PDA

View Full Version : The Oracle of Death, Part II


maradine
2013-09-20, 01:44 AM
9-25-2013 Update: All commands are now available via PM. This should be your preferred query method unless you are in an active discussion on the result and need to display it in the channel.

The Oracle of Death is in live alpha. She sits in waiting, her only purpose to answer your questions. What does live alpha mean?


Shit breaks.
Functionality is minimal.
Anything may change for any or no reason.


But, in the mean time, get your daily stat comparisons by your own hand. How do you commune with the Oracle? Connect your IRC client to irc.slashnet.org, channel #planetside2.

What can you do with the Oracle?

!oracle list types
An enumeration of all of PS2's item categories, filtered for things that can actually show up in a kill log. Listed for convenience in next post.

!oracle list weapons [type_id]
An enumeration of all of the weapons for that particular type. Not currently broken out by faction, but that's next.

!oracle ask weapon [weapon_id]
The things you're used to - kills, uniques, kpu, etc. By default, the time period will be the last complete 24 hour period ending in 06:30 UTC - PS2's kill nadir, when Oracle maintenance is run. If you actually know what period you're looking for, you can also run . . .

!oracle ask weapon [weapon_id] [period_id]

Want to know which time period is which?

!oracle list periods

Perhaps you just want to cut to the chase and get everything the Oracle knows about a weapon. In that case:

!oracle dump weapon [weapon_id]


An important note: I am not currently optimized for message passing, and the server will throttle the Oracle. If lots of people make requests at the same time, especially for lists, there will be a delay as the IRC server queues out response lines. Do not ask twice - you will be answered. Slashnet may or may not be the Oracle's final home. A server with liberal flood whitelists might be ideal.

/me coughs politely

So! Have fun, say hi to the FKPK guys, ask anything. The Oracle awaits.

maradine
2013-09-20, 01:47 AM
Type Enums:


03:03:11 @maradine | !oracle list types
03:03:12 %ORACLE | maradine: Delivering via PM.

03:03:12 ORACLE | Listing Types:
03:03:13 ORACLE | 2: Knife
03:03:13 ORACLE | 3: Pistol
03:03:14 ORACLE | 4: Shotgun
03:03:14 ORACLE | 5: SMG
03:03:15 ORACLE | 6: LMG
03:03:15 ORACLE | 7: Assault Rifle
03:03:16 ORACLE | 8: Carbine
03:03:16 ORACLE | 9: AV MAX (Left)
03:03:17 ORACLE | 10: AI MAX (Left)
03:03:17 ORACLE | 11: Sniper Rifle
03:03:18 ORACLE | 12: Scout Rifle
03:03:18 ORACLE | 13: Rocket Launcher
03:03:19 ORACLE | 14: Heavy Gun
03:03:20 ORACLE | 16: Flak MAX
03:03:20 ORACLE | 17: Grenade
03:03:21 ORACLE | 18: Explosive
03:03:22 ORACLE | 19: Battle Rifle
03:03:22 ORACLE | 20: AA MAX (Right)
03:03:24 ORACLE | 21: AV MAX (Right)
03:03:24 ORACLE | 22: AI MAX (Right)
03:03:26 ORACLE | 23: AA MAX (Left)
03:03:36 ORACLE | 109: Flash Primary Weapon
03:03:37 ORACLE | 110: Galaxy Left Weapon
03:03:37 ORACLE | 111: Galaxy Tail Weapon
03:03:39 ORACLE | 112: Galaxy Right Weapon
03:03:40 ORACLE | 113: Galaxy Top Weapon
03:03:41 ORACLE | 114: Harasser Top Gunner
03:03:41 ORACLE | 115: Liberator Belly Weapon
03:03:42 ORACLE | 116: Liberator Nose Cannon
03:03:44 ORACLE | 117: Liberator Tail Weapon
03:03:45 ORACLE | 118: Lightning Primary Weapon
03:03:46 ORACLE | 119: Magrider Gunner Weapon
03:03:46 ORACLE | 120: Magrider Primary Weapon
03:03:47 ORACLE | 121: Mosquito Nose Cannon
03:03:49 ORACLE | 122: Mosquito Wing Mount
03:03:50 ORACLE | 123: Prowler Gunner Weapon
03:03:51 ORACLE | 124: Prowler Primary Weapon
03:03:53 ORACLE | 125: Reaver Nose Cannon
03:03:54 ORACLE | 126: Reaver Wing Mount
03:03:54 ORACLE | 127: Scythe Nose Cannon
03:03:55 ORACLE | 128: Scythe Wing Mount
03:03:56 ORACLE | 129: Sunderer Front Gunner
03:03:56 ORACLE | 130: Sunderer Rear Gunner
03:03:57 ORACLE | 131: Vanguard Gunner Weapon
03:03:59 ORACLE | 132: Vanguard Primary Weapon

ChipMHazard
2013-09-20, 06:27 AM
How long before she gains consciousness and KILLS US ALL!
Eh, sorry about that.

This looks neat:D

Taramafor
2013-09-20, 06:33 AM
How long before she gains consciousness and KILLS US ALL!


When the name is changed to Skynet. :evil:

MerlO
2013-09-20, 10:14 AM
I seem to be getting null on q3kpu for every query I make...

Thanks for this maradine, very cool...

maradine
2013-09-20, 11:14 AM
Good catch - fixed.

Stomps
2013-09-21, 02:50 AM
Only seems to do weapons with a period added, otherwise any straight 'ask weapon nnnn' query errors with 'Was that a legit weapon id and period? I didn't get anything back.'

Also how to do group calls? Eg stats on all shotguns, or all pump shotguns but without entering them all individually query line by query line?

Also theres no faction tags on things like liberator belly guns, no idea which is TR/NC/VS after a 'list weapons 115' query

And tank mines/c4 only have one entry for all

just minor problems tho, looks good so far on the whole, nice idea doing it through an ircbot

MerlO
2013-09-21, 09:36 AM
Would it be possible to get a datestamp on the output when querying a previous period - for looking at trends over time? I guess you will get to adding querying of multiple periods...

Thanks again.

maradine
2013-09-21, 01:21 PM
Yeah, I will. I actually added '!oracle list periods' yesterday, but it's outputting in epoch time at the moment - useful for me and SOE, not for anyone else. As an aside, Java date/time handling is by far the worst part of the default language. TL;DR, I'll get it pretty this morning.

As to Stomps's list:


It's also possible your weapon id was off. Default periods was working yesterday. I'll take a look and see if I fucked up something.
Group calls are next on the feature list. Haven't figured out if I want to add sorting syntax or not, but it should be done this weekend.
As mentioned above, I have it, I'm just not presenting it yet. Will fix shortly.
What you see is what you get for explosives. That's how SOE records them. There's a real-time resolve workaround I can do for weapons like that since I'm sitting in the raw stream, but it's pretty low on the priority list.


Thanks for the feedback!

Mordelicius
2013-09-21, 02:44 PM
Oh mighty Oracle, I beseech thee, share to us your knowledge and understanding of everything PS2.

Is it possible to get Headshot data between factions? Such as specific/general infantry weapon headshots?

maradine
2013-09-21, 03:33 PM
Yes. Let's take a look at whether that would be interesting or not:

Period 7, September 20th, All Servers
id 24004: EM4 Longshot
====
Kills: 14,918
Kills that were headshots: 11,662, 78%
Uniques: 1,338
KPU: 11.15
AVGBR: 44.7

id 25004: RAMS .50M
====
Kills: 18,304
Kills that were headshots: 14,322, 78%
Uniques: 1,656
KPU: 11.05
AVGBR: 45.3

id 26004: Parallax VX3
====
Kills: 15,678
Kills that were headshots: 12,538, 80%
Uniques: 1,310
KPU: 11.97
AVGBR: 47.1

I'd call that a hell of a control.

If headshot data is interesting to you, I'll add it to the manifest. It's a simple additional summation in the aggregator over rows it already has to touch every night.

Stomps
2013-09-21, 08:13 PM
I did try a few different ones, they all errored until i put in a period (5) and they then completed

Id be interested in headshots over 50% for automatic weapons (cheaters to report)

Something ive not seen mentioned yet, and not something we can easily pull at the user level, what is the user weighted overall Kpu for all weapons, all vehicles and all universally combined?

maradine
2013-09-21, 08:35 PM
I don't do per-user stats. I have the data, but that's not the direction I'm approaching this from. For the record, there are guys inside SOE running that exact logic looking for outliers.

I popped in this morning to check on the query engine - everything's reporting back successfully.

Stomps
2013-09-22, 02:59 AM
I don't do per-user stats. I have the data, but that's not the direction I'm approaching this from. For the record, there are guys inside SOE running that exact logic looking for outliers.

I popped in this morning to check on the query engine - everything's reporting back successfully.

1. No perhaps I didnt explain it fully, I mean im interested in knowing what is the global all game all weapon average KPU, a single figure averaging kpus for every weapon in the game (or broken out to three figures, the global all figure again but showing its make up of global soldier weapons + global vehicle weapons) to find a single overall Kpu - for us to find that at our user level (using the ircbot) would involve pulling kpus from every weapon class and computing the answer from those when yourself would maybe have that on your fingertips (or at least be able to run the query much faster)?

