.DAT Mining: 4/7/2014 Update - Findings and Discussion

[Darkmagi]

This was "tested" already with the rieve harp... Why would be different with this harp?... Only I can think those who tested where really bored and had no interest in any other weapon to use the items in the harp, already having daurdabla.

Btw those ppl who understand the packets things, is possible to make the game think you have Clarion all the time(no idea but someone using ashita did enable widescan for all the jobs before SE gave us that ability) or makes it thinks you got daurdabla99 equipped.... Or the mythic dagger....

As far as I know, you can't just give yourself buffs/gear without having it, but you were able to use set blu spells on any job as well as make items sell for whatever gil you wanted at vendors, so who knows!

[Sechs]

Edit:

I need to learn to read.

[kjLotus]

As far as I know, you can't just give yourself buffs/gear without having it, but you were able to use set blu spells on any job as well as make items sell for whatever gil you wanted at vendors, so who knows!

Status Effects, statistic modifiers and equipment are not decided by the client in any way - you cannot change what the server thinks your status effects are (it does not ask the client for its status effects)

[dasva]

That idea falls under "fucking stupid" and "really fucking stupid". A 90 daur bard constantly having 4 songs up would set off the reddest of red flags, redder than the reddest red times infinity. Hacking widescan is hardly detectable, whereas having a constant extra song would draw attention very quickly.

But 90 duar brds can 4 songs constantly in a lot of situations legitimately

[Foldypaws]

Plus I can't think of any situation involving anything where that would even be a good idea.

Given the naked-except-for-TH-gear thieves I've seen slowly whittling down Behemoth, I'm pretty sure people would do it there. But yeah, other than that I can't think of any fight where you're completely unworried about time, *and* pops are the limiting factor.

[NynJa]

But 90 duar brds can 4 songs constantly in a lot of situations legitimately

without actually having clarion up, without an endless cycle of not having the 4th song drop?

[Keichan]

Why is it even a debate? We already had evidence from Camaraderie Harp that it wouldn't stack with Daur. Moreover, SE specifically stated that this new harp would be an easier replacement for Daur.

Like others have said, if this did stack with Daur, it would have been a major glitch and fixed right away.

Fine, I did not know that there was a previous test with a reive harp and would have changed my opinions.

Considering there is diminishing returns as more songs (effectiveness per song) are added as well as a 12 (roll/song) cap, alliances couldn't even utilize this many songs at one time anyway. Will stop here.

without actually having clarion up, without an endless cycle of not having the 4th song drop?

They'd have to use clarion at the start, but yes, once the song placeholders are up, you can overwrite them over and over again with Gjaller songs. Only way to lose the 4th song is for one song to get dispelled or you let one of the songs drop.

[Yugl]

People getting pissy because someone tested an item and found results similar to previous testing? Find better shit to argue about morons.

[Trumpy]

The devs used the term "closing the gap" for non dura brd and dura brd. to me closing the gap doesnt mean givin dura brds more shit, that would widen the gap if anything. But the testing has been done and its been found not to add anything extra so why is there still discussion?

[dasva]

without actually having clarion up, without an endless cycle of not having the 4th song drop?

Well if it was up constantly the 4th song wouldn't have dropped now would it

[Snprphnx]

Can't we just agree that the new harp is equal to a 90 Daurdabla, for a properly geared BRD?

[Prothescar]

several pages of overly verbose shit saying what everyone already knows and people taking offense to said shit, great. it's a great harp for people who have a bit of money to spend and not a lot of time. only thing a 90 Daur would be better for is like... horde lullaby because accuracy and radius and like 6 extra seconds on troub or something.

there you have it, a succinct explanation of where things stand that would easily avoid pages worth of bickering.

[Kohan]

With that said, let's not have any more of the bickering. There are other things to talk about, and you're scaring off all the customers.

[shepardG]

Kohan, my girl, great idea:

Martel or anyone who would know:

What's best way to test this shield out? the HP cap on reprisal is still the same, so that is "enhanced". Would it just be like adamas probably? functionallity of an Ochain +1 with reprisal, less than an Ochain without reprisal?

[Martel]

I still question the claimed block rate of Adamas with reprisal. I never saw any conclusive testing. And my personal tests on adamas showed increased reprisal spikes dmg. Which further increases my doubts about the block rate. Two different enhancements seems unlikely, unless reprisal spikes dmg is somehow linked to the block rat e potency.

