[Frost][5.2] 10 Minute Parse / T15 4PC Problems

[zomgDPS]

Question though, any idea why my recount # of ice lances is higher than the counted FoF procs in the combat log?

You were spamming Ice Lance, a few got through which weren't FoFed.

In fact, you spam yer keys quite a lot lol :P You get quite a few "SPELL_FAILED" events due to another action in progress.

I think I might keep building this parser to be a playstyle analysis tool. Like it will take your combat log from a raid or what not, then produce some interesting data about HOW you play, not really about your performance. It should be fun to see how 'spammy' some people are and how others are pretty cool under pressure and to see what impact that has on performance.


OOOH the possibilities, but never enough time

---------- Post added 2013-03-01 at 01:29 PM ----------

And on a side note, I really appreciate you backing me. It takes a while to do these tests

Back in the day, we in the theorycrafting community (this was back when I used to still primarily work over at the EJ forums) we had a mage who was our "Master of the Field Test". Zaldinar was his name. I believe, he stopped playing WoW (or maybe just gave up on theorycrafting?).


He helped us all out by being the one who would go out there and do the type of stuff you are doing. We have to remember, back then we weren't told what the coefficients of the spells were, how they scaled, or really anything about them. Pretty much everything we got was by doing real testing, the type you are doing now Akraen.
Due to this, I have had a long standing respect for 'fieldcrafters', i.e. theorycrafters who were the guys who went out there into the field and actually did something to bring the rest of us data. The entire theorycrafting process is worthless without them.

The sad part is, Lhivera knew Zaldinar too, and I'm quite sure he respected him as well (we all did). So for him to just blatantly turn away someone who is now doing the same just goes to show me how far Lhiv is from his original self. It is sad, though at the same time, it is ever more important to wrestle back control of your class, lest you face the same damnation that the MVP does.

[Akraen]

So true, so true.

I got my button mashing habit from tanking as a bear druid for 6 years, haha. Helps me focus if I'm always pressing something. I'll see if I can get your code goin' and I will have time to do some tests this evening.

[Deleted]

@Zomg: I think you really over-estimating Lhiv importance. I always feel better with a dummies than with SimC, and RaidBot over all.
Dummies don't lies !!!!

[Wuga]

Question though, any idea why my recount # of ice lances is higher than the counted FoF procs in the combat log?

There are separate events for procing fof when you have zero stacks versus when you have one stack... I slightly modified the script above to account for this, which gives closer results (apparently you did double mash it once):

Linecount: 55711
# of Frostbolts cast: 3200
# of FoF procs: 506

For reference:

if ($#ARGV != 0 ) {
	print "Gotta type the name of the parsefile broseph...\n";
	print "$#ARGV";
	exit;
}
$filename=$ARGV[0];
$numFrostbolts = 0;
$numFoFprocs = 0;


$event1 = "SPELL_CAST_SUCCESS";
$event2 = "SPELL_AURA_APPLIED";
$event3 = "SPELL_AURA_APPLIED_DOSE";

$spell1 = "\"Frostbolt\"";
$spell2 = "\"Fingers of Frost\"";

$actor = "\"Akraen\"";
$testActor = "\"Spiderr-Brill(EU)\"";
$lineCount = 0;


open (FILE, $filename) or die $!;

 while (my $lines = <FILE>) {
	
	$lineCount++;
	$line = $lines;

	@first = split(',', $line);
	@second = split(' ', $first[0]);
	
	#check event first
	if( $second[2] ne $event1 && $second[2] ne $event2 && $second[2] ne $event3 ) {
		next;
	}
	
	#next check the actor (logs pick up all sorts of randomness)
	if( $first[2] ne $actor ) {
		next;
	}
	
	if($second[2] eq $event1) {
		if( $first[10] eq $spell1 ) { 
			$numFrostbolts++; 
		}
	}
	
	if($second[2] eq $event2 || $second[2] eq $event3 ) {
		if( $first[10] eq $spell2 ) { 
			$numFoFprocs++; 
		}
	}
	
 }
 close (FILE);

 print "---------\n";
 print "Output\n";
 print "Linecount: $lineCount\n";
 print "# of Frostbolts cast: $numFrostbolts\n";
 print "# of FoF procs:	$numFoFprocs\n";

 exit;

[zomgDPS]

There are separate events for procing fof when you have zero stacks versus when you have one stack... I slightly modified the script above to account for this

Aaah, good catch.