2. For me its still making an error without a period. Here are my steps, ill copy the exact lines out from my irc entry bar along with the relevant part of their replies needed to test. (brackets are my notes)

!oracle list weapons 4 (shotguns)
ORACLE 39001: TRS-12 Uppercut (15 shotguns returned including this one)
!oracle ask weapon 39001 (to retrieve data for the uppercut, note no period used)
ORACLE MyIrcName: Was that a legit weapon id and period? I didn't get anything back. (error here)
!oracle ask weapon 39001 5 (same line with added period)
ORACLE MyIrcName: TRS-12 Uppercut - kills: 3240 uniques: 315 kpu: 10.3 avgbr: 59.6 q1kpu: 8.2 q2kpu: 9.7 q3kpu: 12.2 q4kpu: 9.1 (completed query)

hth

MerlO
2013-09-22, 03:27 AM
2. For me its still making an error without a period. Here are my steps, ill copy the exact lines out from my irc entry bar along with the relevant part of their replies needed to test. (brackets are my notes)


hth

Now it's happening to me too. From what I can tell it's related to: 1,2,3 character weapon IDs work without period, 4,5 characters don't.

Strange that yesterday I queried a 4 character weapon ID and it worked...

maradine
2013-09-22, 11:28 AM
!oracle ask weapon 39001 (to retrieve data for the uppercut, note no period used)
ORACLE MyIrcName: Was that a legit weapon id and period? I didn't get anything back. (error here)


15:24:17 @maradine | !oracle ask weapon 39001
15:24:18 ORACLE | maradine: TRS-12 Uppercut - kills: 4621 uniques: 355 kpu: 13.0 avgbr: 68.7 q1kpu: 6.2 q2kpu: 7.6 q3kpu: 12.4 q4kpu: 25.5

I can see the failed queries up in scrollback, so something's definitely up. It's a bug hunt!
edit: I think I see whats happening here - all of the failed examples are happening during nightly maintenance. That shouldn't cause them to fail (not in that failure mode, anyway), but now I know when to watch.

edit2: tricky tricky tricky. I call the assembly's attention to the nightly aggregation procedure:

-- --------------------------------------------------------------------------------
-- Routine DDL
-- Note: comments before and after the routine body will not be stored by the server
-- --------------------------------------------------------------------------------
DELIMITER $$

CREATE DEFINER=`fkpk`@`%` PROCEDURE `kills_nightly`(out deleted_rows int, out time_elapsed int)
BEGIN
DECLARE start_stamp int;
DECLARE end_stamp int;
DECLARE period_id int;

SELECT UNIX_TIMESTAMP(NOW()) into start_stamp;

-- get the timestamps
CALL fkpk.get_time_brackets(@front, @back, @week);

-- create a new time period

INSERT INTO fkpk.v2_time_periods (start_time, end_time, is_daily) VALUES (@front, @back, 1);
SELECT last_insert_id() into period_id;


-- run aggregates over timestamps
CALL fkpk.aggregate_weapons(period_id, @front, @back);

-- delete before weekly
DELETE FROM fkpk.v2_kills
WHERE timestamp < @week;
SELECT row_count() INTO deleted_rows;

SELECT UNIX_TIMESTAMP(NOW()) into end_stamp;
SET time_elapsed = end_stamp - start_stamp;

END

Parts, in essence:

1. Figure out what time it is.
2. Build a new time period.
3. Aggregate every weapon for that time period.
4. Clean out any raw kill data older than a week.

Parts one and two are essentially instant. Parts 3 and 4 each take about an hour each for an RDS instance of this class. Now to the request:

public static HashMap<String, String> getKillAggregate(Properties props, int id) throws SQLException {
//figure out most revent daily and then run more granular call;

Connection conn = null;
ResultSet rs = null;
PreparedStatement stmt = null;

try {
conn = getConnection(props);
String sqlString = "SELECT id "+
"FROM fkpk.v2_time_periods "+
"WHERE is_daily = 1 "+
"ORDER BY end_time DESC "+
"LIMIT 1;";
stmt = conn.prepareStatement(sqlString);
rs = stmt.executeQuery();
rs.next();
int period = rs.getInt(1);
return getKillAggregate(props, id, period);
} catch (SQLException ex) { blah blah blah

Typos aside, a period-less call queries the database for the most recent period and then runs the more specific call with that period.

You are asking the Oracle for an aggregate immediately after the time period has been built, but before the aggregator has fought it's way to the weapon id in question. Fun!

A few possible simple fixes for this logic error. I'll probably just not write out the period until after the aggregator's done; I just need to make sure I'm not relying on that row being written already elsewhere.

Good catch, guys!

edit3: Your reward is a better period query:

15:58:54 @maradine | !oracle list periods
15:58:56 ORACLE | Listing Periods:
15:58:56 ORACLE | id: 8 start: 2013-09-21T06:31:40.000Z end: 2013-09-22T06:31:40.000Z daily? true
15:58:57 ORACLE | id: 7 start: 2013-09-20T06:31:38.000Z end: 2013-09-21T06:31:38.000Z daily? true
15:58:57 ORACLE | id: 6 start: 2013-09-19T06:31:36.000Z end: 2013-09-20T06:31:36.000Z daily? true
15:58:58 ORACLE | id: 5 start: 2013-09-18T06:31:34.000Z end: 2013-09-19T06:31:34.000Z daily? true
15:58:58 ORACLE | id: 4 start: 2013-09-17T06:31:32.000Z end: 2013-09-18T06:31:32.000Z daily? true
15:58:59 ORACLE | id: 3 start: 2013-09-16T06:35:30.000Z end: 2013-09-17T06:35:30.000Z daily? true
15:59:00 ORACLE | id: 2 start: 2013-09-15T06:31:28.000Z end: 2013-09-16T06:31:28.000Z daily? true
15:59:02 ORACLE | id: 1 start: 2013-09-14T06:31:26.000Z end: 2013-09-15T06:31:26.000Z daily? true

MGP
2013-09-23, 08:50 AM
It's been a week.
Let's see fresh ESAV statistics, shall we?

maradine
2013-09-23, 12:27 PM
You done been taught to fish. Go get some.

MerlO
2013-09-24, 11:19 AM
Hope this works, ESAV Launchers:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFotUDlzV0RTRjZ0bXlia3JlajZnZ Wc&usp=sharing

maradine
2013-09-24, 11:28 AM
Beautiful :)

MerlO
2013-09-24, 11:40 AM
Beautiful :)
Yeah :)

I took the liberty of adding in the before GU 8/12/13 Launcher sheet for KPU comparison, from the old Oracle thread, thanks for that.

MerlO
2013-09-24, 11:56 AM
Harasser ESAV:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dGxYX2xyNG53TERHRjk3a2xTS21JR 2c&usp=sharing

maradine
2013-09-24, 12:55 PM
Can't do a KPU comparison - KPU tends to increase over time frame. This is one of the reasons the Oracle is now standardized on 24 hour time blocks. The prior set was a 125 hour single pull, and you ran the average of a bunch of 24 hour pulls. Apple to oranges, in this case. What I can do is run a modern 125 hour set if you really want to see the changes.

Plaqueis
2013-09-24, 04:16 PM
Hope this works, ESAV Launchers:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFotUDlzV0RTRjZ0bXlia3JlajZnZ Wc&usp=sharing

Harasser ESAV:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dGxYX2xyNG53TERHRjk3a2xTS21JR 2c&usp=sharing

Lmao.. no surprises there...

MerlO
2013-09-24, 08:51 PM
Can't do a KPU comparison - KPU tends to increase over time frame. This is one of the reasons the Oracle is now standardized on 24 hour time blocks. The prior set was a 125 hour single pull, and you ran the average of a bunch of 24 hour pulls. Apple to oranges, in this case. What I can do is run a modern 125 hour set if you really want to see the changes.

Thanks, it would be great if you could do this for the ESAV launchers to really see what effect the Striker changes had. Are you able to match the modern 125 hour pull to be the same time as the prior data, given the weekend KPU shifts?

maradine
2013-09-24, 08:53 PM
Yep! I keep all the timestamps. I'll run it during kill nadir.

MerlO
2013-09-25, 03:09 AM
Harasser ESAI:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dGN1M3BvY19kQkZNMmxpOENCRnBvR 0E&usp=sharing

MGP
2013-09-25, 03:18 AM
Harasser ESAI:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dGN1M3BvY19kQkZNMmxpOENCRnBvR 0E&usp=sharing

Can you post AI MAXes plz plz?

MerlO
2013-09-25, 12:18 PM
Can you post AI MAXes plz plz?

Started on it, but the #planetside2 channel spam doing queries is pretty crazy, need maradine to arrange a separate channel before going any further. :)

maradine
2013-09-25, 01:16 PM
I'm redoing the interface so you can ask both publicly (for illustrative calls) and privately (for walls'o'text). I'd also like to get the avg queries done today, but that's wishful thinking. Today is a cluster.

mrmrmrj
2013-09-25, 01:29 PM
Harasser ESAI:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dGN1M3BvY19kQkZNMmxpOENCRnBvR 0E&usp=sharing

Marauder pwns.

