Some things I'd be interested in seeing added to the PBF timeline with a time, date, and PK ratio stamp.

When class choices are made and what they are. For example, warrior specs, shifter shapefoci, thief skills, shaman/paladin dedication, warrior legacies, etc.

When edges are taken and what they are. Specifically this would be interesting for when someone takes a edge and it appears to make their pk ratio change significantly or whatever.

When skills are perfected. I realize this is a lot of info, but cutting it down to the "important" ones is probably more work than shoving them all in there. I'd like to see is someone perfected their defenses at 15 and then wrecked shop for the next 20 levels, or if they just leveled up naturally and still did so. Would also be interesting to see when someone perfects something really tough like flurry or evade and how it affects their ratio from that point forward.

Just things I'd be interested in seeing personally.

>Some things I'd be interested in seeing added to the PBF
>timeline with a time, date, and PK ratio stamp.
>
>When class choices are made and what they are. For example,
>warrior specs, shifter shapefoci, thief skills, shaman/paladin
>dedication, warrior legacies, etc.

I'm fine with this and have changed the code to add this to the timeline history.

>When edges are taken and what they are. Specifically this
>would be interesting for when someone takes a edge and it
>appears to make their pk ratio change significantly or
>whatever.
>
>When skills are perfected. I realize this is a lot of info,
>but cutting it down to the "important" ones is probably more
>work than shoving them all in there. I'd like to see is
>someone perfected their defenses at 15 and then wrecked shop
>for the next 20 levels, or if they just leveled up naturally
>and still did so. Would also be interesting to see when
>someone perfects something really tough like flurry or evade
>and how it affects their ratio from that point forward.

I'm not a fan of these...correlation does not imply causation.

Beyond that, history has a finite amount of space (that's why certain powerhouse characters don't get all their PK wins recorded), and this just washes out what I think is important data.

So long, and thanks for all the fish!

It was always more data than I really wanted, but lacking a good way to whittle it down, I figured asking for everything was the right way to go.

Any chance of getting PK wins/losses broken out by group size:

You can "sort of" calculate this from looking at the "PK wins" and "PK deaths" pages, but since those get truncated the data's not complete.

I'm envisioning something like:

PK Wins by Group Size:
1: 123
2: 40
3: 8
3+: 2

PK Losses by Group Size:
1: 50
2: 20
3: 10
3+: 3

If that's too much trouble (which I can understand) then how about just breaking out "solo wins" and "solo losses"?

>If that's too much trouble (which I can understand) then how
>about just breaking out "solo wins" and "solo losses"?

That was already documented on the gank-o-meter page.

>Any chance of getting PK wins/losses broken out by group
>size:
>
>You can "sort of" calculate this from looking at the "PK wins"
>and "PK deaths" pages, but since those get truncated the
>data's not complete.
>
>I'm envisioning something like:
>
>PK Wins by Group Size:
>1: 123
>2: 40
>3: 8
>3+: 2
>
>PK Losses by Group Size:
>1: 50
>2: 20
>3: 10
>3+: 3

I've updated the gank-o-meter page to show solo, group of 2, group of 3, group of 4+ for both kills and deaths.

You can see the layout here on Taerg's republished PBF

http://forums.carrionfields.com/dc/dcboard.php?az=show_mesg&forum=31&topic_id=39774&mesg_id=39794&page=

So long, and thanks for all the fish!

rock on

I was looking through Taerg's PBF, and I am thinking that this entry on his PK wins, Taerg gets an assist credit here, would count as a PK win with a group of 4+ since those stats include PK wins and assists.

Oct 28, 2011|Lv 51|The Eastern Road|Odrallag vs 6: Arrgakk (0%), Gdot (22%), Taerg (0%), Bolzhor (0%), Jolka (24%), Allysia (53%, reverberating blast)

I then looked in Gdot's and saw this.
Oct 28, 2011|Lv 51|BattleRager Village|Bolzhor vs 4: Gdot (45%, nova), Odrallag (1%), Jolka (24%), Allysia (28%)
Oct 28, 2011|Lv 51|BattleRager Village|Taerg vs 4: Gdot (7%), Odrallag (0%), Jolka (47%, nova), Allysia (45%)
Oct 28, 2011|Lv 51|Galadon|Odrallag vs 6: Arrgakk (0%), Gdot (22%), Taerg (0%), Bolzhor (0%), Jolka (24%), Allysia (53%, reverberating blast)

