Skip to content
Maintankadin

Improved tanking simulator

AergisAergis's avatar
Tue Apr 24, 2007 2:56 pm
sweet, thanks, i'll work that in. Need to do a range for damage too. I have been working with a static damage per hit number, but a range is more realistic.
_________________
Arcand
Wed Apr 25, 2007 1:31 am
I've still got the log on my work computer; I'll pick the numbers apart and let you know if it's uniform or normally distributed or what.
_________________
Scarlet Crusade - 0/46/15 - 11/11 Kara, 2/2 GL, 4/6 ZA
hiro
Wed Apr 25, 2007 6:40 am
Can you post that file anywhere? ATM we're going to get crushed a lot just due to that casting bug, but I'd love to see the average response time of a tank on the ground, what kind of gap we are talking about between HS expiring and being re-applied.

That won't be going away and won't be improved at all by the current improved HS.
hiro
Wed Apr 25, 2007 7:26 am
Yup, lag and human timing errors. If you have perfect inhuman timing, and the lag isn;t a factor (the cooldown is client side) each one should be ten seconds after the next:

21:13
21:23
21:33
21:43
21:53
22:03
22:13
22:23
etc etc

However, it is FAR more likely it looks like:

21:13
21:23
21:34
21:44
21:55
22:05
22:16
22:26
etc etc

In the above example, there was a consistant small lag of about half a second. I suspect real tank parses will show a much higher average, probably near a second. I'm trying to get a handle on exactly how big a gap we are looking at if they never give us the overlap warriors get.
Arcand
Wed Apr 25, 2007 8:28 am
I'll start a separate thread for it; it's not the whole fight. At one point I take two crushes in a row, one from the casting bug (Curator hits me at the exact millisecond that I hit him) and a second one right after that, I guess because I screwed up and let HS stay down.

Another odd thing - in several places I take normal hits while HS is up...
_________________
Scarlet Crusade - 0/46/15 - 11/11 Kara, 2/2 GL, 4/6 ZA
kurros
Wed Apr 25, 2007 9:43 am
Is there a way to make the simulator run X times and then show the averaged results?
Torakka
Thu Apr 26, 2007 6:28 am
Wrong, every 15% of non-special melee attacks made by 3 or more levels higher mob than you are Crushing unless you can push them off the table. So for example with 10% miss, 15% dodge and 15% parry but not counting block rest 60% of attacks made against you succeed. They are divided to normal hits (45%) and Crushing Blows (15%) and if you just count succesfull attacks it look like CBs had a 25% chance. In reality you should of course have enough block and/or higher miss/dodge/parry to avoid CBs completely.
_________________
Arcand
Thu Apr 26, 2007 8:31 am
Here's the hit I was thinking of, and there's no mystery after all. I cast Consecration, and he normal-hit me 0.109 seconds later. Sure will be nice when that fix goes in. Smile

Torakka
Thu Apr 26, 2007 10:27 am
Kathryn, yea I know that. I was just using that scenario without any block at all as an example to demonstrate that Crushing Blows occur 15% of all attacks made, succesfull or not, if they aren't pushed off.
_________________
Torakka
Thu Apr 26, 2007 11:01 am
I was referring to this: