Jump to content


Matchmaking


  • Please log in to reply
8 replies to this topic

Generalissimus_Lucius #1 Posted 05 February 2018 - 04:50 PM

    Munifex

  • Closed Alpha Gladiator
  • 61
  • Member since:
    01-08-2018

I played 1022 battles, i won 45% of it. 

I know it is impossible to create a matchmaking-system, that makes it 50% - so 45% is not ..."that" bad.. but it is bad.

Since i went to T5 battles I most of the time faced T7 and and few 6er units, but never T4.

Therefore it is safe to assume to say you should make matchmaking much more range fights from 1 Tier under to  one from above your own.

Or not?



DanieliusVas #2 Posted 05 February 2018 - 05:42 PM

    Optio Centuriae

  • Closed Alpha Gladiator
  • 158
  • Member since:
    10-08-2017
It's not easy to do that.

 "If numbers are what matters, all Greece cannot match a small part of that army; but if courage is what counts, this number is sufficient."

 

 -Leonidas I 


Raevaks #3 Posted 05 February 2018 - 07:37 PM

    Munifex

  • Closed Alpha Gladiator
  • 11
  • Member since:
    05-15-2017

I played ~2000 battles and I won 60% of those battles. A big part of battle that I won was with T8 against T9-T10 (because the time I was the most active was when T8 archers were bugged and can't be up to T9) and I have more than 60% win rate with all my other T5 commanders against T6-T7.

The problem is that if you do that, it won't change anything, players who knows how to play (high win rate) will have more games against higher tier if I understand your concept and bad players will play more games against lower tier.

So, when a high win rate's player will find a match against higher tier, those higher tier will be the bad players, and they will win even more...



RTKgotterfunkenMD #4 Posted 05 February 2018 - 08:56 PM

    Munifex

  • Closed Alpha Gladiator
  • 88
  • Member since:
    12-18-2016
A good matchmaking would be a dream, but i don't think they game achieve it, World of Warships EU have daily 30k players online and even so matching for certain tiers (I am look at you tier 5 and 8) are completely horrible and borderline unplayable unless you are a Super Unicum player, i have almost 2.5k matchs in that game and can't really play those kinda of matchmaking. Consequently with Arena's population fixing matchmaking is impossible, unfortunately it is sad but true.

macpla_EU #5 Posted 05 February 2018 - 10:12 PM

    Munifex

  • Closed Alpha Gladiator
  • 23
  • Member since:
    02-05-2018

[edited because most of the msg in this form was lost and the thing which was published simply had no sense]. 

 

In short. Do your best. Compare your local engagement not a win/lose ratio of a whole battles. Always play for your team benefits, and ppl will more likely want to party with you.


Edited by macpla_EU, 05 February 2018 - 10:21 PM.


Im_Melliana_I_burn_you #6 Posted 06 February 2018 - 01:11 AM

    Optio Centuriae

  • Closed Alpha Gladiator
  • 236
  • Member since:
    09-20-2011

View PostRaevaks, on 05 February 2018 - 08:37 PM, said:

I played ~2000 battles and I won 60% of those battles. A big part of battle that I won was with T8 against T9-T10 (because the time I was the most active was when T8 archers were bugged and can't be up to T9) and I have more than 60% win rate with all my other T5 commanders against T6-T7.

The problem is that if you do that, it won't change anything, players who knows how to play (high win rate) will have more games against higher tier if I understand your concept and bad players will play more games against lower tier.

So, when a high win rate's player will find a match against higher tier, those higher tier will be the bad players, and they will win even more...

 

LOL high win rate dont mean you are good. You just have tpo play as 4 with archer and greek inf and it s easy win. Abusing op pick as premade give you great win rate but no it dont mean you are good.
CATAPHRACTS FAN CLUB MEMBER

PaiNzzz #7 Posted 06 February 2018 - 08:12 AM

    Optio Centuriae

  • Closed Alpha Gladiator
  • 197
  • Member since:
    07-23-2011

Well, if you are playing against higher tiers, it means that someone (tip, your enemies) plays against lower tiers. Sample size is kind of small, so win rate might not be representative of your skill. 5% seems to me a large number to pin on the matchmaker.

 

That's how matchmaking already works. One tier lower than you to one tier higher than you ends up with two tiers spread. It just happens that the game is not centered on you, so you sometimes end up being another's -1 to a +2 tier game. You sometimes end up as someone's +1 and are top tier. And sometimes you end up as actually you, being mid tier.

 

My experience from tier 5 mm was very good. Enemies ranging from tier 3 to tier 7, but most usually being 4-6.



what4_ #8 Posted 07 February 2018 - 09:13 PM

    Optio Centuriae

  • Closed Alpha Gladiator
  • 216
  • Member since:
    03-25-2017

I have to admit, that MM seems to get worse. Not only in tier, but also in units. 

I played a few battles low tier and it happened too often that one team had all the archers / all the cav while other team got none. As it was tier 1 (all players if I remember correctly) I´m wondering why this is. Is MM not taking Units into consideration?


"Nuts!" - Anthony McAuliffe


PaiNzzz #9 Posted 08 February 2018 - 08:32 AM

    Optio Centuriae

  • Closed Alpha Gladiator
  • 197
  • Member since:
    07-23-2011

View Postwhat4_, on 07 February 2018 - 11:13 PM, said:

I have to admit, that MM seems to get worse. Not only in tier, but also in units. 

I played a few battles low tier and it happened too often that one team had all the archers / all the cav while other team got none. As it was tier 1 (all players if I remember correctly) I´m wondering why this is. Is MM not taking Units into consideration?

 

Yes, that is bad. I only meant that my experience was good tier-wise, not balance wise. So, the matchmaking tool found 10 cavalry players. Why the heck would it give 8 of them to the one team? Same with every unit.




1 user(s) are reading this topic

0 members, 0 guests, 0 anonymous users