Jump to content


How MoE is calculated


  • Please log in to reply
33 replies to this topic

LordMuffin #1 Posted 09 January 2017 - 02:52 PM

    General

  • Player
  • 44842 battles
  • 9,298
  • [-MM] -MM
  • Member since:
    06-21-2011
Since weenis claimed MoE was calculated in a different way then I thought, I tried to find support to this in some ways. So here is what I found.
 
There seems to be a few different ideas how to calculate MoE.
 
1: The Wotlabs formula. being this one:
Supported on Wotlabs forums: 
http://forum.wotlabs...xcellence-work/
 
And are supposed to have a source somewhere within WG according to this:
http://forum.wotlabs...#comment-269863
And especially this quite:
Spoiler
But I can't get those 2 links to that forum to work. So I can't see the first hand information.
 
2: WG website that claims it is the combined damage and spotting.
Spoiler
3: Then we have the console version (which should be same as PC version.
Relevant part:
Spoiler
Claiming it is average damage + assisted damage. This article makes no difference between spotting and tracking. Thus making me think it just combines them. As it is never mentioned, and that whole article is quite thorough.
 
4: But I did, of course, ask support aswell what they thought.
And this is what happened there:
 
My question:
Spoiler

first answer.

Spoiler

I did of course reopen, since I don't really trust WG support that much in these matters mentioning that other WG source at WGNA and the wotlabs formula

Spoiler

Answer:

Spoiler

Which didn't really provide much anyway, but I reopened again, since he gave an opportunity to get a specialist on this.

Spoiler

Mentioning wotlabs again and their take on it. While also casually namedropping dakillzor, since he is in favour of wotlabs formula AND is working for WG in some way with commentating (I think).

Answer:

Spoiler

 

So WG support are supporting the idea that it is average damage and assisting (which is a combined value of spotting and tracking).
 
I also tried to get my latest answer to wotlabs forums to see what they would say. But my account there can't create threads nor post in existing ones.
 
My humble opinion on how MoE is calculated is now same as WG support and that claimed specialist of WG.

 


Edited by LordMuffin, 09 January 2017 - 03:11 PM.


Moppedhupe #2 Posted 09 January 2017 - 02:53 PM

    Lieutenant

  • Player
  • 27332 battles
  • 1,694
  • [PZK] PZK
  • Member since:
    03-20-2013
:popcorn:

brumbarr #3 Posted 09 January 2017 - 03:04 PM

    Brigadier

  • Player
  • 35143 battles
  • 4,540
  • [GO-IN] GO-IN
  • Member since:
    07-30-2012
Educate the masses muffin! My guess is you are using the forum post on wotlabs as a source? or you did your own calculations?

LordMuffin #4 Posted 09 January 2017 - 03:09 PM

    General

  • Player
  • 44842 battles
  • 9,298
  • [-MM] -MM
  • Member since:
    06-21-2011

View Postbrumbarr, on 09 January 2017 - 03:04 PM, said:

Educate the masses muffin! My guess is you are using the forum post on wotlabs as a source? or you did your own calculations?

 

done

 

I wrote everything down first. pressed post. and only a small part of what I wrote was posted. so I had to edit and redo everything. But now it seems to work.

 


Edited by LordMuffin, 09 January 2017 - 03:10 PM.


Press2ForSkill #5 Posted 09 January 2017 - 03:27 PM

    Warrant Officer

  • Player
  • 10961 battles
  • 689
  • [E-5OM] E-5OM
  • Member since:
    06-01-2016
nice, al though WG support is pretty damn stupid, but i trust that "specialist" i guess ;D just a few days ago i asked if my record experience in a battle on my service record is influenced by a premium account or not, and they clearly said "nope"... but just today i did ace tanker on my T29 and this was my record exp 2300+ somewhere exp WITH A PREMIUM ACCOUNT! and this is the exp it shows at my record score also, so support was stupid. speaking of T29, i just got my 2nd mark after 133 battles on it :) and i use the 90mm gun and not the 105mm, but good to know this formula, hopefully the specialist aint stupid :D

Joggaman #6 Posted 09 January 2017 - 03:28 PM

    Major General

  • Player
  • 20743 battles
  • 5,265
  • [WJDE] WJDE
  • Member since:
    05-23-2011
Thx for sharing!

