Several days ago I decided to swap my gear optimization from crit fire to haste frost. As part of this I downloaded and ran Simcraft.

Surprisingly, Simcraft weighted int at something like 5.27 and haste at around 2.50. Based on those results 1 point of int was worth more than 2 points of haste. At the time I got these results I had gemmed for max haste.

Based on those results, I then swapped my gems from pure haste to int/haste. This raised my int but lowered my haste.

I then ran Simcraft again... and got the opposite result. Now that my haste was lower, haste was worth more than before. 1 point of int was no longer worth more than 2 points of haste.

It looks like that optimum result for me, at least according to Simcraft, would be to have some pure haste gems and some int-haste gems. I didn't want to spend thousands of gold trying random combos and guessing though, so for now I'll just leave most of my gear int/haste.

At the risk of sounding dumb what is an RPPM trinket?

RPPM refers to the system currently being used for passive procs on trinkets and enchants (e.g., jade spirit). Thus, an RPPM trinket is basically any trinket with a passive proc.

In contrast, a trinket like shock-charger medallion, with an active proc, does not use the RPPM system obviously because you control the proc.

Blizzard is definitely showing a bias right now to passive procs (my guess is that it's easier to balance that way), and you will note that your trinkets (shado, yu'lon) and mine (hydra, woo) are all RRPM trinkets.

Here's the kicker: proc rate is affected by haste! From AMR:

The only thing that can make your trinket proc more often is Haste. This includes Haste Rating from gear, the 5% Haste raid buff, and Heroism. Attack speed buffs do not count, only Haste. So if you have 20% Haste, you increase the chance to proc by 20%. So if you have an RPPM of 1 (1 proc per minute) and 20% Haste, your trinket would proc 20% more often. So that would be 1.2 times per minute. It’s that simple!

For these reasons, haste becomes even more valuable for a frost mage when they obtain a RRPM trinket. What's nice for frost mages is that haste is already the most valued stat.

Several days ago I decided to swap my gear optimization from crit fire to haste frost. As part of this I downloaded and ran Simcraft.

Surprisingly, Simcraft weighted int at something like 5.27 and haste at around 2.50. Based on those results 1 point of int was worth more than 2 points of haste. At the time I got these results I had gemmed for max haste.

Based on those results, I then swapped my gems from pure haste to int/haste. This raised my int but lowered my haste.

I then ran Simcraft again... and got the opposite result. Now that my haste was lower, haste was worth more than before. 1 point of int was no longer worth more than 2 points of haste.

It looks like that optimum result for me, at least according to Simcraft, would be to have some pure haste gems and some int-haste gems. I didn't want to spend thousands of gold trying random combos and guessing though, so for now I'll just leave most of my gear int/haste.

OMG, been there, done that, HATED the

**gold**results.

Here's what I think is BETTER:

1) Go to a training dummy and sim yourself for 10 minutes.

2) Do changes that you want to test out

3) Go to a training dummy and sim yourself for 10 minutes

Which is higher: 1 or 3?

That's how I ultimately convinced myself that my 502 wooshly trinket was indeed better than my 522 shado-pan trinket.

P.S. Nice to see ya Mistwynd!

I then ran Simcraft again... and got the opposite result. Now that my haste was lower, haste was worth more than before. 1 point of int was no longer worth more than 2 points of haste.

That's exactly right. As any stat increases, the value of adding more of that stat decreases relative to the other stats, and vice-versa.

Look at it this way: if you add enough intellect to increase damage by 1%, you've just increased the value of:

- Haste: every additional spell haste gives you is now worth more

- Crit: every additional crit you get from it is now worth more

- Mastery: every spellcast affected by Mastery is now worth more

The only stat whose value does not increase as you add more intellect is...intellect. And thus the value of all the other stats goes up, while the value of intellect is unchanged, and intellect is therefore worth less than it was before relative to the other stats.

The same is true for every stat. So it's important to make incremental changes and test the results for new weights.

Several days ago I decided to swap my gear optimization from crit fire to haste frost. As part of this I downloaded and ran Simcraft.

Surprisingly, Simcraft weighted int at something like 5.27 and haste at around 2.50. Based on those results 1 point of int was worth more than 2 points of haste. At the time I got these results I had gemmed for max haste.

Based on those results, I then swapped my gems from pure haste to int/haste. This raised my int but lowered my haste.

I then ran Simcraft again... and got the opposite result. Now that my haste was lower, haste was worth more than before. 1 point of int was no longer worth more than 2 points of haste.

It looks like that optimum result for me, at least according to Simcraft, would be to have some pure haste gems and some int-haste gems. I didn't want to spend thousands of gold trying random combos and guessing though, so for now I'll just leave most of my gear int/haste.

OMG, been there, done that, HATED thegoldresults.

Here's what I think is BETTER:

1) Go to a training dummy and sim yourself for 10 minutes.

2) Do changes that you want to test out

3) Go to a training dummy and sim yourself for 10 minutes

Which is higher: 1 or 3?

That's how I ultimately convinced myself that my 502 wooshly trinket was indeed better than my 522 shado-pan trinket.

