Jump to content


How does XVM efficiency rating work?

Confused

This topic has been archived.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

jediwomble #1 Posted 05 May 2012 - 06:23 PM

    Lance-corporal

  • Player
  • 26470 battles
  • 69
  • Member since:
    11-04-2011
Does anyone actually know how exactly the XVM efficiency rating is calculated?

The wot-news site states that it is based on the following critieria:
Average destroyed in the battle: Average damage in the battle: Average spotted in the battle: Average defense points in the battle: Average captured points in the battle:

But how do these numbers relate to the actual efficiency rating, in what propotion etc?

Also, the dossier app tool seems to include hit accuracy rate in its calculation of efficiency. Is this just a totally different rating system, or what?

Any help appreciated. Thanks.

Nectaria #2 Posted 05 May 2012 - 06:28 PM

    First Sergeant

  • Player
  • 9371 battles
  • 4,334
  • Member since:
    10-23-2011
Basically capping/defending > damage done > survivability/spotting > everything else.

It's not perfect, but you're 99% sure a green player will do good and a red player will do 99% bad.

Edited by Nectaria, 05 May 2012 - 06:29 PM.


TRUST_NATO #3 Posted 05 May 2012 - 06:37 PM

    Lance-corporal

  • Player
  • 10153 battles
  • 27
  • Member since:
    05-02-2011
It really is that simple, and has a surprisingly high accuracy level :Smile_great:

kwk75l48 #4 Posted 05 May 2012 - 06:55 PM

    First Sergeant

  • Beta Tester
  • 17827 battles
  • 2,642
  • [TDC] TDC
  • Member since:
    11-28-2010
Thats as clear as mud,  many believe, many don't, many cannot be bothered.

the mystical mathametical system is unknown to me.

but generally your rating will go up with more dmg more kills more cap and def points and more spots increasing you lowest type of score seems to boost it more.

dawdler #5 Posted 05 May 2012 - 07:03 PM

    Senior Sergeant

  • Player
  • 6103 battles
  • 755
  • Member since:
    05-09-2011

View PostNectaria, on 05 May 2012 - 06:28 PM, said:

It's not perfect, but you're 99% sure a green player will do good and a red player will do 99% bad.
And yet its all random at the same time. According to XVM, I am a mediocre player, not even close to the ones that brag about their 1500+, 2000+ rating.

Still I got ace tanker with the M3 Lee. HA, IN YOUR FACE XVM!!!

jkksir #6 Posted 05 May 2012 - 07:04 PM

    Sergeant

  • Beta Tester
  • 27285 battles
  • 273
  • [DAC] DAC
  • Member since:
    08-28-2010
Well i've heard that caping and defending pushes up your rating but i don't know for how much to be exact.For ex 3 days ago i had 1444 rating and yesterday it got up to 1449 and i didn't defend or caped that much,just the normal drill:kill some tanks and cap when it was necessary.But the thing is it is easier to raise your efficiency when you have a low number of games.I've see player with 17k-20k games that have under 700 rating and that is just plain stupididy ,those player realy suck and they dont have any excuse like ,,MM hates me'' or ,,I always get the bad teams'',heck everybody gets bad teams but i don't have that complex.

So in a nuttshell the efficiency rating is more influenced at a low number of games(caping ,damaging ,spotting ,anything) and with the rise of the number count of games it will be harder to pump it up.

jediwomble #7 Posted 05 May 2012 - 07:11 PM

    Lance-corporal

  • Player
  • 26470 battles
  • 69
  • Member since:
    11-04-2011
Replys so far confirm my suspicion: Many people use XVM, but no one actually knows exactly how it is calculated. Therefore its probably meaningless.

