take damage done and divide it by the number of seconds with the new added downtime. Boom now i have a parse that includes downtime. There is no need to agrue over who is parsing correct and who is not. Just simply know whether the person you compare with is using downtime or not and adjust so you cam compare correct numbers. There is only one wrong way to parse, and thats not parsing at all.

---

That is one way to see it, but only works if you know the exact downtime to the seconds.

Deaths is also a downtime to factor in, if any.

But I can agree that parses should be compared with the same conditions.

Just re-watched my stream of this weeks 6-9 clear and our numbers for T9 at various points in the fight are this: (9:40 total encounter duration, all unmerged, so you can see the large drop between phases. This was also near 5 1/2 months of clearing for us, so there's a decent balance of familiarity as well as slacking!)

DPS right before the double meteor stream occurs: (To get an idea of DPS before any significant downtime starts occuring)

464(SMN) 408(MNK) 401(SMN) 380(BRD) 340(combined tanks) 183(combined healers)


DPS When phase 2 starts: (Only ~20 seconds after the above numbers were taken - this gives you an idea of how quickly downtime affects your numbers)

413(SMN) 364(SMN) 358(MNK) 352(BRD) 312(combined tanks) 165(combined healers)


DPS When phase 3 starts: (Nael just landed after casting Mega Flare, for a reference of how phase 2 and it's downtime affects numbers. DPS just before she lands, as well as DPS after the first action on her is made)

308 before she lands >> 278 after first combat action(SMN) 302>>274(combined tanks) 285>>261(BRD) 280>>255(SMN) 278>>251(MNK) 119>>109(combined healers)


DPS at the start of the first divebombs, as well as DPS after the first action is made when she lands:

313 just before divebombs>>304 after combat starts when she lands(SMN) 312>>308(combined tanks) 309>>301(BRD) 300>>294(SMN) 280>>272(MNK) 119>>113(combined healers)


And lastly, DPS when she dies: (We have no more downtime after she lands from the first divebombs. Boss dies right when Lunar Dynamo is cast after the 2nd set of divebombs)

332(combined tanks) 319(SMN) 315(BRD) 300(SMN) 284(MNK) 114(combined healers)

This was with a 9:40 kill time. Basically in the end, don't use T9 to measure DPS. :P

For example, we waited about 10 seconds after killing the second set of golems before we got the 10 second countdown for Mega Flare, meaning it was more of a 20 second countdown rather than 10. Other groups may have a longer wait, and some groups will kill the last golem just as flare goes off, so the downtime difference between these groups is huge.

As a reference, the same (DPS) players above pull on average about 450 to 415 DPS (450 being highest player, 415 being lowest) in T6 with a kill time of 3:30ish, as well as T8 with a kill time of 7:30ish. I recommend using one of those 2 fights to gauge your DPS if it's something you feel matters.

I haven't seen so many illiterate people in such a long time. If you guys ever bothered to read Ooshima's post before the edit, it was painfully obvious what was meant in it. Anyways the solution is simple, just have 1 member with an early cull and 1 member with a long cull and bam you now have more information. The most consistent way to compare DPS across groups in t9 is still unculled + kill time though.

Quote:

---

Originally Posted by OneWingedSora

Dunno but other BLM was 320, BRD was 317, DRG was 270 and other BRD 240.

I know what you are trying to imply, the lower the groups DPS as a whole the better the DPS for me because of the downtime before Megaflare and yeah that's true. That's why T9 should never be used for DPS measurement unless you are parsing phase by phase.

---

That, and triple flaring the add is overkill and stat padding your numbers with useless DPS (which you said you did in another thread).

I'd be very interested to see the proof and to see how much Foe's you get with two bards, because that is obviously going to be a large factor. Same with your kill time.

Quote:

---

Originally Posted by HaroldSaxon

That, and triple flaring the add is overkill and stat padding your numbers with useless DPS (which you said you did in another thread).

I'd be very interested to see the proof and to see how much Foe's you get with two bards, because that is obviously going to be a large factor. Same with your kill time.

---

Foes are up pretty much the entire fight and kill time is around 11-11:30 mins.

Also, triple flaring the add isn't really useless DPS since the other DPS stay on Nael while I triple flare then LB immediately after.

Quote:

---

Originally Posted by OneWingedSora

Foes are up pretty much the entire fight and kill time is around 11-11:30 mins.

Also, triple flaring the add isn't really useless DPS since the other DPS stay on Nael while I triple flare then LB immediately after.

---

Do you have Seleen up also?

Quote:

---

Originally Posted by HaroldSaxon

Do you have Seleen up also?

---

Sadly not.

Sigh... Don't understand why people are arguing whether to measure the temperature of drink by degrees Celsius or degrees Fahrenheit is better. :)

Merged DPS lets you gauge better in between phases how much your doing, aka hitting a sand dummy. Unmerged DPS (aka encounter DPS) measure your DPS over the whole encounter. A good DPS worth their salt will work to maximise both.

What matters is just which stat do you use to compare with which. It's like one person saying I'm 100 degrees, another one says no I'm 212 degrees you suck, but the one at 100 degress is measuring with Celsius and the one at 212 degrees is measuring with Fahrenheit, and for the uninformed it's both equal temperature.

Some like it easy to just take Merged DPS, since it allows you more awareness of which phases are your strengths and use it to compare everything from there on. Some like to use unmerged DPS to see your capability to react to the encounter. Both are not wrong so seriously stop arguing which one is better. Gosh.