Plaqueis
2013-09-25, 01:38 PM
Harasser ESAI:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dGN1M3BvY19kQkZNMmxpOENCRnBvR 0E&usp=sharing

Rofl.. pattern continues.

maradine
2013-09-25, 02:24 PM
The Oracle will now respond to all current commands from the PM window so you don't need to do them in the main channel unless you're trying to make a point. Will update OP.

MerlO
2013-09-26, 12:46 PM
The Oracle will now respond to all current commands from the PM window so you don't need to do them in the main channel unless you're trying to make a point. Will update OP.

It's been working well, sorry for the spam :)

Now i've been grabbing MAX data, it's quite a lot and takes a while to tidy up in spreadsheet. Would this be possible from the channel: give the oracle a batch request list - x weapons * y periods, the oracle does some thinking, and then it DCC a space or comma delimited file back?

mrmrmrj
2013-09-26, 12:51 PM
I really wish the data available to the Oracle tracked slug use in shotguns and MAXes.

maradine
2013-09-26, 01:38 PM
It's been working well, sorry for the spam :)

Now i've been grabbing MAX data, it's quite a lot and takes a while to tidy up in spreadsheet. Would this be possible from the channel: give the oracle a batch request list - x weapons * y periods, the oracle does some thinking, and then it DCC a space or comma delimited file back?

DCC's a giant pain in the ass, but Roy fronted me a snippet for doing batch posts to hastebin, which I will be adapting and integrating shortly. Today, even, assuming I get the time.

I really wish the data available to the Oracle tracked slug use in shotguns and MAXes.


There's a metric ton of things SOE can't or won't provide - we work the best we can with what we're given. :)

MerlO
2013-09-27, 03:47 AM
MAX data!

AI MAX:
NC and TR pretty close considering TR overpop, VS ahead.

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDBEQVFWS3hQQnZETWp1SmRZYk1oS Wc&usp=sharing

http://i.imgur.com/05EhgyT.png

AV MAX:
Need kills broken down by vehicle/infantry here.

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDNzZmtJdGlJUEtadVVkWkU4dkRLd 0E&usp=sharing

http://i.imgur.com/aNzT6A1.png

AA MAX:
Not very interesting...

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dGRYcktHYkthSXJ2MS1RcjlzNzF3N Xc&usp=sharing

http://i.imgur.com/MQEvRkX.png

MGP
2013-09-27, 04:22 AM
I like how Onslaughts are least popular but with highest avgBR and highest KPU. As an Onslaught user myself i feel like an MLGPRO :) So many Mercy users... the forumside syndrome. Everyone was yelling "Mercy OP, Mercy OP!". But the truth is - Mercy sux balls.
Yet TR AI MAX being the worst one makes me a little sad.

Edit: Wow, i thought what ZOE was OP, but THIS? Almost twice as good as TR guns!
inb4 random VS guy saying what it's VS PRO SKILLZ.

maradine
2013-09-27, 12:08 PM
Heh, looks like the cat's out of the bag on Nebulas. Ah well, I enjoyed my hipsterdom while it lasted. :D

phungus
2013-09-30, 12:18 AM
Can you use this to check the deaths by a weapon over a time period? I'm interested because I think the lock on changes, after the last hotfix, will show a significant change to stats, reflected by the fact the Striker is no longer so powerful. There were some good arguments (many I made) and data from the Oracle of Death showing the effects of the striker, I'm curious to see how the changes to the striker rockets (no longer clipping through terrain, loses fire and forget ability and instead requires the heavy to maintain lock until rocket hits) change the data from when the Striker was clearly the best at everything launcher.

MerlO
2013-09-30, 12:34 AM
Can you use this to check the deaths by a weapon over a time period? I'm interested because I think the lock on changes, after the last hotfix, will show a significant change to stats, reflected by the fact the Striker is no longer so powerful. There were some good arguments (many I made) and data from the Oracle of Death showing the effects of the striker, I'm curious to see how the changes to the striker rockets (no longer clipping through terrain, loses fire and forget ability and instead requires the heavy to maintain lock until rocket hits) change the data from when the Striker was clearly the best at everything launcher.

maradine was going to grab a comparable data set to compare before/after the lock on change

MerlO
2013-09-30, 12:37 AM
MBT AV Secondaries from 1st week WDS:

Enforcer and Vulcan look close, surprised that Saron is so far ahead..

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFVmLWg3MkN5YWlncmJYSVlJbzh5Z Xc&usp=sharing

http://i.imgur.com/VWwV6wf.png

MGP
2013-09-30, 02:31 AM
maradine was going to grab a comparable data set to compare before/after the lock on change

They have already reverted the change/bug on Striker, it's now once again fire-and-forget, like the rest of lockons.

phungus
2013-09-30, 03:16 AM
MBT AV Secondaries from 1st week WDS:

Enforcer and Vulcan look close, surprised that Saron is so far ahead..

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFVmLWg3MkN5YWlncmJYSVlJbzh5Z Xc&usp=sharing

http://i.imgur.com/VWwV6wf.png

The reason it gets more kills is because it is good at all duties, so it dominates. For NC and TR there are alot more heavy tanks (and Harrassers) that pull Furys and Maurders.

This stat though helps demonstrate something I keep trying to explain to people who complain about the Vulcan. The Vulcan really isn't OP when compared to the Saron, they are both great.

phungus
2013-09-30, 03:23 AM
They have already reverted the change/bug on Striker, it's now once again fire-and-forget, like the rest of lockons.

Nope, you have to maintain active lock. Just tested it again, I only got hits by keeping the lock, otherwise rocket shot up into the sky and no hit marker or +air defense points.

phungus
2013-09-30, 03:24 AM
maradine was going to grab a comparable data set to compare before/after the lock on change

Need to compare after the last hotfix. Right after the lock on change strikers created even more havok because they caused infinite lock on tone - requiring using flares or muting sounds if you didn't have flares/smoke every time they were used against a target!

MGP
2013-09-30, 03:27 AM
Nope, you have to maintain active lock. Just tested it again, I only got hits by keeping the lock, otherwise rocket shot up into the sky and no hit marker or +air defense points.

Yesterday i was using it "like normal".

maradine
2013-09-30, 12:23 PM
I honestly have no idea what to poll at this point. I've heard a dozen different stories about in what periods what was working as intended or not. :D

MGP
2013-09-30, 02:45 PM
I honestly have no idea what to poll at this point. I've heard a dozen different stories about in what periods what was working as intended or not. :D

Do the lasher/JH/MCG.

phungus
2013-09-30, 03:24 PM
I honestly have no idea what to poll at this point. I've heard a dozen different stories about in what periods what was working as intended or not. :D

The hotfix just happened, so checking Striker data wol't help.

I would like to see a comparrison of all MBT and Harrasser secondaries, by faction. I think things are more even then the last table posted. Ie I'm fairly certain the Sauron's status as an outlier will be made up by a significantly reduced kill count from VS on the Fury (and maurader for TR). Again I feel like the Vulcan and Saron are fine, they are just both very good at what they do (as is the Fury and Maurader for that matter, but because of how the Saron works I don't think there are nearly as many VS who use a fury as there are NC MBTs or Harrassers that load them).

maradine
2013-09-30, 03:53 PM
You have been given all the necessary tools to make that happen - go for it!

derito
2013-09-30, 04:31 PM
Can you have stats per server (the busiest ones at least) ? I'm wondering if there are noticeable differences between them. Should they exist, it might mean servers can have some kind of local metagame. It would need more time to have statistically significant results though.

MerlO
2013-09-30, 08:15 PM
The hotfix just happened, so checking Striker data wol't help.

I would like to see a comparrison of all MBT and Harrasser secondaries, by faction.

All harasser secondaries is next on my list, I plan to grab this current weeks data before posting during the weekend.

As for the Striker being unfixed or not - i'm also confused, but i'll grab that at the end of the week as well and check the daily trend.

maradine
2013-09-30, 09:13 PM
There are a few interesting Striker KPU spikes in the historicals, but we'll need a good understanding of what changed when to make heads or tails of linking cause and effect.

MerlO
2013-10-02, 10:04 AM
Do the lasher/JH/MCG.

Here you go! JH and MCG are close on KPU. Hard to compare with the Lasher given it's different functionality.

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHNCallDSUF5aVhBUVctUXQxUnlIS 2c&usp=sharing

Plaqueis
2013-10-02, 10:46 AM
Here you go! JH and MCG are close on KPU. Hard to compare with the Lasher given it's different functionality.

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHNCallDSUF5aVhBUVctUXQxUnlIS 2c&usp=sharing

Well, it's supposed to be a supportweapon i guess, just like the other 2.. it just fails in comparison.

maradine
2013-10-02, 11:32 AM
It fails as a kill producer - no argument there. But, it's also an excellent bridesmaid weapon, producing vast quantities of assists and area denial to boot. Does that make it equal with the others on balance of traits? I don't think we'll ever have the data to make that call. I wouldn't be unhappy with a power boost, but I'm not unhappy with it as it is now.

mrmrmrj
2013-10-03, 01:49 PM
Not every tool should be judged solely on its kill ability. I love concuss grenades, for example.

Plaqueis
2013-10-03, 07:20 PM
nvm.

