That’s completely unrelated. You can see in the data at it 100% isn’t PPM. If it was PPM, the proc chance with be constant and you wouldn’t have a time dependence.
2H not having 4.5 procs per minute can be explained with an RPPM system.
Is 2h getting 4.5 real procs per minute? If no then where is the unlucky streak protection? If proc chance increases as time goes on then it shouldn’t be getting 1 proc per 30 seconds.
PPM and crits being a trigger for the roll explains it directly.
Weren’t you also the one that said it should be fine and not an issue since km has been normalized with the RPPM system yet here it is 2h being behind, and pretty far behind because the trigger has changed.
I’m talking about averages. RPPM can have lucky and unlucky streaks just like PPM and flat percentage. Streaks are irrelevant, because they are by definition the outliers.
As I pointed out, your claimed percentages do not line up with in-game data, and hence are flat out wrong.
Again, KM does not appear to have the built-in protection, and a number of other RPPM effects do not either. So all it cares about is your last proc attempt.
Again, your claims do not line up with in-game data, and thus your claims are clearly wrong.
<mic drop>
Not all RPPM effects have the bad luck protection, and from testing, KM appears to be one that does not. 2-hander doesn’t get 4.5 procs per minute simply because it doesn’t AA crit often enough for that to happen.
No. No it doesn’t. PPM is not time-dependent. Every single AA crit would have an identical chance to proc. But Bicep’s data clearly shows time-dependency, which clearly demonstrates that it is in fact RPPM.
2 Likes
I’ll go through it here again but I’ve gone through it at another place.
RPPM isn’t 100% known, as in it’s not known how it works. Time since last proc is part of the proc rate at every proc opportunity but I’m not sure if it ever goes to 100%.
Issue with 2H is that it doesn’t have a lot of crit autos. At 0% haste and 10% crit, you get 1.7 crit autos per minute. So even if you get KM on 100% on those crit autos, you’d never get to 4.5 per minute. Reality is that you don’t get 100% proc rate on your KMs so the overall is even lower.
1 Like
They’ve explained it to us in some detail, though. We know that the time since last proc attempt caps at 10s, and for anything with an average proc interval of >10s, that means that that portion of the equation necessarily caps out at less than 100% proc chance. If the given effect also lacks the bad luck protection, which it appears that KM does (probably due to it’s rather high proc rate), then no swing can ever hit 100% proc chance. For KM, no swing could exceed about a 75% proc chance even if it’s been quite a bit longer than 10s since last AA crit.
Yes, it doesn’t have a lot of chances to roll to see if it will proc. The exact issue it has had for 11 years. The issue with RPPM is that it does have protections in it, it has to to be able to see a real proc of 4.5 a minute.
I mean the people saying it’s RPPM isn’t even on the same page so it’s just not a good system for a type of proc like KM which just last patch people were talking about it being a ppm system.
And again, you said, when I brought up all of this even before 2h was announced, that the proc rate would be fine since it has been normalized since it was changed to RPPM, and then we see a giant difference in proc rates between the 2 weapons.
Even the blue post explaining how RPPM works says that the system will get you that number on average no matter the events, if you are a fast attacking rogue or a slower hitting spec like enhancement, it works to get you that number.
So, with both weapons having the same stats, and it’s based on RPPM, why is 2h just so far behind when the RPPM system is not giving 4.5 procs per minute. It’s kind of pointless to have a system working to get you that many procs per minute with protections included, if it doesn’t get you what it’s supposed to.
Not sure if you misunderstood the discussion here. It was about whether or not KM is PPM or RPPM. It’s clearly RPPM.
I just think you need to reach a threshold where you get enough trials to reach the RPPM rate. Changing it to non-crit autos as well would help a great deal with that.
Are you trying to argue that they should change it from RPPM? I’m not sure what your point is. I hope they make changes so that you get the same amount of KM procs regardless of weapon choice.
4 Likes
Not sure why this matters? I learn new stuff about DKs continuously.
I’m saying it’s not RPPM. Even if it was it’s just a worse system than ppm for this type of proc since it’s clearly not working as intended because it isn’t giving you a real proc per minute of 4.5.
What should the unlucky streak protection be? It should come into play after a certain number of seconds of not seeing a proc, which if that calculation ends up being over 1.0 then that value is the multiplier applied to the % chance. Where are these protections at that should be propping up 2h.