P.S. Nice to see ya Mistwynd!

Hrm... I don't know how reliable the results of a single parse are going to be... Way too many variables to take into consideration. You need a larger sample size. Fortunately simulationcraft can simulate your rotation ten thousand times in a matter of seconds.

OMG, been there, done that, HATED thegoldresults.

Here's what I think is BETTER:

1) Go to a training dummy and sim yourself for 10 minutes.

2) Do changes that you want to test out

3) Go to a training dummy and sim yourself for 10 minutes

Which is higher: 1 or 3?

That's how I ultimately convinced myself that my 502 wooshly trinket was indeed better than my 522 shado-pan trinket.

P.S. Nice to see ya Mistwynd!

Hrm... I don't know how reliable the results of a single parse are going to be... Way too many variables to take into consideration. You need a larger sample size. Fortunately simulationcraft can simulate your rotation ten thousand times in a matter of seconds.

I don't deny that for a second.

I typically do both.

My main issue with simcraft for comparing changes is two-fold.

1) I'm not a big fan of the sim they have in place. I really think Nether Tempest should be given higher priority, and they include things like casting fire blast while moving (which I personally never do). I know some people actually edit that whole section, but I'm a little too intimidated doing that as it does dramatically change the sim results.

2) I'm never too sure where to place the skill level or fight-type. 100% perfection doesn't seem to reflect my skill level, but not sure the "fire is hot" setting is more accurate for me.

So at the end of the day, I think neither method is perfect. Like I said, I typically do both before making a final decision, but my own personal bias is towards the results I get on the actual training dummy because I know it reflects my rotation and my skill level.

I then ran Simcraft again... and got the opposite result. Now that my haste was lower, haste was worth more than before. 1 point of int was no longer worth more than 2 points of haste.

That's exactly right. As any stat increases, the value of adding more of that stat decreases relative to the other stats, and vice-versa.

Look at it this way: if you add enough intellect to increase damage by 1%, you've just increased the value of:

- Haste: every additional spell haste gives you is now worth more

- Crit: every additional crit you get from it is now worth more

- Mastery: every spellcast affected by Mastery is now worth more

The only stat whose value does not increase as you add more intellect is...intellect. And thus the value of all the other stats goes up, while the value of intellect is unchanged, and intellect is therefore worth less than it was before relative to the other stats.

The same is true for every stat. So it's important to make incremental changes and test the results for new weights.

Yes, but what is quite frustrating is when the predicted damage result goes down.

Example (using made up numbers), I sim myself in simcraft and it produces my scale weights along a predicted damage number of 140k. Then, I place those scale weights into AMR, change gems/enchants/reforges accordingly, and then sim myself again simcraft. I do obtain new scale weights, yes, but the predicted damage is now 132k rather than 140k!

When I see results like that, it makes me feel that I just wasted a lot of gold.

29.76% Haste

27.89% Crit

21.59% Mastery

I figure once I'm raid buffed I'd be over 28% crit so IL and FFB should always crit on boss and we have a spriest in raid for the 5% haste buff. Just wondering if I should be gemming straight haste or int/haste in yellow.

When it suggests ...

Int go for Int/Haste

Int/Hit go for Haste/Hit

Int/Haste go for Haste

Basically you should prefer the Ice-Veins gemming suggestions over AMR's Int gem heavy suggestions while still maintaining socket bonuses.

Just wondering if I should be gemming straight haste or int/haste in yellow.

Lhevera answered this earlier:

Basically, in a yellow socket, you can have:

160 Int plus no bonus

80 Int plus 160 Haste plus a bonus

320 Haste plus a bonus

In order for option 1 to make sense, option 2 must also make sense.

In order for option 2 to make sense, a point of Int has to be worth as much or more than two points of Haste. That's virtually never true, so that rules out both option 2 and option 1.

That leaves option 3 as the only one that makes sense.

You'd do better by making

*one*of the recommended changes, resimming for new weights, repeat. This lets you zero in gradually with incremental changes.

You'd do better by makingoneof the recommended changes, resimming for new weights, repeat. This lets you zero in gradually with incremental changes.

Interesting.

Any suggestions on which change should be examined first?

For example, would checking a gem change be preferable to checking a reforge change? or checking a reforge that changes 100 of a stat prior to one that changes 200 of a stat?

Just in case someone needed another reason for gemming haste over INT.

I was honestly unaware, and they only added it to the guide on April 17th.

It's just based on your stats though. So for example on Jin Rokh you can park him anywhere, and move into the pool for the 40% damage boost, and he gets it too, even if he's not in the pool himself.

It's just based on your stats though. So for example on Jin Rokh you can park him anywhere, and move into the pool for the 40% damage boost, and he gets it too, even if he's not in the pool himself.

For that fight has anyone tested if he will stay on the boss without despawning if you park him on it and then go over to fly? With the Glyph whenever you fly he goes away. . .

I think you just want the Haste as your base but would still want Potions/Trinket Procs that are Int. Others can correct me if I'm wrong however.

o.o

I thought Visions of Insanity was +500 to all base stats (Int/Str/Spirit/Agi/Str/Stam) and didn't affect secondary stats