## Different in gemming with RoRo

9 posts
• Page

**1**of**1**### Different in gemming with RoRo

Please bare with me on this one as I will try to explain as best as I can.

I've been looking around and reading alot, and checking high end geared ferals to understand how exactly they change their gems when obtaining RoRo. and I was experimenting with this one using Catus alot, but I am not seeing the result that makes me change the gems especially that Catus is showing me different result when I click on the feature "Simulator: Quick compare" thats when I add a + new identical character to mine but I change its gems which therefore changes the whole stats of agility + mastery + crit + haste

Anyway let me get into more detail.

First then I got RoRo trinket, I obviously stacked higher haste, so I changed my yellow gems to an orage "Agility + Haste" and kept red on pure agility gems, and kept blues as Agility + hit.

The druids I seen in armory always use Orange agility + secondaries on reds, and pure secondaries on yellows, and secondary + hit on blues.

so you can see the huge different there in the two scenraios above. both will have RoRO reforges which will make the secondary stats as mastery + 1 = crit = haste

but the different in both scenarios would basicaly be that one will have much higher agility, and a tad lower mastery:crit:haste ratios

on the second scenario you will have obviously lower agility, higher secondaries mastery:crit:haste ratios. <-- which will make RoRo proc be bigger ofcourse.

But then again, when I compare both sims in catus, it kinda either come head to head with higher gear (If I compare the high end geared ferals)

or in my case, I have two ferals one at 570 ilvl, the second at 550 ilvl. if I compare my ferals, the scenario with bigger agility wins, I have screenshots if you want to see.

My question is this, am I crazy. or should I regem with full on seconady stats to make the roro proc bigger, or should I trust my catus result which is telling me that the agility gemming with minor orange agility+secondary on yellows win?

I'm not sure what to do, but the result of bigger agility wins by like alot in my catus results.

I've been looking around and reading alot, and checking high end geared ferals to understand how exactly they change their gems when obtaining RoRo. and I was experimenting with this one using Catus alot, but I am not seeing the result that makes me change the gems especially that Catus is showing me different result when I click on the feature "Simulator: Quick compare" thats when I add a + new identical character to mine but I change its gems which therefore changes the whole stats of agility + mastery + crit + haste

Anyway let me get into more detail.

First then I got RoRo trinket, I obviously stacked higher haste, so I changed my yellow gems to an orage "Agility + Haste" and kept red on pure agility gems, and kept blues as Agility + hit.

The druids I seen in armory always use Orange agility + secondaries on reds, and pure secondaries on yellows, and secondary + hit on blues.

so you can see the huge different there in the two scenraios above. both will have RoRO reforges which will make the secondary stats as mastery + 1 = crit = haste

but the different in both scenarios would basicaly be that one will have much higher agility, and a tad lower mastery:crit:haste ratios

on the second scenario you will have obviously lower agility, higher secondaries mastery:crit:haste ratios. <-- which will make RoRo proc be bigger ofcourse.

But then again, when I compare both sims in catus, it kinda either come head to head with higher gear (If I compare the high end geared ferals)

or in my case, I have two ferals one at 570 ilvl, the second at 550 ilvl. if I compare my ferals, the scenario with bigger agility wins, I have screenshots if you want to see.

My question is this, am I crazy. or should I regem with full on seconady stats to make the roro proc bigger, or should I trust my catus result which is telling me that the agility gemming with minor orange agility+secondary on yellows win?

I'm not sure what to do, but the result of bigger agility wins by like alot in my catus results.

### Re: Different in gemming with RoRo

what version rune do you have?

secondary stats going ahead around 575+, 570 they could be still on pair

use catus to reforge and regem, don't do it manually

secondary stats going ahead around 575+, 570 they could be still on pair

use catus to reforge and regem, don't do it manually

### Re: Different in gemming with RoRo

rune normal right now, wasn't lucky on the heroic version at all in the past weeks.

but you make perfect sense, because thats exactly what I thought, they will shine with 575+ also with the heroic rune.

I always use catus, but for gemming catus only gives to choose one gem in each color, so you can't mix things up like that, so I gota gem manualy with catus, then I use reforge by catus. unless there's another way to gem to the maximum via catus that I am not sure of, please enlightening me.

Also when I sim on high end ferals like Maximum from duality, it clearly show that his secondary gemming is best, so I think the ilvl has somthing to do with it.

but you make perfect sense, because thats exactly what I thought, they will shine with 575+ also with the heroic rune.

I always use catus, but for gemming catus only gives to choose one gem in each color, so you can't mix things up like that, so I gota gem manualy with catus, then I use reforge by catus. unless there's another way to gem to the maximum via catus that I am not sure of, please enlightening me.

Also when I sim on high end ferals like Maximum from duality, it clearly show that his secondary gemming is best, so I think the ilvl has somthing to do with it.

### Re: Different in gemming with RoRo

You change to yellow gems when it sims higher DPS, that's it (it's covered in my guide that's stickied at the top of this forum).

Catus can gem for you automatically when you're using Rune of Re-Origination, just check the box in the reforge section for Rune. You want to use the "yellow" setting vs the "red" setting and see which comes out higher. Check out this guide for specific steps.

