Jump to content


winrate is not a reflection of player skill


  • Please log in to reply
801 replies to this topic

cheapbooks #301 Posted Jun 07 2019 - 17:43

    Captain

  • Players
  • 39114 battles
  • 1,142
  • [CBKS] CBKS
  • Member since:
    08-21-2013

View Postgingerted91, on May 20 2019 - 04:56, said:


Ummm not according to the Hall of Fame in the game...  here is a quick summary

 


Ummm, maybe check that ego mate... looks like you are incorrect.
 


The Hall of Fame shows you under performed in comparison to the top players in three separate months and overall.
 

 

Yes, yes he is.

 

nice try, you sorted the values by a different metric than I did, to make it look like my numbers were wrong.

cheapbooks #302 Posted Jun 07 2019 - 17:44

    Captain

  • Players
  • 39114 battles
  • 1,142
  • [CBKS] CBKS
  • Member since:
    08-21-2013

this is a typical game for me. 0 damage, 0 assisted, 0 blocked by armor:

 

https://www.youtube.com/watch?v=zAZgFI1YRYk

 

yesterday I had a 60% winrate, so I guess I was super-unicum for a day.



1919Eternal #303 Posted Jun 07 2019 - 17:56

    Staff sergeant

  • -Players-
  • 59106 battles
  • 415
  • [PRO_S] PRO_S
  • Member since:
    07-17-2014
I stopped reading after the second point which directly contradicts your thesis. 

moogleslam #304 Posted Jun 07 2019 - 17:57

    Major

  • Players
  • 46421 battles
  • 4,983
  • [GURUS] GURUS
  • Member since:
    12-20-2013

View Postcheapbooks, on Jun 07 2019 - 12:44, said:

this is a typical game for me. 0 damage, 0 assisted, 0 blocked by armor:

 

https://www.youtube.com/watch?v=zAZgFI1YRYk

 

yesterday I had a 60% winrate, so I guess I was super-unicum for a day.

 

I watched this video.  If you have a 999 ping, that doesn't indicate a bad connection; that indicates NO connection!  You said in this video that battles like this, where you have a 999 ping are the reason for your bad stats and that you're not a bad player.  A ping so bad that you're not in control of your tank is obviously going to negatively impact your stats.  However, I've watched quite a few of your videos, and that's far from the only reason you have bad stats.  You also have bad stats because you are a bad player.  That is harsh, but that's just a fact.  Embrace it, and use it as a starting point to try and approve.  I was bad once, super super bad, and once I realized that, I figured out how to improve.

 

Like I've said elsewhere, you need to improve your map awareness, you need to learn where to go for the tank class you're playing, and then work toward learning where to go for the specific tank you're playing.  You need to learn to protect your tank, and not give the enemy free damage all the time.  You need to learn when to use autoaim, or more so, when not to use it.  You need to start taking high percentage shots.  You need to learn when to retreat.  



NeatoMan #305 Posted Jun 07 2019 - 17:58

    Major

  • Players
  • 28198 battles
  • 20,730
  • Member since:
    06-28-2011

"My ping time goes up to 999ms and my tank decides to turn and drive off a cliff. I end the game with 0 damage, 0 assistance damage, minimum exp. This happens to me often because I play from the Philippines to NA server over a cell network. Yes it affects my winrate."

 

But we can't see what hardware or whatever else you are using.  All we see are the end results.  You are that player with the bad connection and the results show it.  It doesn't matter what kind of talent you possess if you can't or won't bring it to bear in your games.  "Skill" is used to describe what you personally bring to battle every game.   Maybe you've taken offense to that term being used to describe it, but it that's pretty much how it applies.



SquishySupreme #306 Posted Jun 07 2019 - 18:08

    First lieutenant

  • -Players-
  • 3672 battles
  • 748
  • Member since:
    10-29-2018

View PostcKy_, on Apr 17 2019 - 13:23, said:

 

Better players perform better in battle and as a result better players win more often. They do not win 100% of the time, because no one player is good enough to carry every single team. Just like how you pointed out that AFK'ers or bots or whatever don't have a 0% winrate, they have a 38% winrate. This shows that despite the AFK player putting a massive handicap on the team, there sometimes are good players that can carry a 14v15 battle. 

