The Odyssey - || Strategy And Discussion ||

Seemed like they just disagreed with that premise.

And it'll tell you if its 10% 12.5% 20% 25% etc

And you get to find out if offensive ja/spells count

And you'll have pictures.

What happened to that FFXIAH strat guide that existed for Ody? The one that was sorted by NM. I can’t seem to find it.

Formerly this thing:
https://www.ffxiah.com/forum/topic/56314/the-odyssey-v15-nm-focused-strategies/

Now this thing:
https://forum.square-enix.com/ffxi/threads/59053

Is what you're referring to?

Oh ya

I forgot it got deleted. RIP

I have no doubt that the wall exists and with WAR + MNK it seems to be about 25% when Tomahawk isn't up. The GEO (club) and BRD (staff) also jump in once the first add spawns but they don't get TP anywhere near as fast as WAR and MNK do. There is no way it requires 20 actions for each WS to prevent -87.5% damage. It's likely that someone else WS's knocks all the others down a tier.

Well would be cool if JA/offensive actions did reduce the wall penalty. I'll put it into practice until I or someone test/verify the validity of it. Would mean jumps could be more strategically placed I guess.

Had no idea the 10s wall thing was a bug. Good to know

You don't need exact data and rigorous testing to glen useful information on how the wall works. I just performed a test on V25 Dhelan Dhe right now with my dancer. The test is very simple and easy to replicate the methodology on any job

For this test I mainhanded twashtar and alternated between pyrrhic kleos and dancing edge. The two weaponskills are very similar. Dancing edge is a 5 hit weaponskill with a 40% dex/40% chr mod and a 1.175 fTP that transfers across all hits, whereas pyrrhic is a 4 hit weaponskill with a 40% dex/40% str mod and a 1.75 fTP that transfers across all hits. Pyrrhic is a bit stronger than dancing edge but not by a whole lot. Since multi attacks affect these weaponskills quite a bit I weaponskilled in my rudra's set to minimize extra swings for consistency.

The trusts I summoned were cornelia, moonbeureaux, cheryukki, and kuppipi. None of those 5 trusts will ever engage, so the only weaponskills used were mine alone. Trusts did not interfere.

I started by alternating between dancing edge and pyrrhic, then at one point I spammed 10 pyrrhics in a row, and then went back to alternating the two. This is the results.



I separated the section where I intentionally walled myself to maximum with 10 pyrrhics in a row. After that I went back to swapping between the two and you can see that pyrrhic unwalled itself naturally and completely (or as completely as it's going to get). So while the specifics of how the wall are still a little elusive, I can say pretty conclusively that you only need TWO separate weaponskills to keep the wall to a minimum.

So, recap, if my understanding is correct, just 2 WS back and forth isn’t subject to the "wall". If you somehow hit the "wall", a round of 5 WS back and forth will reset the “wall”, so the 6th WS is back at full potency.

Thank you

2 back and forth should be losing 10% (why wildfire would be better, for exact percentage) physical ws's aren't the best eyeball

But it's staying at -10% and not compounding resistance

Monberaux doing Samson's Mix also contributing a bit to ws's up/down, another thing someone will inevitably mention.

I agree with this. I'm certain there is still SOME wall in there, but it's effects are so minimal that they don't affect the fight. As long as you have some form of alternation it's easy to keep the wall from building up to the higher tiers of resistance. 10% wall is not an unsurmountable barrier. 50 or 75% though... that would be a f%$#@kin nightmare.

There's testing on long_horned's twitter around December 24th 2022 in sortie.

https://twitter.com/long_horned/status/1606672844783767554
https://twitter.com/long_horned/status/1607130255889952769
https://twitter.com/long_horned/status/1607129158752305153

He tests using Atonement with known damage. (1547)

0 Repeat 0% (1547)
1 Repeat -10% (1392)
2 Repeat -25% (1160)
3 Repeat -40% (928)
4 Repeat -60% (618)
5 Repeat -85% (232)

He does mewing lullaby 5 times and it brings the damage back to 1547.

He then tests Atonement > Mewing Lullaby > Atonement > Mewing Lullaby > Atonement...(repeat)

1547 > Mew > 1392 > Mew > 1160 > Mew > 1160 > Mew > 1160...(repeat)

Next he does 6 Atonements to give it -85% then cycles Atonement/Mews.

232 > Mew > 618 > Mew > 928 > Mew > 1160 > Mew > 1160... (repeat)

Then he tests Atonement > Mewing Lullaby > Nether Blast(?) > Atonement > Mewing Lullaby > Nether Blast...(repeat)

1547 > Mew > Nether > 1392 > Mew > Nether > 1392...(repeat)

He states that non damaging moves such as mewing lullaby or stunga(Shock squall?) don't count if they get resisted.