If you see people gemming agility + secondaries instead of expertise + secondaries it's either because they have too much expertise or because they don't know what they're doing; although it's not a huge difference, expertise is definitely better. As long as you can reforge out of it, any extra expertise directly equates to more secondary stats, so there's no logical reason why you would pick agility instead if you're gemming secondary stats in the first place.

Catus can gem for you automatically when you're using Rune of Re-Origination, just check the box in the reforge section for Rune. You want to use the "yellow" setting vs the "red" setting and see which comes out higher. Check out this guide for specific steps.

If you see people gemming agility + secondaries instead of expertise + secondaries it's either because they have too much expertise or because they don't know what they're doing; although it's not a huge difference, expertise is definitely better. As long as you can reforge out of it, any extra expertise directly equates to more secondary stats, so there's no logical reason why you would pick agility instead if you're gemming secondary stats in the first place.

### Re: Different in gemming with RoRo

aggixx wrote:Catus can gem for you automatically when you're using Rune of Re-Origination, just check the box in the reforge section for Rune. You want to use the "yellow" setting vs the "red" setting and see which comes out higher. Check out this guide for specific steps.

Thx a ton, for that guide !

### Re: Different in gemming with RoRo

If you see people gemming agility + secondaries instead of expertise + secondaries it's either because they have too much expertise or because they don't know what they're doing; although it's not a huge difference, expertise is definitely better. As long as you can reforge out of it, any extra expertise directly equates to more secondary stats, so there's no logical reason why you would pick agility instead if you're gemming secondary stats in the first place.

aggixx your information always helps me, and thanks for the guide I've seen them before, but now I saw them again and digested them better.

I wanted to ask regarding your comment of which I'm quoting. in regard to saying that expertise + secondaries will be better. I would like to ask in regards to this and to Catus, when I scroll down and tick on "Change gem" in the scrolldown menu we choose "yellow" instead of "red" but I also see the option of "Exp/Sec" , Do I choose that instead of yellow for better result? because seeing that when I choose yellow, it doesn't put up Exp + secondary gems in.

thanks

### Re: Different in gemming with RoRo

"Yellow" has access to all gems with secondary stats, this includes: Agi/Hit, Agi/Sec, Exp/Sec, Hit/Sec, and pure Sec

"Yellow (No Agi)" drops Agi/Hit and Agi/Sec

(I guess I should add pure Hit and pure Exp too -- they're in next update)

"Yellow" can find any solution that "Yellow (No Agi)" can, it's just that Catus only keeps track of a finite number of reforgings (hundreds of thousands.) Since Hit/Exp are optimized first, it's possible that "Yellow" prevents most of the Exp/Sec solutions from being selected because the Agi/Sec solutions have much better Hit/Exp. This only happens if your Exp/Sec solutions are sloppy (you have a lot of Expertise already on your gear) and you go way over the cap (like 8%+ Exp) but the overall reforge (weighted score) is better. In this case, "Yellow (No Agi)" would prevent the Agi/Sec solutions from prematurely defeating the sloppier Exp/Sec solutions.

It's also sensitive to your Hit/Exp Range setting. A range of 50, will probably pick more Agi/Sec. A range of 250, will probably pick more Exp/Sec.

This problem arises because we never figured out the Agi vs Sec ratio. It's very difficult to determine because it's dependent on your ilvl, your Rune ilvl, and your current reforge. You could do a Stat Weights sim, with [x] Inert turned-off, with Mastery-Haste-Crit [x] Group'd together.

It's also because Hit/Exp aren't given a weight. IMO, the best way to reforge: you pick the desired level of Hit/Exp (5%, 6%, 7%, 7.5%) and then you optimize the rest accordingly.

For Sten's gear using recommended defaults using Stat Weight sim w/ 20k iter, 900 delta, [_] inert, [_] force cap, [x] Agility [x] Group

12.52 / 5.90 +/- 0.23 = approximately 2:1

For reference, Catus uses 2:1 and I never got around to making this a setting.

Also [549] Rune (since higher proc rate) makes Yellow more dominate.

"Yellow (No Agi)" drops Agi/Hit and Agi/Sec

(I guess I should add pure Hit and pure Exp too -- they're in next update)

"Yellow" can find any solution that "Yellow (No Agi)" can, it's just that Catus only keeps track of a finite number of reforgings (hundreds of thousands.) Since Hit/Exp are optimized first, it's possible that "Yellow" prevents most of the Exp/Sec solutions from being selected because the Agi/Sec solutions have much better Hit/Exp. This only happens if your Exp/Sec solutions are sloppy (you have a lot of Expertise already on your gear) and you go way over the cap (like 8%+ Exp) but the overall reforge (weighted score) is better. In this case, "Yellow (No Agi)" would prevent the Agi/Sec solutions from prematurely defeating the sloppier Exp/Sec solutions.

It's also sensitive to your Hit/Exp Range setting. A range of 50, will probably pick more Agi/Sec. A range of 250, will probably pick more Exp/Sec.

