Jump to content
You need to play a total of 50 battles to post in this section.
gobolino

Collaborative study of MM on Eu server

47 comments in this topic

Recommended Posts

[CHATS]
Privateer
6,718 posts
11,594 battles

Hello Captains,

 

Because of the recuring subject of MatchMaker "cheating" with players, we have decided to make a study of the MM on different tier (6 to 9)

We have decided to keep it simple and to record only the date of the battle, player name, ship type and name, Max and min Tier of the battle, result of the battle and time period of the day (Morning, Afternoon, Evening, Night) and if the game was played in division or Solo.

 

Also we have decided that we need a minimum of 1.000 records before releasing the results and a minimum of 250 battle per ship type.

 

Here are the results for the tier 8 games 

Spoiler

 

Stats
Top Tier 32,10%
Mid Tier 18,30%
Low Tier 49,60%
   
   
Top Tier Minuit 43,48%
Top Tier Matin 36,02%
Top Tier Midi 33,33%
Top Tier Soir 26,35%
   
Top Tier DD 28,78%
Top Tier CA 36,46%
Top Tier BB 32,43%
Top Tier CV 29,71%

 

and erebelow the link to the spreadsheet

 

https://docs.google.com/spreadsheets/d/1hXImdppMWoEU3UKgT3TLRW_kGK84V_x5TtXBsHLRPAI/edit?usp=sharing 

 

 

The work is still in progress (not enough results to be accurate per ship type and time period) but we dont focus the records only on tier 8 anymore. 

 

What we can conclude at the moment is when you hit the "battle" button with a tier 8 ship, you have 1 chance over 3 to be top tier, one over 2 to be low dog and the rest of the time you could be mid tier.

 

One suprise was that CA have a better chance to be top tier and DD less (not enough data on CV to be accurate).

 

You can find the original thread in french here 

I hope it can help.

 

I will keep you informed of the evolution of the study

 

Update 24/11/2018

 

Here are the final resluts of our Study (Tier 8 Ships).

 

 

Stats 1-500  
Top Tier 32,40%    
Mid Tier 22,80%    
Low Tier 44,80%    
Stats 1-1000 Stats 501-1000
Top Tier 32,10% Top Tier 31,59%
Mid Tier 18,30% Mid Tier 13,80%
Low Tier 49,60% Low Tier 54,40%
Stats 1-1500 Stats 1001-1500
Top Tier 33,00% Top Tier 34,80%
Mid Tier 16,20% Mid Tier 12,00%
Low Tier 50,80% Low Tier 53,20%

 

 

We will cary on with new datas because of the modification of the MatchMaker

 

 

 

  • Cool 17

Share this post


Link to post
Share on other sites
Players
4,608 posts
8,081 battles
9 minutes ago, gobolino said:

 

 

Because of the recuring subject of MatchMaker "cheating" with players

 

 

 

and what exactly would that be?

Share this post


Link to post
Share on other sites
[UNICS]
Players
3,739 posts
14,634 battles

Nice! :Smile_Default:

11 minutes ago, Tyrendian89 said:

and what exactly would that be?

Some game theory to get people to invest more time and money would be the most logical answer. ..but here around I guess the answer would be that WG is out to get them.

  • Cool 1

Share this post


Link to post
Share on other sites
Players
1,643 posts
3,193 battles

Also is that study specific to the EU cluster? 

 

You need to look at whats happening on all clusters and compare the results.

Share this post


Link to post
Share on other sites
Players
1,078 posts
8,085 battles

Well this is interesting.

 

On one hand it confirms that T8 MM is a bit rough. But it also breaks a couple assumptions I would've made based on my memory, which is apparently selective.

 

First, I could've sworn you're bottom tier more often than this. Second, I would've assumed it's worse for cruisers (because of shortage of T10 cruisers).

Share this post


Link to post
Share on other sites
[CHATS]
Privateer
6,718 posts
11,594 battles
Il y a 34 minutes, Tyrendian89 a dit :

and what exactly would that be?

