jmolvar10

How Marks of excellence really work

15 posts in this topic

HOW MARKS OF EXCELLENCE REALLY WORK
I know the secret of Marks of excellence there are many theories but most of the are wrong. First everyone who played a select tank in the last 14 days highest damage+damage assisted battle is taken, whether one battle was played by select player or many battles were played. Then all the highest damage+damage assisted battles by each player are taken and averaged. The averaged result is a 100% ratio for Marks of Excellence. Now for how this affects you. Your last 100 battles are taken and put on an exponential moving average (meaning your most recent battle counts more on the scale than a battle 100 battles ago). The formula takes into account Damage and Damage contributed. It is obviously possible to receive a mark in less than 100 battles if you are a above average player. So say for the M6 1 mark is probably around 700+ damage and assisted so if you do 700 damage and assisted combined your ratio will stay the same. If you do 0 damage and assisted at 65% your ratio will fall about (1.0%) and if you do 1400 damage and assisted your ratio will go up to 66% . Another point for the first battle of the day is your ratio could go down even if you have a good battle, because everyone else was playing better forcing your ratio to go down. It is way harder to 3 mark a tank where damage levels are inconsistent such as a spg. On Na server you will notice that there are way fewer spgs with 3 marks of excellence due to the the fact that it is easy to do a lot of damage in an spg but it is hard to have a good consistency.
Damage and Damage Assisted required by tier and type for marks Approximate. Note for spgs less is required Note for extremely rare tanks more is required.
!!APPROXIMATE VALUES!!
Tier 5
1 mark: 500 2 marks: 850 3 marks: 1100
Tier 6
1 mark: 750 2 marks: 1150 3 marks: 1400
Tier 7
1 mark: 1100 2 marks: 1600 3 marks: 2100
Tier 8
1 mark: 1500 2 marks: 2100 3 marks: 2600
Tier 9
1 mark: 1900 2 marks: 2650 3 marks: 3200
Tier 10
1 mark: 2700 2 marks: 3650 3 marks: 4500

Share this post


Link to post
Share on other sites

Generally what it takes to 3 mark a tank is what it takes to 2 mark the tank above it.

Tier 10 second mark req 3600

Tier 9 3 third mark req 3600

Tier 9 2 second mark req 2800

Tier 8 3rd mark req 2800

etc At least that's what I use for rough estimates

Share this post


Link to post
Share on other sites

Can't quite confirm those expected values, they must be for the hardest tank of that tier. 

I had my first mark on all tier 10s at about 2200 DPG, on the Obj 140 even at 1900.

Second mark on the FCM at 1900 DPG, on the T34 with 1700.

There must be some other factor. I heard that the frequency at which you played the last 100 battles also has an impact on the mark. 

Share this post


Link to post
Share on other sites
2 hours ago, Super_Rybak said:

I heard that the frequency at which you played the last 100 battles also has an impact on the mark. 

The DPGs for all players are constantly changing, you'll notice this especially when buffs/nerfs happen. An OP tank that you haven't played in a while but got nerfed since will go up on marks simply because of the DPG values going down, it's a 2week interval of data

I got my Sturer Emil 2nd mark literally the very next game after getting the first mark because the meta has crushed the tank so hard (haven't played it in over a year) since I finished the grind at exactly the first mark. 

 

While it says top 95% of the combined values, it's clearly more like 99% because of how the marks work. People having one good game and leaving it there (multiply this by a lot) will completely rape a mark req because amount of battles have no impact, only players. It's also been suggested that the DPG averages isn't what's setting the mark, but only battles considered "good" over a certain treshold because otherwise the numbers simply wouldn't add up. Blue players (95th percentile) hardly average more than 3k dpg in anything, let alone combined 4,5k for tier 10s as the marks "state" they should just by statistics

 

One thing I've noticed while grinding marks is that even if I'm constantly pulling the required damage per game and averaging it over 100 games I will not get the marks because the final push to 95% requires more than the average damage data on vBAddict. TVP T50/51 was at 4800 combined for the mark and that was my average, but it wouldn't move above 92% unless I started averaging above 5300 for the final part, meaning the marks are more about standalone good sessions than consistent play, arguably making them matter less than they do right now

Share this post


Link to post
Share on other sites

Yeah, I'm not sure I buy the OP assertion that even a zero damage game is counted if its the only battle played by a given player in the 14 day running average.  There has to be some threshold to weed out the bad games from the player pool.  There are a lot more bot level shitters dragging the 14 day average down then 4K dpg rockstars pulling it up.

Share this post


Link to post
Share on other sites
5 hours ago, EthanDuong said:

The theory is nice and all, but do you have any source or proof?

