Perdu Scythe

[Ruin]

Is Perdu Scythe still the best scythe for drk? (In opinion)

Before I waste the IS on it, was wondering if its still the best dmg overall, I know a few new scythe's released last update from wotg NMs. But does Perdu still stay on top?

[Starr]

I think it might be easier to use the search function than ask people to repeat themselves for the 10th time but I dunno.

[Ruin]

Eh its opinion...

Molione's DMG95 Delay501 STR+3
VS
Perdu DMG93 Delay501 ATK14 ACC5

[Vodevil]

I wouldn't say that Perdu was ever definitively on top as far as the non relic endgame Scythes go. The difference between Perdu and the super cheap Tredecim Scythe is rather neglible and can go either way. From what I've read, the new scythe/ring set is in a relatively similar position. In short, use whatever you like better as there is no definitive answer.

[Baha]

I don't understand why the KI crowd dissed Perdu, but goes crazy over Molione set.

Moliones's set
95DMG 501dly, 5acc 7STR, enhance souleater

Perdu + Flame
93DMG 501dly, 5 acc 5STR 14atk

2 DMG 2 STR vs 14 atk? Might be a little better, but when Perdu came out it was all like "waah waah Trece is better". Guess the dismissive people didn't have the assault rank. or something ;x

[Bagel]

I'm no DRK, but speaking as WAR, if they made a crit+ G Axe with slightly higher base dmg instead of 9 less than Perdu, I'd say you were crazy not to use it. In other words, I'm pretty sure I'd use Tred if I was DRK.

[fantasticdan]

Is the enhances souleater set bonus the same extra 2% as on the JSE body/AF head? If it's nothing impressive I'll just keep using tred, lol.

[Raen Ryong]

Perdu Sickle is a LOT faster at hitting than Tredecim Scythe, faster and faster as you get more Haste. This difference is further shown when using Desperate Blows, when you are essentially minizerging at 80% haste most of the time. In a recent parse against another well-geared drk/sam using Tredecim (mamool camp) not only was my accuracy slightly higher due to Perdu, but I did approximately 40 more hits within the duration of the party (a measly 30 minutes), also allowing me to do many extra weaponskills also. My average melee/weaponskill damages were also much higher. There really wasn't much competition.

I don't see why everyone hates on the Perdu Sickle, and it's not the same as a straight +CRIT on WAR; Guillotine cannot critical and if you miss/naturally critical a 13th hit you've just negated Tredecim's advantage. The DMG increase normally doesn't negate the large attack bonus from Perdu Sickle either.

I don't understand where all the hate for Perdu Sickle comes from, but I can understand the argument that Tredecim is nearly as good and doesn't eat your IS etc etc; I however own both and always use Perdu unless I know for sure that I'll have to store TP =/

EDIT: And yes, 2% I believe.

[Kirschy]

Perdu Sickle is a LOT faster at hitting than Tredecim Scythe, faster and faster as you get more Haste. This difference is further shown when using Desperate Blows, when you are essentially minizerging at 80% haste most of the time.

This is actualli wrong. At higher haste builds the difference in delay between Tred and Perdu becomes smaller and smaller. To do the math at 80% in attack speed haste:

at 0%
Tred - 528 Delay
Perdu - 501 Delay
Difference - 27 Delay. Slightly less than half a second faster swinging.

at 80%
Tred - 528 * ( 1-.8 ) = 105.6 Delay
Perdu - 501 * ( 1-.8 ) = 100.2 Delay
Difference - 5.4 Delay. Thats less than 1/10th of a second difference.

Take a well equipped DRK. Turban/Speed/Homamx3 = 20% Haste Eq. 15% Spell, 20% March x2, 10% Hasso. 65% Haste in a good March x2 merit party.

at 65%
Tred - 528 * ( 1.-.65 ) = 184.8 Delay
Perdu - 501 * ( 1-.65 ) = 175.35 Delay
Difference - 9.45 Delay. Thats slightly less than 1/6th of a second difference.

