For The Shinies! A Guide For Thief

Ok, thats +5 gear.. +8TH including job trait.

so assuming +8 total TH. if we want to proc th12; your results are stating that +8total and +12total will get us there at the same rate?

I include the job trait when I say TH +8 in gear. Perhaps I should phrase it as TH +8 WITH gear, but it's pretty standard that TH +8 == TH 8. My apologies if that wasn't clear.

That works. How about the other question?

If the goal is to max proc the NM to say, TH12. Does th8 and th12 provide the same results?

yes, its close enough to not matter

I guess im struggling to understand the maths :P

I see that zero th13/14 procs happened with only wearing th8, versus th14.

Am I looking at it wrong? It looks like only wearing th8 you cannot obtain th levels 13/14.

thats significant?

You have JP test thats shows T14 with TH8.
Its just so rare, that mob usually dies before it happens.

Yes, you are looking at that wrong. Don't take any individual mob kill as absolute. There is variance. Just as flipping a coin 10 times does not yield an even distribution of 5 heads and 5 tails every time, so too does this proces have variance. You don't always get a proc every approximately ~225 hits. Sometime you might not get one in 500 hits. Sometimes you might get two in 50 hits. Luck plays a role here. It's the whole point of random. The Treasure hunter 13 and 14 totals were lucky runs, and they also had larger sample sizes. I hit a TH 14 value once, and it was after 781 swings. The longer you swing at a mob, the more likely you are to see a proc. It's simply volume. More hits is more chances to proc. Notice how after I resumed the testing with my updated TH 14 set the average TH value went down...but so did the total number of swings per mob.

Look at the number of swings in each kill. Not just the final value. And realize how the variance of random number generator plays into that.

Ok.

So we are arguing that we "might as well" wear th8. because the gains of wearing more are deminished and DPS gear outweighs the marginal th gains in wearing more?

I'm just looking for it simplified. Is the result of the test to show us that more TH does NOT increase TH proc, or is it to show us that the increase is nominal?

The results are within margin of error and not big enough sample size to be definite about anything.

40 Mobs is insanely small of a sample size. The data is basically worthless.

The JP data is more ample, but still kind of small when we consider the % chance of this occurring. We can at least conclude that if there is an effect from having higher TH gear, it is incredibly small.

I find this argument moot though because if you are playing to increase TH level there is no reason you couldn't put on any TH+ Pieces you have within reason. What are you worried about the DPS loss from swapping into Empy boots and swapping out your ammo slot / belt ?

The 40 mobs is not the sample size though.
its the 20.000+ attacks (minus maybe 40 form being fist attacks) that the simple size

Misunderstading the data with a factor of 500 will lead to bad conclussions

Not sure why this is so hard to understand.
this is not a per mob test or function.
Its a per swing test of a per swing function

The evaluation you want if something is worth it is no is subjective and situational.
However the test did show that in 20.000+ swings. having TH14 instead of TH8 gave non measurable difference in TH proc rate.

The measure difference was 3.09% difference which is concidered within margen of error on a 20.000 samples test

if you want to bet on those 3% extra proc rate by going in full th14 gear that is of cause your choice

remember its not 3% chance of going from th8 to 14.
its 3% higher chance on the chance of getting just one increase in TH. if you really want to push the a favorable conclusion from the data.

Okay so I need some suggestions for gear for my TP set.