However, I never managed any good reprisal block rate testing, either. So I can't exclude the possibility. Although, how ppl would be able to measure adamas' enhancement without first knowing base reprisal block mystifies me.

But anyway, new shield.

Kohan, my girl, great idea:

Martel or anyone who would know:

What's best way to test this shield out? the HP cap on reprisal is still the same, so that is "enhanced". Would it just be like adamas probably? functionality of an Ochain +1 with reprisal, less than an Ochain without reprisal?

Was that is there supposed to be an "isn't"? The context is kinda conflicting there.

If so, the we can mark off one possibility. Regarding testing methods..

Reprisal testing sucks. hard. There's a lot of obstacles involved in getting a good sample. It's mostly to do with the lack of a good way to parse reprisal up hits vs reprisal down hits. And even if you have the buffs to full time reprisal, when tanking 10+ mobs for speedy data collection, you do often get interrupted. Then non reprisal hits start seeping in... It's frustrating to put so much work into a sample, only to doubt it. And it's tiring. Reprisal use is more difficult to automate, particularly when trying not to let it go down.

shield testing difficulties

Spoiler: show

For more general shield testing you just have to be able to tell the blocks from the non blocks without doubt. For larger size shields this is simple by looking at the dmg taken. There will be a gap between the ranges for blocked and non blocked hits. But for small shields like beatific, the ranges overlap, making it hard to tell was was blocked.

More recently, battlemod has made it possible to see blocked hits via the block animation packet. So that eliminates the need for hits to do dmg at all, and makes testing easier. except.... Kparser and battlemod do not get along. So you're left gathering data via logger. Block rate is not to terribly difficult to determine via a few searches through logs(count block, count hits, etc) but measuring dmg- on block becomes extremely difficult. You'd have to gather the average block and non block dmg taken by hand... for thousands of hits.

However, Thanks to a programmer friend of mine, I think I have a decent way of finally of filtering out non reprisal hits from logs. So expect lots of reprisal testing sometime in the near future.

I wonder if a shield block parser could be made as an addon, kinda like scoreboard..? would be epic. But it's a bit beyond my lua abilities atm.

[Prothescar]

Could probably find a way to filter non reprisal blocks by having the parser ignore any blocks that arent parsed with spike dmg text coming after them, however i dont have the programming knowledge to make that happen or know if it's feasible

[Martel]

Spikes are a valid way of getting blocked hits during reprisal. But you need the non blocked hits too, and only the ones while reprisal is active.

[Foldypaws]

More recently, battlemod has made it possible to see blocked hits via the block animation packet. So that eliminates the need for hits to do dmg at all, and makes testing easier. except.... Kparser and battlemod do not get along. So you're left gathering data via logger. Block rate is not to terribly difficult to determine via a few searches through logs(count block, count hits, etc) but measuring dmg- on block becomes extremely difficult. You'd have to gather the average block and non block dmg taken by hand... for thousands of hits.

Shouldn't be very hard to write a little script to pull out blocked vs non-blocked hits, sum, and average them.

In fact, if you've got logs of the above sort, I'll do it for you.

[Martel]

Shouldn't be very hard to write a little script to pull out blocked vs non-blocked hits, sum, and average them.

In fact, if you've got logs of the above sort, I'll do it for you.

This is pretty much what the log parser my programmer friend wrote does. He just updated it recently to only gather hits/blocks while reprisal is up. So I should be able to get some good reprisal data.

Unfortunately, all my old reprisal tests predate battlemod. And I don't load battle mod during normal shield tests either cause gathering data using Kparser is just easier. So I don't have any reprisal logs lying around.

But I'll go gather a series of reprisal tests as soon as I have the time(probably this weekend) and make a nice big reprisal post.

[dasva]

Spikes are a valid way of getting blocked hits during reprisal. But you need the non blocked hits too, and only the ones while reprisal is active.

Or just have a pocket brd to make sure reprisal is up full time lol. Was thinking of doing something like that to try and test out except I'm still an old fart with windower 3.4 so sorting block and unblocked without spikes is hard. Even with I have no log parsers so still not fun. As far as the adamans testing went iirc it was more of an eyeballing of without reprisal it blocked like normal shields did back in the day and with it it was like ochain. Obviously there is no way in hell you could quantify that but by the same token the next best block rate shield at the time didn't get those kinds of rates with just reprisal... though since it was eyeballing I'd imagine it probably could easily have been near aegis blockrate