Yep i found about this theory on a wot xbox official site where they talked about it. And it also makes sense for many reasons. Firstly there are way less players with 3 marks on arties that is because in the last 14 days the highest battle for damage and damage assisted is very high, but because arties are so inconsistent it is very hard to keep a high damage and damage contributed average. It also makes sense because on the Na server there are only 2 players who have two marked the obj 260 even though there 289 owned on Na server. Based on Wot news Na server stats. This makes sense because it is so difficult. This is because so many good players have it, so the highest damage and damage assisted battles by each player in the last 14 days is very high.

Share this post


Link to post
Share on other sites
11 hours ago, Archaic_One said:

Yeah, I'm not sure I buy the OP assertion that even a zero damage game is counted if its the only battle played by a given player in the 14 day running average.  There has to be some threshold to weed out the bad games from the player pool.  There are a lot more bot level shitters dragging the 14 day average down then 4K dpg rockstars pulling it up.

Yes I am not sure if it is but according the article I read it is. I am not sure if there is a threshold but I think they just take the highest DAMAGE+DAMAGE CONTRIBUTED battle in the last 14 days from each player that played that select tank and then all these values are averaged. This averaged value is 100% ratio.

12 hours ago, Kolni said:

The DPGs for all players are constantly changing, you'll notice this especially when buffs/nerfs happen. An OP tank that you haven't played in a while but got nerfed since will go up on marks simply because of the DPG values going down, it's a 2week interval of data

I got my Sturer Emil 2nd mark literally the very next game after getting the first mark because the meta has crushed the tank so hard (haven't played it in over a year) since I finished the grind at exactly the first mark. 

 

While it says top 95% of the combined values, it's clearly more like 99% because of how the marks work. People having one good game and leaving it there (multiply this by a lot) will completely rape a mark req because amount of battles have no impact, only players. It's also been suggested that the DPG averages isn't what's setting the mark, but only battles considered "good" over a certain treshold because otherwise the numbers simply wouldn't add up. Blue players (95th percentile) hardly average more than 3k dpg in anything, let alone combined 4,5k for tier 10s as the marks "state" they should just by statistics

 

One thing I've noticed while grinding marks is that even if I'm constantly pulling the required damage per game and averaging it over 100 games I will not get the marks because the final push to 95% requires more than the average damage data on vBAddict. TVP T50/51 was at 4800 combined for the mark and that was my average, but it wouldn't move above 92% unless I started averaging above 5300 for the final part, meaning the marks are more about standalone good sessions than consistent play, arguably making them matter less than they do right now

DPG averages isn't setting the mark I stated that in my article. The mark is set by The highest damage+assisted battle by every player on that tank in the last 14 days. Then this number is averaged and this value is the 100% ratio.  It also makes sense because on the Na server there are only 2 players who have two marked the obj 260 even though there 289 owned on Na server. Based on Wot news Na server stats. This makes sense because it is so difficult. This is because so many good players have it, so the highest damage +damage assisted battles by each player in the last 14 days is very high.

Share this post


Link to post
Share on other sites
5 hours ago, Kolni said:

How about finding the proof and posting it here instead of argueing for your proof being correct without anything backing it up?

I was trying to get confirmation. How do you think it works? I dont really have hard proof.

Share this post


Link to post
Share on other sites
12 hours ago, Kolni said:

'I KNOW THE SECRET, THERE ARE MANY THEORIES BUT THEY ARE WRONG'

>no proof

though it does seem to make some sense.

it also accounts for the high requirements- everyone's damage is spikey so the top 5% of players will not consistently get the top 5% of results- even idiots have good games when everything goes right.

Share this post


Link to post
Share on other sites
On 10/26/2016 at 5:37 AM, jmolvar10 said:

Your last 100 battles are taken and put on an exponential moving average (meaning your most recent battle counts more on the scale than a battle 100 battles ago). The formula takes into account Damage and Damage contributed. It is obviously possible to receive a mark in less than 100 battles if you are a above average player.

I don't think WG keeps records of the games at all. There are no "last 100 games" records, or anything like that.

Your MoE value for a certain tank is just a moving figure. On a good game it goes up and on a bad game it goes down, based on DPG and assist/spot damage. If you hit a certain tank specific threshold, you get a mark.

I had first mark on T21 after 20 games and now I'm at 93.6% after 54 games. The better you play, the faster you get them. The "100 last games" is just fluff from WG.

 

Share this post


Link to post
Share on other sites

Create an account or sign in to comment

You need to be a member in order to leave a comment

Create an account

Sign up for a new account in our community. It's easy!

Register a new account

Sign in

Already have an account? Sign in here.

Sign In Now

  • Recently Browsing   0 members

    No registered users viewing this page.