MerlO
2013-10-05, 08:49 PM
Hi Maradine, would it be possible to provide a query for daily total empire kills? Rather than having to query every weapon and summing... :(

Stomps
2013-10-06, 08:23 AM
Dear Oracle, after reading a thread about how NC teamkill each other and TR have a friendly inviting manner, would it be possible to test this by sampling how many times players of each faction are teamkilled and if possible, run a second time for soldiers teamkilled more than once by the same person (also in the api).

Thank you O oracle!

maradine
2013-10-06, 09:55 PM
Hi Maradine, would it be possible to provide a query for daily total empire kills? Rather than having to query every weapon and summing... :(

Yeah, that shouldn't be a big deal. I'll take a look this week.

Dear Oracle, after reading a thread about how NC teamkill each other and TR have a friendly inviting manner, would it be possible to test this by sampling how many times players of each faction are teamkilled and if possible, run a second time for soldiers teamkilled more than once by the same person (also in the api).

Thank you O oracle!

Hmm.

The first half of that is pretty simple, though it would require a new index to return in any reasonable amount of time. The second half would require some interesting caching work, since you don't get that sort of data outright. Essentially, I'd need to throw every apparent TK into memcache with a attacker->victim key, set a reasonable expiry, and watch for duplicate insertions. Possible? Yeah. October? Probably not. On the list, though. It would definitely be an interesting pull, especially by server.

Stomps
2013-10-07, 03:28 PM
Hmm.

The first half of that is pretty simple, though it would require a new index to return in any reasonable amount of time. The second half would require some interesting caching work, since you don't get that sort of data outright. Essentially, I'd need to throw every apparent TK into memcache with a attacker->victim key, set a reasonable expiry, and watch for duplicate insertions. Possible? Yeah. October? Probably not. On the list, though. It would definitely be an interesting pull, especially by server.

Thanks for considering it, what i noticed was on peoples killfeeds they have the kills/deaths listed by number, would this help?

Example (stats from some random I checked out randomly and happened to still be open on a tab)

https://players.planetside2.com/#!/5428100110516136785/killboard/?view=kills

MerlO
2013-10-11, 03:33 AM
Updated Phoenix/Striker/Lancer stats:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFotUDlzV0RTRjZ0bXlia3JlajZnZ Wc&usp=sharing

Key dates:
pre 9/15 - Striker lockon changed
9/19 - ESRL on sale
9/27 - Warpgate rotation / hotfix changes lock on behavior?

Daily KPU trend:
http://i.imgur.com/B6nFqLi.png

The Phoenix and Striker seem relatively stable KPU wise. It's not apparent to me, from this data, that the 9/27 hotfix has significantly changed the effectiveness of the Striker pre-hotfix. The Lancer KPU looks to be dropping - possibly the lancer not as effective in the Indar SE warpgate? (i've never used the Striker or Lancer)

MerlO
2013-10-11, 11:03 AM
Harasser bonanza! All harasser weapon data for the last few weeks:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dE4yblJ1cUEtdEVNNUk0OVhVekRqZ lE&usp=sharing

So much death by TR:
http://i.imgur.com/4XleaDy.png

NC/VS Harasser Furys don't kill enough to keep pace with the Marauder-H.

Comparison with MBT ES Secondaries:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFVmLWg3MkN5YWlncmJYSVlJbzh5Z Xc&usp=sharing

http://i.imgur.com/mLwZhp9.png

The harasser platform puts the Vulcan-H and Marauder-H way ahead of MBT versions - the Vulcan-H even does ~10% less damage than Vulcan. If only there was kill data broken down by Vehicle/Infantry kills. Is it possible Maradine? :)

maradine
2013-10-11, 12:28 PM
Not exactly, but something similar could be synthesized. Oracle's heartbeat is realtime kill data, but there is a roughly equivalent stream of vehicle destruction data that I can run in parallel. On could be subtracted out of the other to get an approximation of what you're looking for, but it wouldn't be exact.

Consider - a Prowler is destroyed at 1381508422 by a Saron-HRB. A TR trooper is killed at 1381508422 by a Saron on the same continent. Was he in the vehicle? At the same fight? Even remotely involved? For things like the (Higby-plz)Ranger, where there are handful of events a day, things line up cleanly. Popular weapons start making assumptions that skew the numbers. Still, it would be close.

Roadmap-wise, I have two large priorities:

Merge kill data with the weapon_stat live stream to produce up-to-the-minute usage times.
Start doing something with the vehicle destruction stream (ex, what is the most dangerous Scythe killer?)


Both are waiting on changes from the API dev team. While I'm waiting, I'm doing the pastebin work to make people like Merl0's lives easier. I also have a few private requests for data feeds outstanding, so "Powered by the Oracle" may be a thing in the intermediate future.

You know, between the normal pile of feasting, fighting, and fornication over here.

Stay tuned!

Boomzor
2013-10-11, 12:36 PM
(ex, what is the most dangerous Scythe killer?)

Stay tuned!

Clearly it would be static objects, like the ground, trees and the occasional wall.

OP, I tell you ;)

SerethiX
2013-10-11, 05:59 PM
Clearly it would be static objects, like the ground, trees and the occasional wall.

OP, I tell you ;)

So fucking true!

SerethiX - www.serethi.de

maradine
2013-10-11, 06:04 PM
With enough time, I could extract vehicular suicide, but it wouldn't tell us anything we don't already know - nerf fucking trees.

Mordelicius
2013-10-12, 03:18 AM
Harasser bonanza! All harasser weapon data for the last few weeks:

https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dE4yblJ1cUEtdEVNNUk0OVhVekRqZ lE&usp=sharing

So much death by TR:
http://i.imgur.com/4XleaDy.png

NC/VS Harasser Furys don't kill enough to keep pace with the Marauder-H.

Comparison with MBT ES Secondaries:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFVmLWg3MkN5YWlncmJYSVlJbzh5Z Xc&usp=sharing

http://i.imgur.com/mLwZhp9.png

The harasser platform puts the Vulcan-H and Marauder-H way ahead of MBT versions - the Vulcan-H even does ~10% less damage than Vulcan. If only there was kill data broken down by Vehicle/Infantry kills. Is it possible Maradine? :)
TY lmao. We've been saying the Marauder and Vulcan for Harassers are the main cause of the current faction imbalance (read our posts). No graphs are really needed to initially figure this out, just for confirmation.

If Devs actually check their live servers instead of just testing on the VR Room, they would have noticed this by now and fixed it even before the WDS Preseason. But no, they have no inkling. Offering solutions to everything except the actual problem (faction imbalance).

All it takes is several hits from a Harasser Marauder and you'd know it's way OP, taking down clumps of infantry from so far away like they are chucking ranged C4s.

As for the Harasser Vulcan, it's very effective against EVERYTHING. They kill infantry with ease. They melt armor and turrets like butter. They even beat ESFs with impunity. I've seen a Reaver tried chasing this thing. They just do a slight donut til the Reaver over shoot, and fire back. Next thing you know, the Reaver is smoking and being chased, forced to land and still get caught when it does.

A broken, OP Harasser + broken OP weapons is a horrible combo.

Vulcans shouldn't be that effective against armor. And this thing should be very weak at long range. Harassers are too fast and too tough to kill to be wielding this.

Marauders shouldn't be that strong against infantry. This things are like long ranged, spammable C4 projectiles. The splash damage and radius should be severely nerfed. Again, the damage dropoff should be very pronounced for this weapon as well.

One can't give a vehicles strong anti-infantry capabilites. That's like giving infantry a gun that just destroys tanks. As it is, the TR has counters for everything. Their infantry has strong counter against vehicles in the form of strikers. Their vehicles are too effective against infantry. Hence their infantry can fight infantry and vehicles, while their vehicles can take on vehicles and infantry.

Whereas a ZOE, OP as it, is only strong against infantry.

maradine
2013-10-12, 05:38 PM
Go make a thread about it - we try to keep it about the numbers and the acquisition of numbers in here; not so much the implications.

maradine
2013-10-16, 01:55 PM
Good news! The badasses over that SOE API team have added the necessary bits and pieces to make realtime KPH possible. Getting on it.

MerlO
2013-10-17, 12:04 AM
Sounds good! Lets see how well it compares with KPU..

Omnimon
2013-10-17, 12:48 AM
Massive Post : Got almost all the weapons under the sun save for

Grenades ( of all types)
Max Punch
Flash / Galaxy / Sunderer / Lib tail and nose weapons

This has all the other weapons ( in general) :

https://docs.google.com/spreadsheet/ccc?key=0AgxmYUzGgCZ1dGV6XzBGTE02WGlLaHJzRUo5UzVXc mc&usp=drive_web#gid=29

maradine
2013-10-17, 01:57 AM
Nice work, brother. Next time will be much easier. <3

maradine
2013-10-17, 01:36 PM
Sooner than you think!

OK spreadsheet nerds, I'm pleased to announce that the Oracle now supports data dumps via pastebin for both public and private requests.

