Unbalanced Matchmaking

[Juanx]

20 hours ago, Merlin851526 said:

Just a few example of what I'm enduring these past few days and weeks. Now I know it's somewhat of an old argument as to whether MM is broken or not, and no doubt the usual "get gud" fraternity will emerge. However I just don't know how much more of this I can endure. I don't expect to win every game I play but I do expect to have a fighting chance and thats not currently happening and hasn't been happening for quite some time. Take this game last night for example. The MM put our CV captain with a win rate of 46% up against the enemy CV captain with a win rate of 83.68% . Moreover the enemy CV captain almost exclusively plays CV,  we lost of course and very quickly as well.

 

And this one from today A Div of Conquerors just went merrily about the map burning all in sundry down. I actually caught that Worchester broadside at roughly 11 Km and the shots that didn't miss over penned such was my RNG. 

 

 

A big dose of L2P seems to be needed...

[Floofz]

Two teams, one wins one loses. So before the game starts you have a 50/50 chance of winning or losing without knowing any additional information. If you apply that to other forms of randomness, like a deck of cards for example. Open a fully mixed deck of cards, now draw a card, is it red or black? Now draw a card again. Do that 20 times in a row. It is highly mathematicly improbable that youre going to draw 20 red cards in a row. Yet this happens in WoWS, over and over and over again. We constantly see (or atleast that is the case on my account) monumental winning or losing streaks, for days. Ive had weeks where Ive won more or less every game I played. And this past month I have a below 30% winrate with days where Ive lost 20-30 games in a row. Its very very rarely the case that you win some and you lose some, you either win all or lose all.

The only factor in this matchmaking RNG that can possibe affect the outcome of this is you yourself as a player. However. in the past few weeks I can count on my hand the number of games Ive played where Ive clearly been in the better team, the team that wouldve won no matter what I did. Instead, the games that I have eventually managed to win have been a very very hard fight and very very nearly a defeat. This means that Ive been put on the worse team in 90 of 100 games roughly speaking, especially since Ive been top of my team in xp in the same 90 games (and I have battle resultscreens to prove it believe me).

 

So because of these things I am completely convinced that MM is not random. There is some sort of balancing function in play here that WG isnt telling us about. 

[Floofz]

6 hours ago, Cazbol said:

I'm still waiting for someone to post that the matchmaker must be broken because the poster is winning too much.

 

It's typically human to attribute our successes to our own ability, while our failures must be due to some external factors.  Those external factors may include gods, demons and software that's developed specifically with you in mind to ensure your loss.

 

As has been said repeatedly, the WoWS matchmaker is trying to match the teams in terms of ship tiers and class.  It's absolutely not trying to match the teams in terms of player skill.

 

The fact that you played well but still lost is completely normal.  Long losing streaks are perfectly normal, just like long winning streaks.

 

Suspecting gaming companies of being crafty in their monetization can be considered healthy scepticism, but coming to the conclusion that the matchmaker is favouring better paying customers should require something to support it.  My experience with the game contradicts that conspiracy theory.  I spent more on the game at the start, mainly on port slots because I kept all my ships and one gets new ships quickly at the start.  My win rate was poor.  Once I hit tier 7 in the German and French lines, my win rate was around 47%, if I remember correctly.  That's when I decided to back to he lower tiers and learn to use each ship at each tier, instead of rushing through some grind.  Over the course of a year my win rate is crept up to 54%.  My spending has decreased, which is negatively correlated to my increased win rate.  Instead, the win rate seems to be affected by the time I've spent on the game trying to improve myself.

 

You read what you want to read. I have stated before that I had an abnormaly high winrate in my first 1000 games. It felt like I simply couldnt lose no matter how badly I played. I have because of both this and the insane jumps up and down in losing and win streaks in this game been convinced that there is a balancingsystem in play that is working extremely badly.

 

This is coming from someone who played WoT for many years and the win/lose statistics in that game isnt even nearly as odd and "unrandom".

[antean]

On ‎12‎/‎1‎/‎2018 at 12:52 PM, KHETTIFER said:

my friend who spends a buttload on this gets the absolute opposite

Here's your answer.

[antean]

4 minutes ago, AdmiralDing3Ling said:

There is some sort of balancing function in play here that WG isnt telling us about. 

I'd believe you if you had said 'unbalancing' function, lol.

[antean]

7 minutes ago, AdmiralDing3Ling said:

Instead, the games that I have eventually managed to win have been a very very hard fight and very very nearly a defeat. This means that Ive been put on the worse team in 90 of 100 games roughly speaking, especially since Ive been top of my team in xp in the same 90 games (and I have battle resultscreens to prove it believe me).