This is my current one
http://www.ffxiah.com/item-sets/378471

A few things, working in ginsing? the ammo from the tenzen fight, just havent got it yet. I know I need to upgrade the pants to +2, will once the new months ambu rolls around for the +1 voucher. The cape isn't done but I'm working on the TP one 20dex/30acc/20atk/10stp/-10%pdt

As far as offhand goes whats the next upgrade past this? I have a Ternion +1 and was told its better than shijo if the DW is a dead stat(isnt at the moment because I dont have enough Job points) but is the 2% more TA better than 15 dex, 4% crit and 5% crit damage after being upgraded? Seems like shijo is just forever best outside of needed the extra acc from Tern. Also is that my best option for an offhand aside from REMA? Also any other suggestions for other pieces that are not TOO hard to get? The TP sets in the updated guide stop at mid level and don't go beyond that. Any input would be appreciated! TIA! :D

(also tauret in mainhand dunno why the gearset put it on OH)

the gearsets dont seem to put weapons in the correct hands most of the time, i've noticed. it would also help others that can help you to say what your augments are. like you could have 10 dw or 10 stp or w/e on your cape, and that would affect the advice.

You have to like... mess with it to get weapons into the right slot. It's kinda funky.

You can add multiple weapons to a set, until they appear in the right order.

IE, add Tauret, then add a Shijo, if it pops infront of Tauret, add another shijo, and another shijo and another until it pops behind tauret. Then delete all the other shijo if you want to.

Just a note. If we assume that each TH level is progressively harder to obtain, than this is a function of swings and mobs. Right? You can't say over 20k swings got me only TH10 5 times but then disregard how many mobs you fought. All that matters is the number of swings up until that level of TH, the rest are negligible for these calculations.

The test should be per mob, per TH level, per amount of swings.

Testing is so hard(tedious) to actually do, and do in an ideal setup so wouldn't wish it on anyone.

But at end of day if 3% is within margin of error for 20,000 attacks/rounds, what would the number of swings for it to be outside be? 200k? 2M? Idk how to determine that, but its such a low chance to increase TH I don't see a test actually being done that is outside of margin of error.

If the goal is just to see if TH over 8 matters when the TH on mob is higher than 8, focusing on results at only TH9(or maybe TH10) probably would be easiest as it will have more data and we are looking at one standard "base" rate.

The other thread looks like it is coming up with an addon that could make things easier.

If you could just gather data while people do their every day thiefing things, it could be pretty easy to get enough data to prove something over a large period of time.

An add-on or parser modification would be immensely useful. Otherwise, the only way to test this would be use Feint/SA/TA on mobs that die within 2 hits using 1H dagger only.

This assuming Feint/SA/TA is a straight multiplication of proc chance and only need to test TH8->TH9

Correct.
and that Valid point has been brought up.
the fact that more swings landed on a mob and thereby incruede a TH upgrade from the more swings to a levle with lesse procrate mesna more swings on a lower procrate.

but that is not a sample size issue that is a limitation of the test.


now we could argue that 3% increase in procrate when the proracte of the swing goes down due to the higher TH could indicate mayab a possible increased procrate that was hidden in the test.
but thats not proof... yet

Which is something i plan ontesting tomorrow

optimal setup would be
600swing on a mob each time so everyting is equal

In short: valid argument but we need to test it before drawing conclussion on that part.

All this talk about TH procing and all I ever do is tag th and straight back to my DPS set...

seems like a waste of energy if you ask me...

opinions and value is subjective and doesn't relly belong to a debate on metric and objective analyzes.

but i would tend to agree with you.

It's always situational. I see it both ways. Sometimes I don't give a flyin *** about the parse and go for TH, sometimes I care about killing it faster and fitting more kills into a time frame.

It's merely time management. Sometimes it's optimal to go more kills per hour and sometimes it's more optimal to stack as high as possible because you can't get more kills per hour. It's most important to know when to do what.

Just questioning my self here

im under the impression for a DPS oriented setup mina twashatart + tp bonus dagger or ternion+1 (replaced with the oddysey dagger) is the optimal DPS set

However BG guidke still states that Anus/twatstar seems to be the more optimal approuch

I dont have the spreadsheet so anyone that could clarify this for me . Thank you

Aeneas is behind Twashtar/Gleti's and significantly behind Twashtar/Cento.

What would an updated Tauret Evisceration set look like with the new armor? Also, how does the Gleti knife fit in?

Simon, can you share your AE set please?