 

:Smile_Default:

 

Some players believe that WG "adjust" the MM just for them in order to put them in a difficult situation because :

- They dont put money in the game

- They've put money in the game

- They have a bad WR

- WG dont like them

- Earth is flat

 

and so on.

 

Most of the time it's because they have had a good series of games (good MM, top tier and so on) so WG need to put the pressure on them to make them loose. :cap_old:

 

The other point was that I want to study the impact of the MM rules on various tiers and ship classes.

  • Cool 5

Share this post


Link to post
Share on other sites
[UNICS]
Players
3,739 posts
14,634 battles
2 minutes ago, gobolino said:

The other point was that I want to study the impact of the MM rules on various tiers and ship classes.

Next could be what impact MM rules have on CC vs common plebs :Smile_hiding:

Share this post


Link to post
Share on other sites
[ADRIA]
Players
4,945 posts
7,657 battles
35 minutes ago, gobolino said:

Some players believe that WG "adjust" the MM just for them

- Bcuz they are special snowflakes who, obviously, are the center of the universe, so WG is very interested in screwing around with exactly their account

 

 

  • Cool 1

Share this post


Link to post
Share on other sites
[CHATS]
Privateer
6,718 posts
11,594 battles
Il y a 48 minutes, loppantorkel a dit :

Next could be what impact MM rules have on CC vs common plebs :Smile_hiding:

:Smile_teethhappy:

 

True.... but I'm ot the only one who contribute to the recording, we have all type of players, some old sailors, some very good players, some average or less than average, some with premium times and so on.

Share this post


Link to post
Share on other sites
[UNICS]
Players
3,739 posts
14,634 battles
1 hour ago, gobolino said:

Some players believe that WG "adjust" the MM just for them in order to put them in a difficult situation because :

- They dont put money in the game

- They've put money in the game

- They have a bad WR

- WG dont like them

- Earth is flat

 

and so on.

 

Most of the time it's because they have had a good series of games (good MM, top tier and so on) so WG need to put the pressure on them to make them loose. :cap_old:

 

The other point was that I want to study the impact of the MM rules on various tiers and ship classes.

 

11 minutes ago, gobolino said:

True.... but I'm ot the only one who contribute to the recording, we have all type of players, some old sailors, some very good players, some average or less than average, some with premium times and so on.

It's interesting data, but so far it seems you've only focused on the tier 8 MM, not anything around WG 'adjusting' the MM to cheat some players, which you touch upon in the OP. It looks like you've gone into the 'other point' exclusively this far. Will you continue into more tinfoil-hatty territory in the evolution of your study?

Share this post


Link to post
Share on other sites
[CHATS]
Privateer
6,718 posts
11,594 battles

At the moment we have not enough data on a individual base (less than 300 records) so we cannot have an accurate answer on that point. it will come later with the comparison of the results of the individual

 

We havea already worked on that but the results are not very conclusive as you can read it below

 

Spoiler
Stats        Gobolino        Juesty_Tylor             Global
Top Tier 33,33% 31,79% 32,19%
Mid Tier 19,30% 15,43% 18,00%
Low Tier 47,37% 52,78% 49,81%
WR 52,19% 68,83% 60,67%
       
WR global (players) 51,28% 49,74%  
WR Top Tier 46,05% 68,93% 62,72%
WR Mid Tier 50,00% 80,00% 62,96%
WR Low Tier 57,41% 65,50% 58,51%
       
Top Tier Night 0 45,00% 39,39%
Top Tier Morning 37,88% 27,27% 35,71%
Top Tier Afternoon 34,53% 37,40% 33,69%
Top Tier Evening 13,04% 25,93% 26,45%
       
Top Tier DD 33,71% 26,67% 28,82%
Top Tier CA 39,29% 45,00% 35,84%
Top Tier BB 28,92% 35,00% 32,44%
Top Tier CV 0 30,17% 31,40%

 

Division

     
Top Tier Solo 33,33% 32,08% 32,32%
Top Tier Duo 0 36,36% 36,36%
Top Tier Trio 0 25,00% 23,81%

 

