Simcrafting Theralion's Mirror 4.2 Simcraft

[Blackmorgrim]

To be honest, I don't think this bug is as big as it sounds.

It might settle the issue of the haste/mastery balance between 1601 and 2589.

But beyond 2589, most WoL reports are showing near 100% ES uptime anyway.

I don't even expect our BiS to change in the slightest.

Right now, I'm just laughing at all the people freaking out... and apparently taking down Shadowpriest.com in the process.

L2gear ppl... its not that hard! ;)

[Kilee]

I think there will be some small changes similar to what Twintop is saying, but I think you're right that in the end they will be very minor and probably will not change anything about our final sets. (This in itself because of the limitations of what is available.)

I can honestly say, though, that ES falls off about once per battle, perhaps more. Never for more than a few seconds, and certainly not bad enough that I worry about it. But a simulation that isn't looking at ES would reliably cast a few dots during this down time, and not refresh them again.

I expect a true "Simulation" to show these errors, or else - quite frankly - it's a bad simulation. I'm going to be looking for this behavior in the next version.

In the past I assumed Simulationcraft wasn't picking up anomolies like this because it was single-targetting, or because it was somehow smoothing out the RNG of Orb generation. After this fix, what I'm expecting to see is ES falling every now and then, especially at low haste levels. I'm also expecting to find a lower threshold of haste where ES is down for long periods of time.

As Twin says, this should drive mastery's value down, at least at lower gear levels anyway.

I'm also expecting the opening rotation of the sim to make a real, noticeable difference in the total dps now. Because before with the rotation it was doing, there was no way ES would go up as fast as it appeared to.

I also would like to go back and re-do my graphs with this fix and see if there's any interesting changes in the stat graphs going from iLevel 333 to 397.

So anyway there's a lot of exciting things to look forward to, for me. Maybe there won't be many changes from a practical standpoint, but I enjoy seeing how the relationships of our spells work in more detail.

[Kilee]

Also lol at shadowpriest going down... I half-wonder if the past few days of "debate" took the site down because of "spectators" checking up on it.

[Schisme]

Also lol at shadowpriest going down... I half-wonder if the past few days of "debate" took the site down because of "spectators" checking up on it.

Kilee, it's obviously because of my revelation about Nina abusing blackholes for his own Shadow Priestery benefits that forced them to take action to cover up the conspiracy. :)

But on topic, I've experienced the same thing... that ES has 95% uptime or higher in my fights and I don't know that I ever have DoTs fall off or are recast in that downtime so MF is the only thing affected in that period.

[Kilee]

Althor, if you do come back and read this, I have a few requests/questions about Simulationcraft.

1. I would like to see Simulationcraft correctly handling the clipping of a 2nd tick of Mind Flay in favor of refreshing dots and casting Mind Blast. There's a command for interrupting spells, but when I use it, it seems to interrupt EVERY Mind Flay, which shouldn't be happening. I certainly don't think it's interrupting Mind Flay in an intelligent way.

2. I would like to see Simulationcraft able to recognize and handle the refreshing of dots whenever procs happen. This may be easy to do now, yet require hard-coding the action list in some way that I'm not aware of.

3. In addition to #1, I would like to see Simulationcraft able to handle casting a Mind Flay on top of another Mind Flay and "continuing the channel" for a result of a 4-tick Mind Flay. This would only work if #1 were modeled correctly, as not modeling #1 correctly would make this a loss in DPS. I realize that there's probably very little difference in PP values between extending ticks vs re-casting them, so this is a very minor request.

My concern really is more of getting Simulationcraft to behave as we cast in the actual game. Thus becoming more of a true Simulation than current.

On a personal level I would like for Simulationcraft to model clipping of Mind Flay so that there is some kind of imperical proof that we should/could be doing it in game. The other stuff is largely unimportant.

[eleusis]

Does this mean that currently this trinket is really under valued?

Also what impact does this bug have on current normal pp values? is it even significant in any way and if so what stat weights would change?

[Revulva]

Yes, this will definitely have an effect on any stat weight theory which has been done using SimulationCraft. Pretty significant flaw, but, who knows, maybe the "real" up-time will be so close to 100% that our results won't change TOO much. I would still have to think that Mastery's value will be slightly reduced.

[Arlee]

Yes, this will definitely have an effect on any stat weight theory which has been done using SimulationCraft. Pretty significant flaw, but, who knows, maybe the "real" up-time will be so close to 100% that our results won't change TOO much. I would still have to think that Mastery's value will be slightly reduced.

Idk... At higher gear levels ES up time seems to be pretty close to 100% anyways, so it might not make much of a difference there. At lower gear levels... well mastery has never scored very high anyways, so I am not entirely sure it will actually make that much of a difference. This big difference would be if after this it could model mastery procs better (not overwriting a proc'ed es until you have to etc) which would give a better feel for how important those are. Idk if that can be done or even would happen as a result of this fix though.

[Revulva]

Where does one find the beta versions of SimC? I can only see the current release in the downloads.

[Kilee]

Interesting. I can't wait to play with it.