When you factor in a fast-paced merit party where mobs are dying fast, the delay difference between the scythes become negligable. Whats important to factor are the following:

The delay differences affecting TP gain in a 6-hit setup.
2 DMG + Crit+% vs Acc+5 Attack+14 ( I believe )

I personalli have never been able to parse the two or compare them in my setup. I've been asked plenty of times which is better, and honestly I just don't know. They both seem acceptable endgame Scythe and both parse very well.

[Raen Ryong]

Hmm, that's odd. It does make mathematical sense like that, but I don't know how else to attribute the ~40 hit difference, especially since I was casting Dread Spikes more (with hasso/seigan down obviously)... I was just basing it off the fact that lower delays get more % benefit from haste? Or did I do the math wrong? <_<;

[Tulun]

As it's been said, the reason why Tred is better is simply because of the 6-hit build. For merit situations, you should sub sam, and Tred needs very little beyond /sam to achieve the 6-hit build. Perdu requires extra gear that may be less than ideal or hard to obtain. 501 delay is a curse, not a blessing.

I think Perdu will shine more on HNM, where you'll want that extra acc and attack, assuming you go DRK to HNM.. and assuming you can even melee it without it owning you with TP moves.

[Raen Ryong]

You can 6hit with Perdu, or 5hit + Guillo.

I do it with StoreTP from Rajas + Brutal alone <_<

EDIT: on /sam sub of course.

The whole "x hit" thing isn't an issue unless you miss a Guillo; you can always get a Meditate in on the first "100% TP round" in any fight situation.

[Kirschy]

It's why people always denounced the delay -3% strap in high haste builds. Haste isn't cumlative /w lower delays. The extra hits could easily come from swapping to the next target consistantly faster, although it's hard to say for sure why you parsed more hits.

As for Store TP/6hit build on Scythe. I honestly don't know the numbers in my head. My friend is a career BLM but leveling DRK. He asked which Scythe he should use, and I know he won't have a Rajas which might be a factor in which Scythe he ends up /w.

[Godlywargod]

Perdu Sickle is a LOT faster at hitting than Tredecim Scythe, faster and faster as you get more Haste. This difference is further shown when using Desperate Blows, when you are essentially minizerging at 80% haste most of the time.

This is actualli wrong. At higher haste builds the difference in delay between Tred and Perdu becomes smaller and smaller. To do the math at 80% in attack speed haste:

at 0%
Tred - 528 Delay
Perdu - 501 Delay
Difference - 27 Delay. Slightly less than half a second faster swinging.

at 80%
Tred - 528 * ( 1-.8 ) = 105.6 Delay
Perdu - 501 * ( 1-.8 ) = 100.2 Delay
Difference - 5.4 Delay. Thats less than 1/10th of a second difference.

Take a well equipped DRK. Turban/Speed/Homamx3 = 20% Haste Eq. 15% Spell, 20% March x2, 10% Hasso. 65% Haste in a good March x2 merit party.

at 65%
Tred - 528 * ( 1.-.65 ) = 184.8 Delay
Perdu - 501 * ( 1-.65 ) = 175.35 Delay
Difference - 9.45 Delay. Thats slightly less than 1/6th of a second difference.

When you factor in a fast-paced merit party where mobs are dying fast, the delay difference between the scythes become negligable. Whats important to factor are the following:

The delay differences affecting TP gain in a 6-hit setup.
2 DMG + Crit+% vs Acc+5 Attack+14 ( I believe )

I personalli have never been able to parse the two or compare them in my setup. I've been asked plenty of times which is better, and honestly I just don't know. They both seem acceptable endgame Scythe and both parse very well.

501/528 = ~.949

175.35/184.8 = ~.949

100.2/105.6 = ~.949

same gap percentage wise.

(means you'll still get the same % more hits no matter the haste)

[Ryda]

Perdu Sickle is a LOT faster at hitting than Tredecim Scythe, faster and faster as you get more Haste. This difference is further shown when using Desperate Blows, when you are essentially minizerging at 80% haste most of the time.

