Didn't see your question actually answered for the macro so here is what I use and it works flawlessly.

#showtooltip
/cast [mod:alt,@mouseover][mod:shift,@focus][] Judgement
/startattack

#showtooltip
/cast [mod:ctrl,@mouseover][mod:shift,@focus][] Exorcism
/startattack

Same for Rebuke/HoJ but for keeping the stack on 2 targets this works great. Where you noticed the stack not being visible on the focus target some addons don't hook into the combat log to monitor that. In which case you want to submit a ticket to the author of the addon.

Avalanche doesn't work quite like you suggest. For starters, testing during beta revealed that it's 5 ppm off of melee attacks. (test #14). In fact, it seems from the testing that's been done that the new Cataclysm enchants (Landslide, Avalanche) are 5 ppm on melee attacks.

However, Avalanche also can proc from spells, including Censure ticks. It's not a simple 5 ppm from spells though; it uses a different mechanic. Our testing suggested that it is a flat 20% proc chance with a 10-second internal cooldown.

I wouldn't be surprised at all if that worked out to around 12 ppm for an average ret rotation, but it's not a simple 6ppm/6ppm from melee/spells.

Well, that's exactly what it works out to. So, unless your Avalanche procs are hitting for ~1.1k...

If that data holds up over long periods of testing, does that bring the value of avalanche down to the same level as hurricane as the choice of weapon enchant until landslide is a bit more affordable? It would seem by those numbers they're both around 100dps give or take, unless hurricane figures have dipped too with the patch

You seem to have a false understanding of what "ppm" means. ppm refers to the average procs per minute WHILE AUTOATTACKING. It's just a different way of saying that the proc chance is lower with a faster weapon so you average about the same amount of procs with a slow and a fast weapon, again only while autoattacking. Using your full arsenal of styles and spells, a 6 ppm enchant procs way more often than 6 times per minute. (e.g. avalanche procs almost 20 times per minute)

"Exactly?" I'd love to see the parses you used to confirm that it's exactly 12 procs every minute. Because that's not how PPM enchants work, for starters. Also, you noted yourself that the actual DPS in a raid seemed to vary from 120 to 180 DPS from parse to parse. I have trouble believing that difference was due to buffs or lucky crits.

The point is that we have evidence that Avalanche isn't a flat PPM mechanic at all, and PPM enchants don't work the way you described. Because of those two errors, your modeling isn't going to be accurate. An enchant with 6 ppm that procs off of melee and spells will act differently than one that procs 5ppm melee and 20%/10ICD from spells (and different yet from an enchant that obeys the incorrect PPM model you're suggesting in your post). The proc rate of the second one will vary significantly with rotation type, because the chance to proc from spells and melee attacks is different.

Let's do some napkin math to see whether Avalanche really behaves as you suggest. A 6 PPM enchant on a 3.8-speed weapon has a 6*3.8/60=0.38, or 38% chance to proc. We'll assume around 25% haste for raid buffs and gear, so that drops your swing timer to almost exactly 3 seconds. You'll get 20 melee attacks per minute at that speed, resulting in 20*0.38=7.6 procs.

In addition, you have 40 GCDs per minute (ignoring hasted GCDs for the moment) in which to cast melee or spell attacks. Assuming you fill all of them, that's 40*0.38=15.2 more procs. So we're now up to 22.8 procs per minute, or 190 DPS at a minimum assuming 500 damage per proc and no crits or raid buffs. Even if we drop ability usage to 35 (i.e. 5 empty GCDs per minute, probably an over-estimate) that's still 13.3 procs, for a total of 20.9 procs or 174 DPS. Both of those are considerably higher than your observed values; with raid buffs and crits included we'd expect to be seeing parses with 220+ DPS from Avalanche.

Let's try the other model. 5 ppm is a 31.67% chance per cast, or 6.3 procs from auto-attacks. Let's say that out of your 40 GCDs you get 13 CS's, 10 TV's, 6 J's, 4 Exorcisms, 4 Holy Wraths, and 3 empty GCDs (or Cons, which doesn't seem to proc it anyway). Breakdown here is dictated by cooldowns where applicable and rough ability usage ratios from a random magmaw parse. With that breakdown, you have 29 melee abilities and 8 spell-like abilities. The 29 melee give you another 9.2 procs. The 8 spell-like abilities, assuming they're equally spaced at 7.5 seconds, will overlap with the internal cooldown, netting you only 4 chances to proc at 20% each, or 4*0.2=~0.8 procs. That's a total of around 16.3 procs per minute, or around 136 DPS before raid buffs and crits. Right in the ballpark of what you're observing.