N° of Records

     
Night Battles 0 20 33
Morning Battles 66 11 238
Afternoon Battles 139 131 469
Evening Battles 23 162 310
       
Batailles DD 89 45 288
Batailles CA 56 20 293
Batailles BB 83 80 262
Batailles CV 0 179 207

 

Night : 0.00 to 6.00 AM

Morning : 6.00 AM to 12.00

Afternoon : 12.00 to 6.00 PM

Evening : 6.00 PM to 12.00 PM

 

 

 

  • Cool 2

Share this post


Link to post
Share on other sites
Players
535 posts
6,957 battles

Well this is absolutely normal in free to play game - game makes you want to live T8 ASAP and by that buying dublones to exchange it  for freeXP to get to T10 where you will be top tier all the time and where you can not play without premium time or 5000dub perma camo for 1 ship ( and all of potatos and other bots need perma + premium to be in positive credit balance --- that's why WG loves potatos, they pay a lot )...

 

The only problem I have is that in last 6 month or more MM is giving ( in at least 70% of games ) all potatoes in one team what result in fast lose or fast win = great frustration,, I would just like fromn WG that MM gives equal number of "less skilled" players to both team,,,  so if MM take in 24 players 10-12 with "less skill" ( le say with low av damage, or low PR, or low E/B or low win rate or low all together ) tahn put 5 in one team and 5 on other team or 6/6 or 7/7  DO NOT put like you are doing now 8 in one team and than 2 on other ....

 

And don't give me now carry harder bcs that is not possible anymore, 2 days ago I had battle in Republique with 250k dam, hi cal, confederate, Kraken, and 2 more heroic achh., 3 mil pot dam and we lost with 3 ships difference, than yesterday me and 2 of my teammates had together over 500k dam and over 10 mil pot.dam on us and we manage to lost with 3 ship difference , my teammate in 2 day 2 time 1850xp in lost, every day there is a many situation like this - players are just dying like flies, don't know from what when you take all pot dam on you but yes they manage to die probably from 150-200.000 pot dam....   There can not be all in one team almost every game, that is not fun for winning and especially for losing side .

Share this post


Link to post
Share on other sites
[UNICS]
Players
3,739 posts
14,634 battles
1 hour ago, VooDooZG_Nervozni_Purger said:

The only problem I have is that in last 6 month or more MM is giving ( in at least 70% of games ) all potatoes in one team what result in fast lose or fast win = great frustration,,

Are these numbers based on your feelings? Doesn't hold up as well as actual data...

Share this post


Link to post
Share on other sites
Players
535 posts
6,957 battles
31 minutes ago, loppantorkel said:

Are these numbers based on your feelings? Doesn't hold up as well as actual data...

do you want me to take SS next 20 battles - I will if you want..

 

this is 4 from last few ( like 7-8 ) I have SS 

 

1 - shot-18.08.04.jpg

2 - shot-18.08.04.jpg

3 - shot-18.08.04.jpg

4 shot-18.08.03.jpg

Share this post


Link to post
Share on other sites
[PONY]
Beta Tester
5,191 posts
8,714 battles

This screens show not what happend, and who of the Team is good or bad. They just show what EP they got, but not if one Team got more or less bad players.

Share this post


Link to post
Share on other sites
[UNICS]
Players
3,739 posts
14,634 battles
3 hours ago, gobolino said:

At the moment we have not enough data on a individual base (less than 300 records) so we cannot have an accurate answer on that point. it will come later with the comparison of the results of the individual

 

We havea already worked on that but the results are not very conclusive as you can read it below

 

  Reveal hidden contents
Stats        Gobolino        Juesty_Tylor             Global
Top Tier 33,33% 31,79% 32,19%
Mid Tier 19,30% 15,43% 18,00%
Low Tier 47,37% 52,78% 49,81%
WR 52,19% 68,83% 60,67%
       
WR global (players) 51,28% 49,74%  
WR Top Tier 46,05% 68,93% 62,72%
WR Mid Tier 50,00% 80,00% 62,96%
WR Low Tier 57,41% 65,50% 58,51%
       