Rather than:
!oracle ask weapon 34002

. . . you can instead:

!oracle dump weapon 34002

. . . which will produce an output something like this:
http://pastebin.com/qx2arUei

. . . which contains a text wall like this:
35002 Lancer VS22 periods 1-33
period, kills, uniques, kpu, avgbr, q1kpu, q2kpu, q3kpu, q4kpu
1,695,318,2.2,58.4,1.9,1.9,2.4,2.3
2,1623,667,2.4,55.4,2.3,2.6,2.4,2.3
3,1398,579,2.4,57.2,2.2,2.3,2.3,2.8
4,1229,521,2.4,58.5,2.1,2.3,2.1,2.8
5,1157,505,2.3,56.4,2.0,2.1,2.4,2.6
6,1759,719,2.4,53.1,2.6,2.3,2.5,2.6
7,1685,681,2.5,53.6,2.4,2.4,2.4,2.7
8,2297,819,2.8,56.7,2.2,2.4,3.1,3.3
9,1900,711,2.7,57.7,2.1,2.3,2.9,3.4
10,1285,546,2.4,57.5,2.3,2.1,2.3,2.8
11,1238,545,2.3,56.5,2.1,2.1,2.3,2.4
12,1185,537,2.2,53.4,2.4,2.2,2.2,2.1
13,1270,576,2.2,57.0,2.2,2.0,2.1,2.6
14,1477,614,2.4,58.5,2.5,2.1,2.4,2.8
15,1685,671,2.5,54.9,2.7,2.4,2.5,2.6
16,1276,596,2.1,54.2,2.1,2.1,2.0,2.5
17,940,434,2.2,53.3,2.3,2.2,2.1,2.1
18,969,444,2.2,57.5,2.2,2.0,2.1,2.6
19,933,424,2.2,54.5,1.8,2.3,2.2,2.2
20,1089,467,2.3,56.1,2.4,2.1,2.5,2.4
21,1372,586,2.3,55.8,2.4,2.3,2.3,2.6
22,1391,611,2.3,59.0,1.6,2.1,2.3,2.8
23,1144,522,2.2,57.6,1.8,2.0,2.2,2.7
24,891,405,2.2,57.8,1.8,2.2,2.0,2.6
25,769,396,1.9,53.7,2.0,1.8,2.0,2.1
26,925,412,2.2,57.9,1.6,2.1,2.3,2.7
27,1003,460,2.2,59.0,1.7,2.3,2.0,2.6
28,1101,485,2.3,56.8,2.3,2.2,2.2,2.5
29,1214,551,2.2,58.1,1.6,2.0,2.3,2.6
30,1345,552,2.4,58.0,2.0,2.4,2.5,2.6
31,847,408,2.1,56.1,1.8,2.1,2.0,2.3
32,866,426,2.0,56.6,1.5,2.2,2.2,1.9
33,722,352,2.1,57.8,2.3,2.0,2.0,2.1

Then you can do that thing you do.

Next - playtime!

MerlO
2013-10-18, 01:51 AM
Excellent! :)

CanadianWolvie
2013-10-24, 01:09 PM
From the now apparently defunct thread:

Short answer: no - I'm not tracking that. This should be reasonably extractable with simple API calls to the existing SOE data source, tho.

May I please be taught how to do that? Is there a tutorial or something?

MerlO
2013-10-24, 09:51 PM
Up to date stats of the files i've posted earlier in this thread.

ESAV Launchers:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFotUDlzV0RTRjZ0bXlia3JlajZnZ Wc&usp=sharing

AI MAX:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDBEQVFWS3hQQnZETWp1SmRZYk1oS Wc&usp=sharing

AV MAX:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDNzZmtJdGlJUEtadVVkWkU4dkRLd 0E&usp=sharing

AA MAX:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dGRYcktHYkthSXJ2MS1RcjlzNzF3N Xc&usp=sharing

ES Heavy Assault Guns:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHNCallDSUF5aVhBUVctUXQxUnlIS 2c&usp=sharing

All Harasser:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dE4yblJ1cUEtdEVNNUk0OVhVekRqZ lE&usp=sharing

MBT Primary:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHE4aC0tZTFrdmVFMjF0Rkhrb0VIR FE&usp=sharing

MBT Secondary:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFVmLWg3MkN5YWlncmJYSVlJbzh5Z Xc&usp=sharing

ChipMHazard
2013-10-25, 08:03 AM
Many thanks for the up to date data:D

ChipMHazard
2013-10-28, 07:31 AM
I've been asked if it's possible to get data on vehicle kills for MBT weapons and AV weapons?

Plaqueis
2013-10-28, 10:34 AM
Merlo, could you please pull the Heavy Assault (LMG's atleast, maybe shottys too) stats whenever you have time? Thanks in advance.

maradine
2013-10-28, 02:04 PM
I've been asked if it's possible to get data on vehicle kills for MBT weapons and AV weapons?

Not yet. Requires a new collector for the vehicle destruction stream, which I'll implement right after the playtime stream. Super high on the priority list.

MerlO
2013-10-28, 11:38 PM
Merlo, could you please pull the Heavy Assault (LMG's atleast, maybe shottys too) stats whenever you have time? Thanks in advance.

If you need it now, 3 weeks worth is here:
https://docs.google.com/spreadsheet/ccc?key=0AgxmYUzGgCZ1dElKYXcyR0NTMHYtMlBuNDJzSm11d lE&usp=drive_web#gid=28

I believe this sheet is a copy of Omnimon's data in post #74. #74 Omnimon link no longer works for me.

Later this week I might look at grabbing all periods for LMG so far up to date, if thats what you're after? To be honest i'm not very interested in the general infantry stats, I think it is reasonably balanced - no elephants in the room so to speak. That being said if the devs make changes, then i'll grab it.

maradine
2013-10-28, 11:46 PM
I think he was looking for a merged LMG / HA Special set, which I don't think we've done since the last nuke and pave. I could be wrong though, I'm starting to lose track myself! :)

Plaqueis
2013-10-29, 05:27 AM
If you need it now, 3 weeks worth is here:
https://docs.google.com/spreadsheet/ccc?key=0AgxmYUzGgCZ1dElKYXcyR0NTMHYtMlBuNDJzSm11d lE&usp=drive_web#gid=28

I believe this sheet is a copy of Omnimon's data in post #74. #74 Omnimon link no longer works for me.

Later this week I might look at grabbing all periods for LMG so far up to date, if thats what you're after? To be honest i'm not very interested in the general infantry stats, I think it is reasonably balanced - no elephants in the room so to speak. That being said if the devs make changes, then i'll grab it.

I ment the empire specific HA weapons, like Orion, T9's, Gauss and the rest. Im just interested to see if some claims in the MCG thread are true.

I tried getting them myself and failed... :P

maradine
2013-10-29, 11:49 AM
I'll gin one up today over lunch if MerlO can't get to it - been a while since I've consumed my own product, as it were.

SerethiX
2013-10-29, 11:50 AM
I'll gin one up today over lunch if MerlO can't get to it - been a while since I've consumed my own product, as it were.

Lmao xD

Gesendet von meinem HTC One X mit Tapatalk

Plaqueis
2013-10-29, 12:43 PM
I'll gin one up today over lunch if MerlO can't get to it - been a while since I've consumed my own product, as it were.

Cool. Thanks in advance.

maradine
2013-10-29, 01:50 PM
Here's a big pile. No time for pretty pictures today, sadly.

https://docs.google.com/spreadsheet/ccc?key=0Arp9yneyL3DndDlfQjRWc1ljM3BjOHJrZjkxZkdlN UE&usp=sharing

qazz
2013-10-29, 02:25 PM
Here's a big pile. No time for pretty pictures today, sadly.

https://docs.google.com/spreadsheet/ccc?key=0Arp9yneyL3DndDlfQjRWc1ljM3BjOHJrZjkxZkdlN UE&usp=sharing

awesome work on the oracle so far! do you have any idea when vehicle kills might be included in the data dumps?

also, it seems the sva is only one of the best lmgs because it is really easy to use. there's a lot less variation among the quartiles than for other weapons, resulting in a high average, but a normal-high q4kpu.

why the hell is the mswr and gs22 performance so low when compared with the orion though? mhmm.

maradine
2013-10-29, 03:36 PM
As per comment to Chip, it's languishing behind the playtime feature, which I'm maybe 50% through. November is a good estimate, but I promise nothing so as to not disappoint.

ChipMHazard
2013-10-29, 03:38 PM
As per comment to Chip, it's languishing behind the playtime feature, which I'm maybe 50% through. November is a good estimate, but I promise nothing so as to not disappoint.

Fair enough, I'm sure the person who asked me is happy to wait:p

Stomps
2013-10-30, 05:53 PM
Here's a big pile. No time for pretty pictures today, sadly.

https://docs.google.com/spreadsheet/ccc?key=0Arp9yneyL3DndDlfQjRWc1ljM3BjOHJrZjkxZkdlN UE&usp=sharing

fyi minor formatting problem, the top line is glitched on 4 charts (flare,sva88,pulsar,polaris) making it so it doesnt show the averages on the first line

what i wanted to ask was whats going on with the NS15MP? why are the kpus swinging so wildly?

MerlO
2013-10-30, 10:42 PM
what i wanted to ask was whats going on with the NS15MP? why are the kpus swinging so wildly?

Few uniques - i.e one guy has a long play session, only with the NS15MP, and gets a lot of kills. You can tell what BR that player was - for example period 19 has a Q3KPU of 208!

maradine
2013-10-31, 02:34 PM
Few uniques - i.e one guy has a long play session, only with the NS15MP, and gets a lot of kills. You can tell what BR that player was - for example period 19 has a Q3KPU of 208!