It may not be quite so blatant as what you refer to here, DingELing, but the types of battle results you refer to is, unfortunately, very prevalent.

Sure, most like to be part of a winning roflstomp as opposed to being pretty much instantly done in but they are both symptomatic of WG MM garbage, quite frankly.

&, of course, it's simply 'co-incidental' that half one's wins are close battles as opposed to the many overwhelming losses (roflstomps) one sees.

Note: All WG will say is the MM is working as 'intended'. Here WG is actually telling the truth.

[ColonelPete]

24 minutes ago, AdmiralDing3Ling said:

Two teams, one wins one loses. So before the game starts you have a 50/50 chance of winning or losing without knowing any additional information. If you apply that to other forms of randomness, like a deck of cards for example. Open a fully mixed deck of cards, now draw a card, is it red or black? Now draw a card again. Do that 20 times in a row. It is highly mathematicly improbable that youre going to draw 20 red cards in a row. Yet this happens in WoWS, over and over and over again. We constantly see (or atleast that is the case on my account) monumental winning or losing streaks, for days. Ive had weeks where Ive won more or less every game I played. And this past month I have a below 30% winrate with days where Ive lost 20-30 games in a row. Its very very rarely the case that you win some and you lose some, you either win all or lose all.

The only factor in this matchmaking RNG that can possibe affect the outcome of this is you yourself as a player. However. in the past few weeks I can count on my hand the number of games Ive played where Ive clearly been in the better team, the team that wouldve won no matter what I did. Instead, the games that I have eventually managed to win have been a very very hard fight and very very nearly a defeat. This means that Ive been put on the worse team in 90 of 100 games roughly speaking, especially since Ive been top of my team in xp in the same 90 games (and I have battle resultscreens to prove it believe me).

 

So because of these things I am completely convinced that MM is not random. There is some sort of balancing function in play here that WG isnt telling us about. 

You are missing an important factor: the player.

How well do you suppose one plays when one has lost five games in a row? As good as during the first battle? Better? Worse? What does that mean for the chances of success?

 

And no, I do not believe you that you lost 20+ games in a row.

And no, you did not give ONE reason why WG should do that.

[Floofz]

16 minutes ago, ColonelPete said:

You are missing an important factor: the player.

How well do you suppose one plays when one has lost five games in a row? As good as during the first battle? Better? Worse? What does that mean for the chances of success?

 

And no, I do not believe you that you lost 20+ games in a row.

And no, you did not give ONE reason why WG should do that.

 

Did I? "The only factor in this matchmaking RNG that can possibe affect the outcome of this is you yourself as a player."

Maybe you should read the whole post if you want to comment.

 

Well youre happy to believe whatever you want but my clanmates can tell you otherwise. Infact even winning 3 in 20 games is unlikely if youre constantly top on your team. That is to say if the MM was totally random.

 

The reason as far as I can see would be to balance matches and winrates. And if that is the case its a noble cause, but obviously it isnt working as it should. To be fair, the reason behind it is at this point completely irrelevant. Because statisticly and mathematicly its very very unlikely given these results that the MM is totally random.

 

Just FYI. I have a second account with under 1000 battles on it and 74% winrate.

[DFens_666]

17 minutes ago, AdmiralDing3Ling said:

The reason as far as I can see would be to balance matches and winrates.

 

Then they are failing miserably, since i got 80% in my last 40 games

I mean, there are many of those which were hard to win (even mostly playing 3x div) and some of them we lost even tho it was winable, but not with those teammates.

Today 5 games

- first 2 losses, couldnt carry those, idiots were running to the border chasing 1 ship couldnt kill that ofc, or 4 ships on my side dieing all within 2 minutes.

- 3rd game could have gone either way it was 1v1 DD in the end, while i gunboated a Farragut in my Mutsuki He was shooting AP tho.

- 4th game pretty easy win

- 5th game could have gone either way, both teams tried to throw (posted that in greatest gaming achievements + replay because its hilariously bad)

[Sark74]

I haven't read all the comments, but I've had games where some teams are made to lose. There is no other explanation!. You get into a position where you think you will be a tactical advantage to your team, only to look up and your 4 ships down already. Your teams DD's all died sat broadside in smoke or over extended and called everyone a noob for not helping them, Everyone here must have been in those teams at some point. I've come to the conclusion that the team with the the best CV or DD captains will generally win. The Poor CV player will get overwhelmed or attack the wrong ships or lesser skilled DD die to sitting broadside in smoke, Using guns while they should be trying to get concealment back..  I'd like to see a game where if u have 5 BB, 4 Cruisers and 3 DD's all of the BB's, cruisers and DD's have similar Personal ratings.. or all 3 Add up to the same number.. if u get a DD captain with 400 PR up against a 1400.. its clear to me who will come out on top :(  If people really want a better/fairer match making, they will have to suffer longer queue's is it worth it? 

