Ok, fair point, but it is directly from the Frost DK theorycrafting community. But I guess a random level 10 alt knows DK mechanics better than them, right?
Wait, so…the spell data specifies that it’s RPPM, but you’re still trying to maintain that it’s not?
Just for reference, I wrote that article.
Heh, look at that, so you did. You’re also not Kelliste, so my apologies. I inferred you were on the other side of this contention.
I’m really not. I’ve yet to find anything suggesting it’s PPM but I don’t dismiss anything I’m not 100% sure about. I don’t know about anything else that works with PPM and I’m not quite sure to test if it does or not.
Actually, side note on that. AMR lists the RPPM as also being increased by crit chance (which would be unique amongst RPPM mechanics, to my knowledge, but I guess it would kinda make sense). You also state that the RPPM is increased by crit, though don’t go into any details. However, you state in the article that you need some minimum level of crit to reach the necessary threshold. Is the RPPM scaled by crit chance as well? And if so, wouldn’t it be impossible to hit said threshold, since the procs per minute would rise at the same rate as your crits per minute (at least, at any given haste level)?
As for testing it, the easiest way would be to us only a single 1-hander, and then walk up and AA a dummy until the first critical strike, then walk away and wait at least 10 seconds until doing it again. Track how many of these single crits proc. Now equip two 1-handers (or just the one, honestly) and afk attack for a few minutes, and track how many procs you get compared to crits. If they differ considerably, it’s RPPM, since the time spent not attacking balloons the proc chance (if it’s 4.5 RPPM, in theory it should have about a 75% proc chance on the first crit after 10+ seconds of no crits). If they’re the same, it’s a fixed percentage (though whether that means it’s a flat percentage or some version of classic PPM would take testing with different speed weapons). Ideally, this would be without gear and with statless weapons, to avoid stats affecting it. You should also probably try to remain in combat for the duration, just in case the reset-to-zero-on-combat effect happens on target dummies in addition to raid bosses. If I find some time, I’ll see if I can test it myself.
My current knowledge is that it scales as 4.5 x crit x haste. Minimum crit is required to reach a level sufficient to hit the RPPM value pretty much. It’s the exact same argument as to why 2H has such low proc rate.
Actually, ya, thinking about it again, I had it modeled wrong in my head. Was thinking the RPPM scaled at the same rate as crits, but that’s not the case. An increase from 5% crit to 10% crit, for example, doubles the number of crits per minute, but would only increase the crit multiplier from 1.05 to 1.1, or +4.76% more procs. From that, you would need ~10.8% crit in order for your average crits per minute to equal your modifier RPPM.
The problem is, even that’s not enough. The RPPM equation caps the “time since last proc attempt” at 10s, so your average crit interval can’t exceed that (which would require 13.0% crit chance). However, even if your average crit interval equaled 10s, that means roughly half the time it takes longer than that interval between crits, wasting procs.
Honestly, I just don’t understand how they expected this to work. RPPM isn’t designed for such a large band between procable events. 2h in particular would require a 36% crit chance just to reach an average crit interval of 10s.
Unclear if they want to balance it atm, we’ll see
Ya, I mean, Blizzard’s approach to Frost weapon balance has historically been a bit “eh, whatever, who cares”.
Honestly, if they kept the existing crit scaling on the RPPM rate, but made it RPPM from all attacks, KM would be in a massively better spot, and would work for both DW and 2h. 2h would have a 27% chance per AA, increased by crit, and DW would have a 9.75% chance per AA, increased by crit. Both would see an average of the expected modified RPPM. There’s little reason, in my opinion, to have it only trigger from AA crits when the RPPM itself scales with crit chance.
hahahaha 
I heard that mic drop even louder.
Why do you think it’s real procs per minute? It doesn’t even remotely work that way.
How would it give you a chance per crit? Did you really just ask that? Let’s see, use the ppm % and have it only trigger off of crits instead of just autos. The ppm system just gives you a % per hit, or in this case per crit, as you increase your crit and haste it procs more often doesn’t it? The exact thing that the ppm system is supposed to do. Increase the number of chances for the % to roll and you increase the number of procs.
I’ve looked at data that fits the 4.5 x crit x haste formula rather nicely. It’s strange to me that they wouldn’t use RPPM since that’s the baseline for so much.
I’ve heard that they initially changed it to RPPM in WoD, going to get a proper source on that.
I’ll do some tests tomorrow to see if I can prove if it’s either one of the two.
There was an article written explaining the different types of procs. ICD which is once it procs it can’t proc again until the cd is finished. PPM is a % which is calculated off of ppm, 60s and base weapon speed. RPPM is that many procs per minute regardless of what you do. So if something has an RPPM of 4.5 you get 4.5 procs per second. If it has an RPPM of 1 you get 1 proc per minute. That’s the simple explanation. I’ll look for the article later because it even explains the formula of what goes into RPPM, it has timers of when the fight starts and checks if you have gotten procs.