Exactly. For weapons with few or no consistent users, the numbers swing wildly and can be largely ignored.

fyi minor formatting problem, the top line is glitched on 4 charts (flare,sva88,pulsar,polaris) making it so it doesnt show the averages on the first line

Sorry about that; fixed.

maradine
2013-11-01, 12:27 AM
The good news:

Real-time playtime tracking is live! I'm about to push the commits to github, for those of you following along at home. Kills Per Hour is about to be a real thing!

The bad news:

Doing this efficiently inside a reasonable hardware budget means a ton of caching, and based on PS2's player velocity, it looks like it will take a few days for the cache to warm up enough to produce accurate data. So, don't expect the oracle to spit it out quite yet. A few interesting bits of trivia from the live stream, though:

The three longest individual uninterrupted usage updates today:

The Force-Blade: equipped in a 6,418 second stretch by Jboych.
The Chainblade: equipped in a 6,314 second stretch by Artaxin.
The T7 Mini-Chaingun: also equipped in a 6,314 second stretch by Artaxin.

You gotta go a bit further down the list to find a Mag-cutter. :)

Mana AV turrets id501 and id502 produce negative playtime numbers. Yeah, I got nothin'.

So, good times! Stay tuned!

ChipMHazard
2013-11-01, 12:32 PM
Sad to say that people have started abusing the oracle over at the official forum.

maradine
2013-11-01, 12:38 PM
That wasn't unexpected. PSU and Reddit have been very mature about using it as argument fodder, which I appreciate.

Once I get the vehicle destruction pulls running, based on what I'm reading in the prelims, I expect a whole new round of utter hysteria . . . unless those rumored incoming vehicle weapon changes do what I think they're going to do.

MerlO
2013-11-01, 10:33 PM
The three longest individual uninterrupted usage updates today:

The Force-Blade: equipped in a 6,418 second stretch by Jboych.
The Chainblade: equipped in a 6,314 second stretch by Artaxin.
The T7 Mini-Chaingun: also equipped in a 6,314 second stretch by Artaxin.



Warpgate afk-ing?

maradine
2013-11-02, 12:36 AM
More likely the save event covers everything that was in a loadout during playtime. This explains a lot about the early PSU weapon stats - and also why everyone seems to be hucking around medkits. :)

Filtering all that out, you still get some interesting bits. Now that the cache has been cooking for a while, I can see all sorts of neat stuff, like metAXAmike running around with a Lynx for 25.74 consecutive hours. Ironman.

AlexMNC
2013-11-26, 11:08 AM
Hey Maradine, I love what you did making the Oracle of Death, but I'm having trouble. I went to the IRC chat and it doesn't seem that the Oracle is available. I've made queries to it before, and it worked, so it's not that I'm doing something wrong. I very much liked being able to just have that information at my fingertips.

Phaden
2013-11-26, 11:26 AM
He turned it off a while back, it was eating up resources and personal time. There is a thread somewhere here about it. http://www.planetside-universe.com/showthread.php?t=56993

AlexMNC
2013-11-26, 12:23 PM
Oh, that's disappointing :(

I wanted to see how the increase in frame rate with O:MFG affected the NC weapon's performance since low frame rate is more punishing to low rate of fire weapons.

maradine
2013-11-27, 10:25 PM
Probably pretty substantially, I'd wager. Alas, it shall remain a mystery. :)

MerlO
2013-12-28, 10:54 AM
Updated some of the files to compare pre / post PU02. Only a weeks worth of PU02 data compared with 6weeks from Sept-Nov.

Harasser:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dE4yblJ1cUEtdEVNNUk0OVhVekRqZ lE&usp=sharing
http://i.imgur.com/0id479v.png
Drops in KPU especially TR after the comprehensive nerfs to the harasser. Interestingly VS PPA is the same/better.


MAX AV:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDNzZmtJdGlJUEtadVVkWkU4dkRLd 0E&usp=sharing
http://i.imgur.com/94M654t.png
Pounders! :)

MAX AI:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDBEQVFWS3hQQnZETWp1SmRZYk1oS Wc&usp=sharing
Smaller changes here - KPU small increase for NC/TR, maybe due to nanoweave changes. VS KPU drops with ZOE changes.


Sniper Rifles:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dEN6dkViNDI3OERrRkl0V1A4dTU2c kE&usp=sharing
I'm not an infil player, but it seems that bolt action rifles do better now after all the changes...

DviddLeff
2013-12-29, 05:25 PM
Awesome, nice work MerlO.

Any chance you could check out the MBT and ESF data as they got some specific balancing in the patch?

Although hell, what didn't get changed!

MerlO
2013-12-29, 08:41 PM
I'll try and find some time to add more weapons over the next week :)

DviddLeff
2013-12-30, 05:41 AM
Cheers MerlO!

ChipMHazard
2013-12-30, 06:52 AM
Aye, cheers MerlO!:D

MGP
2014-01-01, 03:40 PM
Updated some of the files to compare pre / post PU02.

Fire still hot. Nights still dark. TR MAXes still suck... nothing new.

DviddLeff
2014-01-02, 05:11 AM
Except for the Pounder, which is better (KPU wise) than every other MAX in the game...

Falcon_br
2014-01-03, 12:01 AM
As I imagined, all TR harasser weapons have been nerfed into oblivion.

The more strange part is that a PPA-H harasser kill enemy harassers maybe in one clip, it is just a lot fast!

While the Vulcan and the Marauder has little effect on enemy harassers. It is better to use a Kobalt or the Basilisk.

I hate notice that sunderer convoys with repair and Basilisks are a normal sight on the field right now, I also see several Harassers from the TR using Basilisks! He Basilisks is now the best anti air, anti vehicle and anti infantry weapon on the game from almost all ranges (fury with the buff is still better at close range), it is even better then the Vulcan on the harasser right now!

But I think that the harasser nerf was so hard that 2 infantry with annihilators can one shot kill enemy harassers, almost all lock on lanchers can do that, and my harasser have a maximum level armor!

Forgot to mention the fractures, they need now about 18 shots to kill infantry, a lot more with infantry using flak armor!
Pounders are now better then fractures in all scenarios, even on long range engagements.
Now the TR is the only faction that canĀ“t kill enemy infantry with Max anti tank weapons. Comet, Falcon and Raven can still do it.

MerlO
2014-01-03, 09:00 PM
As I imagined, all TR harasser weapons have been nerfed into oblivion.

The more strange part is that a PPA-H harasser kill enemy harassers maybe in one clip, it is just a lot fast!

Well the harasser does look more balanced now between factions, although i'll give you that the PPA-H may be stronger.

Be careful not to read too much into the data :)

Is the Reaper really the best Assault Rifle now? Highly likely a lot of high BR NC medics have gone back to it because of the buffs and are now playing it a lot, also to Auraxium it. In a few weeks the numbers should stabilize.

Assault Rifles:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDBXakR2ZnUtZG5UU0Eybnc1cmowS Gc&usp=sharing

LMG:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHJZNEZPVDB1Ykl3WmVTY3BxaVE4Y lE&usp=sharing

Carbines:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHp5X3VERW5WT2Y0OC1ld09zV184Y VE&usp=sharing

ESF:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dF9NcGJNVzl5WndWNlNHNkF3RVdac mc&usp=sharing

MBT Primary:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHE4aC0tZTFrdmVFMjF0Rkhrb0VIR FE&usp=sharing

MBT Secondary:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFVmLWg3MkN5YWlncmJYSVlJbzh5Z Xc&usp=sharing

This will probably be it for now, on vacation and haven't been collecting data this week.

ChipMHazard
2014-01-04, 05:47 AM
Be careful not to read too much into the data :)

This will probably be it for now, on vacation and haven't been collecting data this week.

Aye, never take statistics at face value.

The Oracle, she be a demanding mistress:p

Plaqueis
2014-01-04, 02:58 PM
The Oracle, she be a demanding mistress:p

http://fc09.deviantart.net/fs20/f/2007/232/f/7/Jack_Sparrow_by_gavwoodhouse.jpg

That's pirateslang right there.. :D

Ketadine
2014-01-10, 05:53 AM
As I imagined, all TR harasser weapons have been nerfed into oblivion.

[...]

What bugs me the most is the huge range nerf of the Vulcan , 95%. Sure, 200 m was a lot, but 10m for max range damage and a shotgun spread pattern if you hold the trigger ?! It's absurd!

SonofSkz
2014-01-13, 10:32 AM
Did anyone ever figure out if it is possible to get roadkill stats? I'd just like to see it so maybe we can get a weed whacker or something added to the magrider.

Falcon_br
2014-01-14, 01:14 PM
What bugs me the most is the huge range nerf of the Vulcan , 95%. Sure, 200 m was a lot, but 10m for max range damage and a shotgun spread pattern if you hold the trigger ?! It's absurd!

I no longer use TR vehicle empire weapons.

The marauder is now the worst AI gun, unable to kill harassers, like all the other AI guns can do. Even the Fury is better in all stats.

The vulcan is good only on melee range, if you can kill an infantry with it, he can toss a c4 in you. With the range nerd all other AV weapons has a better DPS above 30m then the vulcan, getting close to make your vulcan useful will make you dead.

Fractures no longer kills infantry, while all other AV max weapons does. It also no longer kill vehicles.