You also said that km procs shouldn’t matter since they have been normalized, yet on alpha it’s exactly what I said would happen.
Within limits. If you don’t have enough procable events, it can very easily fall short. The problem is, RPPM was designed with relatively low proc rate (1-2 RPPM) effects and relatively high proc even frequency (attack and/or spell hit). Having a high RPPM (4.5) effect that procs off low-frequency events (AA crits) strains the capabilities of the system. That’s why I’ve been arguing that having it trigger off AA crits is both useless and counter-productive.
This. If it was all AA attacks, the maximal time between procable events would be ~3.6s. Even at 4.5 RPPM, that’s plenty of events per minute to reach the average. The problem is, with AA crits, it’s very easy to go longer than 10s without an AA crit, especially with a 2-hander (but it’s possible even while dual-wielding), and 10s is the cap on the time-since-last-procable-event variable. Any time it takes >10s between AA crits, you’re losing procs, because that extra time is simply wasted.
Explain the time-dependency, then. PPM cannot exhibit time-dependent behavior. So please, explain why the data is showing considerable and very obvious time-dependency.
That protection isn’t applied to all RPPM effects, and it is well known that this is the case. Most high-RPPM effects lack it, because they frankly don’t need it. If 10s since last procable attack already puts you at >50% proc chance (and for KM, it puts you at 75%), the bad luck protection is simply overkill.
2 Likes
What time dependency? You mean the post with hardly information and a graph? How did they come to this conclusion? All they basically said, this is my hypothesis, I looked at some logs, and this graph I made is what I found.
How did they get to that point? I just read it again and read it probably 5 times prior.
Took 10 logs. Matched each KM event with the correct crit auto. For each crit auto, counted time since last KM and whether it generated a new KM or not. Aggregate for 1 second interval and calculate the chance that a crit auto gives a KM proc if it occurs at the specific interval.
3 Likes
“Lalalala I can’t hear you because your data doesn’t agree with my belief structure lalalala”
Ya, ok, you’ve just boxed yourself in rather nicely. Fake news and all that. If the evidence doesn’t agree with your devoutly held belief, clearly the evidence is wrong. Where’s your data?
Oh right, you don’t have any. Your entire position is simply supported by your own insistence that it’s not RPPM.
5 Likes
Actually, the protection is there, it’s just that the threshold is not met. Elemental force was one example. It has an RPPM of 10. You have to go 9 seconds without a proc for this protection to activate. Hidden Satyr is at 2.5 RPPM and you need 36 seconds of no proc for the protection to come into play. With Elemental Force being 10 RPPM, that’s a proc every 6 seconds on average, 3 seconds lower than where the protection would take place.
No, there was no work shown. They just said it again, I took 10 logs and this is what I found. Where is the work, aggregated what? The information isn’t there.
I can send all the details tomorrow sure. You think I’m making the numbers up or something?
3 Likes
That’s…always the case? Like, the protection, if active, takes effect when the time since the last proc is >150% of the average interval between procs (defined as 60 / RPPM).
However, Elemental Force needs no such protection. It already hits 100% chance to proc if the time since last procable attack is >= 6s. The issue with KM is that the average interval between procs is >10s, which means the non-bad-luck protection piece cannot hit 100% like it can for Elemental Force.
And the work you’ve shown is simply insisting it isn’t RPPM. You have zero legitimacy in this discussion, Kelliste. You are not an authority on anything at all, and have blatantly misunderstood how RPPM works this entire time to boot. Please, by all mean, submit evidence for why you think it is PPM, evidence that cannot be explained by RPPM, just like the time-dependency data cannot be explained by PPM.
It disagrees with her world view. Even if you linked the 10 logs and posted completely reproducible steps to get the data, she’d still call it “fake news”. There are just some people in this world that think that personal beliefs >>> evidence.
5 Likes
Then neither do you or the other person since nothing has been shown from you guys either.
Plus, no, if the work is shown I have absolutely no issues saying I am wrong. They were able to do it before with the calculation of ability damage yet here it’s just not there.
I have also explained that ppm is consistent with what we see in game. You even asked how it can be ppm if it’s based off of crits, which crits are just the trigger for the roll. It’s not hard at all to understand.
This is a trend you’ll notice.
3 Likes