Pages: 1 2 3 4 5 6 [7] 8 9 10 11 12 13 14 .. 14 :: one page |
|
Author |
Thread Statistics | Show CCP posts - 9 post(s) |
Thead Enco
Killing is Business Get Off My Lawn
51
|
Posted - 2014.01.17 02:20:00 -
[181] - Quote
CCP Phantom wrote:I would like to remind everyone to please stay constructive. It is of course valid to ask for the reasons of a change, nothing wrong with that! Constructive feedback is always welcome.
Yes please do tell................9 pages later and still crickets.............
"Any man who must say 'I am the king' is no true king."
Tywin Lannister-á |
Aebe Amraen
Dreddit Test Alliance Please Ignore
25
|
Posted - 2014.01.17 02:22:00 -
[182] - Quote
Pinky Hops wrote: [snip]
Well, to be fair those aren't exactly well scaled graphs. The last one (overall DPS) is particularly offensive in that regard.
Eg he talks about an 8.5% damage difference - but nowhere on those graphs do I see what would visually represent 8.5% - that's a pretty big indicator that things are scaled funny/selectively.
The last graph is intentionally scaled from 0 to 1, which is the only reasonable scale when discussing a ratio which is bounded between 0 and 1. Anything else would be misleading.
The other two graphs are scaled according to the range of the function.
If you can suggest a better way to scale them, please do. |
Pinky Hops
Spartan's DNA
296
|
Posted - 2014.01.17 02:25:00 -
[183] - Quote
Aebe Amraen wrote:Pinky Hops wrote: [snip]
Well, to be fair those aren't exactly well scaled graphs. The last one (overall DPS) is particularly offensive in that regard.
Eg he talks about an 8.5% damage difference - but nowhere on those graphs do I see what would visually represent 8.5% - that's a pretty big indicator that things are scaled funny/selectively.
The last graph is intentionally scaled from 0 to 1, which is the only reasonable scale when discussing a ratio which is bounded between 0 and 1. Anything else would be misleading.
Erm. What?
Considering the actual plots don't even traverse below .85 on the y axis I would start by not rendering anything below that so that you can see finer resolution where it's relevant - as in relative gains/losses (if you're using Excel -- right click the Y-axis and click format axis, and force the minimum to a number closer to .8)
Again - you mentioned an 8.5% damage increase in a situation. Where do I see that 8.5% difference visually with these graphs? |
Tyberius Franklin
Federal Navy Academy Gallente Federation
967
|
Posted - 2014.01.17 02:53:00 -
[184] - Quote
Pinky Hops wrote:Aebe Amraen wrote:Pinky Hops wrote: [snip]
Well, to be fair those aren't exactly well scaled graphs. The last one (overall DPS) is particularly offensive in that regard.
Eg he talks about an 8.5% damage difference - but nowhere on those graphs do I see what would visually represent 8.5% - that's a pretty big indicator that things are scaled funny/selectively.
The last graph is intentionally scaled from 0 to 1, which is the only reasonable scale when discussing a ratio which is bounded between 0 and 1. Anything else would be misleading. Erm. What? Considering the actual plots don't even traverse below .85 on the y axis I would start by not rendering anything below that so that you can see finer resolution where it's relevant - as in relative gains/losses (if you're using Excel -- right click the Y-axis and click format axis, and force the minimum to a number closer to .8) Again - you mentioned an 8.5% damage increase in a situation. Where do I see that 8.5% difference visually with these graphs? Is it supposed to be emphasized by that poorly scaled slight hump where 85% of the graph is white space? Because his goal is to demonstrate the magnitude of the DPS change visually as well. A scale from 0.8-1.0 doesn't necessarily do that since it doesn't show a total range of possible reduction to compare to the actual data plotted. |
Pinky Hops
Spartan's DNA
297
|
Posted - 2014.01.17 02:56:00 -
[185] - Quote
Tyberius Franklin wrote:Pinky Hops wrote:Aebe Amraen wrote:Pinky Hops wrote: [snip]
Well, to be fair those aren't exactly well scaled graphs. The last one (overall DPS) is particularly offensive in that regard.
