| Pages: 1 2 3 4 5 6 7 8 9 10 11 [12] 13 14 15 16 17 18 19 20 .. 22 :: one page |
|
|
| Author |
Thread Statistics | Show CCP posts - 4 post(s) |
Benny Ohu
Chaotic Tranquility
Warp to Cyno.
4390
   |
Posted - 2015.04.01 20:14:19 -
[331] - Quote
Jenshae Chiroptera wrote:Benny Ohu wrote:are we assuming that ccp only looked at fifteen days and never thought of looking at a longer period of time
or are we assuming that ccp looked at a longer period of time, saw data that disagreed with what they saw in fifteen days and decided to present the fifteen day data at fanfest anyway
"after having redefined sandbox contrary to the popular understanding of the word, i find eve is not a sandbox" Taking the 15 days of people's play is too limited and ludicrous. That this was even used, even run through the database makes CCP look very questionable. That people are clutching onto it is even more ridiculous.
you're assuming that the limited example designed to demonstrate the importance of getting proper data at the beginning of a talk is the entirety of the data ccp collected
is that correct
or are we suggesting that ccp rise presented a conclusion he knew was false |
Mag's
the united
19220
|
Posted - 2015.04.01 20:32:32 -
[332] - Quote
Benny Ohu wrote:Jenshae Chiroptera wrote:Benny Ohu wrote:are we assuming that ccp only looked at fifteen days and never thought of looking at a longer period of time
or are we assuming that ccp looked at a longer period of time, saw data that disagreed with what they saw in fifteen days and decided to present the fifteen day data at fanfest anyway
"after having redefined sandbox contrary to the popular understanding of the word, i find eve is not a sandbox" Taking the 15 days of people's play is too limited and ludicrous. That this was even used, even run through the database makes CCP look very questionable. That people are clutching onto it is even more ridiculous. you're assuming that the limited example designed to demonstrate the importance of getting proper data at the beginning of a talk is the entirety of the data ccp collected is that correct or are we suggesting that ccp rise presented a conclusion he knew was false e: and knowingly presented misleading data to support the false conclusion Why would they ever call him out for lying, when simply ignoring and denying the information is easier?
**Destination SkillQueue:- **
It's like assuming the lions will ignore you in the savannah, if you're small, fat and look helpless.
|
KIller Wabbit
The Scope
Gallente Federation
916
|
Posted - 2015.04.01 21:55:18 -
[333] - Quote
Jenshae Chiroptera wrote:For newbies about
86% - don't die
13% - die legally
1% - die to ganks.
Two things not analysed:
- How many died in Low Sec?
- How many died in their first battleship?
How about how many of the 13% were tricked into a corp under war dec and then mowed down?
How about how many were ganked in the next 15 days of their first month?
CCP does know how tip toe around the stats.
---
Signature ganked by CCP |
Jonah Gravenstein
Machiavellian Space Bastards
23392
|
Posted - 2015.04.01 22:09:06 -
[334] - Quote
Jenshae Chiroptera wrote: There is data, which can be turned into information and there is propeganda. Guess which one you are clutching to your breast and screaming that it is the truth?  So CCP Rise's presentation was propaganda? I'm pretty sure CCP leave propaganda to the marketing department to screw up.
We're trusting CCP Rise's presentation to be an accurate reflection of the current state of affairs, as a CCP employee his factual presentation should be regarded as authoritative regardless of what it said.
If it had said the exact opposite then I'm pretty sure that many among the criminal and merc elements in highsec would be looking to change their playstyle slightly to accommodate CCPs wish for newbie retention.
They've got to be the right type of newbies obviously; we don't want to be invaded by people thinking it's WoW in space, they can get bent.
Civilized behaviour is knowing that violence is barbaric, but paying other people to do it is business.
Nil mortifi sine lucre.
|
Jenshae Chiroptera
1195
|
Posted - 2015.04.02 00:13:13 -
[335] - Quote
Benny Ohu wrote:Jenshae Chiroptera wrote:Taking the 15 days of people's play is too limited and ludicrous.
That this was even used, even run through the database makes CCP look very questionable. That people are clutching onto it is even more ridiculous. you're assuming that the limited example designed to demonstrate the importance of getting proper data at the beginning of a talk is the entirety of the data ccp collected. Example:
Someone goes and eats loads of junk food, takes no exercise, almost dies of a heart attack and proclaims, "You see! I was right about healthy eating and training!"
Then I think, "Well obviously. That was plainly stupid to begin with; why did you just waste everyone's time?" I then leads me to question how that person's thought processes work to begin with that they present that.
CCP - Building ant hills and magnifying glasses for fat kids.
.
High Sec needs a stepping stone to other areas of space, where they can grow.
Fozzie is treating a symptom.
|
Pok Nibin
Filial Pariahs
670
|
Posted - 2015.04.02 00:21:28 -
[336] - Quote
*offers a new dead horse to beat upon*
The right to free speech doesn't automatically carry with it the right to be taken seriously.
|
Jenshae Chiroptera
1195
|
Posted - 2015.04.02 00:22:56 -
[337] - Quote
Pok Nibin wrote:*offers a new dead horse to beat upon* Thank you. I will use it as bait and keep shooting the vultures.
CCP - Building ant hills and magnifying glasses for fat kids.
.
High Sec needs a stepping stone to other areas of space, where they can grow.
Fozzie is treating a symptom.
|
Scipio Artelius
The Vendunari
End of Life
35351
|
Posted - 2015.04.02 00:32:38 -
[338] - Quote
Jenshae Chiroptera wrote:Example:
Someone goes and eats loads of junk food, takes no exercise, almost dies of a heart attack and proclaims, "You see! I was right about healthy eating and training!"
Then I think, "Well obviously. That was plainly stupid to begin with; why did you just waste everyone's time?" It then leads me to question how that person's thought processes work to begin with that they present that.
Sure, if you take a single data point as the basis of a decision.
If you took the data from 80,000 people all doing what they do and found that among those, the unhealthy ones were more likely to have a heart attack, that has significant value for decision making.
Same here. If the data from 80,000 different users shows that those that remain the most isolated are the ones most likely to leave the game early, that has significant value to making decisions about the future of the game. If that data also shows that the population looked at also wasn't negatively impacted by ganking, that also has value for decision making.
Come Win At Eve - Join The Vendunari
|
Jenshae Chiroptera
1195
|
Posted - 2015.04.02 00:43:43 -
[339] - Quote
Scipio Artelius wrote:If you took the data from 80,000 people It doesn't matter if the data is from one person, 80 000 or twenty million. It is useless if you ask the wrong questions of the wrong people. 
CCP - Building ant hills and magnifying glasses for fat kids.
.
High Sec needs a stepping stone to other areas of space, where they can grow.
Fozzie is treating a symptom.
|
Scipio Artelius
The Vendunari
End of Life
35352
|
Posted - 2015.04.02 00:47:49 -
[340] - Quote
Jenshae Chiroptera wrote:Scipio Artelius wrote:If you took the data from 80,000 people It doesn't matter if the data is from one person, 80 000 or twenty million. It is useless if you ask the wrong questions of the wrong people.
What? You don't believe a population study on the risk of heart attack is better than a single data point (it's the example you used. No one else invented it)?
As to the CCP data. There is nothing wrong with it for the purpose it was used for by Rise. You have not once provided any reasonable argument as to why it is inadequate for the scope of its use in that presentation.
Not one. That won't change after this many pages. It's just all circular at this point.
Come Win At Eve - Join The Vendunari
|
|
|
Kaarous Aldurald
Glorious Revolutionary Armed Forces of Highsec
CODE.
12388
|
Posted - 2015.04.02 00:53:03 -
[341] - Quote
Scipio Artelius wrote:You have not once provided any reasonable argument as to why it is inadequate for the scope of its use in that presentation.
Yeah they have. They don't want it to be true.
Oh, you said reasonable.
"Verily, I have often laughed at the weaklings who thought themselves good because they had no claws."
One of ours, ten of theirs.
|
Jenshae Chiroptera
1195
|
Posted - 2015.04.02 00:56:05 -
[342] - Quote
Scipio Artelius wrote:What? You don't believe a population study on the risk of heart attack is better than a single data point (it's the example you used. No one else invented it)?. Let's go and ask 80 000 kids that are five years old about the heart attacks they have had. I am sssuuuuure they have had loads of them!
CCP - Building ant hills and magnifying glasses for fat kids.
.
High Sec needs a stepping stone to other areas of space, where they can grow.
Fozzie is treating a symptom.
|
Scipio Artelius
The Vendunari
End of Life
35352
|
Posted - 2015.04.02 01:00:07 -
[343] - Quote
Jenshae Chiroptera wrote:Scipio Artelius wrote:What? You don't believe a population study on the risk of heart attack is better than a single data point (it's the example you used. No one else invented it)?. Let's go and ask 80 000 kids that are five years old about the heart attacks they have had. I am sssuuuuure they have had loads of them!
Go back and read your own example. They are your words. No one else's. It's your example.
Here, I'll requote it for you:
Example:
Someone goes and eats loads of junk food, takes no exercise, almost dies of a heart attack and proclaims, "You see! I was right about healthy eating and training!"
