Updates? In my ruke LJ?

[RKenshin]

I would be interested to see a parse with hagun and with a more common food like say coeurl subs. As he stated a few times he is very near the damage cap with Relic. I think if hagun were used the attack difference with berzerk would result in more than a 9 damage per hit average.

The relative difference should still be about the same, but yeah I'd still like to see a well-done comparison with Hagun. A few people I know are planning to give it a good shot in the future, so we'll see!

According to the delay math though, Hagun should get slightly more of an advantage.

Indeed, the jumping would get tiresome rather quickly, especially without merited timers. I'm fully merited and have pretty much everything save Ares's Gauntlets, Hecatomb Cap, Mars's Ring, and Gungnir, so I could probably do the tests in your stead. I haven't done much in the way of hard testing things like you have outside of math theory, but I imagine I would just need a sufficiently large sample size of Jumps, High Jumps, and melee swings (control) with and without Drachen Greaves +1 keeping stats at a consistent baseline, and repeating at a noticeably increased level for each stat I'd want to test, correct? What sample size would you recommend for such a test?

I'm capped on both jump merits too, but just missing those greaves. I'm also only 1/5 on Ares, hoping to get that further soon!

As for a sample size, it's hard to say. I know this isn't very helpful but basically if you get an extreme amount of variation, you'll need a large sample size. Otherwise, smaller is OK. But it should at least be big enough to establish a high/low and an average for both situations. I'd probably shoot for 50 jumps each, but in reality ~20 should also be OK for a decent comparison.

[Sanjiyan]

Do you think any of the results from this study could be applied to other melee/DRGs? Specifically MNK/DRG is what I have in mind.

I've know MNK/DRG to be quite viable and a fun alternative to MNK/WAR for a long time now, and it's definitely better than MNK/NIN, but recently I've met some more resistance to /DRG as a subjob. My reasoning for MNK/DRG is similar to yours, more haste (capped without using feet slot which allows dune boots useage), more accuracy (allows easier use of meat dishes), and jump/high jumps for more TP. It's less overall haste compared to SAM due to hasso, and loss of suvivability (over /NIN at least, no seigan for MNKs), but still better than /WAR because of high jump. I'm also not sure if any pieces of usukane plays into the equation, since I don't have usukane. However, MNKs do have high VIT for better jumps, not to mention kick attack procs on them which is not influenced by higher DA. Like SAM, MNK also has to trade berserk for /DRG. Then again, loss of defense from berserk is inconsequential if counterstance is used full time, though I don't think a MNK/WAR or MNK/DRG should use counterstance full time.

[Sylphet]

/nerd mode on

For anyone who wants to perform tests there is no set number that will prove to be "enough". If you want to be scientific about it you'll need enough to get not only an average but an average with a standard deviation and the differences should be analyzed for statistical significance using something like a t-test. This sample size can be as small as three if the difference is statistically significant. Most of the time it will need to be larger. In the scientific world statistical significance can be measured in several ways but the most common method is a p-value as determined by a (for example) t-test. Anything that is less than or equal to a t-test score of 0.05 shows statistical significance. That means that the probability of getting those results by chance (the null hypothesis) is 5%. Here's a link to a site that will perform a student's t-test using your raw sample numbers for you.

http://www.physics.csbsju.edu/stats/t-test.html

For instance (these values are completely made up)
Jump damage using gear set 1 =
350
320
345
375

Jump damage using gear set 2 =
500
530
450
375

gear set 1: mean = 348 +/- 50
gear set 2: mean = 464 +/- 68

Plug those sets in and you get a p-value of 0.017, which means that there is a statistical significance to those means, which allows you to say with high confidence that the variables (in this case gear) were responsible for the differences seen.

ok sorry for the science lesson, continue

[yrpaine]

*wonders who bet ruke he couldn't sam's across all servers to try /drg*

[AkashiXI]

I don't even have /drg leveled or SAM @ 75 but I read the entire thing (even all the formulas), and that was spectacular, Ruke;o.

[Nickolias]

Have you truly found Alky's to be more beneficial than AF+1 hands on SAM, even given the new findings?

[Max™]

I'd think that with the silly ass accuracy of 1 hit WSes, coupled with gorget, the -13 Dex for 4 Str isn't that bad of a swap, since it isn't crit based.

Though, AF +1 hands for Jump/High jump might not be a bad idea, more Str towards fStr, Dex for landing crits... oh! Bibiki shell is nice for it too.

[GSDragoon]

Where did you get the -33% hate with high jump from?

[shirak]

Upon usage, hit or miss the user will lose 33% of his/her enmity accumulated during the course of battle.

Equipping Wyrm Brais increases the amount of hate that is shed. (from 33% to 50%)

-wiki

[Harutaru]

*watches the price of wyvern earrings soar*

[Ayn]

Get back in your drum.

[GSDragoon]

Upon usage, hit or miss the user will lose 33% of his/her enmity accumulated during the course of battle.

[quote:2xw5ygt5]Equipping Wyrm Brais increases the amount of hate that is shed. (from 33% to 50%)

-wiki[/quote:2xw5ygt5]

Ok... but I still haven't seen any data backing the information up.

[Kaeko]

If there is a way to force a 'miss' on high jump I can test this for you really easily.

Blind Pots are 100% proc like Para Potion maybe?

Also, awesome LJ post =P

[GSDragoon]

If there is a way to force a 'miss' on high jump I can test this for you really easily.

Blind Pots are 100% proc like Para Potion maybe?

Also, awesome LJ post =P

Jump with a weapon you have little or no skill in on a higher level mob?

[Sylphet]

Also, awesome LJ post =P

::starts a Kaeko LJ update fan club::

Seriously, since we're gushing over LJs here I look forward to yours probably more than anyone else's and it has improved my knowledge of FFXI game mechanics more than any other. Thanks

Ruke's is ok too.

[aurik]

If there is a way to force a 'miss' on high jump I can test this for you really easily.

Blind Pots are 100% proc like Para Potion maybe?

Also, awesome LJ post =P

Blind pots are ridiculously strong. If they're not -100%, they're like, -256 acc or something. Blind pots + weapon you don't have skill in = 0 accuracy.

[Kirschy]

Interesting read for sure~

As for Jumps. SA stacks with high jump and jump. This lowers the pdif into a very small range and makes your accuracy more precise with fewer tests. (Before the 3.0 cap was removed, Jump was shown to have a 50% VIT mod.) I remember hearing High Jump had an AGI mod, but was never tested... Should be fairly easy to though. (I'll prob do it tonite or tomorrow if no one else does.)

If Blind pots have an enormous static accuracy down, I can try also try tests on High Jump enmity reduction.

[Kaeko]

I couldn't get high jump to miss lvl0 mobs even with lvl0 skill and blinding pots.

Have to use a higher level mob.

[Max™]

The Kraken in Qufim?

Static level 40 mob.

[Tempestwolf]

Damnit Ruke, now people are going to sub DRG other then me =x /angst.
On further note, those same people wont /tell me going "Lol/DRG".
Yay LJ though.