This is actualli wrong. At higher haste builds the difference in delay between Tred and Perdu becomes smaller and smaller. To do the math at 80% in attack speed haste:

at 0%
Tred - 528 Delay
Perdu - 501 Delay
Difference - 27 Delay. Slightly less than half a second faster swinging.

at 80%
Tred - 528 * ( 1-.8 ) = 105.6 Delay
Perdu - 501 * ( 1-.8 ) = 100.2 Delay
Difference - 5.4 Delay. Thats less than 1/10th of a second difference.

Take a well equipped DRK. Turban/Speed/Homamx3 = 20% Haste Eq. 15% Spell, 20% March x2, 10% Hasso. 65% Haste in a good March x2 merit party.

at 65%
Tred - 528 * ( 1.-.65 ) = 184.8 Delay
Perdu - 501 * ( 1-.65 ) = 175.35 Delay
Difference - 9.45 Delay. Thats slightly less than 1/6th of a second difference.

When you factor in a fast-paced merit party where mobs are dying fast, the delay difference between the scythes become negligable. Whats important to factor are the following:

The delay differences affecting TP gain in a 6-hit setup.
2 DMG + Crit+% vs Acc+5 Attack+14 ( I believe )

I personalli have never been able to parse the two or compare them in my setup. I've been asked plenty of times which is better, and honestly I just don't know. They both seem acceptable endgame Scythe and both parse very well.

Being familiar with your position as the master haste math I hesitate here...but I have to disagree.

I don't like when terms like raw delay difference and difference in time between swings are used because it doesn't provide useful information unless contextualized by the pre-modification delay or time to swing.

I think fundamentally people care about the % increase in their damage and those numbers used don't directly relate to that question. I can say "I now have one second less between my hits" but that statement could mean something totally different if i was previously at 2 seconds between hits or 4 seconds between hits. In the first case I doubled my DoT, in the second I increased it by 33%....very different results with the same language used. The same is true of using delay...what is better -10 delay or -5? You dont know unless I tell you that the original number is the same in both cases. if your delay starts at 100, subtracting 10 is a small change....if you start at 20 subtracting 5 is huge.

This leads to accurate, but misleading to the average reader, statements like "At higher haste builds the difference in delay between Tred and Perdu becomes smaller and smaller." Its true because the bare value of delay does become smaller, but that isn't the relevant number. And the implication that the difference in the delay helps you less that I'm sure everyone read into that is wrong also. I think the most acurate way to say it would be that no matter what your haste is, Perdu is about 5% faster than Tred...you will hit about 5% more often.

I guess I'm just making a point about the terms people use to describe these things and how they often imply something incorrect.

[Ryda]

501/528 = ~.949

175.35/184.8 = ~.949

100.2/105.6 = ~.949

same gap percentage wise.

(means you'll still get the same % more hits no matter the haste)

Wow....or i could have just said that
I talk too much.

[Kirschy]

While I mostly agree /w what youre saying, I think you have to visualize the actual context in which it's used. I was talking about this the other week /w someone, so let me copy paste the silly diagram:

Lets say I wanted to test how 5% haste affected my SAM, in a 2 SAM pt~
Let's say we're both using 300 Delay weapons. (300/60 = 5 seconds per swing.)
SAM A /w no Haste has 300 Delay(5 Seconds)
SAM B /w 5 Haste has .95 * 300 = 285 Delay(4.75 seconds per swing)
Now imagine it took 20 hits total to kill the mob

SAM A:
|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|
SAM B:
|-----|-----|-----|----|-----|-----|-----|----|-----|-----|

Even though SAM B had 5% haste, it didn't help him get any extra swings or TP in this scenario. 5% haste and saving a quarter of a second on the attack swing, wouldn't help unless you attacked the mob 20 times (In which case you'd swing a 21st time when the non-hasted person would hit 20 times.)

Just imagine if you will instead of 5% haste, 5% difference in direct delay and the argument still holds. In a good TP burn party, you don't get to swing enough to make that small delay difference matter. Switching your target half a second faster than someone else ends up making a bigger difference than the small delay. This is how I see it.

