Jump to content


What is WN7 rating?

WN7 Rating

This topic has been archived. This means that you cannot reply to this topic.
73 replies to this topic

sancheng #1 Posted Mar 25 2013 - 23:29

    Private

  • Players
  • 7684 battles
  • 1
  • Member since:
    11-09-2012
Which shows how good you are, efficiency or WN7? What's the difference and which one is a better indicator?

NavySnipers #2 Posted Mar 25 2013 - 23:40

    Major

  • Players
  • 20340 battles
  • 11,888
  • [LABS] LABS
  • Member since:
    10-18-2011
WN7 is far superior to eff.

AR_15 #3 Posted Mar 25 2013 - 23:41

    Sergeant

  • Players
  • 19568 battles
  • 233
  • [BULBA] BULBA
  • Member since:
    12-17-2011
Recently many people have asked about the WN rating. We recently switched from WN6 to the 7th version, which brought few but important changes. It’s better if we go to the beginning, so we can have a good picture of what WN is and isn’t.
The WN rating was first published by its author, Praetor77, on the NA forums on the 28th of November, 2012. The main goal is to eventually replace Efficiency, which is a widely flawed formula (prior to the recent update, which actually did not help much at all). The WN rating was created using advanced mathematics to create an accurate formula, using Win Rate as a proxy to accurately determine the weight each stat would have in the formula. The Efficiency formula was the basis for an analysis to figure out what was wrong with it and create an improved formula with those problems fixed.
WN is short for Weighed and Normalized, and implements various ways to deal with statpadding, and in the end it turned out the only way to pad your WN rating was to be actually good at the game.
Key points of the WN Rating are:
Damage is scaled according to your average tier and is the most important stat in the formula
Players with a considerable number of battles who have an average tier lower than 4 are heavily penalized for sealclubbing
Cap points are not counted towards your rating, since standing inside a circle requires no skill
To compensate for things like cap used to lure the enemy out and crucial decisions not recorded in the stats, the win rate counts for up to 10% of your rating
Average Defense points is capped at 2.2 to prevent padding
One of the most important characteristics of the WN Rating is the open development format, meaning any player can talk directly to Praetor through the thread above and suggest modifications, which will be tested and, if successful, implemented. By having this open development model, it effectively eliminates biases which closed formulas such as Efficiency have.
After the rating was released, WoTLabs was the first website to implement it, keeping it up to date every time a new version is launched. Although many people dislike the fast evolution of the rating (having gone through several changes and versions in only 4 months), this means that the formula rapidly grows more accurate. It has spread to the point where the WN Rating is set to become the standard rating used on the XVM mod, although it will probably come with some modifications.
Common Misconceptions
Despite having been based on advanced algorithms, the WN Rating did not pass without heavy criticism, although most, if not all, of this criticism turned out to be misconceptions.
One common complaint was that, if it was made to correlate with win rate, then we could just use win rate. Unfortunately win rate can be easily padded by platooning with good players. The WN Rating separates those padded players by using their actual stats. A veteran player with a low WN Rating but a high win rate has been heavily padded.
Another common criticism is that we should stop caring about statistics and just play the game, since statistics eventually lead to mockery. Unfortunately World of Tanks is a competitive game, and as in all competitive games, there must be a measuring stick in order to know if you are improving and how much you are improving, otherwise we might as well go play Farmville. Shooting tanks for the heck of it gets boring after a while.
Other players criticize the formula for not taking things like spotting damage into account. This can hardly be pinned on the WN Rating, since Wargaming has not released this information publicly. The WN Rating can only work with what it has available. To account for those invisible stats that help win games, Win Rate has been added to the formula.


source: http://blog.wotlabs.net/page/3/

This is all you need to know about WN7~

Edited by AR_15, Mar 25 2013 - 23:42.


DBrimstone #4 Posted Mar 25 2013 - 23:41

    First lieutenant

  • Beta Testers
  • 10258 battles
  • 597
  • Member since:
    12-31-2010
wn7 is a little bit better but the efficiency formula has been readjusted and is still a good indicator. The old efficiency formula used to put way too much emphasis on capping and resetting. The new one still emphasizes it more than wn7, so it is a little easier to pad it, but it's not bad. Look up the formula's, they're fairly easy to find.