But even that aside, if you believe there are no children that have heart disease, you would be wrong. So even for that population, where the incidence is very low, a population study is better than an individual data point, for broad decision making.
Come Win At Eve - Join The Vendunari
|
Jenshae Chiroptera
1196
|
Posted - 2015.04.02 03:42:03 -
[344] - Quote
Scipio Artelius wrote:Blah Are you deliberately being dense? The orginal example pertains to developers that run silly queries against the database.
As to children having heart disease, is might be somewhat similar to newbies being "killed unlawfully" within 15 days.
"Look mummy! I am eating mud!"
"Why are you showing me this? Why are your proud of being so mentally deficient?"
CCP - Building ant hills and magnifying glasses for fat kids.
.
High Sec needs a stepping stone to other areas of space, where they can grow.
Fozzie is treating a symptom.
|
Lienzo
Amanuensis
69
|
Posted - 2015.04.02 04:02:23 -
[345] - Quote
Well, I don't really think mechanics are at the heart of it, and looking at graphs of correlations or covariances probably doesn't tell us much.
Mainly, I think it is the social element that makes or breaks a subscription.
If we were to look at a mechanic, we might look at killboards, as many players are more afraid of them than they are of losing ships. They fear that if they lose the ship, they'll somehow "look stupid," and consequently don't "undock" or rather simply avoid engagement. We might make the hypothesis that the more they avoid engagement with other players, the more likely they are to leave the game.
How do you address that? Perhaps a bit of judicious carebearism aimed at propping up the ego.
One thing we might do with APIs is ascribe a record of losses to a corporate entity rather than individual players. Your record lives and dies by the lifespan of that corporation. European countries are big on scrubbing search history results these days, and maybe EVE could benefit in the long run by some player "bill of rights" as regards how they shape their persona.
Some people really really like their killboards of course, and as such perhaps such APIs could be linked to a numerical character ID. If players want to register their IDs with that out of game killboard service, then they can do so. Their own records in their kill history will still be limited to corporate centric records, so even if they manually copy it out, they cannot manually force another player to exist on an out of game service. Alternately, if we find that players look for ways to circumvent this, then we allow players to cycle their keys, or give the option to issue them new ones when they join new corps. Ideally, these API tags would be multi-part tags, identifying both corporate identity as well as the individual, allowing those who do hold those kill mails to retain their record in categorical form.
I know the purists are raging at this point. However, I will focus on one practical consideration, which is solo pvpers. Many of them like to look up their pending opponents as a form of free intel. Well, it probably isn't too obsequious to expect people to establish their reputations in the game, rather than out of it. Gaining intel on corporation habits and activity is more vague, but it injects more uncertainty into the equation.
Is this a fair outcome? No. Grizzled old vets like me can't really be embarrassed by our shoddy and lackluster pvp performance. It's not that easy to be rid of us.  However, I can see how it affects a range of intermediate players negatively. If anything, it errs a bit more on freedom and less on consequence. It takes the burden of PR consideration off of the shoulder of more timorous players, and puts it more squarely on the shoulders of specialized corporate flakkies and propaganda spinners.
If not being able to wring the tear sponge for longer than a moment results in more risk taking behavior overall, I think we can all make the calculation of what is being gained versus what is being lost. |
Black Pedro
Yammerschooner
790
|
Posted - 2015.04.02 05:49:49 -
[346] - Quote
Jenshae Chiroptera wrote:Scipio Artelius wrote:Blah Are you deliberately being dense? The orginal example pertains to developers that run silly queries against the database.
Asking the question "how many new players leave after being exploded by another player" is a silly question to ask when trying to assess why new players don't convert from trial to subscriber?
CCP formulated a hypothesis - that new players were not staying because they were being killed by existing players - and tested that by querying over 80 000 data points from their database. They examined the correlation between deaths to other players during the trial and retention rates to test this hypothesis. Surprisingly, they found not only was there no data to support this hypothesis, they found that deaths to other players had the opposite effect.
Sure, this doesn't establish conclusively that one factor causes the other, but certainly it is not a silly question to ask nor a silly way to go about testing that hypothesis. The result is strong evidence that losing a ship to another player (during the trial) is not a predictor of whether a player will stay with the game. Therefore, CCP learned that this is not an area they need to fix and can focus their efforts on other aspects of the trial that do correlate with players not subscribing.
Why is that so difficult to accept? |
Mag's
the united
19223
|
Posted - 2015.04.02 06:13:32 -
[347] - Quote
Black Pedro wrote:Why is that so difficult to accept? Simple. Because it doesn't fit their narrative.