LordMuffin #7 Posted 09 January 2017 - 03:30 PM

    General

  • Player
  • 44842 battles
  • 9,298
  • [-MM] -MM
  • Member since:
    06-21-2011

View PostTerblanche77, on 09 January 2017 - 03:27 PM, said:

nice, al though WG support is pretty damn stupid, but i trust that "specialist" i guess ;D just a few days ago i asked if my record experience in a battle on my service record is influenced by a premium account or not, and they clearly said "nope"... but just today i did ace tanker on my T29 and this was my record exp 2300+ somewhere exp WITH A PREMIUM ACCOUNT! and this is the exp it shows at my record score also, so support was stupid. speaking of T29, i just got my 2nd mark after 133 battles on it :) and i use the 90mm gun and not the 105mm, but good to know this formula, hopefully the specialist aint stupid :D

 

You have to reopen a few times, you can now reopen that ticket with your evidence and see what happens.

brumbarr #8 Posted 09 January 2017 - 03:31 PM

    Brigadier

  • Player
  • 35143 battles
  • 4,540
  • [GO-IN] GO-IN
  • Member since:
    07-30-2012
hmm, that is weird, since I see from experience  that tracking doesnt count. If I have a game with 2000 dmg 1000 spot and 1000 track, it doesnt go up as much as if I would have gotten 2000 dmg + 2000 spot. I see it especially well when the tracking makes the difference in going up or down,lets say you wanan 3 mark the BC,  3K dmg and you get 1K spot and 1K track, I noticed you go down in this case , which shouldnt happen it they where all counted. 
My experience is 3 marking 26 tanks, so I have seen a lot of those numbers.

I still believe in the wotlabs formula.

LordMuffin #9 Posted 09 January 2017 - 03:47 PM

    General

  • Player
  • 44842 battles
  • 9,298
  • [-MM] -MM
  • Member since:
    06-21-2011

View Postbrumbarr, on 09 January 2017 - 03:31 PM, said:

hmm, that is weird, since I see from experience  that tracking doesnt count. If I have a game with 2000 dmg 1000 spot and 1000 track, it doesnt go up as much as if I would have gotten 2000 dmg + 2000 spot. I see it especially well when the tracking makes the difference in going up or down,lets say you wanan 3 mark the BC,  3K dmg and you get 1K spot and 1K track, I noticed you go down in this case , which shouldnt happen it they where all counted. 
My experience is 3 marking 26 tanks, so I have seen a lot of those numbers.

I still believe in the wotlabs formula.

 

I don't like it either. That WG specialist claims one thing anf Wotlabs another.

 

I have 3marked a fare share of tanks.

And to me it seems the different jumps to similar result is more due to which result gets removed in the other end.

I think I have gained more MoE points for a low dmg battle+combined then a higher. Also that I gain different amount of MoE for roughly same dmg. When I 3 marked ISU it could increase by a different amount for same 3k dmg 0 assist battles. 



Etre_ #10 Posted 09 January 2017 - 03:59 PM

    Second Lieutenant

  • Player
  • 45172 battles
  • 1,014
  • [WEBOB] WEBOB
  • Member since:
    04-21-2014

Well , assist is a combination of tracking and spotting. Skill4ltul just sums damage and assisted. I don't know if for convenience or he really thinks it works like that. 

 

 

In my opinion, has to be a formula derived from the one calculating the XP.

 

That is:

- damage done on your own spotting is counted 100%

- damage done on others spotting is counted just 50%

- assist damage is counted 50%

 

Kills don't matter. 


Edited by Etre_, 09 January 2017 - 04:29 PM.


weenis #11 Posted 09 January 2017 - 03:59 PM

    Warrant Officer

  • Player
  • 25683 battles
  • 708
  • Member since:
    09-02-2011

View PostLordMuffin, on 09 January 2017 - 02:47 PM, said:

 

I don't like it either. That WG specialist claims one thing anf Wotlabs another.

 

I have 3marked a fare share of tanks.

And to me it seems the different jumps to similar result is more due to which result gets removed in the other end.

I think I have gained more MoE points for a low dmg battle+combined then a higher. Also that I gain different amount of MoE for roughly same dmg. When I 3 marked ISU it could increase by a different amount for same 3k dmg 0 assist battles. 

 

Interesting stuff. As mentioned I still stand behind the wotlabs formula simply because if it was both spotting + damage, certain tanks could easily require 6k combined.

 

As for the same result increasing or decreasing your MOE, this is easily explained as it's comparing it not only to the game you played, but also the game you played exactly 100 battles ago (as it's a rolling average). This is of course what you already mentioned, but let's consider this: 

 

Let X be the first game in your series (whether it's your 1st, 100th or 1000th battle in a certain tank). Let's ignore assistance for this example.

1)Game X            0 damage

2)Game X+1        3k damage

3)Game X+99      4k damage

4)Game X+100    4k damage

 

The 4k damage you did in 3) will have a much greater impact on your MOE progress than the 4k you did in 4). This is because when averaged, it's a significant increase. This also works vice versa with bad results of course.


Edited by weenis, 09 January 2017 - 04:01 PM.


