EPTOfficial 'what is Dota anymore' discussion - Page 561
| Forum Index > Dota 2 General |
|
ahswtini
Northern Ireland22213 Posts
| ||
|
andyrau
13015 Posts
if you take income range of the top 1%, that shit goes from 350k+ a year to hundreds of millions or 1.5 billion if you're notch. | ||
|
Sn0_Man
Tebellong44238 Posts
On February 21 2015 04:11 tehh4ck3r wrote: i would think that most abandoned accounts wouldn't even make it into playing ranked, since they wouldn't have reached level 13 That's why valve's pre-ranked numbers where 4.1k was like 99th percentile don't really match with more recent surveys that only look at ranked or w/e yes | ||
BluemoonSC
SoCal8933 Posts
On February 21 2015 04:14 SKC wrote: The data isn't ranked only. At least it doesnt say so anywhere. The average ranked player should be slightly higher, but still not like 2k higher like some people thought. the data came from a dotabuff employee's comment in another thread. there's a ton of bias in the data like i, and others, have brought up. | ||
|
SKC
Brazil18828 Posts
On February 21 2015 04:18 BluemoonSC wrote: the data came from a dotabuff employee's comment in another thread. there's a ton of bias in the data like i, and others, have brought up. There is literally no bias. You don't have to opt in, you have to opt in for your ID to show. It's easy to find games with 10 anonymous players. Dotabuff can still track the game and its "skill level". | ||
|
Comeh
United States18919 Posts
Would be interesting to see this relationship. | ||
|
BluemoonSC
SoCal8933 Posts
On February 21 2015 04:21 SKC wrote: There is literally no bias. You don't have to opt in, you have to opt in for your ID to show. It's easy to find games with 10 anonymous players. Dotabuff can still track the game and its "skill level". read further down. "High MMR players with like 2-4k games skew the numbers by a lot by playing much on a regular basis. Whereas Low MMR players are likely just casual players who play Dota every once in a while. Looking at Dotabuff's stats and considering that Very High starts at 3.7k, there's quite a decent chance at Valve's original MMR distribution (avg 2250) still holding true." | ||
|
SKC
Brazil18828 Posts
That's not bias in the data. That's the uncertainty from taken one kind of data (games played) and extrapolating it to another kind of data (number of players). The data itself is fine, the conclusion may be biased. But since the bias should go against the conclusion, it's not a big deal. | ||
|
Plansix
United States60190 Posts
| ||
|
Slomo
Germany7198 Posts
| ||
|
SKC
Brazil18828 Posts
On February 21 2015 04:27 Plansix wrote: This is almost as good as "where have all the terran's gone" thread. Where people would use SC2 ranks to prove terrans were the most oppressed race of SC2. All based on a data mining site and extremely limited information. How many games don't show on Dotabuff? Hardly any. Even if it did, it wouldn't be biased towards any side of the argument. Server issues or any data collection issue would affect everyone. The data mining is not limited. | ||
|
teddyoojo
Germany22369 Posts
| ||
|
Plansix
United States60190 Posts
| ||
|
SKC
Brazil18828 Posts
On February 21 2015 04:29 Plansix wrote: The whole debate is silly because ranked match making is an opt in system. I have people on my friends list with 5000 games of dota who have not done their 10 placement matches because they give zero fucks. That and people can hide their information with the push of a button. It is not about ranked matchmaking. Where did you get that from? You cannot hide information. Where did you get that from? http://www.dotabuff.com/matches/1257903286 This is what happens when every single player "hides information". The data is still there. | ||
|
Plansix
United States60190 Posts
On February 21 2015 04:30 SKC wrote: It is not about ranked matchmaking. Where did you get that from? You cannot hide information. Where did you get that from? I am confused as to your argument then. What are you trying to prove? That the data is accurate? What meaning are we trying to discern from the data beyond that the matches happened. Edit: you are right, that match totally happened. | ||
|
SKC
Brazil18828 Posts
On February 21 2015 04:32 Plansix wrote: I am confused as to your argument then. What are you trying to prove? That the data is accurate? What meaning are we trying to discern from the data beyond that the matches happened. The data isnt biased. There is no reason to doubt the %s given. There is no reason to believe the conclusion that the average player should still be in the 2k range if 75% of the matches are made of under 3.1k players. Everyone is throwing around bullshit reasons why the data is not valid or their conclusions are not valid. I'm just trying to say they are bullshit. | ||
|
Plansix
United States60190 Posts
| ||
|
Howie_Dewitt
United States1416 Posts
Alright | ||
|
Sn0_Man
Tebellong44238 Posts
unless ur above at least 5.5k and even then thats doesn't by default make you good it just makes it a possibility | ||
|
Shaella
United States14828 Posts
| ||
| ||