We lost the marauder, the vulcan and the fracture, the only thing we got is that now the pounder has less gravity and the same damage (what is a buff), all tr players that I know are using the pounder, halberts and basiliks. While all the NC and VS are using Enforcer and Saron.

maradine
2014-01-16, 01:23 AM
Imagine being NC. They were out in that cold for months.

Chewy
2014-01-16, 02:46 AM
Imagine being NC. They were out in that cold for months.

The Canister still isn't good next to either the Marauder or PPA in my opinion. At least those 2 get splash damage to make up for misses and high speed misshaps. A shotgun is still a shotgun that has to deal with 2 kinds of spread and needs direct hits to do damage.

I also don't use the ES vehicle weapons that much myself anyway. Only on a Vangaurd do I have an Enforcer and I don't like to pull that vehicle. A Halberd is great on anything and the bassy kicks ass, more so on a Sunderer team. The Enforcer is a good weapon, I just hate aiming with that dot you get and it sounds kinda puny to me.

I think that making an AV weapon not work all that well on infantry is a good thing. I like that the Fracture isn't used for AI anymore and that they can still do real damage to vehicles. One TR MAX with a bit of range will ruin your day as a vehicle driver, but get close and that MAX shouldn't be a hard kill. Fractures are a ranged weapon, not a in the fray weapon. I can't easily take my Ravens into a fray, why should the other factions? Up close is suppose to be for Comets, Falcons, and Pounders and they all do that very well.

Speaking of Pounders. They are just deadly. Can't say more than that, I just don't see any reason to bitch about the Pounders anymore with how good they are now.

All weapons are suppose to be a side-grade. Gain something for loosing another part. If a one weapon is better for you, then great! If it isn't better for you, then try another. There shouldn't be a single weapon in the game that is highly used other than stock gear. If a weapon is overused, then it isn't in its niche yet.


Then again. All of this is my opinion. I have nothing to back any of this up but my gut feeling.

MerlO
2014-01-18, 07:34 AM
Did anyone ever figure out if it is possible to get roadkill stats? I'd just like to see it so maybe we can get a weed whacker or something added to the magrider.

Should be possible, i'll try and add it in.

Old data from the week after PU02 nuked, available here if anyone wants it.
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dG02enplNXJlMjZSUVhGZDBYdUtwd 0E&usp=sharing

Period 2 in this file was the day before PU02.


Now started collecting again after the latest patch :)

MerlO
2014-01-20, 09:28 AM
I've been able to add the API vehicle destruction stream into maradines stat collection system ( the Oracle of Death :) ).

Info and Limitations:

Vehicle suicide - includes world kills - i.e vehicle destroyed in restricted areas such as map edges. Decided not to include self destroying your own vehicle with a weapon under suicide - it's collected under the weapon.
Vehicle collision - i.e where there is no attacker weapon, and attacker id isn't the same as vehicle character id. This includes friendly vehicle collisions as i'm not sure how to get the faction of the attacker id.
Newly created players not saved in system are not included as the attacker weapon id is not available. Vehicle deaths from tutorial,VR training not included.
NS weapons are collected together (C4,Decimator etc..), the problem is not knowing attacker faction id as above.
No attempt is made at normalizing the data for population - lately NC seems to be over popped somewhat.

MerlO
2014-01-20, 09:32 AM
ESF Deaths for the last 2-3 days:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dERMcHJlWkJXYy13WVROc2s2TTNjO Wc&usp=sharing

I've ranked the top 15 weapons responsible for deaths on the summary sheet. Perhaps some further refinement is needed such as combining burster max arms etc...

ESF Suicide % is close across all factions although scythe collision % is a little different. Is the scythe physically larger than the other 2 ESFs and so more likely to have a collision? Or maybe it's due to Indar warpgate position of the canyons somehow?

I'll post other vehicles as I get around to it...

ChipMHazard
2014-01-20, 11:05 AM
People have been using the Scythe to ram other ESFs for some time now, because of the how the landing gear and ramming works. That might account for the sligthly higher collision rate, for when the ramming fails.

snafus
2014-01-20, 07:03 PM
ESF Deaths for the last 2-3 days:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dERMcHJlWkJXYy13WVROc2s2TTNjO Wc&usp=sharing

I've ranked the top 15 weapons responsible for deaths on the summary sheet. Perhaps some further refinement is needed such as combining burster max arms etc...

ESF Suicide % is close across all factions although scythe collision % is a little different. Is the scythe physically larger than the other 2 ESFs and so more likely to have a collision? Or maybe it's due to Indar warpgate position of the canyons somehow?

I'll post other vehicles as I get around to it...
Very interesting but these numbers on suicides and collisions don't surprise me as aircraft always seem to die to terrain more often then enemy fire.

diLLa
2014-01-20, 07:53 PM
At least the numbers show how useless the striker is against air nowadays...

SerethiX
2014-01-21, 01:10 AM
Nerf trees!

Gesendet von meinem HTC One X mit Tapatalk

MerlO
2014-01-21, 03:22 AM
Nerf trees!

And canyons and desert :)

Plaqueis
2014-01-21, 09:52 AM
At least the numbers show how useless the striker is against air nowadays...

Yet it still outperforms its counterparts (Phoenix/Lancer) by a wide margin..

ChipMHazard
2014-01-21, 10:06 AM
Yet it still outperforms its counterparts (Phoenix/Lancer) by a wide margin..

Against air? Naturally.

diLLa
2014-01-21, 12:34 PM
Yet it still outperforms its counterparts (Phoenix/Lancer) by a wide margin..

The Striker was specifically made to lock onto vehicles. The lancer and phoenix can be used against infantry, with maxes included. Naturally it outperforms the other ES launchers against air. It's performance however is awful..

Plaqueis
2014-01-21, 10:02 PM
The Striker was specifically made to lock onto vehicles. The lancer and phoenix can be used against infantry, with maxes included. Naturally it outperforms the other ES launchers against air. It's performance however is awful..

Well, can't say nothing about the Phoenix as i haven't used it (yet). But i do have both Lancer and Striker (bought the Striker after the nerf btw), and having some gametime with both, there's no doubt in my mind which one is better, no matter the target (ye you can target infantry with Lancer, but it's pointless as you will lose everytime if they shoot back at you).

I didn't try the pre-nerf Striker at all as it was so ridiculously OP, but it's still far better than Lancer even after the nerf. That said, i wouldn't mind a small range increase for it.

EDIT: There's a way to make Lancer effective, it's just not very practical; find a place where you can shoot from range (high ground on open area where targets have nothing to use as cover (around Dahaka for example). Find 10+ guys that can hit moving targets at visual range. Then lead them all, as in tell what target to shoot and when. When 10 guys hit that same target at same time, it will die no matter what it is. We used to do this every now and then when i was still in Trident, those were the only times i remember where Lancer was worth spawning.

ChipMHazard
2014-01-21, 11:54 PM
That overall effectiveness of the ESPLs seems largely irrelevant.
The stats in question shows ESF deaths and what destroyed them.
Naturally the Striker will outperform both the Phoenix and Lancer because it's a lock-on. Neither the Lancer or Phoenix were designed to take out targets like the ESF, even though it can be done.
If you want to argue that changes are needed overall then those are the wrong stats to go by.

Plaqueis
2014-01-22, 07:46 AM
That overall effectiveness of the ESPLs seems largely irrelevant.
The stats in question shows ESF deaths and what destroyed them.
Naturally the Striker will outperform both the Phoenix and Lancer because it's a lock-on. Neither the Lancer or Phoenix were designed to take out targets like the ESF, even though it can be done.
If you want to argue that changes are needed overall then those are the wrong stats to go by.

Oh i know this, i'm not the one who's whining about the Striker. My point was that it's still quite usable, and more in line with others after the nerf.

MerlO
2014-01-22, 09:37 AM
Sunderer Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDRtblZRYWdWbzFMTkZ0RUhZVUdFS 2c&usp=sharing

No surprises there...

maradine
2014-01-22, 04:03 PM
Nice to see the codebase getting some use. =D

edit: stunned to see the ranger getting some use.

Emperor Newt
2014-01-22, 05:47 PM
The Lancer numbers suprise me. I would have expected it to get way more last hits due to the weapons design.

MerlO
2014-01-24, 05:08 AM
That overall effectiveness of the ESPLs seems largely irrelevant.
The stats in question shows ESF deaths and what destroyed them.
Naturally the Striker will outperform both the Phoenix and Lancer because it's a lock-on. Neither the Lancer or Phoenix were designed to take out targets like the ESF, even though it can be done.
If you want to argue that changes are needed overall then those are the wrong stats to go by.

Fortunately collecting vehicle deaths enables an approximate comparison of weapon performance against vehicles, compared to standard KPU from kills.

ES Rocket Launchers, last 7 days:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFotUDlzV0RTRjZ0bXlia3JlajZnZ Wc&usp=sharing

v_deaths (vKPU) = ground vehicle deaths, a_deaths (aKPU) = aircraft deaths

Note that total kills will not sum to v_deaths + a_deaths, basically a vehicle can be destroyed with no one in it :).

vKPU should be understood as deaths per unique. vKPU is calculated from unique killers - not exactly perfect but does give some basis for comparison across weapons.