Eg he talks about an 8.5% damage difference - but nowhere on those graphs do I see what would visually represent 8.5% - that's a pretty big indicator that things are scaled funny/selectively.
The last graph is intentionally scaled from 0 to 1, which is the only reasonable scale when discussing a ratio which is bounded between 0 and 1. Anything else would be misleading. Erm. What? Considering the actual plots don't even traverse below .85 on the y axis I would start by not rendering anything below that so that you can see finer resolution where it's relevant - as in relative gains/losses (if you're using Excel -- right click the Y-axis and click format axis, and force the minimum to a number closer to .8) Again - you mentioned an 8.5% damage increase in a situation. Where do I see that 8.5% difference visually with these graphs? Is it supposed to be emphasized by that poorly scaled slight hump where 85% of the graph is white space? Because his goal is to demonstrate the magnitude of the DPS change visually as well. A scale from 0.8-1.0 doesn't necessarily do that since it doesn't show a total range of possible reduction to compare to the actual data plotted.
You're right.
They could have just scaled the all the damages right down to zero.
What was I thinking. |
Tyberius Franklin
Federal Navy Academy Gallente Federation
967
|
Posted - 2014.01.17 03:05:00 -
[186] - Quote
Pinky Hops wrote:Tyberius Franklin wrote:Pinky Hops wrote:Aebe Amraen wrote:Pinky Hops wrote: [snip]
Well, to be fair those aren't exactly well scaled graphs. The last one (overall DPS) is particularly offensive in that regard.
Eg he talks about an 8.5% damage difference - but nowhere on those graphs do I see what would visually represent 8.5% - that's a pretty big indicator that things are scaled funny/selectively.
The last graph is intentionally scaled from 0 to 1, which is the only reasonable scale when discussing a ratio which is bounded between 0 and 1. Anything else would be misleading. Erm. What? Considering the actual plots don't even traverse below .85 on the y axis I would start by not rendering anything below that so that you can see finer resolution where it's relevant - as in relative gains/losses (if you're using Excel -- right click the Y-axis and click format axis, and force the minimum to a number closer to .8) Again - you mentioned an 8.5% damage increase in a situation. Where do I see that 8.5% difference visually with these graphs? Is it supposed to be emphasized by that poorly scaled slight hump where 85% of the graph is white space? Because his goal is to demonstrate the magnitude of the DPS change visually as well. A scale from 0.8-1.0 doesn't necessarily do that since it doesn't show a total range of possible reduction to compare to the actual data plotted. You're right. They could have just scaled the all the damages right down to zero. What was I thinking. You can bet that if he adjusted the axis as you suggest, someone would ignore the label and claim they had. |
Aebe Amraen
Dreddit Test Alliance Please Ignore
27
|
Posted - 2014.01.17 03:27:00 -
[187] - Quote
Pinky Hops wrote:Aebe Amraen wrote:Pinky Hops wrote: [snip]
Well, to be fair those aren't exactly well scaled graphs. The last one (overall DPS) is particularly offensive in that regard.
Eg he talks about an 8.5% damage difference - but nowhere on those graphs do I see what would visually represent 8.5% - that's a pretty big indicator that things are scaled funny/selectively.
The last graph is intentionally scaled from 0 to 1, which is the only reasonable scale when discussing a ratio which is bounded between 0 and 1. Anything else would be misleading. Erm. What? Considering the actual plots don't even traverse below .85 on the y axis I would start by not rendering anything below that so that you can see finer resolution where it's relevant - as in relative gains/losses (if you're using Excel -- right click the Y-axis and click format axis, and force the minimum to a number closer to .8) Again - you mentioned an 8.5% damage increase in a situation. Where do I see that 8.5% difference visually with these graphs? Is it supposed to be emphasized by that poorly scaled slight hump where 85% of the graph is white space?
