Pages: 1 [2] :: one page |
|
Author |
Thread Statistics | Show CCP posts - 9 post(s) |
Vincent Athena
V.I.C.E.
1549
|
Posted - 2013.01.30 16:05:00 -
[31] - Quote
CCP Masterplan wrote:Those were indeed my words. Ultimately whatever we do to increase capacity, there will probably always be more players ready to consume it all. The goal is to manage the situation as best as possible when this happens (eg TiDi) Not quite true. There are only so many active accounts. If they could all log in, all go to one system and all fight, without lag or TiDi, then you would have adequate hardware. http://vincentoneve.wordpress.com/ |
Traidir
Hedion University Amarr Empire
47
|
Posted - 2013.01.30 18:34:00 -
[32] - Quote
The kill ratio graph should be the RvB flag.
Also, the Average Damage graph does not show potentially avoided damage as via speed tanking does it? Minmatar ships would probably fare far better by that rubric.
Meanwhile, the Punisher has 5% bonus/lvl to armor resistances which would make it also appear lower on the chart compared to the "Potential" damage
The Incursus is only so high because of it's 10% bonus/lvl to Armor Repairer effectiveness... |
Illectroculus Defined
GoonWaffe Goonswarm Federation
101
|
Posted - 2013.01.30 22:21:00 -
[33] - Quote
" I can promise you that weGÇÖll have some wonderful time-lapse statistics of New Eden to show you on Fanfest. "
What a Brilliant Idea..... http://www.youtube.com/watch?v=WDR5TaP_VqY
I mean if you run out of spare time you can always show my movie. |
|
CCP Quant
C C P C C P Alliance
8
|
Posted - 2013.01.31 01:53:00 -
[34] - Quote
Traidir wrote:The kill ratio graph should be the RvB flag. Also, the Average Damage graph does not show potentially avoided damage as via speed tanking does it? Minmatar ships would probably fare far better by that rubric. Meanwhile, the Punisher has 5% bonus/lvl to armor resistances which would make it also appear lower on the chart compared to the "Potential" damage The Incursus is only so high because of it's 10% bonus/lvl to Armor Repairer effectiveness...
You are absolutely right, the damage/tank was from the raw kill logs/reports. Although possible, it could get a little time consuming to take speed tanking into account
Illectroculus Defined wrote:" I can promise you that weGÇÖll have some wonderful time-lapse statistics of New Eden to show you on Fanfest. " What a Brilliant Idea..... http://www.youtube.com/watch?v=WDR5TaP_VqY I mean if you run out of spare time you can always show my movie.
Brilliant work in some cases we might be able to provide data on more accessible formats (csv,json, yaml) to reduce the data mining overhead of you guys, i.e. scraping data together from killboards, api, etc. So if you're planning on making more it won't hurt to send me a line. |
|
Max Therion
Jita Ikami Bank
4
|
Posted - 2013.01.31 01:57:00 -
[35] - Quote
Is CCP using Hortonworks? Hmmmm.... I know of this. www.hortonworks.com |
iskflakes
313
|
Posted - 2013.01.31 03:41:00 -
[36] - Quote
So, when are you going to stop running the entire 3000 man fight on a single python thread? - |
deepos
Blue Republic RvB - BLUE Republic
5
|
Posted - 2013.01.31 09:07:00 -
[37] - Quote
Awesome graphs!
Thanks a lot quant
Poinen must burn was an amazing event !
|
Steve Ronuken
Fuzzwork Enterprises
1111
|
Posted - 2013.01.31 10:29:00 -
[38] - Quote
iskflakes wrote:So, when are you going to stop running the entire 3000 man fight on a single python thread?
Well, how happy would you be about the async thread managing a titan getting a doomsday off on you, where as before it would have been alphaed before that shot got off?
On a more serious note, looks like the next big thing will be a dispatcher that works on a different thread. Which will allow at least 2 cpus to do work. FuzzWork Enterprises http://www.fuzzwork.co.uk/ Blueprint calculator, invention chance calculator, isk/m3 Ore chart-á and other 'useful' utilities.As well as mysql and CSV/XLS conversions of the Static Data Extract. |
Deerin
Murientor Tribe Defiant Legacy
56
|
Posted - 2013.01.31 12:49:00 -
[39] - Quote
Very nice numbers thank you.
Do you have number of logouts(disconnects, crashers connection losses) in asakai during that time? Can you compare it to same data from uemon (if possible)
...and although I have no idea how hardware switching works, will it ever be possible to make the switch on-the fly. I mean is there light at the end of the tunnel or shall we just forget about its possibility? |
Spurty
V0LTA Verge of Collapse
783
|
Posted - 2013.01.31 17:48:00 -
[40] - Quote
More graphs with 4 scales of darkish grey please!