Top Tier Night 0 45,00% 39,39%
Top Tier Morning 37,88% 27,27% 35,71%
Top Tier Afternoon 34,53% 37,40% 33,69%
Top Tier Evening 13,04% 25,93% 26,45%
       
Top Tier DD 33,71% 26,67% 28,82%
Top Tier CA 39,29% 45,00% 35,84%
Top Tier BB 28,92% 35,00% 32,44%
Top Tier CV 0 30,17% 31,40%

 

Division

     
Top Tier Solo 33,33% 32,08% 32,32%
Top Tier Duo 0 36,36% 36,36%
Top Tier Trio 0 25,00% 23,81%

 

N° of Records

     
Night Battles 0 20 33
Morning Battles 66 11 238
Afternoon Battles 139 131 469
Evening Battles 23 162 310
       
Batailles DD 89 45 288
Batailles CA 56 20 293
Batailles BB 83 80 262
Batailles CV 0 179 207

 

Night : 0.00 to 6.00 AM

Morning : 6.00 AM to 12.00

Afternoon : 12.00 to 6.00 PM

Evening : 6.00 PM to 12.00 PM

 

 

 

Thanks! Still interesting. One thing to maybe consider, if you're trying to prove or disprove a point is to maybe take into account what ships are played. Juesty is going very heavy on GZ for instance. Might skew things a bit depending on what you're going for.

Share this post


Link to post
Share on other sites
[CHATS]
Privateer
6,718 posts
11,594 battles

Thats one of the problem we have with the data collection, we dont have enough CV players.

 

We are recording the Ship name for this purpose (not only the ship type) but in fact the MM rules according WG declaration dont look at the ship ame, but only at his rank and type (and nation for the DD's if I remember)

Spoiler

Navire Batailles %
Admiral Hipper 24 2,27%
Akizuki 45 4,26%
Alabama 22 2,08%
Amagi 25 2,37%
ARP Takao 4 0,38%
Asashio 19 1,80%
Atago 0 0,00%
Baltimore 22 2,08%
Benson 28 2,65%
Bismarck 21 1,99%
Chapayev 13 1,23%
Charles Martel 61 5,78%
Cleveland 112 10,61%
Edinburgh 0 0,00%
Enterprise 3 0,28%
Gascogne 21 1,99%
Graf Zeppelin 181 17,14%
HSF Harekaze 32 3,03%
Hsienyang 0 0,00%
Kagero 44 4,17%
Kidd 20 1,89%
Kiev 23 2,18%
Kii 1 0,09%
Lexington 0 0,00%
Loyang 20 1,89%
Massachusetts 32 3,03%
Mikhail Kutuzov 9 0,85%
Mogami 39 3,69%
Monarch 26 2,46%
North Carolina 13 1,23%
Ognevoi 20 1,89%
Prinz Eugen 10 0,95%
Richelieu 21 1,99%
Roma 20 1,89%
Shōkaku 28 2,65%
Tirpitz 59 5,59%
Z-23 38 3,60%
TOTAL 1056 100,00%

 

Share this post


Link to post
Share on other sites
[ALONE]
Players
1,417 posts
10,924 battles
6 hours ago, gobolino said:

 

Also we have decided that we need a minimum of 1.000 records before releasing the results and a minimum of 250 battle per ship type.

 

 

1 hour ago, 159Hunter said:

Would you be interested in more data? 

 

If one keeps archiving replays on disk (it takes some space, but not terribly much), there is non-encrypted header in each of them, and I am still thinking that one day I will finally make use of this ;)

 

I really hope the API will still understand old ship IDs... :Smile_bajan2:

 

Share this post


Link to post
Share on other sites
[CHATS]
Privateer
6,718 posts
11,594 battles
Il y a 25 minutes, EdiJo a dit :

 

 

If one keeps archiving replays on disk (it takes some space, but not terribly much), there is non-encrypted header in each of them, and I am still thinking that one day I will finally make use of this ;)

 

I really hope the API will still understand old ship IDs... :Smile_bajan2:

 

 

I've just wiped out my HDD of all replays accumulated since the beginning of the games :cap_fainting:I've discovered MxStats too late.