UrQuan #12 Posted 09 January 2017 - 04:01 PM

    Brigadier

  • Player
  • 17631 battles
  • 4,970
  • [T-D-U] T-D-U
  • Member since:
    08-19-2011

View PostLordMuffin, on 09 January 2017 - 03:47 PM, said:

 

I don't like it either. That WG specialist claims one thing anf Wotlabs another.

 

I have 3marked a fare share of tanks.

And to me it seems the different jumps to similar result is more due to which result gets removed in the other end.

I think I have gained more MoE points for a low dmg battle+combined then a higher. Also that I gain different amount of MoE for roughly same dmg. When I 3 marked ISU it could increase by a different amount for same 3k dmg 0 assist battles. 

 

There is something fishy on the 3 mark calculation tho. Even a 0 damage game (no spotting/damage!) can grant you a 3rd MoE, even tho that should be impossible.

I know the adage of 'Get proof or Get out!' so watch this play of Circon for the Hellcat 3rd mark: (6 min vid)

For those who think Unicums never fail, I warn you, this video will dispel that myth.


Edited by UrQuan, 09 January 2017 - 04:03 PM.


weenis #13 Posted 09 January 2017 - 04:04 PM

    Warrant Officer

  • Player
  • 25683 battles
  • 708
  • Member since:
    09-02-2011

View PostUrQuan, on 09 January 2017 - 03:01 PM, said:

 

There is something fishy on the 3 mark calculation tho. Even a 0 damage game (no spotting/damage!) can grant you a 3rd MoE, even tho that should be impossible.

I know the adage of 'Get proof or Get out!' so watch this play of Circon for the Hellcat 3rd mark:

 

If you haven't played in a long time, the values for a third mark might go down significantly. With the hellcat in particular, it obviously got nerfed quite hard, so it's expected that the rolling average you need is lower now than it was when it was OP.

 

Therefore, if Circon hadn't played it in a while, he might already have an average (based on last 100 games if WG can be believed) over the required % for a third mark. Thus a 0damage,0assist game can give you the third mark.



LordMuffin #14 Posted 09 January 2017 - 04:13 PM

    General

  • Player
  • 44842 battles
  • 9,298
  • [-MM] -MM
  • Member since:
    06-21-2011

View Postweenis, on 09 January 2017 - 03:59 PM, said:

 

Interesting stuff. As mentioned I still stand behind the wotlabs formula simply because if it was both spotting + damage, certain tanks could easily require 6k combined.

 

As for the same result increasing or decreasing your MOE, this is easily explained as it's comparing it not only to the game you played, but also the game you played exactly 100 battles ago (as it's a rolling average). This is of course what you already mentioned, but let's consider this: 

 

Let X be the first game in your series (whether it's your 1st, 100th or 1000th battle in a certain tank). Let's ignore assistance for this example.

1)Game X            0 damage

2)Game X+1        3k damage

3)Game X+99      4k damage

4)Game X+100    4k damage

 

The 4k damage you did in 3) will have a much greater impact on your MOE progress than the 4k you did in 4). This is because when averaged, it's a significant increase. This also works vice versa with bad results of course.

 

I know. This is why looking at you % at a per-game basis and trying to estimate if it is both tracking and assist or just higher value that is counted is hard. Since you don't know which game you remove from your sample.

 

If I could get to that source wotlabs have, the WGNA guy. It would be very good. Or at least get to read whar he posted on NA forums. But I don't get access there.

 

 

It is disturbing to me that 2 reliable sources are claiming two different andwerd. 

 



Long_Range_Sniper #15 Posted 09 January 2017 - 04:23 PM

    Lieutenant General

  • Player
  • 25280 battles
  • 6,184
  • [T-D-U] T-D-U
  • Member since:
    04-04-2011

View Postweenis, on 09 January 2017 - 03:04 PM, said:

 

If you haven't played in a long time, the values for a third mark might go down significantly.

 

Which leads me to a question.

 

In the WoTlabs source LM has quoted in the OP, it states,

 

Generally, new battles are TWICE as important for the average as they would be if this was a straight average of the last 100 battles (thus the 2 factor in 2/101). As more battles accumulate the older ones start being less and less important. The 2 factor is used to make the average more responsive to changes, so you dont need to wait 100 good games to see an improvement, it is quite a bit more responsive than that. A 100 factor would make the newest battle as important as all your previous battles combined (which would suck if it is 0 game...). So a new battle that is good can balance 2 shitty older battles, and a shitty new battle cancels 2 great ones. 

 

A few great games in a row, without any bad games, would bump the average the most. However, this group of good games would be less and less useful the more you play after that. 

 

So if you buy a new tank, free XP to the top gun, and play 100 games shooting premium ammunition you've more chance of getting MOE's if that's your thing.