Trolling aside (no, really 4 scales of darkish grey, MY EYES!), very much appreciate the breakdown / analysis!
Feeds the inner nerd.
--- GÇ£If you think this Universe is bad, you should see some of the others.GÇ¥ GÇò Philip K. **** |
|
Illectroculus Defined
GoonWaffe Goonswarm Federation
101
|
Posted - 2013.01.31 18:15:00 -
[41] - Quote
CCP Quant wrote:Illectroculus Defined wrote:" I can promise you that weGÇÖll have some wonderful time-lapse statistics of New Eden to show you on Fanfest. " What a Brilliant Idea..... http://www.youtube.com/watch?v=WDR5TaP_VqYI mean if you run out of spare time you can always show my movie. Brilliant work in some cases we might be able to provide data on more accessible formats (csv,json, yaml) to reduce the data mining overhead of you guys, i.e. scraping data together from killboards, api, etc. So if you're planning on making more it won't hurt to send me a line.
In that case, we really need to talk, because I have huge holes in my API history and would really like to update a few of these things to tell some stories.
|
Dinsdale Pirannha
Pirannha Corp
982
|
Posted - 2013.01.31 21:10:00 -
[42] - Quote
I find it quite enlightening that a CCP employee has the time to dive deep into the numbers to glean highly detailed data, and then present it in some very illuminating graphics about 2 battles, yet no dev / employee has the slightest inclination to end a 110 page threadnaught on economic capabilities of null vs high sec, but posting the results, without graphics, of some very basic SQL commands.
You know, stuff like how many mfg slots exist in high sec/low sec/null sec, how many supercaps were built in the last 7 months, how many T2 items are built in null sec vs high sec.
You know, stuff that would end any propaganda in a few minutes, instead of the ISD shutting down a thread after 110 pages.
|
Brandon Syne
Science and Trade Institute Caldari State
20
|
Posted - 2013.01.31 21:29:00 -
[43] - Quote
http://cdn1.eveonline.com/www/newssystem/media/64150/1/AsakaiEffectiveDamageByWeapon.png
Are doomsdays so insignificant that they aren't included here?
Or perhaps that is the real reason why "smartbomb" came so high up. |
Fronkfurter McSheebleton
The Graduates RAZOR Alliance
184
|
Posted - 2013.01.31 23:19:00 -
[44] - Quote
Brandon Syne wrote:http://cdn1.eveonline.com/www/newssystem/media/64150/1/AsakaiEffectiveDamageByWeapon.png
Are doomsdays so insignificant that they aren't included here?
Or perhaps that is the real reason why "smartbomb" came so high up. It was lowsec, bro....no DDs. thhief ghabmoef |
NARDAC
NIGHTMARE FACTORY INDUSTRIES Claimed.
18
|
Posted - 2013.01.31 23:26:00 -
[45] - Quote
So, to cut down lag in future big fights like this: 1) No undocking in a rookie ship from a station in a solar system that is under TiDi. Undock in pod, okay. Go get a real ship and come back, getting traffic controlled like everyone else, okay. But no going back in a free rookie ship.
Are all undock's in TiDi system traffic controlled like gate jumps?
2) No jumping into a system under TiDi in a rookie ship.
3) This one would be harder, but no warping to grid were ammo is being traded, in pod, in a solar system under TiDi.
I know people want to lookie loo on a big fight like this, but the people in the big expensive ships, and cutting down their lag, should take priority over accommodating sightseers. , in my opinion. |
Tarvos Telesto
Blood Fanatics
770
|
Posted - 2013.02.01 00:34:00 -
[46] - Quote
Nice graphs, thanks CCP for cool blog.
Ps. Here stil more than 1000 titans waiting to become part of history, oh wait, they never come to usage because of broken moon pasive incomes and broken sov mechanc war mechanic. EvE isn't game, its style of living. |
Vaerah Vahrokha
Vahrokh Consulting
3764
|
Posted - 2013.02.01 08:40:00 -
[47] - Quote
CCP Quant, I really loved the blog up to the last bit!
Now, could you also do something similar for the EvE markets?