Whether it's the same in Odyssey I can't say for sure but I would expect so. I think it may just be counting repeat WS in the last 6 WS and walling them accordingly.

e.g.
WS1(0%) > WS2(0%) > WS1(-10%) > WS2(-10%) > WS1(-25%) > WS2(-25%) then stays at 25%

WS1(0%) > WS2(0%) > WS3(0%) > WS1 (-10%) > WS2 (-10%) > WS3 (-10%) then stays at 10%

The problem with physical WS as a test is that you can't really discern if you're taking a 25% penalty or a 10% penalty or even no penalty with much reliability. Given long_horned's test, which is much more controlled, it seems likely the same is happening with Melliny's test. But, before long_horned's was posted, it would have been very hard to conclude 25% from Melliny's data alone.

This seems to fit every sample.

Other considerations would be what else counts, is it only blood pacts? Quick draw? Trust WS? Resisted shock squall doesn't count, but does a 0 damage WS count[you could feasibly throw out random ws to erode wall]? Is a skillchain counted as it's own entry? It seems unlikely that magic counts if that is the criteria, since it'd be highly visible.

Holy cow. Data driven, meaningful discussion on ffxiah. I like it. We didn't have to call anyone stupid or disparage them either.

And it took less than 15 minutes

Exactly which part of this do you think took less than 15 minutes...?

(Had it been wildfire, it wouldve been cleaner)

I wasn’t trying to pin down exact numbers with my test, merely show the difference between the model thorny proposed on the previous page and what I’ve been seeing in practice. Not only is it not easy to eyeball 10 or 25% reduction on a physical weapon skill, but multi hits like the ones I used have an inherent variance based on not just random pdif but multi hit procs. The data on long_horns twitter is much more definitive for nailing down specific values.

The testing methodology he and I used was the same process though, and my numbers supported his as well.



This does appear to be the way it works

In case of multihit WS or multiattack procs on WS, is it safe to assume the WS wall applies to the overall WS damage (i.e. each hit) and not just the first?
And then there's TP transfer WS and those who don't TP transfer, the behaviour could be different against the WS wall.

For instance if the WS wall only applies to the first hit, or additional hits exclusively for TP transfer ones, it would explain why the difference seems smaller than it should for some people, wouldn't it not?

It’s the whole weaponskill. Just look at my numbers. When I did 10 pyrrhics in a row the last one dropped to less than 400 damage. That’s clearly getting hit with the 87.5% penalty on every hit. If only the first hit was affected pyrrhic, smite, etc would be nearly unaffected by the wall, and that isn’t how it works. You can see how drastic the latter pyrrhics in the 10 consecutive uses dropped in damage compared to the rest. The difference is huge. There isn’t even any room for speculation here. The data flat out screams every hit is affected.

This is exactly how I'd say it works from eyeballing. MNK WAR do about 75% of normal ws damage if just using 2 WS on Ngai. If you add in a 3rd like RDM doing BH it's 90% of normal ws damage.

So... does this mean WAR MNK should alternate 2 WS each for max damage, or are their 2nd best ws too much weaker than Judgment and VS that it isn't worth it?

I tried Arebati V20 on RNG yesterday for the first time, with R1 Foma. My average WS was 15k only spamming Last Stand. I didn't check the COR's gear but I was still doing 65%+ of the damage on the parse. The DPS felt incredibly slow with the timer. I don't have a DPS monitor plugin, but if I'm shooting and WSing at max speed, what kind of average WS would the RNG need to be pulling to clear if it can be calculated this way?

I'm working on my gear but I will not get anni R15 any time soon (something like 180M total from scratch now?). Would it be possible to clear with R15 Foma on RNG only? I could always go BRD, but in case another shout comes up that needs RNG.

I'll address this question first. The answer is yes. I've cleared V20 Arebati on Rng with R15 Fomalhaut. In fact, my first ever clear on V20 (back before V25 existed) was using fomalhaut. So it's definitely doable. Back then the strategy was rune tank, bard, rng, cor, igeo, sch with the bard kiting the pig and the scholar did caper emissaries to transfer hate to the rune when I reached the hate cap. Since then people have adopted the Pld, Run, igeo, cor, brd, rng strategy instead. If hate is an issue on V20 with fomalhaut only and you find yourself reaching the hate cap because you don't have coronach, you may have to consider the former setup (it's really challenging though)


Speaking from personal experience, if your average last stand damage is 15K on ranger something is wrong. Since you mentioned that your fomalhaut is only rank 1 I suspect that you have some gear improvements to make, but it's also possible that buffs were wrong. I highly suggest you use the following buffs

Bard -- SV NITRO Honor march, 3 minuets, and swift etude
Igeo -- Geo-fury, Indi-Agi, and you can entrust whatever
Cor -- Crooked chaos roll, samurai roll

You should be capping -enmity down naturally just with capped hover shot stacks, and a lot of ranger's best mid shot gear has -enmity down too (ikenga's), so dirge is generally a wasted roll. If the corsair reaches the hate cap in that fight something's gone wrong so it's just a matter of managing your own hate mainly (I highly recommend you look into building an annihalator for the record. Coronach's static enmity is really really important in this fight).