You can use it to figure out different things you could improve on. Mine started to go up when I noticed that I don't survive enough and I don't reset cap near often enough. I started to focus on not dying needlessly and making sure to watch the mini-map a little more so I knew when to go reset cap. I started watching my WN7 about 1,000 games back, and living longer so I can get those last couple crucial kills and resetting to extend the game, has probably led to 30-40 more wins in that time. That's about a 3-4% increase in win rate from what I would have had. Note: overall win rate is 55%, last 1,000 games is about 59%

Overdile_Spittle #5 Posted Mar 25 2013 - 23:42

    First lieutenant

  • Players
  • 34113 battles
  • 580
  • [XOXO] XOXO
  • Member since:
    07-29-2011
Here is the thread you are looking for

http://forum.worldof...w-does-it-work/

DJYC21215 #6 Posted Mar 25 2013 - 23:43

    Captain

  • Players
  • 13275 battles
  • 1,229
  • Member since:
    07-18-2012
Wn7 Is superior, I run that in my XVM instead of Efficiency.

GollumsFish #7 Posted Mar 26 2013 - 00:41

    Staff sergeant

  • Players
  • 7401 battles
  • 388
  • [GANDP] GANDP
  • Member since:
    11-20-2012

View PostDJYC21215, on Mar 25 2013 - 23:43, said:

Wn7 Is superior, I run that in my XVM instead of Efficiency.

How do you do that? Does XVM offer the option or something?

Praetor77 #8 Posted Mar 26 2013 - 13:34

    Captain

  • Players
  • 29723 battles
  • 1,695
  • [BADGR] BADGR
  • Member since:
    05-01-2011
By default XVM uses WN7 but transformed into a scale of 0-99. It´s called "xwn". If you replace every instance of xwn in the config with "wn", you will see normal 0-3000 scale of WN7.

By the way the first post in the WN7 thread now holds a pretty detailed explanation of how it works.

EchoHntr #9 Posted May 16 2013 - 17:33

    Sergeant

  • Veteran Testers
  • 13485 battles
  • 117
  • [VPG] VPG
  • Member since:
    07-29-2011

View PostAR_15, on Mar 25 2013 - 23:41, said:

Key points of the WN Rating are:
Damage is scaled according to your average tier and is the most important stat in the formula
Players with a considerable number of battles who have an average tier lower than 4 are heavily penalized for sealclubbing
...
One common complaint was that, if it was made to correlate with win rate, then we could just use win rate. Unfortunately win rate can be easily padded by platooning with good players.

So isn't the unicum clans that do UC every night and do gobs of damage from the natural capabilities of the T10 tanks (surrounding themselves with unicum players) doing the same thing as those padding by platooning?  

I could never do the amount of damage solo in my T10 tanks in pubs I see on some of these players.  Granted I don't pretend to be as good as them, and I have much in my game to improve - thanks to stat feedback - but I suspect there's a bit of stat "gaming" going on as well just by "surrounding oneself with good people".   I suppose this is no different than what professional sports teams do as well....

Rexxie #10 Posted May 16 2013 - 18:08

    Major

  • Players
  • 29224 battles
  • 4,292
  • [PBKAC] PBKAC
  • Member since:
    03-05-2011

View PostEchoHntr, on May 16 2013 - 17:33, said:

So isn't the unicum clans that do UC every night and do gobs of damage from the natural capabilities of the T10 tanks (surrounding themselves with unicum players) doing the same thing as those padding by platooning?  

I could never do the amount of damage solo in my T10 tanks in pubs I see on some of these players.  Granted I don't pretend to be as good as them, and I have much in my game to improve - thanks to stat feedback - but I suspect there's a bit of stat "gaming" going on as well just by "surrounding oneself with good people".   I suppose this is no different than what professional sports teams do as well....

I don't quite agree - the better players I surround myself with, the worst damage I'm actually going to dish out. Surrounding yourself with unicums (like in CW) is going to be very painful for anything outside of winrate. As someone who performs are unicum level myself, my best bet at astronomical damage numbers is actually the opposite of what you're suggesting - I want my teammates to be just barely capable. The worse they do, the more slack I can pick up, and the higher my damage counter will roll.

Ralinos #11 Posted May 17 2013 - 15:18

    Corporal

  • Players
  • 19131 battles
  • 52
  • [WA] WA
  • Member since:
    10-06-2011

View PostAR_15, on Mar 25 2013 - 23:41, said:

source: http://blog.wotlabs.net/page/3/

This is all you need to know about WN7~

good to know the WN rating is flawed too.

1)W/L is not a good basis for stats B/C you have no control over the other 14 people on the team.
2)furthermore penalizing people for playing low tier tanks is just petty.
3)buying your way to the top IS stat padding.
also Capping is a good strategy, & it raises the overall EXP for the whole team( so does having more tanks alive at the end of battle.).