You having worked in bank and being called Quant... come on... nudges... you know you want to do it! Auditing | Collateral holding and insurance | Consulting | PLEX for Good Charity
Twitter channel |
AxeMinister
Science and Trade Institute Caldari State
0
|
Posted - 2013.02.01 12:51:00 -
[48] - Quote
"Now letGÇÖs see the accumulated amount of damage applied from 2:00 to 8:00:"
Just wanted to point out how badly gimped missiles have become since the nerf. That graph clearly shows this. |
Traidir
Hedion University Amarr Empire
51
|
Posted - 2013.02.01 17:38:00 -
[49] - Quote
In the Effective Damage by Weapon chart, could you normalize the damage by how many weapons of a given type were present?
For example, if 10 ships are using lasers and 5 are using projectiles, it's little wonder that lasers are 100% more total damage.
It might be good to separate them by size class too. |
|
CCP Quant
C C P C C P Alliance
9
|
Posted - 2013.02.01 22:40:00 -
[50] - Quote
Dinsdale Pirannha wrote:I find it quite enlightening that a CCP employee has the time to dive deep into the numbers to glean highly detailed data
I actually didn't have the time! I wrote most of it in the middle of a night and am already behind schedule with some projects because of it It just had to be done
Spurty wrote:More graphs with 4 scales of darkish grey please!
Trolling aside (no, really 4 scales of darkish grey, MY EYES!), very much appreciate the breakdown / analysis!
Feeds the inner nerd.
I only colored the top 4 for emphasis, I thought the rest was negligible so I grayed them out. The legend is sorted top to bottom so it's possible to determine e.g. Projectiles by counting layers
Deerin wrote:Very nice numbers thank you.
Do you have number of logouts(disconnects, crashers connection losses) in asakai during that time? Can you compare it to same data from uemon (if possible)
...and although I have no idea how hardware switching works, will it ever be possible to make the switch on-the fly. I mean is there light at the end of the tunnel or shall we just forget about its possibility? This is definitely a part of the ocean of data available
Nope sorry, in-house framework directly on top of Hadoop.
Vaerah Vahrokha wrote:CCP Quant, I really loved the blog up to the last bit! Now, could you also do something similar for the EvE markets? You having worked in bank and being called Quant... come on... nudges... you know you want to do it! Oh absolutely! I talked to Dr. Eyj+¦ about it and in his opinion we should be doing what I initially suggested; providing you guys with all the data you need (within some limits of course and approved by our IA) to make your own analysis work easier. I was about to give you a link to a great example of community made market analysis, amazingly it turned out you are the author of that post! Dr. Eyj+¦ pointed that post out to me as a good read and I'll say I'm impressed, great work!
For reference: Link 1: https://forums.eveonline.com/default.aspx?g=posts&t=49927 Link 2: http://community.eveonline.com/ingameboard.asp?a=topic&threadID=1375714
AxeMinister wrote:"Now letGÇÖs see the accumulated amount of damage applied from 2:00 to 8:00:"
Just wanted to point out how badly gimped missiles have become since the nerf. That graph clearly shows this.
I'd rather say that in situations like this, missiles are clearly not used much. The graph doesn't include information on how many ships were fielded using each of the weapon platforms. As Traidir points out here below, it would be interesing to normalize these numbers by their usage.
Traidir wrote:In the Effective Damage by Weapon chart, could you normalize the damage by how many weapons of a given type were present? For example, if 10 ships are using lasers and 5 are using projectiles, it's little wonder that lasers are 100% more total damage. It might be good to separate them by size class too.
The data is there, again somewhere in the ocean of all the data available, we'll just have to dig it up . I can try to see if it can be done quickly, but to be honest the time-schedule really fills up quickly here :) this dev blog was mostly written in the middle of a night to get it done |
|
|
Vaerah Vahrokha
Vahrokh Consulting
3769
|
Posted - 2013.02.02 02:11:00 -
[51] - Quote
CCP Quant wrote:Vaerah Vahrokha wrote:CCP Quant, I really loved the blog up to the last bit! Now, could you also do something similar for the EvE markets? You having worked in bank and being called Quant... come on... nudges... you know you want to do it! Oh absolutely! I talked to Dr. Eyj+¦ about it and in his opinion we should be doing what I initially suggested; providing you guys with all the data you need (within some limits of course and approved by our IA) to make your own analysis work easier. I was about to give you a link to a great example of community made market analysis, amazingly it turned out you are the author of that post! Dr. Eyj+¦ pointed that post out to me as a good read and I'll say I'm impressed, great work! For reference: Link 1: https://forums.eveonline.com/default.aspx?g=posts&t=49927Link 2: http://community.eveonline.com/ingameboard.asp?a=topic&threadID=1375714
Thank you, I am humbled and even blushing reading this This will certainly push me into working harder in this unique - I believe - initiative to try and observe real life markets on one side, EvE markets on the other and draw conclusions about where they follow the same laws (very often) and where they differ (due to the smaller EvE liquidity and participants).