Scaling the graph as you suggest would be misleading, since depending on how much whitespace I leave below the curve I can make an arbitrarily small difference seem HUUUUGE. Which I'm sure would match some people's agenda here.
If you want a differently-scaled graph, the only reasonable alternative is to graph a different thing--e.g., graph the % of damage lost, rather than the ratio of post-nerf to pre-nerf damage. This graph would reasonable be scaled from 0 to x%, without a misleading lack of whitespace like you suggest.
FYI, in real life I am a statistician, and I suspect I spend a lot more time reading and thinking about data visualization than you do. The scaling you suggest would draw gasps of horror from statisticians everywhere.
Edit: Typo |
I am disposable
Republic University Minmatar Republic
65
|
Posted - 2014.01.17 03:40:00 -
[188] - Quote
So I guess this is kind of like a Phoenix buff. |
Pinky Hops
Spartan's DNA
302
|
Posted - 2014.01.17 03:49:00 -
[189] - Quote
Aebe Amraen wrote:FYI, in real life I am a statistician, and I suspect I spend a lot more time reading and thinking about data visualization than you do. The scaling you suggest would draw gasps of horror from statisticians everywhere.
Edit: Typo
What is this, some kind of urination contest?
I too come from a technical background -- but never have I met a statistician worthy of his salt who emphasized white space or worried about the psychological interpretations of his data.
I believe the profession you are discussing is called marketing.
Draw grasps of horror from statisticians? What? It's like scaling the change in value of NASDAQ based on the concept that it could go to 0 at any moment. We're looking at relative change -- not some weird marketing ploy. |
HiddenPorpoise
BG-1 The Craniac
151
|
Posted - 2014.01.17 03:50:00 -
[190] - Quote
I am disposable wrote:So I guess this is kind of like a Phoenix buff. If their TP fleet support wasn't just nerfed, it could be. |
|
I am disposable
Republic University Minmatar Republic
65
|
Posted - 2014.01.17 03:51:00 -
[191] - Quote
HiddenPorpoise wrote:I am disposable wrote:So I guess this is kind of like a Phoenix buff. If their TP fleet support wasn't just nerfed, it could be.
Good point.
The TP nerf is truly baffling. |
Red Teufel
Mafia Redux Phobia.
344
|
Posted - 2014.01.17 03:55:00 -
[192] - Quote
wow how stupid...x of shame ccp holy cow.. |
Rahne Sentro
Mafia Redux Phobia.
25
|
Posted - 2014.01.17 04:00:00 -
[193] - Quote
This is an awful idea and CCP seriously needs to reconsider what road they're going down here... |
Aebe Amraen
Dreddit Test Alliance Please Ignore
27
|
Posted - 2014.01.17 04:05:00 -
[194] - Quote
Pinky Hops wrote:Aebe Amraen wrote:FYI, in real life I am a statistician, and I suspect I spend a lot more time reading and thinking about data visualization than you do. The scaling you suggest would draw gasps of horror from statisticians everywhere.
Edit: Typo What is this, some kind of urination contest? I too come from a technical background -- but never have I met a statistician worthy of his salt who emphasized white space or worried about the psychological interpretations of his data. I believe the profession you are discussing is called marketing. Draw grasps of horror from statisticians? What? It's like scaling the change in value of NASDAQ based on the concept that it could go to 0 at any moment. We're looking at relative change -- not some weird marketing ploy.
You must not have met very many good statisticians. A good place to start in the literature would be with a classic like Tufte's "The Visual Display of Quantitative Information" (link)--which, by the way, heavily emphasizes the importance of whitespace. |
Pinky Hops
Spartan's DNA
303
|
Posted - 2014.01.17 04:14:00 -
[195] - Quote
Aebe Amraen wrote:ou must not have met very many good statisticians. A good place to start in the literature would be with a classic like Tufte's "The Visual Display of Quantitative Information" ( link)--which, by the way, heavily emphasizes the importance of whitespace.
trying too hard 0/10 troll.
thread is about dreadnoughts and titans |
Grath Telkin
Sniggerdly Pandemic Legion
2119
|
Posted - 2014.01.17 04:17:00 -
[196] - Quote
@Fozzie/Rise/BalanceTeam
I am frankly shocked you think capital tracking needed to be nerfed unless your undeclared goal to balance titans is to just get them all to quit via boredom.
