Anyone else noticed their gnomes off of juju turned melee? X___x
Anyone else noticed their gnomes off of juju turned melee? X___x
Haste procs from the talisman do indeed increase Real PPM.
You must be unlucky then, because even with mild luck you get ~20% uptime on H bad juju, that equates 1655 agi. H talisman has 1657. If you get even more lucky then switching from bad juju to the talisman would result in you trading agility for stats, which isn't good.
Last edited by Tehstool; 2013-05-13 at 11:53 AM.
What math are you using to determine Bloodlust? Does simcraft even have it correct? The only way I can see simcraft doing it properly is having an average of the worst possible stack of bloodlust consistently at 1 stack, and another at 2 stacks and another at 3...an so on and then taking the average. There have been times on the dummy alone where I have horrible luck and very few stacks, and other times I get consistent 5 stacks. As the other person mentioned haste proc won't increase the actual chance 20% stays 20% (example) but you gain more attack chances. And based on the RPPM formula the uptime on Bloodlust is around 70%+ depending on gear. So how long of that 70% is 5 stacks, how long is it 1 stack?
I feel like there's a bunch of people immediately throwing out Bloodlust because they heard from people it wasn't good, without even testing or trying it themselves.
The way Simulationcraft works is that it runs tens of thousands of simulations, each ideally representing what you'd see in-game in that time frame. It then averages those thousands of iterations to give you the result that you see. RPPM is coded in so that in each simulation the simulated abilities have approximately the proc rate on the trinkets you'd expect to see in-game. In my experience, the uptimes on the trinkets closely match what I see on average. So basically, Simulationcraft takes everything into consideration. If you're looking at the results page and only see something like 68% uptime, then know that it simply doesn't show you the average uptime of each stack. It's not set up to track that. However, the stacking effect is 100% simulated regardless.
And even if it didn't model RPPM perfectly, when comparing two RPPM trinkets, any defect in the modelling is going to affect both. There's no reason it would disproportionately affect Talisman of Bloodlust, especially to as great of a degree as it shows. It is much more realistic to conclude that Talisman of Bloodlust is the inferior trinket. It's not a word of mouth thing, it's pretty clear.
Last edited by Kennyloggins; 2013-05-13 at 03:01 PM.
Helter Skelter settings, both 541 trinkets, 10000 iterations:
Bloodlust: 160k, centered DPS Distrubtion
Bad Juju: 157k, shifted slightly left with higher top end
This is exactly what I am talking about in my post above. If you manage to get the best procs ever with Bad Juju, then it is by far the best trinket. I am not seeing the 7k difference. On patchwerk settings the range was 3k, and favored bad juju.
I am not disagreeing or agreeing with you, but I am merely providing information based on my tests, observations and actual play.
Just for nitpicking about what Libertarian said, Armor Pen wasn't removed because it was "too mathy," it was removed because it was too good and every class that was a majority physical class (hunters, warriors, combat rogues, ferals, etc) stacked it to the exclusion of all else. It got to the point where people were turning down item upgrades because they didn't have armor pen on them, which is a situation blizzard wanted to avoid.
re: Bloodlust, I'm inclined to believe what Simcraft says. Haste is good, and for some people beats crit, but the scaling for SV always shows agility as ~3 times better than the highest secondary. Bloodlust may give you more procs on an agility trinket/scope, but I can't imagine it would be so many more that it outweighs having a second RPPM agility trinket.
Nevermind the fact that the Helter Skelter setting is pretty much useless in any real analysis since it's both totally unrealistic (hell, I don't even know exactly what it is simulating) and it can't be used as a baseline like the Patchwerk setting can, your results are not what I get. You are either using a lower ilvl version of Bad Juju or something else, because even with Helter Skelter I still sim Bad Juju as 6k DPS ahead (which is actually a greater relative difference than with the Patchwerk setting).
So basically it's simming on and off dps. Not sure of the frequency, but it just doesn't really happen that much in this tier.Fight Style
The two options here are Patchwerk or Helter Skelter. Patchwerk is a boss in Naxxramas whose name has become synonymous with fights that expect no movement from DPSers, allowing them to concentrate completely on their rotations. A helter skelter simulation is one where damage is prevent periodically to simulate fights where you are forced to move or are otherwise unable to do damage.
Patchwerk sims don't closely model many actual fights but they do give a good baseline for things like stat weights and makes a good default.
---------- Post added 2013-05-13 at 05:03 PM ----------
Also: I'm a bit of a noob when it comes to simcraft. How do you change gear sets with ease in simcraft? Do you have to change it in the code or is there some other way to do it that I'm not looking at?
Gives you a much more realistic view of what your dps is on most bosses, for things such as minor downtime (deterrence, error, whatever) or target swapping. You can't expect people to play 100% patchwerk on every single boss and pull 185k dps, so 160k is much more realistic for someone in 533 item level. So when you use actual real boss sims the two trinkets are much closer together with bloodlust and juju being practically even. The only difference really is that bad juju has a much wider range of low and high while bloodlust is much more tighter from the tests I did. This is the main thing I'm trying to get at. You don't want the RNG factor of having to hope that you get perfect procs from Bad juju. While it is nice, when you do, it is horrible when you don't. Which is exactly why heroic hunters are going haste now, to improve the RNG.