Auditing | Collateral holding and insurance | Consulting | PLEX for Good Charity
Twitter channel |
PVDNS77
Red Federation RvB - RED Federation
22
|
Posted - 2013.02.02 15:54:00 -
[52] - Quote
thank you very much for the stats. |
Dinsdale Pirannha
Pirannha Corp
991
|
Posted - 2013.02.03 07:06:00 -
[53] - Quote
CCP Quant wrote:Dinsdale Pirannha wrote:I find it quite enlightening that a CCP employee has the time to dive deep into the numbers to glean highly detailed data
I actually didn't have the time! I wrote most of it in the middle of a night and am already behind schedule with some projects because of it It just had to be done ta available, we'll just have to dig it up . I can try to see if it can be done quickly, but to be honest the time-schedule really fills up quickly here :) this dev blog was mostly written in the middle of a night to get it done
How about running some important SQL commands, like how many supercaps have been built in the past 7 months, how many mfg slots exits in high sec stations vs how many are available in active (turned on) POS's?
Things like this will define the next expansion, not a breakdown of how many ships of a certain type were lost in 2 themepark fights.
|
rodyas
Tie Fighters Inc
1037
|
Posted - 2013.02.03 08:01:00 -
[54] - Quote
Dinsdale Pirannha wrote:CCP Quant wrote:Dinsdale Pirannha wrote:I find it quite enlightening that a CCP employee has the time to dive deep into the numbers to glean highly detailed data
I actually didn't have the time! I wrote most of it in the middle of a night and am already behind schedule with some projects because of it It just had to be done ta available, we'll just have to dig it up . I can try to see if it can be done quickly, but to be honest the time-schedule really fills up quickly here :) this dev blog was mostly written in the middle of a night to get it done How about running some important SQL commands, like how many supercaps have been built in the past 7 months, how many mfg slots exits in high sec stations vs how many are available in active (turned on) POS's? Things like this will define the next expansion, not a breakdown of how many ships of a certain type were lost in 2 themepark fights.
Probably enough Supercaps to ruin any dreams, you might have had. Signature removed for inappropriate language - CCP Eterne |
Veshta Yoshida
PIE Inc. Praetoria Imperialis Excubitoris
516
|
Posted - 2013.02.03 22:19:00 -
[55] - Quote
Was there lag in Poinen?
Something tells me that the null fat-cats had about as much fun playing with 10 minute weapon cycles as they did 'back in the day' with drops/black screens/etc. while the RvB blokes got to have what they are in the world to do: Have fun.
Still saddens me that with all the ISK brought into the Asa fight the death-toll was lower than most pirate nests on week nights (well, maybe weekend nights) .. take away the 3 bulk values of the Titan's and eliminate the local mail/loot-whoring frigates and the whole thing comes off as a dud.
Here is a concept: Null throws a few billion at RvB to sponsor an upgrade of the next FFA to cruisers and get in on the fighting. CCP switches off sentries in system to allow outsiders (read: null folk) to engage on equal terms .. everyone has a laugh. |
Remiel Pollard
Intergalactic Trade and Harm
1327
|
Posted - 2013.02.05 12:16:00 -
[56] - Quote
lol @ last graph - "Titan oops jump"
You don't scare me. I've been to Jita. |
Buhhdust Princess
Mind Games. Suddenly Spaceships.
456
|
Posted - 2013.02.13 14:19:00 -
[57] - Quote
Well, as one of the first people in that brawl, I have to say;
1. My PC is obviously awesome, I was dualboxing during that, both accounts with TIDI in full HD graphics on the highest settings, and indeed, it was absolutely outstandingly beautiful watching the massacre unfold.
2. The TIDI did it's job just fine, thank god CCP introduced this as it works. ALTHOUGH i noticed a LOT of module lag on top of the tidi, this doesn't help us at all, especially when everything is 10% the usual speed anyway.
Sry to bring this post back up, just wanted to drop in my 2 cents from that fight as i was there start to end, I had fun, but would've had more fun If i could've made a video of it with my modules working as intended during tidi. -Buhhd |
Hazen Koraka
FutureTech Industrial Inc.
177
|
Posted - 2013.02.14 08:28:00 -
[58] - Quote
iskflakes wrote:So, when are you going to stop running the entire 3000 man fight on a single python thread?
I lol'ed Exploration is Random. Random is Random... or is it?! http://docs.python.org/2/library/random.html |
|
|
|
Pages: 1 [2] :: one page |
First page | Previous page | Next page | Last page |