DAWDLER: Its no harder to get Ace tanker on the M3 Lee than any other tank. The acheivement awards are based on your ability with a tank, measured against all other users of that tank.

grayfoxCode #8 Posted 05 May 2012 - 08:28 PM

    Junior Sergeant

  • Player
  • 17528 battles
  • 160
  • [TAC] TAC
  • Member since:
    05-23-2011
Doesn't help with the calculation and you may already know this but here it is anyway.

http://wot-news.com/...p/stat/calc/en/

danjobsen #9 Posted 05 May 2012 - 08:35 PM

    Junior Sergeant

  • Player
  • 27325 battles
  • 122
  • [FSC-1] FSC-1
  • Member since:
    10-10-2011
found this somewhere (even people not knowing PHP can understand the formula):


$efficiency = ($stats['DAMAGE'] * (10 / $avg_tier) * (0.15 + 2 * $avg_tier / 100) + $stats['KILLS'] * (0.35 - 2 * $avg_tier / 100) * 1000 + $stats['SPOTTED'] * 200 + $stats['CAPTURE'] * 150 + $stats['DEFENSE'] * 150) / $battles;

Bobbydancer #10 Posted 05 May 2012 - 08:35 PM

    Senior Sergeant

  • Player
  • 39359 battles
  • 942
  • [GGG] GGG
  • Member since:
    04-05-2011
ER=(frags/#battles)x(350-(tier x 20)) + ((damage done/#battles)x(0,2+(1,5/tier))) + (200 x spotted/#battles) + ((capture % + Defence %) x 150 / #battles)

helli #11 Posted 05 May 2012 - 08:36 PM

    Sergeant

  • Player
  • 2333 battles
  • 355
  • Member since:
    02-06-2011

View Postjediwomble, on 05 May 2012 - 07:11 PM, said:

Replys so far confirm my suspicion: Many people use XVM, but no one actually knows exactly how it is calculated. Therefore its probably meaningless.

DAWDLER: Its no harder to get Ace tanker on the M3 Lee than any other tank. The acheivement awards are based on your ability with a tank, measured against all other users of that tank.

I looked around for the exact calculation too, didn't find it yet.

It's obvious people using xvm rate it as (sufficiently) accurate but I too am interested how exactly it's calculated.

But let's not confuse the average person with things like logic and inquisive scepticism. After all who heard of a undisclosed and/or, for all we know, pseudo scientific/mathematical method to sort people into bad and good ever causing problems?


Wait, why is my german guilt sense tingling all of a sudden???


P.S.: And yes Ace tank isn't something to brag about, sure some people never reach it but still all you need is just one lucky game where everything falls in place.

Bobbydancer #12 Posted 05 May 2012 - 08:49 PM

    Senior Sergeant

  • Player
  • 39359 battles
  • 942
  • [GGG] GGG
  • Member since:
    04-05-2011
Read up then, two guys knew it and replayed aat same time. Still since i know its formula i dont use the xvm any more. I have tried it earlier and i've also tested it's prediction value of the game outcome wich is inpossible to confirm statisticly. Teams with good EFF rating facing teams with poor just as often loose as opposite.

But why the hell care for efficiency rating and win rate as well, it's just bullsh. Why even care for winning the game. Winning is hardly rewarded in the game. Only damage done and taken has significant impact on you're game economy. Atleast when you are finished grinding. When you are grinding some of the xp factors come to relevans. But after you pass tier five it's credits you need and credits are aproxematly equal to damage.

helli #13 Posted 05 May 2012 - 09:43 PM

    Sergeant

  • Player
  • 2333 battles
  • 355
  • Member since:
    02-06-2011

View PostBobbydancer, on 05 May 2012 - 08:49 PM, said:

Read up then, two guys knew it and replayed aat same time.

Yeah, as I was typing.... slowly as always.... well that was the purpose of this thread wasn't it? :)

And I agree winrate + eff. is pointless, I'm just interested in the eff. calculation (what is weighted how much, the tier coefficient, etc....).