Yea.. I had a feeling there was another event call, but I didn't have time to check the repository of events over at wowpedia (as I said, I was coding in the kitchen :P).

Thanks for the catch man.

(n.b. sorry for the second1 second2 bullsh!te variable names. I really did do it very quickly I really should get it up to style if I want to release public code)

[Wuga]

I should add that aggregating the two logs gives a total of 996 procs based on the parser out of 6400 frostbolts, or 15.6%. This is about 10.7 standard deviations below 21%. The odds of this occurring given a 21% true proc rate are approximately 1 in 200,000,000,000,000,000,000,000,000 (200 trillion trillion).

To give you a sense of how "unlucky" you are, I simulated this thing one billion times (unfortunately, I don't have enough memory for 200 trillion trillion samples)... the black mass is the distribution of procs in 6400 casts assuming 21%, and the red line is the observed proc rate.

[zomgDPS]

For the final stage of this mystery, let us get some logs of a good number of frostbolts done WITHOUT the set bonus.

When we have those results, we will have everything we need to take this to Blizz with incontrovertible proof of the issue. We can do it ourselves too, no need to deal with the MVP's bullshite.

Get to it Akraen + others!! :P

[Deleted]

I just wish they already follow this discussion and they are working for resolving it.
Good night all (23h18 here)

[Akraen]

I am testing it without the bonus on Anasterian now. I'll post logs in a couple hours once I've done a few thousand.

1 in 200 trillion trillion is too hard to conceptualize. I think the point is definitely proven.

---------- Post added 2013-03-01 at 07:49 PM ----------

Good news! Case closed.

Just got confirmation from a dev on the PTR.

It has been fixed for next release, and will go back up to 21%.

It was not turned on in their internal testing from the set items, but now it has been corrected.

Thanks to all who were involved in this useful test

[Leviatron]

My fingers are so happy that the issue was resolved. I just can't imagine what Ak's fingers look like now >.>

[Deleted]

@AkraenStormrage @nathyiel Bug has been found and fixed for 5.2 release! Set bonus was not working. Devs send huge thanks for your testing!

It's always good to have some thanks, especially for Akraen who did most testing.

[Deleted]

Well done Akraen!

[Deleted]

Also, if dev make it for Release, so 5.2 will be launch this next week

[Sw1tch]

Maybe now they can get back to fixing the spec!

[Deleted]

Good work boys(or girls) !

Though this just give some major insight as to how blizzard is doing their "balance" etc, the entire ptr it seems they have been testing and scaling frost numbers against a tier15 set bonus not turned on? seriously ?....

With how close the 3 different specs seems to be(with this bug) and how the mechanics of the fights are working, i would make the assumption that frost will be the spec to go to until we hit the stat cap issues, and then transfer into fire?. Or we can expect a quick nerf to frost to compensate for 6% extra FoF..

[Methusula]

Good work boys(or girls) !

Though this just give some major insight as to how blizzard is doing their "balance" etc, the entire ptr it seems they have been testing and scaling frost numbers against a tier15 set bonus not turned on? seriously ?....

With how close the 3 different specs seems to be(with this bug) and how the mechanics of the fights are working, i would make the assumption that frost will be the spec to go to until we hit the stat cap issues, and then transfer into fire?. Or we can expect a quick nerf to frost to compensate for 6% extra FoF..

I'm leaning toward the former. My initial instinct was to say that they would nerf it, but the more I thought about it, the more I was convinced they would leave it. It's a PvE issue so there shouldn't be much if any, impact on PvP from this and I don't really see an issue with Frost being high early in the tier since it will quickly taper off.

[pyrostorm9001]

Considering most PvE requests of frosts damage being made more competitive with fire/arcane have been met with "we think frost is already competitive" It could be their own internal testing had the set turned on and ours was not.

[Akraen]

They tested it with the set bonus "active" without having the actual gear equipped. Therefore it was noticed that the bonus itself was not active.

That's all - no need to read into it further. We were balanced around it being 21%, but it wasn't there on actual characters wearing the bonus. So it is balanced.

It's good news, frost numbers will be higher than what's been seen on the PTR and certainly higher than sims. Just hang in there.

[Deleted]

Do you really think higher than sims? I guess it makes no difference for a simulatior if a bonus is active or not, because it assumes that it works. But I like the idea that the 4p comes on top of the sim results.

[Deleted]

Highter than Sim, I don't no but more than on PTR, it's for sure.
Better to play, it's certain.