 

"a player with 0 skill is still able to passively scout" Then why doesn't he? Oh yeah, because he isn't skilled enough at the game to know how to effectively passive scout, he might not even know what passive scouting is. 

 

I will confirm that passive scouting is incredibly hard to do, even if you know exactly what to do and where to go.

The real reason it doesn't work is statistics by definition is analyzing sets of data and determining patterns.
If you have a random player and tank list every match, only one death per match, and it is a different set of tanks and players every time, there is no set of data being generated.
You are basically flipping a coin - and as we know, past performance is in no way indicative of future results for singular events.

WN8 is much closer, though it is biased towards the number of matches you have played, and still incorporates win rates in the formula.

sargev55 #307 Posted Jun 07 2019 - 18:15

    Captain

  • Players
  • 60638 battles
  • 1,446
  • [_CIA_] _CIA_
  • Member since:
    02-22-2012

Currently I have accidentally gotten 3 or 4 tanks in the top ten in Hall of Fame, as well as 8 or nine in the top 100 on the server.  Those tanks also have good win rates, I wish I could figure out the correlation.



shaggy996 #308 Posted Jun 07 2019 - 18:28

    Captain

  • Players
  • 47773 battles
  • 1,506
  • [ALERT] ALERT
  • Member since:
    11-17-2012
So 16 pages of bad players avoiding responsibility for their own game play. 

_Tsavo_ #309 Posted Jun 07 2019 - 18:35

    Major

  • Beta Testers
  • 46490 battles
  • 19,868
  • [RELIC] RELIC
  • Member since:
    02-16-2011

View Postcheapbooks, on Jun 07 2019 - 11:44, said:

this is a typical game for me. 0 damage, 0 assisted, 0 blocked by armor

 

Finally, some honesty!



sargev55 #310 Posted Jun 07 2019 - 18:37

    Captain

  • Players
  • 60638 battles
  • 1,446
  • [_CIA_] _CIA_
  • Member since:
    02-22-2012

View Postshaggy996, on Jun 07 2019 - 17:28, said:

So 16 pages of bad players avoiding responsibility for their own game play.

 


Shaggy, apparently you were just super lucky with match making for 45k battles ;)

Blackstone #311 Posted Jun 07 2019 - 18:41

    Major

  • -Players-
  • 21876 battles
  • 2,268
  • Member since:
    04-19-2011

View Postcheapbooks, on Jun 07 2019 - 11:44, said:

this is a typical game for me. 0 damage, 0 assisted, 0 blocked by armor:

 

https://www.youtube.com/watch?v=zAZgFI1YRYk

 

yesterday I had a 60% winrate, so I guess I was super-unicum for a day.

 

GREAT! YOU'VE JUST WON...

Spoiler

Now, please stop. Your ignorance is showing....



SquishySupreme #312 Posted Jun 07 2019 - 19:13

    First lieutenant

  • -Players-
  • 3672 battles
  • 748
  • Member since:
    10-29-2018

View Postshaggy996, on Jun 07 2019 - 18:28, said:

So 16 pages of bad players avoiding responsibility for their own game play. 

 

I don't know about them, but the math simply does not work for random battles.  Other modes, like Clan Wars, possibly.
- Random people
- Random skill levels
- Random tanks
- Random map
- Random numbers of each tier tank/MM
- One death per game
- Results not carried over/no consecutive rounds.
- RNG added to shell damage and aiming that we don't know the exact formulas for.
- Match two teams of this against each other, find a winner.
- Assign a player's individual skill based on which team won.  <-- this part is not mathematically possible given the enormous amount of random variables.

I can look outside my window and see dozens of casinos built on people not understanding how statistics actually works, but perhaps that all is a mirage in the distance...

_Tsavo_ #313 Posted Jun 07 2019 - 19:18

    Major

  • Beta Testers
  • 46490 battles
  • 19,868
  • [RELIC] RELIC
  • Member since:
    02-16-2011

View PostSquishySupreme, on Jun 07 2019 - 13:13, said:

 

I don't know about them, but the math simply does not work for random battles.  Other modes, like Clan Wars, possibly.
- Random people
- Random skill levels
- Random tanks
- Random map
- Random numbers of each tier tank/MM
- One death per game
- Results not carried over/no consecutive rounds.
- RNG added to shell damage and aiming that we don't know the exact formulas for.
- Match two teams of this against each other, find a winner.
- Assign a player's individual skill based on which team won.  <-- this part is not mathematically possible given the enormous amount of random variables.