Urmo345 #14 Posted 06 May 2012 - 01:56 AM

    Senior Sergeant

  • Player
  • 6243 battles
  • 774
  • [EES4] EES4
  • Member since:
    04-05-2012
It is seriously flawed. Period. But we don't have better. Yet.

Edited by Urmo345, 06 May 2012 - 01:56 AM.


Greenwing #15 Posted 06 May 2012 - 02:18 AM

    Sergeant

  • Player
  • 13631 battles
  • 286
  • [GOT] GOT
  • Member since:
    01-29-2011

View PostBobbydancer, on 05 May 2012 - 08:49 PM, said:

Teams with good EFF rating facing teams with poor just as often loose as opposite.
I kept track of it (predict the outcome of a game before it started) and with the mod I can predict the outcome of about 80% of the games correctly (when I only use the start-screen without the mod - so only mm tank-selections it's about 60%)

XVM does not tell you if a player will play good in the current game, it does give a good idea though. A 45% player will on average make a lot more mistakes then a 55% player.

Bobbydancer #16 Posted 06 May 2012 - 10:01 PM

    Senior Sergeant

  • Player
  • 39359 battles
  • 942
  • [GGG] GGG
  • Member since:
    04-05-2011

View PostGreenwing, on 06 May 2012 - 02:18 AM, said:

I kept track of it (predict the outcome of a game before it started) and with the mod I can predict the outcome of about 80% of the games correctly (when I only use the start-screen without the mod - so only mm tank-selections it's about 60%)

XVM does not tell you if a player will play good in the current game, it does give a good idea though. A 45% player will on average make a lot more mistakes then a 55% player.

I realy like to look into you're data. Cause when i collected data i didn't find any prediction value at all.

Greenwing #17 Posted 07 May 2012 - 03:30 AM

    Sergeant

  • Player
  • 13631 battles
  • 286
  • [GOT] GOT
  • Member since:
    01-29-2011

View PostBobbydancer, on 06 May 2012 - 10:01 PM, said:

I realy like to look into you're data. Cause when i collected data i didn't find any prediction value at all.
I just used pen and paper so the list would actually be rather pointless (it's just a paper with a lot of dashes in a few colomns on it). I have to say I only collected the data using my IS-3 + T34 because at that level I know pretty accurately which (combination of) tanks are good and I also only used complete randoms (platooning will probably influence the data by a lot)

For both cases I used about 100 battles (which I know is rather low to get good predictions)

The reason for my higher numbers is probably because I don't only look at the stats or the tanks but also at the combination. A 45% player in an IS-7 can often be kept alive by following him and keeping his back safe. A 45% in a T34 will most probably die in 2 minutes ;)

Edited by Greenwing, 07 May 2012 - 03:31 AM.


Sharpe #18 Posted 07 May 2012 - 04:41 AM

    First Sergeant

  • Beta Tester
  • 18824 battles
  • 1,115
  • [FAME] FAME
  • Member since:
    09-11-2010
Don't bother with efficiency... it puts too much stock on capturing/defending the base points.

Caan #19 Posted 07 May 2012 - 09:49 AM

    Junior Sergeant

  • Player
  • 26259 battles
  • 154
  • [MOOSE] MOOSE
  • Member since:
    05-15-2011
It's an indication which is right 7 out of 10 times. But it's weighted to much towards capping/defending.

aletsan #20 Posted 07 May 2012 - 12:19 PM

    Sergeant

  • Player
  • 9680 battles
  • 433
  • Member since:
    05-03-2011
I don't care if eff stat is easily boosted using capping/defending.
In the end, these are only 2 ways for a win.
And sometimes, particularly in the case of small fast light tanks, they reflect real situation.

The only thing that I don't like is that it only shows the total stats.
I would prefer to know the stats of a player is with the tank he brought in battle.
His total stats is an added bonus.

It's good to know that a player has 5K games, 55% win ratio, and 1200 eff.
But if I know that he only played a particular tank only 10 games would really help me adjust my play.