 



weenis #16 Posted 09 January 2017 - 04:27 PM

    Warrant Officer

  • Player
  • 25683 battles
  • 708
  • Member since:
    09-02-2011

View PostLong_Range_Sniper, on 09 January 2017 - 03:23 PM, said:

 

Which leads me to a question.

 

In the WoTlabs source LM has quoted in the OP, it states,

 

Generally, new battles are TWICE as important for the average as they would be if this was a straight average of the last 100 battles (thus the 2 factor in 2/101). As more battles accumulate the older ones start being less and less important. The 2 factor is used to make the average more responsive to changes, so you dont need to wait 100 good games to see an improvement, it is quite a bit more responsive than that. A 100 factor would make the newest battle as important as all your previous battles combined (which would suck if it is 0 game...). So a new battle that is good can balance 2 shitty older battles, and a shitty new battle cancels 2 great ones. 

 

A few great games in a row, without any bad games, would bump the average the most. However, this group of good games would be less and less useful the more you play after that. 

 

So if you buy a new tank, free XP to the top gun, and play 100 games shooting premium ammunition you've more chance of getting MOE's if that's your thing.

 

 

Exactly. But then there are other confusing factors as well, such as the ability to get the third mark in less than 100 games. I never looked too much into it, I just play to the best of my ability, and get the mark eventually, whether it's the wotlabs formula or a secret wg formula.

 

Would be nice if they knew what they were talking about though. The fact that it had to get escalated about 4 times before someone gave a generic answer as opposed to a formula leaves me suspicious.



tajj7 #17 Posted 09 January 2017 - 04:29 PM

    Field Marshal

  • Player
  • 22555 battles
  • 13,341
  • [-MM] -MM
  • Member since:
    03-30-2014

View PostUrQuan, on 09 January 2017 - 03:01 PM, said:

 

There is something fishy on the 3 mark calculation tho. Even a 0 damage game (no spotting/damage!) can grant you a 3rd MoE, even tho that should be impossible.

I know the adage of 'Get proof or Get out!' so watch this play of Circon for the Hellcat 3rd mark: (6 min vid)

 

It's a moving average though, so that is probably why it's 100 games, so every new game that is your 100th, that would replace your 1st game, which might also have been a zero damage game.

 

And like Weenis said, if you haven't played in a while the needed average could have dropped a lot and you could already be way above it but you just need to play a game to basically 'refresh' the system and get your marks as you can only hit 3 marks by playing a game.

 

It works in reverse as well, for example pre- 9.17 I had something like 84% on M4 Rev, I then played it a day or so after the patch went live, I had a game with I think about 3k ish damage some spotting, my MOE plummeted from 84% to 77% after that game, because obviously after the patch and the tank getting buffed everyone had rushed to play it, probably decent players and shafted the values so my updated average suddenly became a lot worse.

 

I'm also pretty sure (this is on WOTlabs) that it is basically your average V the BEST results averaged of other players playing the tank in the last 14 days, so you are basically playing against other people's best games averaged out, not all their games which is why it's so hard. 

 

 



L0ddy #18 Posted 09 January 2017 - 04:33 PM

    Captain

  • Player
  • 77554 battles
  • 2,325
  • [TWD] TWD
  • Member since:
    04-24-2011

View PostLordMuffin, on 09 January 2017 - 02:47 PM, said:

I have 3marked a fare share of tanks.

And to me it seems the different jumps to similar result is more due to which result gets removed in the other end.

 

 

 

I believe moe works in that way too, if you keep a damage & assist record after every battle for 100 battles then you'll see it happen in the figures.

 

:)



LordMuffin #19 Posted 09 January 2017 - 04:40 PM

    General

  • Player
  • 44842 battles
  • 9,298
  • [-MM] -MM
  • Member since:
    06-21-2011

View Postweenis, on 09 January 2017 - 04:27 PM, said:

 

Exactly. But then there are other confusing factors as well, such as the ability to get the third mark in less than 100 games. I never looked too much into it, I just play to the best of my ability, and get the mark eventually, whether it's the wotlabs formula or a secret wg formula.

 

Would be nice if they knew what they were talking about though. The fact that it had to get escalated about 4 times before someone gave a generic answer as opposed to a formula leaves me suspicious.

 

I will reopen again and ask if a formula could be had.

 

I tried a little bit to get a formula for pen and dmg RNG, but that was a secret formula not made for public. They only said it was a gaussian distribution.

 

Ususlly I play my best and eventuslly get 3 mar ks aswell. But I like to know how stuff works.  



eldrak #20 Posted 09 January 2017 - 04:43 PM

    Warrant Officer

  • Player
  • 42367 battles
  • 825
  • [RNG_M] RNG_M
  • Member since:
    01-27-2011

Who volunteers to play slightly over 100 battles with one tank during one day to test how it actually works?

 

Good thread btw Muffin, many thanks.






1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users