**Destination SkillQueue:- **
It's like assuming the lions will ignore you in the savannah, if you're small, fat and look helpless.
|
Scipio Artelius
The Vendunari
End of Life
35363
|
Posted - 2015.04.02 06:30:46 -
[348] - Quote
Jenshae Chiroptera wrote:Scipio Artelius wrote:Blah Are you deliberately being dense?
No, I'm just naturally that way I guess.
Quote:The orginal example pertains to developers that run silly queries against the database.
Excuse me for being dense, but I still have seen no evidence, counter data or reasonable argument to demonstrate why the data presented as it was, was not relevant within the scope that it was used.
My own denseness I'm sure, but maybe you could explain in simple terms why that data, presented as it was, was not relevant to the point Rise made?
Even one reasonable counter argument. Not personal insults or anything, just the topic and the data presented for the purpose it was presented.
I may be dense, but I am open minded, as I suspect are many people on the forum. Happy to see and understand why the data is wrong, if only it can be explained.
Come Win At Eve - Join The Vendunari
|
Azda Ja
Green Skull LLC
3543
|
Posted - 2015.04.02 06:38:27 -
[349] - Quote
Scipio Artelius wrote:
Even one reasonable counter argument. Not personal insults or anything, just the topic and the data presented for the purpose it was presented.
Now now Scipio, let's be reasonable here.
"I only lose ships when I fly with Azda." - Barry Cuttlefish
|
Jonah Gravenstein
Machiavellian Space Bastards
23399
|
Posted - 2015.04.02 08:00:42 -
[350] - Quote
Azda Ja wrote:Scipio Artelius wrote:
Even one reasonable counter argument. Not personal insults or anything, just the topic and the data presented for the purpose it was presented.
Now now Scipio, let's be reasonable here. This is GD, being reasonable has no place here 
Civilized behaviour is knowing that violence is barbaric, but paying other people to do it is business.
Nil mortifi sine lucre.
|
|
|
Eve Solecist
Science and Trade Institute
Caldari State
118
|
Posted - 2015.04.02 08:05:00 -
[351] - Quote
Jonah Gravenstein wrote:Azda Ja wrote:Scipio Artelius wrote:
Even one reasonable counter argument. Not personal insults or anything, just the topic and the data presented for the purpose it was presented.
Now now Scipio, let's be reasonable here. This is GD, being reasonable has no place here
Jonah, nice to see your scared face around again!
"Please do not file support tickets to ask if your support ticket will be answered soon." - Actual Quote.
|
Eve Solecist
Science and Trade Institute
Caldari State
119
|
Posted - 2015.04.02 08:10:47 -
[352] - Quote
Black Pedro wrote:Jenshae Chiroptera wrote:Scipio Artelius wrote:Blah Are you deliberately being dense? The orginal example pertains to developers that run silly queries against the database. Asking the question "how many new players leave after being exploded by another player" is a silly question to ask when trying to assess why new players don't convert from trial to subscriber? CCP formulated a hypothesis - that new players were not staying because they were being killed by existing players - and tested that by querying over 80 000 data points from their database. They examined the correlation between deaths to other players during the trial and retention rates to test this hypothesis. Surprisingly, they found not only was there no data to support this hypothesis, they found that deaths to other players had the opposite effect. Sure, this doesn't establish conclusively that one factor causes the other, but certainly it is not a silly question to ask nor a silly way to go about testing that hypothesis. The result is strong evidence that losing a ship to another player (during the trial) is not a predictor of whether a player will stay with the game. Therefore, CCP learned that this is not an area they need to fix and can focus their efforts on other aspects of the trial that do correlate with players not subscribing. Why is that so difficult to accept?
For those who don't want to read through...
He says that CCP checked if recruiting by shooting was a thing ...
... noticed that it was ...
... noticed that it actually WORKS ...
... as it has for manymanymany years.
If only the degenerates realised that they have no ground to speak.
"Please do not file support tickets to ask if your support ticket will be answered soon." - Actual Quote.
|
Eve Solecist
Science and Trade Institute
Caldari State
126
|
Posted - 2015.04.02 11:11:54 -
[353] - Quote
Well, I guess the truth has ended the thread of the degenerates.
"Please do not file support tickets to ask if your support ticket will be answered soon." - Actual Quote.
|
Eli Stan
Center for Advanced Studies
Gallente Federation
102
|
Posted - 2015.04.02 15:00:36 -
[354] - Quote
Eve Solecist wrote:For those who don't want to read through...
He says that CCP checked if recruiting by shooting was a thing ...
... noticed that it was ...
... noticed that it actually WORKS ...
... as it has for manymanymany years.
