TW matchmaking is awful

[Maiden6666_4334]

Product: Star Wars: Galaxy of Heroes
Platform:Android Phone
What type of device are you experiencing the issue with? LG Phone, PC
OS Version 6.0
Ally Code: (Find it here: http://bit.ly/AllyCode ) 678-943-849
What type of issue do you have? Other
How often does the bug occur? Every time (100%)
Summarize your bug It's not a bug, but the Territory Wars matchmaking this time was awful.
Steps: How can we find the bug ourselves? Screenshots
Connection Type Wifi
Please select your region South America
Country Argentina

‌

The matchmaking in this TW was awful. The other guild is vast superior, in terms of GP as of playable characters. We don't even get the 117M because one guy didn't join the TW, so it's even worse. What's the deal with this? Is it a joke or what? Get serious CG

[GladOS-013 · Accepted solution]

@Maiden6666_4334 

Apologies if the way I phrased it was confusing to you.  Those variables that I was referring to (in your first quote) meant actually how difficult the match will be, not whether those were used in the actual matchmaking.  It appears that the matchmaking algorithm is 99% (if not 100%) based on active GP.  

If the other guild had 47 players signed up, then their GP should be very close to 114M.  They are at 121M with 50, and an average of about 2.5M.  So subtract 3 players (~7M) from 121M = 114M.  We can't know exactly which players weren't signed up, unless you review all the combat logs from the war, and write down which 47 players were active.  

Speaking of roster variabiliy - if one guild has 50 Trayas and 50 Jedi Revan vs another guild with 0 of each - even if the GPs are equal, that is a very lopsided matchup.  However, those sorts of matchups can happen - which leads me to believe that the matchmaking algorithm doesn't take the roster into account when creating the match.  However, after the guilds have been paired up, the compositions of their rosters will very much impact which guild wins, alongside the other variables like strategy and contribution levels.  

I'm not assuming, I'm no contradicting, and I'm not lying.  Perhaps my word choice was confusing to you, and for that I apologize.  

[GladOS-013]

Hi @Maiden6666_4334 and thanks for the report!

What you're describing isn't a bug, unless you believe the algorithm is not functioning correctly. It sounds like you're providing feedback about how the algorithm works, which belongs on the Feedback Forum:
https://forums.galaxy-of-heroes.starwars.ea.com/categories/feedback-home

Secondly, a few million GP is actually not very significant over 100M GP rankings. That being said, you also don't know how many of the opponent guild registered for this TW. If your guild only has 48 players signed up, then you should have an active GP of about 115M. If they only had 46-47 players sign up, then they should also be approximately 115M.
Just because their top player has a higher GP than your top player does not mean this is a poor matchup.

There are many variables, including the rosters of all players in your guild, squad deployment strategies, attack strategies, and contribution levels, that all can impact the matchup. Given that the average arena rank for both guilds is pretty similar, this looks like a fairly even matchup.

[Maiden6666_4334]

Well, that sounded very Carrie like. Saying absolutely nothing with a bunch of words. Clearly you don't care at all and answer as a way of courtesy. A few million GP is actually VERY significant, but I'm not a whale so why care at all. The variables seem to include "they have 2 Darth Revan, 1 Malak, like 30 JKR and we only have 15 Revans". Yeah, the rooster is clearly a variable, a variable for the super good algorithm. Pfff, your answer is a joke

[Maiden6666_4334]

Also, should you read the posts in feedback (something that takes, dunno, five minutes?) you would notice THAT ALL THE MATCHMAKING IN EVERYTHING SUCKS. Not only TW, but GA. But hey, let's nerf Anakin as a solution, right?

[GladOS-013]

You don’t know how many players they have signed up, so you are assuming they have 4M higher go than you. Usually matchmaking puts two guilds together within a few 100k. EA doesn’t share all the matchmaking details, but it seems to be 9o% purely based on GP.

And yes, I see lots of people complaining on the forums. I also know that players who are happy with their matchups don’t feel the need to post their happiness on the forums - it’s mostly where people go to complain. And usually players complaining about GA matchups are the ones who haven’t invested their resources as effectively as most players at their GP level, so they see a higher percentage of difficult matchups. I get some good matchups, and some challenging matchups, but I understand that within a large enough population, I am likely not the best roster in my bracket, so will sometimes get paired up with a better player.

[Maiden6666_4334]

"There are many variables, including the rosters of all players in your guild, squad deployment strategies, attack strategies, and contribution levels, that all can impact the matchup. Given that the average arena rank for both guilds is pretty similar, this looks like a fairly even matchup."
Your own words, then... "And usually players complaining about GA matchups are the ones who haven’t invested their resources as effectively as most players at their GP level"
So, the rosters are not a variable at all, or, most likely, you don't have the slightest idea of what you are saying and the algorithm is totally random but they won't let you say it. And the other guild had 47 players who signed up. My guild, minus the dude who didn't played, was 114M. The other one was still above that 'cos all the high GP ones entered. So, don't assume, don't contradict yourself and don't lie saying that the rosters are taking into account right before saying that is not a factor

[GladOS-013 · Accepted solution]

@Maiden6666_4334 

Apologies if the way I phrased it was confusing to you.  Those variables that I was referring to (in your first quote) meant actually how difficult the match will be, not whether those were used in the actual matchmaking.  It appears that the matchmaking algorithm is 99% (if not 100%) based on active GP.  

If the other guild had 47 players signed up, then their GP should be very close to 114M.  They are at 121M with 50, and an average of about 2.5M.  So subtract 3 players (~7M) from 121M = 114M.  We can't know exactly which players weren't signed up, unless you review all the combat logs from the war, and write down which 47 players were active.  

Speaking of roster variabiliy - if one guild has 50 Trayas and 50 Jedi Revan vs another guild with 0 of each - even if the GPs are equal, that is a very lopsided matchup.  However, those sorts of matchups can happen - which leads me to believe that the matchmaking algorithm doesn't take the roster into account when creating the match.  However, after the guilds have been paired up, the compositions of their rosters will very much impact which guild wins, alongside the other variables like strategy and contribution levels.  

I'm not assuming, I'm no contradicting, and I'm not lying.  Perhaps my word choice was confusing to you, and for that I apologize.  

[Maiden6666_4334]

Apologies accepted, it's okay, you don't control the matchmaking. Can't say you helped me, but I can't say that you didn't try either. Thanks for taking your time and answering, best of luck