Or really for that matter, a dread, I mean they're just starting to see an upswing in use, so you instantly do everything you can to stomp on that?
Malcanis - Without drone assign, the slowcat doctrine will wither and die.
|
Jafit McJafitson
Dreddit Test Alliance Please Ignore
469
|
Posted - 2014.01.17 04:24:00 -
[197] - Quote
Thanks for steering the thread back in the right direction.
We need more outrage and fewer facts. More Graths and fewer graphs. |
Aebe Amraen
Dreddit Test Alliance Please Ignore
31
|
Posted - 2014.01.17 04:26:00 -
[198] - Quote
Pinky Hops wrote:Aebe Amraen wrote:ou must not have met very many good statisticians. A good place to start in the literature would be with a classic like Tufte's "The Visual Display of Quantitative Information" ( link)--which, by the way, heavily emphasizes the importance of whitespace. trying too hard 0/10 troll. thread is about dreadnoughts and titans
> Be Pinky Hops > Lose argument > Accuse opponent of being troll |
Jack Tronic
borkedLabs
147
|
Posted - 2014.01.17 04:35:00 -
[199] - Quote
Thead Enco wrote:CCP Phantom wrote:I would like to remind everyone to please stay constructive. It is of course valid to ask for the reasons of a change, nothing wrong with that! Constructive feedback is always welcome.
Yes please do tell................9 pages later and still crickets.............
No you see clearly the reason is a single module THAT YOU CAN ONLY OVERHEAT FOR A EXTREMELY LIMITED PERIOD OF TIME is going to make dreadnoughts OP so they must be nerfed.
Meanwhile carriers and domi sentry blobs get to roam free with the broken drone assist mechanic and the broken drone damage bug that allows you to instapop ships. |
Pinky Hops
Spartan's DNA
303
|
Posted - 2014.01.17 04:42:00 -
[200] - Quote
Aebe Amraen wrote:Pinky Hops wrote:Aebe Amraen wrote:ou must not have met very many good statisticians. A good place to start in the literature would be with a classic like Tufte's "The Visual Display of Quantitative Information" ( link)--which, by the way, heavily emphasizes the importance of whitespace. trying too hard 0/10 troll. thread is about dreadnoughts and titans > Be Pinky Hops > Lose argument > Accuse opponent of being troll
An argument about what?
Credentials and what your particular statistics bible is?
EVE is real...apparently. |
|
Caleb Seremshur
Gladiators of Rage Fidelas Constans
194
|
Posted - 2014.01.17 04:43:00 -
[201] - Quote
Kat Ayclism wrote:Why?
Seriously- this is the question we need answered first and foremost.
waa waa waa - big PL supercap blobber wants his guns to track subcaps again.
it is getting done because otherwise you could start tracking subcaps again which is not the purpose of a dread. fin LP store weapon cost rebalance |
Caleb Seremshur
Gladiators of Rage Fidelas Constans
194
|
Posted - 2014.01.17 04:44:00 -
[202] - Quote
Pinky Hops wrote:
An argument about what?
Credentials and what your particular statistics bible is?
EVE is real...apparently.
It's real enough that an economics professor is employed by CCP to keep certain aspects of the game in check.
Just saiyan LP store weapon cost rebalance |
Hanna Cyrus
Paranocxium Brotherhood Of Silent Space
39
|
Posted - 2014.01.17 04:55:00 -
[203] - Quote
I,m sorry to say this is no good idea. A dread should be possible to hit a carrier, with tracking nerf + TP nerf, i think my carrier can then speed tank a dread? A battle is longer than a few seconds, maybe a few hours (thx to tidi). How much nanite paste is needed then? And wenn i'm repping my TC i can't use it.