Finally Jolka's
Oct 28, 2011|Lv 51|BattleRager Village|Bolzhor vs 4: Gdot (45%, nova), Odrallag (1%), Jolka (24%), Allysia (28%)
Oct 28, 2011|Lv 51|BattleRager Village|Taerg vs 4: Gdot (7%), Odrallag (0%), Jolka (47%, nova), Allysia (45%)
Oct 28, 2011|Lv 51|The Outpost of Tir-Talath|Odrallag vs 6: Arrgakk (0%), Gdot (22%), Taerg (0%), Bolzhor (0%), Jolka (24%), Allysia (53%, reverberating blast)

Aside- I'm not sure why the location differs for three PBFs unless the location is determined by where that character is at the time the person dies- so maybe Gdot was in galadon, Jolka in Tir Talath, and Taerg was on Eastern when Allysia probably killed Odrallag in the battlerager village- but Allysia doesn't have this PK win listed to tell for sure and Odrallag has no PBF. I am guessing the location is only accurate for PK deaths or if you are the one that lands the killing blow.

CALCULATIONS
Using that information I did the following little calculations.
I've always found it somewhat odd in the PK stats when these happen because it appears to me to be a skewing of numbers. My guess is Allysia slept some of them, Odrallag attacks them all to wake them, that generates him being involved in their deaths even though they flee instantly, and then when he dies they are included in having aided in his death because they might have attacked him after Odrallag's attack to wake them. There may have been attacks involved all of which were blocked/dodged/parried so no damage was done but still triggered inclusion on the stats.

So Odrallag, a battlerager dies defending the village along with three other battleragers. He came in and woke them up to save their lives but still three of them die. It states that he died to a group of six. Three of the six are his fellow Battleragers who all have 0 %s and I imagine did next to nothing to ensure he died and may have even been dead at the time he died. This seems like a rather adverse impact for all six of these people, but especially for those three battleragers. I use a group count system when looking at gank-o-meters. So your total is everyone ever involved in any PK win or assist with you (who are always listedon the stats), divided by your total of assists + wins.

To use Allysia as an example:
208 wins and 134 assists
gsize is 2.66
We can estimate from those numbers that her group count is roughly 909 (909 /342 = 2.657)
Just removing these five from her group count she goes to 904/342 which = 2.64.
I'll acknowledge Allysia was old as hell, had a lot of PKs so these instances become less relevant when you have that old a char with that many pks (though I would be curious if I had all the PK wins to see just how much it would drop by eliminating these sorts of instances from all 342 win/assists since she was a bard and so likely more prone to generating them then other classes- maybe it goes to 2.60 or something, I honestly have no clue.)

I then looked at Jolka as an example. A much lower pool of wins & assists than Allysia to mitigate this inflated group count with.
23 wins & 30 assists. Had a 2.68.
That means Jolka's group count is 142 against 53.
If you take out the +5 to the count, Jolka's gank-o-meter is now 137/53 which equals 2.58. This is more statisically relevant I would think.

Then I did Gdot,
Gdot goes from 273 to 268, and the gsize goes from an impressive 3.0 to a 2.94.

Finally Taerg
154 wins+assists. Gsize 1.78 Group count = 274.
For Taerg I take out an assist because he loses the assist not being included. He also loses the full 6 from the group count.
268/153. His new gank is 1.75.

Ultimately, because three nexus people attack this group of four somehow the three nexus people ended up all getting hit with a +5 to their total group count for PK wins. Odrallag gets hit with an additional 8 group count to be offset by 2 assists. Taerg, Bolzhor and Argakk all get a 6 count hit to be offest by 1 assist.

The problem is there is no way to actually counter the inflated group count, it just has to be mitigated by having lower group counts in subsequent PKs, 1 to 1 being the lowest for a solo PK. So unless you can rack up enough solo pks, you're going to have the inflated number.

What looks more like a 3 v. 1 (possibly a 3 v. 4 if the three Rift people were fighting all the Battleragers which seems plausible given it is three area spell/song people) ends up showing as a 6 v. 1. Now look at how to lower the gank meters for the nexus chars. If they had only those three pks, they would have a 14 group count to a 3 instead of a group count of 9. To get to a gank o meter of 2 should have taken three more individual pks. (9/3 becomes 12/6)
However, now they need 8 solo pks to get a group count of 22 against 11 wins/assists to get to that same gank o meter of 2. I feel like when you go over 2, anti-ganging will hit you harder in terms of finding openings, but then again, I played Allysia who had a high gsize per kill so maybe it was because I had over a 2.5 that I noticed it more.