EDIT: Just to add, although I've done a lot of research in Haste. I can often visualize things wrong. I'm far from always right.

[Ryda]

While I mostly agree /w what youre saying, I think you have to visualize the actual context in which it's used. I was talking about this the other week /w someone, so let me copy paste the silly diagram:

Lets say I wanted to test how 5% haste affected my SAM, in a 2 SAM pt~
Let's say we're both using 300 Delay weapons. (300/60 = 5 seconds per swing.)
SAM A /w no Haste has 300 Delay(5 Seconds)
SAM B /w 5 Haste has .95 * 300 = 285 Delay(4.75 seconds per swing)
Now imagine it took 20 hits total to kill the mob

SAM A:
|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|
SAM B:
|-----|-----|-----|----|-----|-----|-----|----|-----|-----|

Even though SAM B had 5% haste, it didn't help him get any extra swings or TP in this scenario. 5% haste and saving a quarter of a second on the attack swing, wouldn't help unless you attacked the mob 20 times (In which case you'd swing a 21st time when the non-hasted person would hit 20 times.)

Just imagine if you will instead of 5% haste, 5% difference in direct delay and the argument still holds. In a good TP burn party, you don't get to swing enough to make that small delay difference matter. Switching your target half a second faster than someone else ends up making a bigger difference than the small delay. This is how I see it.

This seems like a completely different point, but ill go ahead and disagree with this also.

Imagine a mob with 200 HP, and 2 melee both hit for 50 every time. In this case it would take a huge amount of haste for on person to hit more than the other. They would both hit for 50..the mob would have 100 hp left...then the quicker guy would hit once...the mob would still be alive...and the the second guy kills. same damage 100 each.

But this isn't how parties work. Everything doesnt line up perfectly like that because with variation in times to reach mobs, variations in damage done and such...you cant get everything nice and neat and sterilized of randomness.

Instead look at it like this.

SAM A:
|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|
SAM B:
|-----|-----|-----|----|-----|-----|-----|----|-----|-----|

If the mob dies----------------------------------------------^ Here: no difference realized
If it dies--------------------------------------------------^ Here: one full hit extra realized

When you get down to the end there are no fractions of hits, either you hit last, or the other guy does...a full hit.
If i have 1% haste, on one fight i might get 100% of a hit extra...but because its only 1% haste, the odds that this will occur are very small. After 100 hits, i *will* get one extra hit...after 50 hits i have a 50% chance of getting a full extra hit....at 10 hits i have a 10% chance of getting a full extra hit. If one hit for me is...lets say, 100 damage...then I get 1% increase in my damage regardless. Whether i add 100 damage *always* after 100 hits...or 100 damage 50% of the time after 50 hits...it still is a 1% increase.

If someone misunderstands this it will be because their model is overly sterilized and does not represent the random nature of a fight of nearly any length. This randomness leads to the point of mob death being randomly placed at some point between those hit intervals shown visually above. The width of the gap between 2 players Xth hit represents the likelyhood the mob death will occur at a point favorable to the quicker person, and because even though that gap gets smaller as the mobs dies faster, his relative benefit is constant because the benefit of a full swing increases as the mob dies faster also.

Now someone say this in 3 sentances and make me feel silly again.

[Ryda]

Now imagine it took 20 hits total to kill the mob

SAM A:
|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|
SAM B:
|-----|-----|-----|----|-----|-----|-----|----|-----|-----|

Ok, maybe I can make that point without talking so much...like this:

Now imagine it took 19 hits total to kill the mob

SAM A:
|-----|-----|-----|-----|-----|-----|-----|-----|-----|-----|
SAM B:
|-----|-----|-----|----|-----|-----|-----|----|-----|-----|

[Kirschy]

Yea that makes sense. I didn't think of it like that. It's hard for me to conceptualize that such small differences in actual delay can realli mean much when mobs are dying fast, WS go off mid-delay, and people switch to targets at different speeds.