I don't think that it will be game breaking, if someone that skilled it, can hit a bit better for a few seconds in a fight.
|
James Amril-Kesh
4S Corporation Goonswarm Federation
8551
|
Posted - 2014.01.17 05:28:00 -
[204] - Quote
Jack Tronic wrote:Meanwhile carriers and domi sentry blobs get to roam free with the broken drone assist mechanic and the drone negative damage bug that allows you to instapop ships. I'm sure they're working to fix it. They know about it for one (a CSM rep said so) and it wouldn't really get a thread, just a mention in the notes of whatever patch fixes it. My EVE Videos |
Spc One
The Chodak Void Alliance
214
|
Posted - 2014.01.17 06:20:00 -
[205] - Quote
CCP Fozzie wrote: In Rubicon 1.1 the tracking speed of all Capital Turrets will decrease by 5%.
only 5% ? You need to do it like 40% .. it will be better this way so XL guns can be even more useless.
|
Jasmine Assasin
Caldari Provisions Caldari State
148
|
Posted - 2014.01.17 06:34:00 -
[206] - Quote
gascanu wrote:you guys just don't get it, do you? this is ccp way of trying to boost the Phoenix ; p.s. by nerfing all other dreads to the phoenix lvl
If that's the case then they still have a lot of work ahead of them.
|
Aebe Amraen
Dreddit Test Alliance Please Ignore
31
|
Posted - 2014.01.17 06:42:00 -
[207] - Quote
Here's another way to look at the tracking nerf: Link.
Given a dreadnaught that does an average of 10,000 dps under ideal conditions--this is the number that EFT/EveHQ/Pyfa will give you as the dps--this is the amount of dps lost after the 5% tracking nerf, varying with the target's radial velocity. It peaks just below 1, where you are losing about 3.6% of your paper DPS due to the nerf (4331 dps pre-nerf, 3975 dps post-nerf).
TL;DR: The 5% tracking nerf never makes you lose more than 3.6% of your paper dps. |
DocWeed's youknowwut
Space Meerkats
0
|
Posted - 2014.01.17 06:50:00 -
[208] - Quote
The fact that both the CFC and N3/PL are both going OMFGWTF R U DERING CCP?!?!?!?!? has me concerned |
Meyr
SiN Corp Black Core Alliance
245
|
Posted - 2014.01.17 07:08:00 -
[209] - Quote
CCP Phantom wrote:I would like to remind everyone to please stay constructive. It is of course valid to ask for the reasons of a change, nothing wrong with that! Constructive feedback is always welcome.
We ARE being constructive with our feedback - you're just sending it to the wrong department.
The 'thought process' behind this is, to say the least, incomprehensible, and has the appearance of being completely arbitrary.
So, yeah, we'd kinda like to know where Fozzie got the idea that this was necessary, 'cause, sure as hell, none of us can conceive of it. |
Meyr
SiN Corp Black Core Alliance
245
|
Posted - 2014.01.17 07:09:00 -
[210] - Quote
Aebe Amraen wrote:Here's another way to look at the tracking nerf: Link. Given a dreadnaught that does an average of 10,000 dps under ideal conditions--this is the number that EFT/EveHQ/Pyfa will give you as the dps--this is the amount of dps lost after the 5% tracking nerf, varying with the target's radial velocity. It peaks just below 1, where you are losing about 3.6% of your paper DPS due to the nerf (4331 dps pre-nerf, 3975 dps post-nerf). TL;DR: The 5% tracking nerf never makes you lose more than 3.6% of your paper dps.
And now that the EFT-warrior crowd has been heard from... |
|
|
|
|
Pages: 1 2 3 4 5 6 [7] 8 9 10 11 12 13 14 .. 14 :: one page |
First page | Previous page | Next page | Last page |