THF Savage Build For Multi-damage Types In Odyssey

Looking for more job flexibility in segment farming, and considering building a Savage Blade setup for my THF to give it flexibility between piercing and slashing damage. Curious- is it worth/smart to offhand Centovente in such a build, or is the faster TP gain and accuracy benefits from offhanding Blurred+1 a better move.

Centovente all the way. This is my savage blade set. Pretty much BiS all around

ItemSet 381127

You could swap out nyame for relic +3 if you dont have it ranked yet, but this is essentially the build you're aiming for. Ideally you would have 30 strength on the cape, but honestly theres no need to build another if you dont want to. The standard rudras cape with 30 dex and 10% wds works just fine. That's what I use myself and I have no issues at all.

I've found in a short (30 minute) events where you usually have a bard getting in 1 (if not 2 on a nice wild card) 1hour set of songs and being buffed to the *** gills, you're getting tp so fast that I would assume that centovente offhand would be my go to. Do I have math? no. but I use tp bonus offhand with blu with no struggle with accuracy and again, would assume that it shouldn't be a problem.

Now, if this is just you going in solo with Thief and are relying solely on trusts for your buffs, disregard everything I've typed.

Yeah this is for 6man party setting for farming segments in Sheol:C. I use Thibron for my offhand in Savage Builds on RDM, as it can benefit from Temper II for multiattack paired with several pieces of Malignance Gear for accuracy/STP/survival- much the way a lot of BLUs I know will use Tizona to get their multiattack paired with Malignance (with the same benefits I see on RDM).

But if I'm going to build a Savage set on THF, I won't have those two advantages for getting multiattack and was concerned I might need something like Blurred+1 offhand to be competitive in TP gain compared to those situations. Glad to get info from those who have done it vs me who's just theorizing.

On that same track - Thief (mastered) naturally has 19% triple attack (while rdm has 0, at 500enhancing magic set you have 20% TA) so its basically the same.

Dammit I always forget about the gifts along the way. THANK YOU!~

While this is true, good melee RDMs have 600+ enhancing skill for Temper II (640 is the highest reachable tier right now, I believe) for 30-34% TA without gear.

Yeah my Temper2 set sits at 600+, but not the true cap (I've been lazy with the UNM sword for +enhancing skill), so it is more....but 19% is a damn good start and allows for a better mix of malig/TripleAttack gear than I was thinking would be needed- in particular when just a few key items available to THF can bring that in line right quick (such as Gere Ring and using an Adhemar+1 body to get the needed DW+some TA to boot.)

Do you have a stacked set for it as well?

Swap seething bomblet +1 for yetshila +1 and Nyame helm for artifact hat +3. Those are the only changes you'd make for a stacked set.

Definitely dig the set- some of us (myself included!) will have to make adjustments, as in my case not all my Nyame is WSD path, but in most cases those swaps are pretty obvious.

I did have a question- you have Aurglemir Orb+1 winning against Seething Bomblet+1? Seething+1 has more attack, definitely more accuracy, and a lot more STR, losing DEX and VIT (no factor for an unstacked Savage Blade) and STP.

Raw stats if no Nyame. I would go 5/5 Relic +3 if I didn't have nyame myelf (in fact that's what I did use before I started ranking it). Plunderer's attire +3 has the highest raw stats all around, with loads of attack, accuracy, and str / mnd stat mods. Nyame is better still, but plunderer's +3 works just fine.



You're right. I just checked my set in game and that's what I have set for the ammo. I threw together the itemset from memory and overlooked that one. Most of our weaponskills are dex mod, but savage is str/mnd. That's an oversight on my part. Itemset updated accordingly. Thanks for pointing it out.

R0 Gleti probably better (than relic+3), sameish base stats and the pdl

Lower mind, some higher str, attack dependent ofc

EDIT: Never mind, you edited that post to specify relic +3. I thought you were referencing nyame when I wrote this. R20 Nyame easily outdamages R20 gleti's, but good point on the gleti's versus relic comparison. If you don't have nyame ranked up yet gletis is a fine alternative to relic.

At this point I'm basically using naegling the vast majority of the time. SU5, vajra, twash, tauret.... Just sb the whole game.

Not at attack cap, which you might actually be at while segment farming.

At attack cap, Gleti's body destroys Nyame. Also Crepuscular Pebble wins with Seething. With Gleti's and Pebble in, PDL starts building diminishing returns, but at this point Gleti's Legs are still significant gain over Nyame Legs, but not as obvious as Gleti's body (First big PDL piece adds the most, since there is almost no diminishing returns for PDL on THF, since it has no PDL on neck like some jobs, so there is only Pebble).

stackable % bonuses dont have diminishing returns. they have constant returns. your base damage +5% = x
your base damage with +50 pdl, 5% more pdl add exactly the same amount as if you had no pdl. Its just less of an increase over the +50 set vs the +0 set.

Diminishing returns means adding +5% would get you less added damage when you have +50 already than +5% would with 0 PDL.

It may have opportunity cost. But it cannot have diminishing returns. Thats impossible for a % based increase to anything.

Is that a bait? :)

Obviously it has a diminishing returns. You are completely wrong here. % increase would be if you would add 5% of CURRENT value, not the 5% of base all the time.

So:
+0.05PDL
+O.0525PDL
+0.055125
etc.