 

just my thoughts tho.. 

 

 

[ColonelPete]

1 hour ago, AdmiralDing3Ling said:

 

Did I? "The only factor in this matchmaking RNG that can possibe affect the outcome of this is you yourself as a player."

Maybe you should read the whole post if you want to comment.

 

Well youre happy to believe whatever you want but my clanmates can tell you otherwise. Infact even winning 3 in 20 games is unlikely if youre constantly top on your team. That is to say if the MM was totally random.

 

The reason as far as I can see would be to balance matches and winrates. And if that is the case its a noble cause, but obviously it isnt working as it should. To be fair, the reason behind it is at this point completely irrelevant. Because statisticly and mathematicly its very very unlikely given these results that the MM is totally random.

 

Just FYI. I have a second account with under 1000 battles on it and 74% winrate.

That is not what I meant...

It is about a player playing inconsistent and not at a fixed "skill level". Therefore you cannot compare the game results with card drawing, where the chance of getting one colour increases as you pull the other colour out of the deck. That is very unlikely to happen to a player, especially when he loses.

 

Could you give mathematic proof that the results are not random? You know that there are statistic tests for that... And your results over the last weeks do not even look a slight bit perculiar.

And if you cannot, do not claim stuff you have no idea of.

 

And what do you do different at your second account? Starting from scratch with all your experience? Hardly surprising....

[Floofz]

15 minutes ago, ColonelPete said:

That is not what I meant...

It is about a player playing inconsistent and not at a fixed "skill level". Therefore you cannot compare the game results with card drawing, where the chance of getting one colour increases as you pull the other colour out of the deck. That is very unlikely to happen to a player, especially when he loses.

 

Could you give mathematic proof that the results are not random? You know that there are statistic tests for that... And your results over the last weeks do not even look a slight bit perculiar.

And if you cannot, do not claim stuff you have no idea of.

 

And what do you do different at your second account? Starting from scratch with all your experience? Hardly surprising....

 

That is why I very carefully stated that as long as Im top on my team in a majority of my defeats then its not very likely that the defeats are due to my poor input. And no, I never ever damage farm, I always play for the objective.

You could have an infinite deck of cards, as long as its well shuffled its not likely to bring out the same color that many times in a row. And as I said, even if there is an odd red card every now and then, its not likely to get 80 black cards out of 100 draws. And that is exactly what I had a few weeks ago, in 50 battles I had a 16% winrate.

Even if I now have 45-50% winrate in total in the past month, the way Ive gotten that percentage is way weird in itself. One day I win almost all games and the next day I lose almost all games. Theres no balance, its always one extreme to the other.

 

Also Ive been very careful to not say that these are facts or that I have conclusive evidence. I said that its not probable, and its not likely, and because of it I am convinced. It ofcourse will take more statistics and research than Im willing to commit to in order to prove this.

[ColonelPete]

Your stats do not reflect what you claim...

And you do not understand what random means.

[Floofz]

31 minutes ago, ColonelPete said:

Your stats do not reflect what you claim...

And you do not understand what random means.

 

My stats doesnt show such things, it just shows my average winrate over a set number of days. It doesnt show day per day.

 

You dont understand what likeliness and probability mean. Given enough statistics any random number generator with a 50/50 resultchance should overall give between 40% and 60%.

What you dont seem to understand is the smaller numbers. Yes, given 3000 battles my winrate is at 56-57%. But its the day to day winrate which is odd. How can the winrate jump 60-70% from day to day? And not only once but constantly.

For example today I won all games I played except for one and that was very close. The day before that I lost almost all games I played.

[ColonelPete]

You can look at you daily games for the past 30 days.

With low number of games, a high volatility is normal. That is what random means and what you fail to understand.

Every random number generator can show you that.

 

1 2 2 2 2 2 2 2 1 1 2 2 2 1 1 1 2 1 1 2 2 2 1 2 1 2 2 2 2 2 1 1 2 1 1 1 2 1 1 2 1 2 2 2 2 1 1 2 1 2 2 1 2 1 1 1 2 1 2 2 1 1 1 1 2 2 1 2 2 2 2 2 2 1 2 1 2 2 2 2 2 2 2 1 1 2 2 2 1 1 2 1 2 2 1 1 1 1 2 2

 

https://www.random.org/integers/?num=100&min=1&max=2&col=5&base=10&format=html&rnd=new

 

 

[Armorin]

The random MM is truly a wonder. My guess is that having players with overall win rates of less than 40% is just bad for business. It could be a true statistical anomaly though.

[ColonelPete]

I doubt you have to change your playstile much 

 

Btw. you could just stop playing. Just an idea.

