Reckoning Uptime Graphed

[Questioner]

Recently I have become interested in how to work reckoning into my spec (for farming and aoe tanking) without putting myself at risk to giant parry haste spikes. This would be through reducing the number of points invested in the talent. Naturally, it COULD still happen regardless of how minimal the chance of reckoning proc would occur, however I felt there would be an "acceptable risk" threshold.





* A word about the how the simulation works. The attackers are generated seperately and all start with their own random swing timers. Their attack speed is currently all the same, but I wouldn't mind tweaking it to vary, however I have a feeling the results will be the same.

** For those easily confused, 0.5 attacks per second means an attack every 2 seconds.

*** I do not take into account the player creating parry haste on the attackers, which would increase reckoning uptime.

**** 'miss' => 12, 'dodge' => 18, 'parry' => 16


The most interesting part of these results to me is how close 6% (3 talent points) and 10% ( 5 talent points).

[Dorvan]

How much avoidance do you have in the simulation?

[Questioner]

Screw you Dorvan!


j/k ... when I wrote it I was thinking you could just decrease the number of attacks per second (eg 50% avoidance = running simulation with half as many attacks) but I suppose the rng might play havoc with the uptime (avoidance streaks)...Alright...to the batcave.

*Updated to reflect avoidance*

[Joacimcans]

i cant see the image sorry.

[jere]

In the odd chance that you aren't already doing it this way, I would run it as a Montecarlo simulation. Since you want to capture avoidance streaks with a random number generator, a series of trials would give more accurate data. Just make sure to seed your random number generator properly.

[Questioner]

Its currently the average of 10 runs of 6000 seconds, which would be what you describe. And the image should be fixed. Darn ftp lied to me.

[jere]

Out of curiousity, do you still have the graph that didn't include avoidance. It would be interesting to compare the two extremes and see how much Reckoning uptime differs between the two.

[Questioner]

The graph before the avoidance sim:



I tried to make a little overlay so you could see the difference, but alas I am not INVISUSIRA

[JacobHC]

I tried to make a little overlay so you could see the difference, but alas I am not INVISUSIRA

Ask and ye shall receive. The faint lines are the pre-avoidance sim.


Was too lazy to change the colors, sorry.

[Menoch]

Any chance you could tweak it to do redoubt uptime. Only real change is the 5 charges can get eaten up.

[cordelia]

Ummm...

What do you use for the tankadin's weapon speed? Tank weapon speed should be essential to the uptime calculation, as reckoning procs are based off boss swing timer, yet the uptime of reckoning is based off the tank swing timer.

[Sabindeus]

Ummm...

What do you use for the tankadin's weapon speed? Tank weapon speed should be essential to the uptime calculation, as reckoning procs are based off boss swing timer, yet the uptime of reckoning is based off the tank swing timer.

uh.

I think I see what you're saying... that reckoning charges will run out at the speed of your weapon, but that isn't what is being modeling here. This is to determine how many attacks you'd have to be taking for Reckoning to be up all the time regardless of how fast you expend the charges.

[Dorvan]

uh.

I think I see what you're saying... that reckoning charges will run out at the speed of your weapon, but that isn't what is being modeling here. This is to determine how many attacks you'd have to be taking for Reckoning to be up all the time regardless of how fast you expend the charges.

...but if the goal is to evaluate the threat boost form reckoning, and most Tankadin weapons are 1.6-1.8 speed, it is important to take that into account....it should have a noticeable impact on the evaluation

[cordelia]

Ummm...

What do you use for the tankadin's weapon speed? Tank weapon speed should be essential to the uptime calculation, as reckoning procs are based off boss swing timer, yet the uptime of reckoning is based off the tank swing timer.

uh.

I think I see what you're saying... that reckoning charges will run out at the speed of your weapon, but that isn't what is being modeling here. This is to determine how many attacks you'd have to be taking for Reckoning to be up all the time regardless of how fast you expend the charges.

Huh? That doesn't make any sense to me. The boss attack speed is fixed, according to Questioner: swing time = 2.0, so you're tanking a set # of attacks. I presumed the graph swhoed the % of reckoning uptime, or what percentage of the total 6000 second simulation that reckoning procced. Did I misinterpret the graph?

PS> Damn, this thread, the threat analysis thread, and the SoR vs. SoV thread are making me want to spend a couple days and build a complete tanking Monte Carlo simulator in Matlab. Oooh.. that would waste so much time... but soo tempting.

[Ariashley]

Reckoning uptime is not impacted by tank weapon speed, only by mob weapon speed, since it procs based on incoming damage. Questioner's post (so far) addresses only how much reckoning is up.

I think what Dorvan et al are getting at is really the next step and it sounds like there are 3 next steps:

1) Given the uptime of reckoning, how much more incoming damage/streaks of incoming damage might be caused by boss parry mechanics (the 1 mob scenario)?

2) Given the uptime of reckoning, in a single target boss encounter, how much more threat will it add at various tank weapon speeds? (this bit is affected by weapon speed) This is somewhat complicated to model since added threat will be affected by weapon speed, weapon damage and hit rating and would also scale with the uptime.

3) Since the graph shows us the 4/5 Reckoning and 5/5 Reckoning both result in very close amounts of uptime and that at higher numbers of mobs, 3/5 Reckoning also comes close, is 1 additional point (after 3) valuable enough to warrant spending it?