Here we are NOT doing any % increase, its actually a static increase of 0.05PDL. We can say it's +5% if it's only PDL you add, because base is 1.0, but it's actually just +0.05. If you add another +0.05 having base of 1.5, not 1.0, then it's no longer +5%, but less and it's exactly why it's a diminishing return.

EDIT: It's diminish returns in increasing your damage, just to make it clear.

EDIT: Real life scenario for this would be you have a bakery. you have 10 workers. Each of them can produce 10 breads.
Your production is then 100 breads.
You add a worker. you can now make 110 breads. You production increased by 10%. You add another worker. You can now make 120 breads. Your production however increased by 9.09%. Same thing happens if you add +0.1PDL, then another +0.1PDL.

This is false. You made the calculation wrong. By definition constant returns to scale occurs when an increase in inputs give the same proportion increase in output.

Supposed that you have 10 workers making 100 breads. Adding 1 worker increase the bread production by 10 to a total of 110, hence 10%. Adding another worker increase production by another 10 to a total of 120. A ~9% increase in production for a ~9% increase in inputs (1/11=10/110).

So a 5% of 100% is 5/100 but another 5% is 5/105. So technically your inputs decrease. This is not diminishing returns because you still get the same proportion increase in outputs.

Another way to look at this is to double the inputs and see if the outputs double. Of course, this assume no other constraints. So a 100% increase in inputs leads to 100% increase in outputs. A 20x5% should leads to a 20x5% increase in outputs.

In terms of math, constant returns to scale is a straight line.

f(2x)=2f(x)

Increasing returns to scale is a curve upwards.

f(2x)>2f(x)

You are right and I'm sorry for being a sheep and following this nomenclature (lots of people was wrong for years I guess) for years. Sorry Lordtrey too.

That being said, while naming was wrong everything else is still correct. Your damage will raise by the same static value, but it will be less efficient the more you add, because of higher base. I have no idea now how to call it now :)

I dont know maybe its a US vs EU thing but diminsh returns has overseass always means that you get less and less percentage increase as you add in same amount of improvent over and over again

So you have 100% attack rate you add 5%pts da you now have a 105%pts attack rate which is 5%more
you ad in another 5%pts da you now have an attack rate of 110% which is only ~4.8% improvement.
The returns on adding in 5%pts da has diminished

Adding in 100%pts DA is still doubling the attack rate to 200%

or maybe im misundertanding your argument

Speaking of diminishing returns either in regard to the raw values or the percentage increases are both valid.

In FFXI, most stats are of the latter, but there are a few of the former, like INT for nukes.

I completely agree with this and offer the subject of Stat weights and thresholds demonstrated in other MMO's as the evidence. At the beginning of an expansion or gear tier typically you have 1 Stat to chase no matter what it will always be best, but once you reach a threshold value, the amount that the other stats give (even if not particularly beneficial to your specialization for damage), will still give better dividends than stacking your primary stay will ever be able to provide.

And to relate this back to XI, your primary seeking values are WSC Properties for WS, MAB/PDL(Whichever is applicable), and WSD.

Stacking 100 CHR on a 100% CHR WS, will always give a solid benefit, but at what point would dropping primary Stat for pdl/mab or wsd be worthwhile, it really comes down to a balancing act especially with our horizontal gearing.

Best real game example I could come up with was messing with my warrior for Bumba. Knowing I'm going to be attack capped 5/5 sakpata was the thought, but I instead opted to keep nyame feet and relic head and 3/5 sakpata.

5/5 sakpata would give me 176 Str, 124 mnd, 30 PDL, and 0 wsd in those slots,
3/5 sakpata with relic head, nyame feet, gives 172 Str, +134 mnd, 21 PDL, and 20 wsd in those slots.

I'm not saying I'm right or wrong in this choice, particularly give the fact that most of the rest of the ws set has other forms of wsd. What I'm saying is the act of balancing stats is always in an attempt to gain more than we lose. The question comes down to, is 20 wsd and +6 wsc stats worth -9 PDL. Which is the true higher value. In a weaponskill set only consisting of those 5 pieces I would argue the latter set is the clear winner, but that's not the case in reality.

It's all true, but apparently it's not called diminishing returns. Logic is good but naming is wrong, because as Andret pointed out when you add 5%DA second time, you increase input by 4.76% (0.05 is 4.76% of 1.05) and get the same output.

So performance of adding 5% DA falls, but it's not because of diminishing returns, but because it's just keeps being a smaller input. I have no idea how to call that tho :P

It's just a constant return. Its not a increasing return like values of haste in ffxi nor diminishing like investing in bitcoin.

so, the idea of diminishing returns has always had a shaky relationship with FFXI crowds. Andret's post here kind of illustrates why.

Diminishing returns (in economics) refers to the observable decline in the productivity of a system as you add more labor. In Andret's bakery example, each new worker adds exactly 10 more baked goods, but in reality it works like this.

workers 1-5: each produce 10 baked goods
worker 6: only produces 9 baked goods
worker 7: only produces 6 baked goods, and now worker 5 only produces 9.

That is diminishing returns, adding extra labor doesn't provide an extra static benefit, but diminishes the actual overall productivity.

FFXI community has always looked at increases in terms of total DPS, and defined "diminishing returns" as a represantation of what a bonus gives in terms of your overall DPS. So adding 1 DA when you're already at 99% has diminished gains over adding 1 DA when you're at 0%. It's pretty accepted now within the community that that is what diminishing returns means in FFXI.

Bigger number always better no need big word say something simple