RPPM is usually used for trinkets or things that act like trinkets, KM doesn’t act like a trinket.
So break it down for me. If you had, say, 0% haste and 20% crit chance, with a 3.6s weapon, what would the proc chance per AA crit be? How would the proc chance be computed, give the above stats?
That’s already explained by the RPPM mechanics. RPPM already directly scales with haste (and that applies to all RPPM mechanics, that’s why Infinite Star is so stupid OP for Havoc, because of our insanely high uptime on the ~40-50% bonus haste from Furious Gaze), and apparently KM scales the RPPM proc rate with crit as well.
I’ll say it again, that’s not how RPPM works. You are in no way guaranteed that number of procs “regardless of what you do”. All RPPM does is ensure that your average procs per minute is the same regardless of whether you’re triggering a chance to proc (in most cases, this is an attack or spell cast) once every 5 seconds, or once every 0.5 seconds. PPM can’t do that, because PPM translates to a flat fixed percentage chance to proc on every procable event, regardless of how often it occurs.
You saw this a lot back in vanilla. A 1 PPM effect would end up procing 3+ times per minute, because the proc chance conversion was based on weapon speed, but ability weapon strikes could trigger it as well.
As an example, the Crusader enchant back in vanilla was 1 PPM. If you were using a 3.6s 2-hander, you’d have a 1 * (3.6 / 60) = 6.0% chance per swing to trigger the effect. You’d get in 60/3.6 = 16.67 auto attacks per minute with it, which at 6% chance would give you the expected 1 proc per minute. However, it can also trigger from ability weapon hits, which used the same proc chance. If you used a melee ability on every single GCD over the minute, you’d get another (60/1.5) * 0.06 = 2.4 procs, raising the actual observed average proc rate to 3.4 per minute.
RPPM isolates it from attack rate. If it’s 1 RPPM and you attack every 1.5 seconds, then each attack has a 1.5 * (1/60) = 2.5% chance to proc, and you’d see, on average, (60/1.5) * 0.025 = 1 proc per minute. If you attack every 5 seconds, then each of those attacks would have a 5 * (1/60) = 8.333% chance to proc, and you’d see an average of, again, 60/5 * 0.083333 = 1 proc per minute.
But that’s not the same thing as being guaranteed to see that many procs per minute. At 8.33%, you have a smidge over a 35% chance to go a full minute without any procs at all, and around a 12.5% chance of going two full minutes without a proc. However, you also have around a 19% chance to see at least two procs within a 60s period, and a ~6% chance of seeing at least three.
The article you’re referencing is this one:
https://xpoff.com/threads/how-rppm-works.81584/
It’s still a great source of information on proc mechanics. However, you really need to read the RPPM section more thoroughly. You seem to have a mistaken impression of how it functions.
RPPM is used for many things other than trinkets. For example, Gluttony for Vengeance is an RPPM system. Demonic Appetite for Havoc is RPPM. Sudden Death for both Arms and Fury is RPPM. KM is also flagged in the spell data as being RPPM, so unless you have rather compelling evidence to support it, your simple assertions that it is not RPPM don’t hold much weight.
Haha, I like this Xaedys guy. Very informative. Good read.
Well, I apparently don’t have the crit on my DK to test it, since he’s geared mostly for Unholy, and also terribly geared to boot (he’s only got about 800 crit rating, so ~17% crit chance). However, the data is at least suggestive, though it’s probably to small of a sample size for the observed difference.
Procedure: I equipped only my mainhand weapon, and unequipped my trinkets (my only items with other proc effects). I entered combat with the tanking dummy, since he keeps you in combat permanently until you’re ~80 yards from him, then walked into range of a regular DPS dummy and auto-attacked until I had a critical auto-attack. I immediately backed out of melee range for 15s, and also noted whether KM proced. I did this for a total of 40 AA crits, from which I got 24 KM procs (60.0% proc rate).
I dropped combat, then equipped both of my weapons as normal, then simply auto-attacked the dummy until I got 40 normal AA crits from continuously AAing. I got 19 KM procs in this time (47.5% proc rate).
Unfortunately, it took me 5 minutes and 35 seconds to get that many AA crits due to my low crit chance. 40 crits in 335 seconds is an average interval of 8.375s between AA crits, which is already very close to the 10s RPPM maximum interval for the formula. That said, 10 / 8.375 = 1.194, and 24 / 19 = 1.263, so that at least suggests that the average interval between AA crits factors into the proc rate, even when crit chance is the same (my offhand has haste/vers).
I’d need someone actually geared for Frost, with a lot more crit rating, to replicate this test and see if they see a larger difference due to their much shorter interval between crits during the continuous auto-attack portion of the test. That would also demonstrate rather definitively whether the proc rate scales with crit chance, since if it does, one would expect a noticeably higher proc rate even in the single-crit-and-back-out test.
They could absolutely do this, but people act like they know everything about game design.
Not saying it would be easy but it is absolutely doable