I can look outside my window and see dozens of casinos built on people not understanding how statistics actually works, but perhaps that all is a mirage in the distance...

 

It does, because you're measuring what difference a single variable has in a sea of ever-changing variables: the player.  There are upper and lower limits as you are only able to influence so much, but a single can -and does- have an impact.  That's been established for ages with this game.  The more consistent the player is in either positive or negative ways drives the win rate up or down from the 49% mean.

 

Everything else is just an excuse.



JohnnyReroll #314 Posted Jun 07 2019 - 21:16

    First lieutenant

  • Players
  • 9741 battles
  • 702
  • Member since:
    03-07-2013
Why even bother playing with 999 ping?  On rare occasions my internet flips out and I get 300 ping.  I stop playing for the night... 

_Tsavo_ #315 Posted Jun 08 2019 - 01:31

    Major

  • Beta Testers
  • 46490 battles
  • 19,868
  • [RELIC] RELIC
  • Member since:
    02-16-2011

View Postcheapbooks, on Apr 17 2019 - 14:46, said:

I posted what I posted because I am sick of people telling me that they are a better player than me simply because they have a higher winrate or higher WN8. that is just plain ignorance.

 

I am a better player than you because I have a higher win rate and wn8.  

 

That is just plain truth ;)



Canadian_Mano #316 Posted Jun 08 2019 - 01:36

    First lieutenant

  • -Players-
  • 35992 battles
  • 513
  • Member since:
    01-29-2012

View Post_Tsavo_, on Jun 07 2019 - 20:31, said:

 

I am a better player than you because I have a higher win rate and wn8.  

 

That is just plain truth ;)

 

playing better and winning more doesnt make you a better player, makes you a better hacker! uwu

Guido1212 #317 Posted Jun 08 2019 - 01:48

    Community Contributor

  • Players
  • 81941 battles
  • 8,922
  • [GFLC] GFLC
  • Member since:
    06-11-2011

View Postsargev55, on Jun 07 2019 - 17:15, said:

Currently I have accidentally gotten 3 or 4 tanks in the top ten in Hall of Fame, as well as 8 or nine in the top 100 on the server.  Those tanks also have good win rates, I wish I could figure out the correlation.

 

If I had to guess, and this is a wild assumption.  I'd say luck cheating and rigging, in no particular order.

Boghie #318 Posted Jun 08 2019 - 02:05

    Captain

  • -Players-
  • 8703 battles
  • 1,178
  • [WCTNT] WCTNT
  • Member since:
    04-10-2016

View Postsargev55, on Jun 07 2019 - 12:15, said:

Currently I have accidentally gotten 3 or 4 tanks in the top ten in Hall of Fame, as well as 8 or nine in the top 100 on the server.  Those tanks also have good win rates, I wish I could figure out the correlation.

 

I.  Just.  Can't.  Figure.  It.  Out.

 

But, I had a 75% WR yesterday.  I am truly amazing.  And, modest too:medal:

 


Edited by Boghie, Jun 08 2019 - 02:06.


sargev55 #319 Posted Jun 08 2019 - 04:39

    Captain

  • Players
  • 60638 battles
  • 1,446
  • [_CIA_] _CIA_
  • Member since:
    02-22-2012

View PostBoghie, on Jun 08 2019 - 01:05, said:

 

I.  Just.  Can't.  Figure.  It.  Out.

 

But, I had a 75% WR yesterday.  I am truly amazing.  And, modest too:medal:

 

 


Whats up Boghie, we need to play again at some point, platoon, tourneys, etc
 

sargev55 #320 Posted Jun 08 2019 - 04:48

    Captain

  • Players
  • 60638 battles
  • 1,446
  • [_CIA_] _CIA_
  • Member since:
    02-22-2012

View PostGuido1212, on Jun 08 2019 - 00:48, said:

 

If I had to guess, and this is a wild assumption.  I'd say luck cheating and rigging, in no particular order.

 


But of course you know better Guido 


nvm, sarcasm, too many beers sorry

unless, that was your assumption, omg paranoia  :)

Edited by sargev55, Jun 08 2019 - 04:59.





1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users