And note that in both models, I've ignored Censure ticks. At 20% chance to proc, it'll take an average of about 4 ticks to proc (~12 seconds) and then go on ICD for another 10 seconds. So you're looking at around 3 procs per minute from Censure in both cases, inflating the DPS of both slightly (~25 DPS).

In other words, 120-180 DPS is perfectly reasonable for the already-established model of Avalanche procs. There's no need to reinvent the wheel as a square, because the existing circular wheel already works better.

The reason I bring this up is that the Avalanche model depends very heavily on rotation. As we've seen, the ICD tends to reduce the proc rate from spellcasts from an already low 20% to closer to 10%. In fact, in our early testing (my own and Redcape's) before we realized there was an ICD, we both thought the proc rate was a flat 10%. It was only when we collected 3-hour parses of Avalanche proccing from Censure ticks that we noticed there were no back-to-back Censure-based procs.

Pre-4.0.6, Exorcism was very highly prioritized, leading to a higher relative percentage of spell casts. With 4.0.6, the priority has shifted to a more melee-heavy distribution, which has had the effect of improving Avalanche's average output slightly.

So... the average dps is ~120-160 dps, right?
And the damage is only barely above 500 per proc, correct? So we can estimate ~600-700?
And this would equal out to around 12-14 procs each minute.

So, semantics aside, how was I wrong?

If this is so well established, then why was it not updated in the OP?

Actually, the OP was made even before the spell proc portion was even an issue.

Exemplar's (OP) post on Avalanche weapon enchant.

Correct, I was never apprised that spells functioned to proc Avalanche on a 20% chance with 10 sec ICD. Instead it was modeled under the same PPM mechanic (where cast time was 1.5, divided by 60 sec, multiplied by 5 PPM to provide a % chance). Modeled in this fashion it produced the DPS listed in the OP.

Changing modeling based on this information produces closer to 200 DPS. Of course, this is on a Patchwerk style fight, which is why it would not match the values found in randomly selected logfiles. Raid buffs other than Haste impacting are a non-issue, as to date Avalanche has only proven to increase damage from Communion.

Actually, updated modeling of Hurricane's functionality gives me a bit over 300 averaged Haste, in my gear. With the far greater priority of CS and my new method of modeling permitting a cleaner valuation of Haste, in my gear each point of Haste is worth about .78 DPS. This puts Hurricane ahead of Avalanche by about 30 DPS. For me. In my gear.

Landslide is even tastier post-patch - my gear is showing closer to 490 DPS, now (40 DPS increase).

I would personally put Avalanche and Hurricane about equal or even place Avalanche ahead. Time off target impacts Avalanche less than Hurricane. Haste is nigh useless when only able to use Judgement (and a few hard cast Exo before you run OOM), while those Judge can still proc Avalanche strikes, as can Censure (which would be refreshed by the Judge). This is a personal opinion based on the fact that real bosses aren't Patchwerk.

As I do with stats, I would strongly recommend any given individual test the values for their own gear in a spreadsheet. Values change as your other stats change - this is why I vastly dislike people plugging in some single value for each stat. Example: Self buffed I have more Crit as superior to more Mastery, but add in raid buffs and suddenly more Mastery is superior. Stats are always relative values, they are never absolute values.

Relative values change in relation to each other.

For me, in my gear, the average PPM of Avalanche worked out to be better than Hurricane by roughtly 68 DPS. Game breaking? No. But, I couldn't find a single point where Hurricane was better than Avalanche UNLESS I went with a CS>Filler rotation. For the most part, it was a DPS loss, except when Hurricane procc'd, then going with a 3.28 CS CD to CS>Filler was a net gain in DPS, starting at that point. From there, as my CS CD approached 3.0 secs, I showed improvements.

I'm still playing around with it, but from what I can see so far, Avalanche still beats out Hurricane until we get closer to the magic 3.0 CS.

Edit: Missed one letter which was important for clarification. Sleep is also important.

I don't think its been brought up because I can't imagine it possibly being an upgrade. You must consider that portion of the post you quoted was from months ago before Cataclysm even released. First don't forget Mastery may be our best secondary stat, but STR is significantly our best stat overall.

So anyway, aside from Dragon Bone Warhelm being behind in STR, it *may* be a better helm than our T11 helm if only compared side-by-side. Set bonuses though would put the T11 ahead by far. None of the other tank pieces would pull ahead at all unless you were able to gain a set bonus or maintain your current 4-piece. But I'm pretty sure the BIS right now is 4-piece T11 with Vicious Gladiator's Scaled Gauntlets. Even if you could find some kind of set that would be worth dropping the 4-piece bonus for tank pieces, I imagine the DPS gain would be so marginal that it wouldn't be worth the quality of life loss by the 4-piece that not only makes things more convenient but helps minimize human error.