But i Guess(now KNOW) you would like to penalize stats for ARTY players too. :angry:

Edited by Ralinos, May 17 2013 - 17:02.


XMdead #12 Posted May 17 2013 - 16:23

    Major

  • Players
  • 34618 battles
  • 3,423
  • [LIMBO] LIMBO
  • Member since:
    07-14-2012
@OP:

WN7 is just another measure of skill.  

WN7 is wildly better than the old Efficiency formula and about equivalent in quality as the new Efficiency formula.

Both WN7 and Efficiency can be stat-padded.  At this time I like Efficiency (new formula) a little better, as it measures player actions that help the team win a little better.  

Things I don't like about WN7 and I hope will some day get addressed:
  • The penalty for average tier under 3 is good, but it also has a smaller penalty at tiers 5-6, that strikes me as wrong.  Optimal WN7 can be achieved by playing tiers 8-10... but that means that premium players (since they are the ones that can afford playing at those tiers) get a boost that has nothing to do with skill.  And yes, I have premium and confirmed this when I was doing the OnTrack for IS7 instead of grinding all the "not yet elited" tanks in my garage... my WN7 skyrocketed compared to its previous numbers.
  • The WR component.  This was good 6 months ago when MM was dishing out teams that were fairly balanced on skill and it was reasonable for most non-unicum good players to "carry" the team.  But recently MM is dishing out teams that are so unbalanced that unless you are a unicum you have little hope of carrying the team to victory.  So now you see those players almost exclusively platooning or doing tank companies to maintain their win rates.  Once you are good enough and get accepted into a good TC team... you win most games.  Skews the numbers.
  • Doesn't bother me as much, but WN7 weighs damage heavily. Premium rounds significantly increases damage per game.  This allows people to pay for better stats, whether by directly buying the rounds with gold or by having a premium account that allows them to get it with credits.

BTW, since you are looking for the better measure of skill... look at 60-day numbers, not at the overall.  60-day stats (whether Efficiency or WN7) are a far better measure.

Edited by XMdead, May 17 2013 - 16:35.


CCCPIvan #13 Posted May 17 2013 - 16:34

    Staff sergeant

  • Players
  • 11724 battles
  • 412
  • Member since:
    05-09-2011
I run wn7 in my xvm, I have found it is a better metric of skill.  Which I have seen directly correlate with on battle performance better than eff ever did.

CCCPIvan #14 Posted May 17 2013 - 16:36

    Staff sergeant

  • Players
  • 11724 battles
  • 412
  • Member since:
    05-09-2011

View PostPraetor77, on Mar 26 2013 - 13:34, said:

By default XVM uses WN7 but transformed into a scale of 0-99. It´s called "xwn". If you replace every instance of xwn in the config with "wn", you will see normal 0-3000 scale of WN7.

By the way the first post in the WN7 thread now holds a pretty detailed explanation of how it works.

Is that in the config files, in res mods/8.5?

Or do I have to open it up with something like "notepad++" (some name like that) and change the file directly?

edit: I have it on the scale of 0-99 but I would prefer it to have the full number instead to gauge the skill level better.

Edited by CCCPIvan, May 17 2013 - 16:37.


Ralinos #15 Posted May 17 2013 - 17:02

    Corporal

  • Players
  • 19131 battles
  • 52
  • [WA] WA
  • Member since:
    10-06-2011

View PostXMdead, on May 17 2013 - 16:23, said:


Premium rounds significantly increases damage per game.  This allows people to pay for better stats, whether by directly buying the rounds with gold or by having a premium account that allows them to get it with credits.

BTW, since you are looking for the better measure of skill... look at 60-day numbers, not at the overall.  60-day stats (whether Efficiency or WN7) are a far better measure.

Don't need premium acct to buy premium ammo w/credits. although that being said I think premium ammo is dumb. just make it the standard, since it doesn't need actual money to get anymore. (though you could always use clan gold before, if available)  

&

my opinion on Stats is: Take a cue from CREDIT REPORTS look at all variations & use the worst score you can find as your baseline for improvement. B/C that is the one "XVM harpers" will use to denigrate you. but if you don't use XVM, keep this in mind "the ones who harp on Stats are usually the ones who are only marginally better than you at best."

Edited by Ralinos, May 17 2013 - 17:04.


