Earthquakes

[Dolalin]

Random just means unpredictable. It doesn't mean evenly distributed.

[mefdinkins]

I appreciate the fact that it is not random because no one wants to be on wave 18 of a Ring War or be 85% done with a Plane of Growth clear and have an earthquake?

The problem is people min max so crazily so when the window for the last one gets to 2/3 days left people start acting crazy - it eliminates some of the randomness aspect of it. "What the hell are you doing in EC buying food/water and selling?!!?!? WE HAVE AN EARTHQUAKE COMING PEOPLE?!!?"

[feniin]

Yeah, I don't even want to log in again until July. These next few days are going to test the limits of how much poop a tube sock can really hold.

[m00r5tuD]

Originally Posted by Mytral [You must be logged in to view images. Log in or Register.]

Earthquakes

10/10/2017
10/28/2017

11/16/2017
11/30/2017

12/28/2017
12/31/2017

1/20/2018
1/28/2018

2/11/2018
2/27/2018

3/13/2018
3/29/2018

4/16/2018
4/29/2018

5/23/2018
5/31/2018

6/26/2018
6/??/2018

sooo average second quake is the 29th. Clear you schedules on Friday boys!

also can we stop doing weekday afternoon quakes? i work 4 hours a day and somehow have missed 2 out of the last 3 quakes. if this keeps up i'm gonna have to put in my 2 weeks so my raid attendance doesn't fall too far.

[eudaimonia_etc]

It's obvious at this point, but here's some statistics. Using a Chi-Square test to see if the distribution of earthquakes matches a uniform distribution over the days of the month (i.e. if earthquakes were rolled strictly randomly), we find, with p < 0.00, the earthquakes are not distributed uniformally. It's easy to tell from the density plot as well.

#mypocketprotectorisbig

[You must be logged in to view images. Log in or Register.]

[Dolalin]

Looks like the first one is randomly distributed across the month (that's a decent Gaussian centered on the middle of the month), but the second is weighted to the end of the month then.

15th +/- 14
29th +/- 2 ?

Not sure why the weighting for the second one.

[Baylan295]

Small sample size much?

[branamil]

Classic patches would not have occurred during US workday. Verant would have deployed them off peak hours like beginning of day or end of day

[eudaimonia_etc]

Normally distributed earthquakes nilbog? You devil.

[Dolalin]

If I had to guess it's probably an artifact of how they've coded up the schedule. Maybe they're trying to space out both earthquakes from each other by some minimum amount, but the second earthquake is artificially constrained by the end of the month, so it's clustering there.