 

Also one of the guy make me notice one of te answer of WG to a Q&A

 

Citation

Battletier XI & XII, so T10 ships will be down tiered as well?

They’ve emulated it already (not real testing, just modelling, and statistics). Modelling showed that T10 battles would be assembled very slowly, the matchmaker would need a longer cue, while T8 battles would be assembled very quickly. T8 ships currently are top tier in 55% of the battles, 13% in T9 battles, and 33% in T10 battles. T8 is in a comfortable spot now. If the distribution gets too skewed, they’ll modify it.

 

 

Look like our work is not without interest.

 

The problem is that they mabe refer to the CEI/NA/Asia server but for Eu the are wrong... or they maybe need to introduce Tier 11 & 12 Battletier (tier 9 ships will not like it) 

Share this post


Link to post
Share on other sites
Players
1,643 posts
3,193 battles
7 hours ago, gobolino said:

 

:Smile_Default:

 

Some players believe that WG "adjust" the MM just for them in order to put them in a difficult situation because :

- They dont put money in the game

- They've put money in the game

- They have a bad WR

- WG dont like them

- Earth is flat

 

and so on.

 

Most of the time it's because they have had a good series of games (good MM, top tier and so on) so WG need to put the pressure on them to make them loose. :cap_old:

 

The other point was that I want to study the impact of the MM rules on various tiers and ship classes.

I think most people just remember those games where - OMG I'm the only T8 in a T10 game.

 

Being bottom tier just means you have to be more cautious and play more conservatively. That goes against some peoples preferred aggressive play style.

 

Personaly I just play more aggressively when I'm top tier, and conservatively when I'm bottom tier. But I also adapt my game depending on my team and the enemy, 

Share this post


Link to post
Share on other sites
Players
535 posts
6,957 battles
5 hours ago, Igarigen said:

This screens show not what happend, and who of the Team is good or bad. They just show what EP they got, but not if one Team got more or less bad players.

 

 

Are you try to be stupid or you are or what ???   

 

When team win with 10 ships and you/or enemy have 1-2 ships or none alive and and you win/lose with that difference there something went very wrong - how can you make 10 out of 12 ship difference if there was not many good/very good players in one team and many potato in other, if you put it in any team sport than team that lost by that difference had many bad players, for eq if football team lost 10:2 tahn many players in that team  are bad and 1-2-3 good players can not make difference enough because defence an GK will get gol so easy with every shoot and you will not have enough of normal players that will give you the ball in right time or will do something with your ball,  and so on..

 

2 years, even 1 year ago in this game you would see that difference maybe once in every 30-40 games and now ATM 6 of 10 games is like this ... 

Share this post


Link to post
Share on other sites
[SCRUB]
Players
2,263 posts
13,470 battles
18 hours ago, loppantorkel said:

Are these numbers based on your feelings? Doesn't hold up as well as actual data...

Someone would have historical records of this issue - which is unlikely - but roflstomps seem to be happening more often these days...

Share this post


Link to post
Share on other sites
[UNICS]
Players
3,739 posts
14,634 battles
25 minutes ago, Skyllon said:

Someone would have historical records of this issue - which is unlikely - but roflstomps seem to be happening more often these days...

I don't disagree. Greater diversity between veterans and newcomers since the Beta. The game is becoming more complex with every new thing added. Divisions weren't in at the start. There are plenty of things you can do to actively make the matches more unbalanced. If you're a good player and want a challenge - play non-OP ships, don't play cvs at all, don't division up with other good players, play tiers that are tougher to make a difference. If you want to rofl-stomp, do the opposite.

 

Still 100% potatoes in one team, 70% of the time, the last 6 months...? :Smile_sceptic: Yea, don't exaggerate in a topic where OP actually is trying to make an effort of collecting data.

Share this post


Link to post
Share on other sites

  • Recently Browsing   0 members

    No registered users viewing this page.

×