 

[KHETTIFER]

6 minutes ago, ColonelPete said:

I doubt you have to change your playstile much 

 

Btw. you could just stop playing. Just an idea.

 

That could possibly be true, and considering the screenshots i posted of my typical games, it doesn't say much for this game, "botched abortion" is a pretty accurate description of WoWs.

Why would i choose to quit when i've finally found the formula for having fun in this travesty, see i'm wondering will i be able to get my W/R below 20% before CancerGaming intervene.

[ColonelPete]

I doubt it. A well known troll, who got at least two one week long ingame bans as a result of my tickets, is around 35% WR.

While your plans reveal much about your character, it will not be anything special with this playerbase. It makes no difference, if you play bad on purpose or you do not know better.

The lower the average performance of the playerbase, the better it is for the good players.

[KHETTIFER]

13 hours ago, ColonelPete said:

I doubt it. A well known troll, who got at least two one week long ingame bans as a result of my tickets, is around 35% WR.

While your plans reveal much about your character, it will not be anything special with this playerbase. It makes no difference, if you play bad on purpose or you do not know better.

The lower the average performance of the playerbase, the better it is for the good players.

So WG are reluctant to ban players regardless of what they do, interesting......

The only thing revealed about my character is that like most people with standards, i have a breaking point.

Truth is matchmaking in this game is far from random, it's actually quite consistent, and match outcomes are easily predicted at the loading screen once you have played for a while, no software or program is actually needed either.

In contrast MWO also has bad matchmaking but it truely IS random, making predictions about outcomes just doesn't work 90% of the time, funny that.

 

 

[ColonelPete]

That is what you think, but it reveals much more. 

 

[Cazbol]

Let's do a table showing the probability of getting either a win or a loss streak of a certain length.

This assumes that you are an average player with a 50% win rate.

Streak of:	Probability	1 in X Matches
1	100.0%	1
2	50.0%	2
3	25.0%	4
4	12.5%	8
5	6.3%	16
6	3.1%	32
7	1.6%	64
8	0.8%	128
9	0.4%	256
10	0.2%	512

If you've played a 1000 matches, you're likely to have had about 2 streaks of 10 wins or losses.  A streak of 10 isn't noteworthy at all.  For poorer players, the loss streaks will be even more likely, and vice versa for the more skilled players.

 

NOT having these streaks of wins or losses... that would be a sign of tampering with the random process.

 

[Floofz]

On 12/4/2018 at 3:15 PM, Cazbol said:

Let's do a table showing the probability of getting either a win or a loss streak of a certain length.

This assumes that you are an average player with a 50% win rate.

Streak of:	Probability	1 in X Matches
1	100.0%	1
2	50.0%	2
3	25.0%	4
4	12.5%	8
5	6.3%	16
6	3.1%	32
7	1.6%	64
8	0.8%	128
9	0.4%	256
10	0.2%	512

If you've played a 1000 matches, you're likely to have had about 2 streaks of 10 wins or losses.  A streak of 10 isn't noteworthy at all.  For poorer players, the loss streaks will be even more likely, and vice versa for the more skilled players.

 

NOT having these streaks of wins or losses... that would be a sign of tampering with the random process.

 

 

Ive had far more insane win streaks than Ive had losing streaks. Ive had countless of streaks with 10 wins or more in a row.

I agree that if you never had those streaks then itd be a sign of tampering. HOWEVER. Long streaks of wins or losses are more common than winning one, losing 2, winning 3, losing 1 etc. The streaks can go on for days, weeks or for me even a month last time. That is what feels so fishy. See I dont understand how in a random system it can put me in the worse team for over 70 games in a row. Yes I won a few but that was because of insane carry and not because I was in the better team.

[El2aZeR]

1 minute ago, AdmiralDing3Ling said:

Ive had far more insane win streaks than Ive had losing streaks. Ive had countless of streaks with 10 wins or more in a row.

On 12/4/2018 at 3:15 PM, Cazbol said:

This assumes that you are an average player with a 50% win rate.

 

Read.

[Odo_Toothless]

On 12/3/2018 at 7:02 PM, AdmiralDing3Ling said:

So before the game starts you have a 50/50 chance of winning or losing without knowing any additional information.

In the long run - Yes. Average player have a 50% chance of winning ex. 100 battles. Looking at particular game you have from ex. from 10% to 90% chances to win . It depends how big is difference between skill of teams.

 

For example on tiers VIII-X, poor CV players has about 33% WR. Match him vs good CV driver in 3 man division and such game is lost in about 90% cases. Beeing constantly uptiered in ex. VIII tier BB vs IX/X and "bad luck" to teams (especially DD drivers), your personal performance means almost nothing in case of WR ...