This problem arises because we never figured out the Agi vs Sec ratio. It's very difficult to determine because it's dependent on your ilvl, your Rune ilvl, and your current reforge. You could do a Stat Weights sim, with [x] Inert turned-off, with Mastery-Haste-Crit [x] Group'd together.

It's also because Hit/Exp aren't given a weight. IMO, the best way to reforge: you pick the desired level of Hit/Exp (5%, 6%, 7%, 7.5%) and then you optimize the rest accordingly.

For Sten's gear using recommended defaults using Stat Weight sim w/ 20k iter, 900 delta, [_] inert, [_] force cap, [x] Agility [x] Group

12.52 / 5.90 +/- 0.23 = approximately 2:1

- Code: Select all
`Name DPS Stdev Gain Error Weight Error Delta`

Baseline 434.4K 20.83K 0 147 0.00 0.16

Agility 445.6K 21.08K 11.27K 210 12.52 0.23 +900 Agility

Secondary 439.7K 21.43K 5310 211 5.90 0.23 +300 Crit, +300 Haste, +300 Mastery

For reference, Catus uses 2:1 and I never got around to making this a setting.

**Spoiler:**show

Also [549] Rune (since higher proc rate) makes Yellow more dominate.

### Re: Different in gemming with RoRo

Not to make this a Catus thread, but...

I added Pure Hit/Exp option and Custom Sec/Agi Ratio.

Example: a value of 1.5 says that 1.5 Sec = 1 Agi.

Note: reforging with a lower value eliminates the Agi/Sec gems (and breaks different bonuses than it normally would.)

I also added "Reforge Map", which explains how each stat is reforged, like Blizzard armory:

I added Pure Hit/Exp option and Custom Sec/Agi Ratio.

Example: a value of 1.5 says that 1.5 Sec = 1 Agi.

Note: reforging with a lower value eliminates the Agi/Sec gems (and breaks different bonuses than it normally would.)

I also added "Reforge Map", which explains how each stat is reforged, like Blizzard armory:

### Re: Different in gemming with RoRo

tl;dr - doesn't matter

After looking at my reforger code, I see that I never really finished fully supporting different weights after I added the different gem modes. Regardless of what gem mode is selected, the reforger is trying to maximize the following function:

In nearby comments, I also have some failed variants:

Mostly this was just bullshitting, however it's not completely off-base:

If we lazily compute this in units of secondaries:

Which is similar to the bullshit equation above (if we divide by a constant):

So I think an improvement is just:

Where _2_ is replaced by the custom "Agi-to-Sec Ratio" if provided, and 0.5 is a constant I pulled out of my butt.

However, when I test this on a bunch of kitties, it's basically the same reforge so I don't think it matters. Minimizing wasted Hit/Exp and achieving 1:1:1 keep the problem pretty restricted. As long as the scoring function rewards higher (Mastery, Haste, Crit, Agility), the weight coefficients don't really matter.

After looking at my reforger code, I see that I never really finished fully supporting different weights after I added the different gem modes. Regardless of what gem mode is selected, the reforger is trying to maximize the following function:

- Code: Select all
`= 1.8 * Mastery + 2.0 * (Haste + Crit + Agility)`

In nearby comments, I also have some failed variants:

- Code: Select all
`//return mastery + 2 * (haste + crit);`

//return mastery + 2 * (haste + crit + agi);

//return 1.5 * agi + haste + crit + mastery;

//return 2 * agi + (mastery + haste + crit);

//return agi + haste + crit;

Mostly this was just bullshitting, however it's not completely off-base:

- Code: Select all
`= (value of agility) + [ (value of mastery/haste/crit during effective downtime) + (value of proc=mastery+2*(haste+crit) during uptime) ]`

If we lazily compute this in units of secondaries:

- Code: Select all
`= (Agi-to-Sec Ratio * Agility) + (Mastery + Haste + Crit) * (1 - Uptime) + (Mastery + 2 * (Haste + Crit)) * Uptime`

= (Agi-to-Sec Ratio * Agility) + (Mastery + (Haste + Crit) * (1 + Uptime))

Which is similar to the bullshit equation above (if we divide by a constant):

- Code: Select all
`= _2_ * Agility + (_1.8_ * Mastery + (Haste + Crit) * (1 + _1_))`

= _1.111_* Agility + (Mastery + (Haste + Crit) * (1 + _0.111_)) // divide by 1.8

So I think an improvement is just:

- Code: Select all
`= _2_ * Agility + (Mastery + (Haste + Crit) * (1 + _0.5_))`

Where _2_ is replaced by the custom "Agi-to-Sec Ratio" if provided, and 0.5 is a constant I pulled out of my butt.

However, when I test this on a bunch of kitties, it's basically the same reforge so I don't think it matters. Minimizing wasted Hit/Exp and achieving 1:1:1 keep the problem pretty restricted. As long as the scoring function rewards higher (Mastery, Haste, Crit, Agility), the weight coefficients don't really matter.

9 posts
• Page

**1**of**1**### Who is online

Users browsing this forum: Bing [Bot], Google [Bot] and 4 guests