Jump to content
Bavor

Question about WN8 and percentiles

Recommended Posts

When I looked at the front page of Wotlabs, it lists a 450 WN8 as the 6th percentile.  So that means that 6% of the players have a 450 WN8 or lower if I'm interpreting that correctly.  However, I noticed that when I did screenshots of the players XVM WN8 stats in my 50 most recent recent battles that were at various times of the day the sub 450 WN8 players made up a lot more than 6% of the player population in tiers 6 through 10.  The sub 450 WN8 players varied from 1 player of the 30 to 13 players of the 30 in the battles.  Most battles had 3 to 8 sub 450 WN8 players in them.  3 of 30 is 10%, 6 of 30 is 20% 8 of 30 is 26.67%.  This seems a lot higher than the 6% listed on the front page.

Is the distribution of WoT players not a normal distribution?  Are the percentiles wrong on the front page of Wotlabs?  Am I interpreting percentiles wrong?  Is 50 battles not a large enough sample size to look at distribution of players?  Does the WN8 info front page of Wotlabs mean that if you are a 450 WN8 you are better than 6% of the players and if you have a 300 WN8 you are better than 0% of the players or does it mean that if you are a 450 WN8 you are better than 20% of the players and if you have a 300 WN8 you are better than 6% of the players?

The reason I looked at this is it seemed to me that in tiers 8 through 10 I've been seeing a lot more sub 450 WN8 players lately.  I was going to compare old screenshots to new ones to see if there was an actual difference.

Link to post
Share on other sites
1 hour ago, Bavor said:

Is the distribution of WoT players not a normal distribution?  Are the percentiles wrong on the front page of Wotlabs?  Am I interpreting percentiles wrong?  Is 50 battles not a large enough sample size to look at distribution of players?

All four actually. That distribution includes a step where all players below 145 WN8 were discarded, and 6% was the mark for the lower bound (300), not the upper bound (450). WN8 isn't a normal distribution, although it would be fairly close if it was contribution-linear.

The observation method won't match anyway except by coincidence, because the percentiles are players rather than player-battles (good players tend to play more battles), and tier distributions are substantially different.

Real percentile for <300 overall WN8 given 2k minimum battles and 30-day activity is 10%, for what it's worth (not much).

Link to post
Share on other sites
On 2/29/2016 at 5:36 PM, RichardNixon said:

All four actually. That distribution includes a step where all players below 145 WN8 were discarded, and 6% was the mark for the lower bound (300), not the upper bound (450). WN8 isn't a normal distribution, although it would be fairly close if it was contribution-linear.

The observation method won't match anyway except by coincidence, because the percentiles are players rather than player-battles (good players tend to play more battles), and tier distributions are substantially different.

Real percentile for <300 overall WN8 given 2k minimum battles and 30-day activity is 10%, for what it's worth (not much).

Thanks that clears it up.  Knowing that the percent shown is the low end also clears things up.

Link to post
Share on other sites
On 29 февраля 2016 г. at 11:59 PM, Bavor said:

When I looked at the front page of Wotlabs, it lists a 450 WN8 as the 6th percentile.  So that means that 6% of the players have a 450 WN8 or lower if I'm interpreting that correctly.

It depends on which ones the players we are considering. This graph for all RU players, for 3 different groups: 1000+ battles, 5000+ battles, 10000+ battles:

56e44b0445579_wn81k-10k.png.5f0888aeb5a6

For wn8 <= 450 we see ~33% players with 1000+ battles, ~10% players with 5000+ battles and ~4% players with 10000+ battles.

If you play on high tiers tanks, you see in battles more players with more battles and so with higher ratings.

 



 

On 29 февраля 2016 г. at 11:59 PM, Bavor said:

Is the distribution of WoT players not a normal distribution?

Graphs for 1000+ battles players:

56e44ee5bc608_raspdiff1.png.5f827c4084f8

56e44ee6eecb6_raspdiff2.png.0688bd6d0c30

"pr" - "Performance Rating" from noobmeter.com

"bs" - rating from armor.kiev.ua

 

Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Recently Browsing   0 members

    No registered users viewing this page.

×
×
  • Create New...