CONCLUSIONS
To me it seems like these sort of things can really hamper an actual gank-o-meter, and I would not care about it at all except that I believe a gank-o-meter has an impact on gameplay, such as ehren and maybe how anti-ganging code works. (It seems like high gank-o meter people can gang less effectively even if they fight 2 v.1) And I am sure the three villagers who have 0%s don't want those on their stats for the negative impacts that accompanies it. Especially Ehren characters who want as close to 1 as possible, things like that just make it all the harder to get down to that when you throw in the actual instance of group fighting that come with being in cabal raids/defense.

I understand that a person does not necessarily have to inflict damage to impact a PK, but a lot of times you will see people involved in PK stats (especially on any class with sleep since allies wake their friends) that are cabal allies or have 0%s. This is often the case when dealing with any class that uses sleep in some form since cabal mates often wake them. Odrallag did everything right to save them, he gets creditted as being in on two 4 v. 1's, and himself gets creditted as being violated by a 6 v.1.

Just curious if there is a way to mitigate this (outside of don't use sleep- don't wake allies), or to have people who are in the same cabal not be included in pk stats like this since their involvement is almost always intended as beneficial over detrimental. I understand some times people do kill each other within the same cabal, but I was thinking if a person trusts the cabal and inflicts no damage on their cabal mate, how could they have possibly aided in that person's PK death when more than likely everything they did was an effort to prevent the death?

I realize that on characters that have more and more pks these instances become less relevant when you have say 100s of pks/assists to balance them out, but I would think it in fact creates a noticeable skewing against classes like bards and necromancers who often will sleep and kill people, even though someone else comes and wakes up the target. And it will also more heavily impact people with far less PKs.

Anyway, that was a lot of math and talking, but it was something I had been thinking about for a while when I look at gank meters, so I thought I would post it while Zulgh was looking at and making changes to this PBF stat.

Too much math, speculation, etc. I browsed it and here's my response.

Yes, if you attack your cabalmate to wake them up and damage them, you go in the list of people who contributed to their death. You actually have to damage them to be on that list, so you may show up as 0%, but you've done damage to them. I have no desire to change this...since you did damage them, thus contributing to their death. If you do no more damage to the player for a few ticks, you'll fall off the damage list, meaning that they died fairly quickly after you woke them up, and thus you're still on the list. The damage history also clears if the player fully heals.

The other way 0% shows up is things like sleep, faerie fire, etc. Offensive things that do no direct damage. They too are part of the contributing factor of their death.

Yes, gank-o-meter comes into play for Ehren stuff.

So long, and thanks for all the fish!

That most of the time there is a zero next to the villager's name is because of how Mercy works. I'd bet if you looked at Ganicus' PBF you would find some of the same instances. I have found that it is very rare that you actually do no damage to someone when you wake them. And as Odrallag, I was generally working with a decent +dam set.

RE: gank-meter
I think players really have to take this with a grain of salt. Any character that does a majority of their PK in a raid situation is almost garaunteed to have a relatively high gank-o-meter. As for how it interacts with the Ehren edges, I think that this is an intended drawback. I won't speculate too much, but for those that weren't around or don't remember, I suggest reading this post. I know it helped me understand the concept a little better.

http://forums.carrionfields.com/dc/dcboard.php?az=show_topic&forum=44&topic_id=303

Nifty!

The PK wins section is any PK in which the guy participated, right? Not just ones where he got the last hit? I was envisioning the latter, but I'm certainly not going to complain.

>Nifty!
>
>The PK wins section is any PK in which the guy participated,
>right? Not just ones where he got the last hit? I was
>envisioning the latter, but I'm certainly not going to
>complain.

Total PK Wins   114 
Total PK Assists   40 
Solo PKs   82 
PKs with a group of 2  39 
PKs with a group of 3  22 
PKs with a group of 4+  11 

So long, and thanks for all the fish!

If I am in a group of three and I am the only one that can harm someone and I kill them, is that still a group of three or a solo? Happened recently and this discussion made me curious. Thanks!

n/t

So long, and thanks for all the fish!

nt

People not to be creative themselves. Most people already know the basics of a cookie cutter character, I don't think dumbing it down any more for somebody really has many benefits at all. We don't need a take this edge at x level section in the pbfs.

Could make the output kind of spammy but it would be nifty to see how different people execute the similar build.

In seeing how or if various edges or perfecting skills made a difference in my ratio, but it would serve the same purpose for other people. It is a bunch of information though. The thought of using it to copy someone else's build wasn't the point.

Joel