If only the degenerates realised that they have no ground to speak.
Was there a causal relationship established?
I thought it was that the players who "got" EVE were more likely to fly around and therefore get shot, not that players who got shot were more likely to "get" EVE.
|
Dots
State Protectorate
Caldari State
5
|
Posted - 2015.04.02 17:49:29 -
[355] - Quote
Eli Stan wrote:Eve Solecist wrote:For those who don't want to read through...
He says that CCP checked if recruiting by shooting was a thing ...
... noticed that it was ...
... noticed that it actually WORKS ...
... as it has for manymanymany years.
If only the degenerates realised that they have no ground to speak. Was there a causal relationship established? I thought it was that the players who "got" EVE were more likely to fly around and therefore get shot, not that players who got shot were more likely to "get" EVE.
They are less likely to leave EVE if the conflict is mutual. They are least likely to leave if nonconsensually shot at. Which is to say these two groups are more likely to keep playing than players who aren't shot at at all.
everything is better with ߦêߦÆߦù-ó on it
New Player Opportunities: a gallery
|
Pohbis
Neo T.E.C.H.
468
|
Posted - 2015.04.02 18:10:18 -
[356] - Quote
Dots wrote:They are less likely to leave EVE if the conflict is mutual. They are least likely to leave if nonconsensually shot at. Which is to say these two groups are more likely to keep playing than players who aren't shot at at all. Correlation does not imply, causation GÇô which is what was asked for.
|
Jenshae Chiroptera
1202
|
Posted - 2015.04.02 18:59:18 -
[357] - Quote
Black Pedro wrote:Jenshae Chiroptera wrote:Scipio Artelius wrote:Blah Are you deliberately being dense? The orginal example pertains to developers that run silly queries against the database. Asking the question "how many new players leave after being exploded by another player" is a silly question to ask when trying to assess why new players don't convert from trial to subscriber? CCP formulated a hypothesis - Yes, a bad hypothesis.
Why not simply asked newbies that are 15 day olds, "Why are you quitting?"
CCP - Building ant hills and magnifying glasses for fat kids.
.
High Sec needs a stepping stone to other areas of space, where they can grow.
Fozzie is treating a symptom.
|
Kaarous Aldurald
Glorious Revolutionary Armed Forces of Highsec
CODE.
12398
|
Posted - 2015.04.02 19:07:17 -
[358] - Quote
Jenshae Chiroptera wrote: Yes, a bad hypothesis.
Why not simply asked newbies that are 15 day olds, "Why are you quitting?"
He really is this ignorant, folks.
What do you think the outgoing survey is for, may I ask? Just to ask them what kinds of Gummi Bears they enjoy?
"Verily, I have often laughed at the weaklings who thought themselves good because they had no claws."
One of ours, ten of theirs.
|
Black Pedro
Yammerschooner
799
|
Posted - 2015.04.02 19:11:02 -
[359] - Quote
Jenshae Chiroptera wrote:Black Pedro wrote:Jenshae Chiroptera wrote:Scipio Artelius wrote:Blah Are you deliberately being dense? The orginal example pertains to developers that run silly queries against the database. Asking the question "how many new players leave after being exploded by another player" is a silly question to ask when trying to assess why new players don't convert from trial to subscriber? CCP formulated a hypothesis - Yes, a bad hypothesis. Why not simply asked newbies that are 15 day olds, "Why are you quitting?"
They did. And less than 1% of the new players said it was because of ship loss or harassment.
Presumably CCP Rise and friends are using the other 99%+ of the answers to fix the NPE as we speak. |
Dots
State Protectorate
Caldari State
5
|
Posted - 2015.04.02 19:27:38 -
[360] - Quote
Pohbis wrote:Dots wrote:They are less likely to leave EVE if the conflict is mutual. They are least likely to leave if nonconsensually shot at. Which is to say these two groups are more likely to keep playing than players who aren't shot at at all. Correlation does not imply, causation GÇô which is what was asked for.
Correlation is a prerequisite for causation. Did either yourself or Eli have any input as to how the data can more exactly pinpoint causation? It's really simply to throw tomatoes all day long, but it just seems more like people are willfully ignoring the 80,000-large sample set because it interferes with their beliefs.
I've yet to see a critic of Rise's study explain in specific terms what is missing or incorrect in the data (and I don't mean creative offtopic analogies).
everything is better with ߦêߦÆߦù-ó on it
New Player Opportunities: a gallery
|
|
|
|
|
| |
|
| Pages: 1 2 3 4 5 6 7 8 9 10 11 [12] 13 14 15 16 17 18 19 20 .. 22 :: one page |
| First page | Previous page | Next page | Last page |