XMdead #16 Posted May 17 2013 - 19:11

    Major

  • Players
  • 34618 battles
  • 3,423
  • [LIMBO] LIMBO
  • Member since:
    07-14-2012

View PostRalinos, on May 17 2013 - 17:02, said:

Don't need premium acct to buy premium ammo w/credits. although that being said I think premium ammo is dumb. just make it the standard, since it doesn't need actual money to get anymore. (though you could always use clan gold before, if available)  
You don't need premium acct to buy the premium ammo... if you are only buying a little.  To use it consistently, you do.

But I agree with you, the split of regular and premium ammo is bad for the game.  I hardly ever use premium ammo in pubs... I normally don't need it, so why pay for it?

Edited by XMdead, May 17 2013 - 19:11.


IanSane #17 Posted May 20 2013 - 04:24

    Corporal

  • Players
  • 15472 battles
  • 36
  • [MYTH_] MYTH_
  • Member since:
    11-10-2011
I have a strong suggestion...guy was bragging about his unicom status both efficiency and WN7...checked his stats....the only tank he has higher than Tier 5 is a Tier 8 which has a bad losing win %.  Easily 2/3rds or more of his battles are in a tier 2 tank which undoubtedly uses gold, all crew at 300% and mods which the average tier 2 player running 50% crews and few or useless mods couldn't hope to defend against.. lets also forget that every player has the maps memorized which newer players wouldn't.  THIS is the kind of player that needs to be erased completely from the ratings.  When you go into a tier 8 and above match 90% of players have at least a 100% crew and know which modules to use so its a relatively level playing field.  Kicking puppies simply to boost your stats is pathetic and should be discouraged.  This guy has never played a single match in 7k battles higher than in a tier 8 and has zero tanks between tier 5 and 7 and has driven just 1 tier 8 (must have free xp'd to it) yet this guy is globaly rated 300 and has a WN7 rating of 1800+ and an efficiency rating >2000.

My suggestion: Give all tanks < Tier 4 zero weight towards affecting stats in any way.  In other words don't even record their stats in sub tier 4 tanks.

Ralinos #18 Posted May 21 2013 - 03:35

    Corporal

  • Players
  • 19131 battles
  • 52
  • [WA] WA
  • Member since:
    10-06-2011

View PostIanSane, on May 20 2013 - 04:24, said:

My suggestion: Give all tanks < Tier 4 zero weight towards affecting stats in any way.  In other words don't even record their stats in sub tier 4 tanks.

excellent suggestion.  +125 lol

unless you meant "greater than" tier 4 not "less than" tier 4.
I get those two symbols confused, so i try not to use them.  :teethhappy:

Edited by Ralinos, May 21 2013 - 03:37.


joelmomma #19 Posted Jun 16 2013 - 18:14

    Sergeant

  • Players
  • 14503 battles
  • 242
  • [DIH] DIH
  • Member since:
    03-10-2013
Here is what I found to be the WN7 formula, according to vbaddict.net:

Formula:
    FRAGS*(350.0 - TIER*20.0)
    +DAMAGE*(0.2 + 1.5 / TIER)
    +(DAMAGE_ASSISTED_RADIO/2)*(0.2 + 1.5 / TIER) (where available, e.g. battle report)
    +(DAMAGE_ASSISTED_TRACK/2)*(0.2 + 1.5 / TIER) (where available, e.g. battle report)
    +200.0*SPOT
    +15.0*CAP
    +15.0*DEF

This formula could use a lot of improvement.  As a Master Degree holder and statistics tutor, I would reccomend a formula without arbitary ratio assignements.  A formula based in standard deviations would be far more accurate.  For example, as in classroom performance each level of course work (or Tier in Wot) would have different difficulties.  Also, each class in a level ( or tank in a tier of Wot) would have different difficulties.  Standard deviations would show which individuals truly excel above and below the norms ( for example 3 standard deviatons above the normal is truly extraordinary).  In a class where everyone gets an "A" grade, 96% performance doesn't mean much at all because it is the normal performance.  In a class like organic chemistry where everyone gets 75%, a grade of 98% would be truly extraordinary.  By identifying standard deviations and using them in a formula you can accurately compare apples to apples, thus defining performance of the extraordinary players.  The current formula is cognitive, but it is as reliable as flipping a coin for judging true performance.  PM me and I would be happy to assist with writing a formula.  Regards!

Red_Ensign #20 Posted Jun 16 2013 - 18:19

    Major

  • Players
  • 2800 battles
  • 2,357
  • Member since:
    02-14-2013

View PostNavySnipers, on Mar 25 2013 - 23:40, said:

WN7 is far superior to eff.

kinda meaningless if you're low tier though.  from what I understand, the fact of being low tier alone gives you a crappy wn7 number (unless you drive a t18 maybe).