About the 3 ESRLs, they seem reasonably balanced to me, no huge differences in vKPU, naturally the striker does better against aircraft. Pretty much what Chip said :)

edit: doesn't include AE versions, in future I should try and get both versions under the one weapon.

ChipMHazard
2014-01-24, 07:12 AM
Aye, those stats match up well with my own experience. A bit surprised that the Striker is still soo popular, but I guess people do like lock-ons.

MerlO
2014-01-25, 09:10 PM
Flash Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFpaNFhQcFF0UG5NOWlweDh5MEZDR VE&usp=sharing

Lightning Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHFIWUhXNWtXcjdjUWN3bDNidFZaQ 2c&usp=sharing

MBT Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHZnOVEyc3dLX2dCUTcwRVZWMDBsW Wc&usp=sharing

ChipMHazard
2014-01-25, 09:37 PM
C-4 sure is a tank killer:p

Falcon_br
2014-01-26, 01:53 AM
Flash Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFpaNFhQcFF0UG5NOWlweDh5MEZDR VE&usp=sharing

Lightning Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHFIWUhXNWtXcjdjUWN3bDNidFZaQ 2c&usp=sharing

MBT Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHZnOVEyc3dLX2dCUTcwRVZWMDBsW Wc&usp=sharing

It is clear that the saron hbr is way better AV weapon then the other faction weapons.
The enforcer ML 85 can still be found on that lists.
But the Vulcan... It is underperforming even the m20 basilisk!

Snoopy
2014-01-26, 03:04 AM
but vehicle suicide > all so, nerf vehicle suicide!

ChipMHazard
2014-01-26, 03:55 AM
but vehicle suicide > all so, nerf vehicle suicide!

Heh aye. It's no suprise to see so many Flashes getting destroyed that way. Flipping around left and right. Hopefully a lot of these vehicle suicides will be reduced when they rework the physics to make vehicles more stable.

Emperor Newt
2014-01-26, 07:55 AM
Also C4 and suicide could relate to all the Lightnings getting stuck somewhere. If I would get 1SC every time that happens to me...

And I would not read all too much into those stats. Especially concerning the Saron. Due to its long range and fast rate of fire it's simply more like to get the last hit and therefore the kill. That doesn't necessarily mean that it's more effective. Just that it's able to get more last hits.

Would it be possible to gather the Coyote stats since the buff? The DasAnfall stats are... "interesting". Especially compared to all the other ESF secondaries. Looks like it's the new farming tool of choice (Again on an ESF. Who would have thought...)

MerlO
2014-01-26, 08:01 AM
It is clear that the saron hbr is way better AV weapon then the other faction weapons.
The enforcer ML 85 can still be found on that lists.
But the Vulcan... It is underperforming even the m20 basilisk!

Yeah I was looking for the Vulcan too...it's not good...KPU / vKPU half the other AV secondaries, aircraft KPU about the same - not even making up for being poor against ground vehicles.

MerlO
2014-01-26, 08:40 AM
Also C4 and suicide could relate to all the Lightnings getting stuck somewhere. If I would get 1SC every time that happens to me...

And I would not read all too much into those stats. Especially concerning the Saron. Due to its long range and fast rate of fire it's simply more like to get the last hit and therefore the kill. That doesn't necessarily mean that it's more effective. Just that it's able to get more last hits.

Would it be possible to gather the Coyote stats since the buff? The DasAnfall stats are... "interesting". Especially compared to all the other ESF secondaries. Looks like it's the new farming tool of choice (Again on an ESF. Who would have thought...)

Regarding the C4 suicides, I made sure that self vehicle kills with weapons aren't included in the suicide deaths - wanted to be able to tell how OP trees are :) . Unfortunately C4 is NS, so the C4 deaths includes: suicides, friendly C4 TKs, enemy C4. Future improvement is to have the faction ID of the attacker so I can split friendly from enemy C4.

I updated the ESF deaths sheet earlier, taking a look at coyote missiles, roughly looks like double the number of ESF deaths for 2 days i've got since the update. If i've got time Monday i'll post some ESF weapon kill stats.

MerlO
2014-01-27, 06:29 AM
Adding the last 3 - Liberator, Galaxy, Harasser. Poor galaxy, dies by suicide more than half the time :(

Liberator Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dF90cTg1LXRvdnVMcExGNjdUWFBNU lE&usp=sharing

Galaxy Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dG5Va1Z5WkJMNFdrbXF6ZFhfTDMya kE&usp=sharing

Harasser Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dElBR2Nnekl3SjdzcUZpWmRHSy1IV nc&usp=sharing


Reposting, to make them all easy to find:

Flash Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFpaNFhQcFF0UG5NOWlweDh5MEZDR VE&usp=sharing

Sunderer Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDRtblZRYWdWbzFMTkZ0RUhZVUdFS 2c&usp=sharing

Lightning Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHFIWUhXNWtXcjdjUWN3bDNidFZaQ 2c&usp=sharing

MBT Deaths
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHZnOVEyc3dLX2dCUTcwRVZWMDBsW Wc&usp=sharing

ESF Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dERMcHJlWkJXYy13WVROc2s2TTNjO Wc&usp=sharing

DirtyBird
2014-01-29, 06:39 PM
They dont call it Flash for nothing, thats about how long it lasts.

:thumbsup: for data.

Falcon_br
2014-02-04, 02:54 PM
Adding the last 3 - Liberator, Galaxy, Harasser. Poor galaxy, dies by suicide more than half the time :(

Liberator Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dF90cTg1LXRvdnVMcExGNjdUWFBNU lE&usp=sharing

Galaxy Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dG5Va1Z5WkJMNFdrbXF6ZFhfTDMya kE&usp=sharing

Harasser Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dElBR2Nnekl3SjdzcUZpWmRHSy1IV nc&usp=sharing


Reposting, to make them all easy to find:

Flash Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dFpaNFhQcFF0UG5NOWlweDh5MEZDR VE&usp=sharing

Sunderer Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dDRtblZRYWdWbzFMTkZ0RUhZVUdFS 2c&usp=sharing

Lightning Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHFIWUhXNWtXcjdjUWN3bDNidFZaQ 2c&usp=sharing

MBT Deaths
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dHZnOVEyc3dLX2dCUTcwRVZWMDBsW Wc&usp=sharing

ESF Deaths:
https://docs.google.com/spreadsheet/ccc?key=0AshlLUNJ_SE8dERMcHJlWkJXYy13WVROc2s2TTNjO Wc&usp=sharing

Finally some data proved I was wrong on something.

Titan-150 AP is the best tank weapon to be used as anti air, but the p2-120 AP is not too far behind. On the Liberator chart.
Maybe it is because we see much more prowlers on the field then Vanguards and the data is making me think I am wrong, while I am not.
When you look into the Galaxy you see that the p2-120 AP is doing better then the Titan-150 AP.
Maybe when we look into the ESF data, the Titan-150 AP will be king. Even more if you see that the Titan-150 HEAT also one shot kill ESF.

MerlO
2014-02-23, 08:55 AM
The oracle lives on at: irc.planetside-universe.com #oracle

If it's not there, then come back later and I should have got it to reconnect, and you can grab your stats :)

The previous commands should all work, but if something is broken let me know.


Some additional commands:

To dump vehicle deaths (a large table, takes a while to generate):
!oracle dump vehicle <faction_id> <vehicle_id>

where faction_id and vehicle_id can be found by:
!oracle list vehicles

Dumping daily all faction kill totals:
!oracle dump faction

As of this post there is a little over a week of data, the first period being the day before the infiltrator update. (period 2 lost due to the downtime)

There's some new additions to this round of collection:
1. Added KPU standard deviation to kill stats
2. Broken down vehicle deaths by BR quartiles
3. Added weapons to breakdown NS weapons by faction - e.g VS C-4, TR NS Decimator

MerlO
2014-02-24, 02:45 AM
Did anyone ever figure out if it is possible to get roadkill stats? I'd just like to see it so maybe we can get a weed whacker or something added to the magrider.

Here you go, the magmower definitely kills more:
https://docs.google.com/spreadsheets/d/1CT6OZMTCp1cvqmd_y_K9T0O1_GmHdnj47etOCKZlVmo/edit?usp=sharing

Note that the roadkills includes friendly and enemy roadkills, so it's certainly possible that the magrider is just better at running friendlies over than enemies :)

Roadkills for other vehicles are largely uninteresting - from a glance I took, the scythe, like the magrider has a higher number among ESFs.

maradine
2014-02-27, 04:41 PM
The oracle lives on at: irc.planetside-universe.com #oracle

/me applauds

Remember to commit source!

MerlO
2014-03-24, 09:21 PM
/thread necro


Also how to do group calls? Eg stats on all shotguns, or all pump shotguns but without entering them all individually query line by query line?


Added a feature to generate kill stats for all weapons of a particular category (type), the output is the standard sum / averaging i've been doing in spreadsheets :
!oracle report <category_id>

where category_id can be found by !oracle list types

e.g !oracle report 4 will output a summary of all shotgun kill stats to date.

Of course raw period data is still needed to track changes over time.

Vehicle deaths are also aggregated in the same manner:
!oracle vehreport <faction_id> <vehicle_id>

MerlO
2014-03-24, 09:25 PM
/me applauds

Remember to commit source!

Added a fork from your source with my changes. Still need to double check that I uploaded everything :)

https://github.com/Merl0/ps2bot