Your ranged attack should hover in the 8k+ range at all times. When bolster goes up you should have 9,999 attack. Do a /checkparam once the fight starts. If your ranged attack is hovering in the 6k or lower range then buffs are wrong and you need to fix them. Arebati has extremely high defenses, and anything less than 7000+ ranged attack probably isn't good enough for that fight. It takes a lot of ranged attack to get damage numbers high enough to take him down. It's normal for the ranger to do 65-70% of the parse damage. But even still your numbers are very low. Once I cap my hover shot stacks I routinely pull 30-32k coronaches on vengence 25 arebati, and coronach is weaker than last stand. The last time I did vengence 20 I did actually use fomalhaut and last stand and the damage numbers still stick in my memory. 45-55k last stands were the norm for me on him once hover shot stacks were capped. V20 only has an innate -20 DT whereas V25 has an innate -25 DT, and V20 also has lower stats in general. So from the sounds of it you have some work to do on your mid shot and weaponskill sets.

I recommend you check out the ranger forms. There's a lot of discussion on the past 10-12 pages or so on Arebati and a lot of up to date gear sets are posted. I've got my midshot and weaponskill sets in there somewhere. There's some discussion on Sarv as well, but plenty of info regarding annihalator, coronach, last stand and general strategy. Much of it is V25, but that's still relevant for V20 since the strategy people use nowadays is often the same as the second KI of the V25 version (Pld healer, rune tank, cor, rng, Igeo, Brd...just with only 1 pig instead of two). It's good to get some practice on V20, because that fight is hellishly intense on 25. I've only cleared the 25 version twice now, though I'm trying to help a third group through it and we've come under 10% for that one. It's doable, but it requires every bit of effort from everyone involved. It's the highest DPS check in the game outside the new master trials.

From the guide:
Perun +1
Nusku Shield
Fomalhaut (Augmented)
Chrono Bullet
Orion Beret +3
Amini Caban +3
Nyame Gauntlets
Nyame Flanchard
Amini Bottillons +3
Fotia Gorget
Moonshade Earring
Amini Earring +2
Regal Ring
Dingir Ring
Fotia Belt
Belenus's Cape
30 AGI, WSD, R.Acc/R.Att

Only difference is I was using Nyame body and feet (R25) since I didn't make any emp+3 yet, and my 2nd earring is only +1. I might have been using Cornelia's Ring instead of Dingir too iirc.

I didn't check buffs or pay much attention as it was just before sleep so can't comment on that. Looking at lots there are lots of 10k, 16-17k, and a 6k (missed 1st hit I guess). I don't know why it was so extremely low with the little difference in gear to this. 50k becoming 15k average with a bit less gear doesn't seem right.

This is correct, and with that gear 15-17K last stands on V20 are way too low. Something is wrong with those figures. Now that I think about it, does vengence 20 even HAVE the weaponskill wall? It doesn't does it? They added the weaponskill wall to v25 fights but the lower vengences don't have it right? In that case I highly recommend you do a checkparam and see what your ranged attack is. If the buffs are wrong or weak and your ranged attack is only like 5k or lower then that's also a posisbility. You need REALLY high ranged attack to overcome arebati's defenses. Strong support is a must for this fight. With that gear I would expect 40k+ last stands on V20 Arebati MINIMUM once you reach capped hover shot stacks. The problem is not that gear set.


It's not. I'm intimately familiar with this fight. You're losing a little bit of damage by not having fomalhaut fully ranked up, but not THAT much damage. Look into the buffs and see what your ranged attack is, and make sure you aren't accidentally resetting your hover shot stacks. You should be doing double that damage on average, or even slightly higher with that gear.

It was just a random pug so nothing I could check. I will make the emp+3 pieces and try again some time, but I couldn't understand why damage was that low either. No WS wall on V20.

I thought not. Your support had to have been doing something wrong then. There's no way buffs were right. That gear set should be pulling 40-50k last stands with proper buffs on V20 arebati. Someone wasn't doing their job correctly and it wasn't you.

Checking log I had Minuet 4+5 and HM, but then Minne5 (wtf) and Dirge. GEO was doing Geo-Fury but seemingly no Indi spell. I thought something was wrong.

I highly suggest you just spend 15 min as a group at the moogle, and let the bard practice (all of you can be afk). Songs might not overwrite the way he want it. Song order/overwriting is finicky.