I believe this warrants a discussion in the General Discussion and NOT in the feedback section. We should be allowed to discuss how exactly the TW algorithm is supposed to work and the "desired outcome"
I want to at least hear an answer from someone on this.
Our guild is now contemplating putting 1 toon on D in protest and let them win, which we know can't be the "evenly matched" outcome you guys want.
At least we get to have fun on offense, since you guys just kittened us on Defense.
my guild got matched with alliance alpha, then wallet warriors 2 weeks in a row, both are like 16m GP over us (alliance pi)
I don't need a detailed explanation. I just need them to see how ridiculously broken it is and promise that they will do something to fix it. Carrie seems like a reasonable person. I really wish she'd jump in here and at least tell us that they are working on it.
I think someone is cut up about 800/1000 less G12 toons. For a 2 page thread i've never seen it mentioned so much.
Argument about lower guilds is not really relevant.
There is so much differences in what ppl have.
Last tw we faced a guild 29mil GP higher (80-109), guess what, they wiped the floor with us but we put up a decent effort.
This time the difference is 6/7mil.
As others, every single tw we have faced a bigger guild since forming in Jan this year.
We won quite a few but also a few have been totally unfair matchup's.
I'm still struggling to see how the matchmaking is done.
For instance on this one they have more jtr's, more zeta talzins, more zeers, more 7* gk's.
So we used to think as we got a few zeta talzins might be why we getting higher guilds but not too sure now.
We only have 22 profiles of theirs to go off but seems a majority of the bigger members are on swgoh.gg.
Our 50 profiles shows 191 G12 toons, their 22 profiles shows 141.
So although cant be confirmed, they probably have more gear 12 than us.
We have 350 zetas, their 22 profiles shows 275, so they probably have more zetas.
So why this matchup in terms of GP seems fairer, they seem to have more of most of the meta than us, more g12 and more zetas.
We however have faced a guild with 15/20mil GP more than us that had TB GP and had a lot less high end meta teams.
Some guilds are obviously overestimated by the "algorithm". Our guild is one of them. We win most matches before the algorithm. But now we lose every matches. I wonder what happened. But I think they will not tell us. This time, we will face a guild with 22 JTR, while we only 6.
May the force with us.
I don't need a detailed explanation. I just need them to see how ridiculously broken it is and promise that they will do something to fix it. Carrie seems like a reasonable person. I really wish she'd jump in here and at least tell us that they are working on it.
I understand that you want them to reassure you of things, but they have stated that.
...Thank you all for your great feedback as we continue to improve the Territory Wars feature. As with all areas of the game, we will continue to monitor the feature to ensure it is meeting target design goals....
I don't need a detailed explanation. I just need them to see how ridiculously broken it is and promise that they will do something to fix it. Carrie seems like a reasonable person. I really wish she'd jump in here and at least tell us that they are working on it.
I understand that you want them to reassure you of things, but they have stated that.
...Thank you all for your great feedback as we continue to improve the Territory Wars feature. As with all areas of the game, we will continue to monitor the feature to ensure it is meeting target design goals....
What we want is for them to explain whether or not they think these matchups are WAI, or do they concur that its broken and they are working on it. Or at least acknowledge that these matchups shouldn't be happening.
The only communication we got so far is "hey look everyone we fixed it (but not really), but we will continue to monitor".
The purpose of this thread was 2 fold.
1.) Receive some communication from CG that these matchup are indeed NOT what they wanted.
3.) and to let CG know that our guild is now using this TW as a practice/sandbox mode and that our guild no longer cares about this aspect of the game (which certainly can't be what they wanted), because of the same algorithm that is supposed to keep things evenly matched.
I am in agreeance with Kyno in that I dont need to, nor do I want to know the algorithim used for matchmaking. As I see it, the system is already exploited enough as is, and adding further details would worsen the problem rather than fix it.
What I would like to see is some type of response for the reports of unfair matchmaking. For those matches and reporting an unfair match it would be reassuring to get a response of “We have investigated your complaint and have concluded you were correct that the system did not give you a fair matchup”. I know that is a long shot but hey, I can dream.
I don't need a detailed explanation. I just need them to see how ridiculously broken it is and promise that they will do something to fix it. Carrie seems like a reasonable person. I really wish she'd jump in here and at least tell us that they are working on it.
I understand that you want them to reassure you of things, but they have stated that.
...Thank you all for your great feedback as we continue to improve the Territory Wars feature. As with all areas of the game, we will continue to monitor the feature to ensure it is meeting target design goals....
What we want is for them to explain whether or not they think these matchups are WAI, or do they concur that its broken and they are working on it. Or at least acknowledge that these matchups shouldn't be happening.
The only communication we got so far is "hey look everyone we fixed it (but not really), but we will continue to monitor".
The purpose of this thread was 2 fold.
1.) Receive some communication from CG that these matchup are indeed NOT what they wanted.
3.) and to let CG know that our guild is now using this TW as a practice/sandbox mode and that our guild no longer cares about this aspect of the game (which certainly can't be what they wanted), because of the same algorithm that is supposed to keep things evenly matched.
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
This also leads down the rabbit hole of each time he answers it gives insight into the algorithm, but let's avoid that discussion for now.
I have flagged him and I know he keeps an eye on these topic, but I dont fully expect you to get a direct answer.
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
And that’s probably where the algorithm is badly designed. Players set the defense teams, players lead the offenses, not algorithms - and the parameters we point to, are the ones determining the outcome. If the algorithm looks at others, it’s missing its intended utility.
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
And that’s probably where the algorithm is badly designed. Players set the defense teams, players lead the offenses, not algorithms - and the parameters we point to, are the ones determining the outcome. If the algorithm looks at others, it’s missing its intended utility.
Sorry, I may be over simplifying, I should have said are not as high a priority. I'm sure they are taken into account.
The problem is players feel that there should be very strick limitations on certain parameters, and in all honesty I have no idea if that's the right approach either. But to say it's just wrong, is not right either. We dont know enough to make that call.
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
And that’s probably where the algorithm is badly designed. Players set the defense teams, players lead the offenses, not algorithms - and the parameters we point to, are the ones determining the outcome. If the algorithm looks at others, it’s missing its intended utility.
Sorry, I may be over simplifying, I should have said are not as high a priority. I'm sure they are taken into account.
The problem is players feel that there should be very strick limitations on certain parameters, and in all honesty I have no idea if that's the right approach either. But to say it's just wrong, is not right either. We dont know enough to make that call.
His point still holds, the algorithm is clearly poorly designed if it's taking into account all of the parameters but ones that are clearly not impacting the outcomes as much are getting higher priority than other parameters.
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
And that’s probably where the algorithm is badly designed. Players set the defense teams, players lead the offenses, not algorithms - and the parameters we point to, are the ones determining the outcome. If the algorithm looks at others, it’s missing its intended utility.
Sorry, I may be over simplifying, I should have said are not as high a priority. I'm sure they are taken into account.
The problem is players feel that there should be very strick limitations on certain parameters, and in all honesty I have no idea if that's the right approach either. But to say it's just wrong, is not right either. We dont know enough to make that call.
His point still holds, the algorithm is clearly poorly designed if it's taking into account all of the parameters but ones that are clearly not impacting the outcomes as much are getting higher priority than other parameters.
From what we have been told they are trying to match on more of a player by player basis, which could be more accurate then a simple numbers match of GP and # of g12 toons.
Not that this is the case here but there is a lot of strategy that goes into TW and to always say "it's not an even match" is not always fair on the system either, I'm sure many guilds could overcome a deficit that they didn't realize. There is a lot to be said for farming strategy also, people can have a bunch of g12 toons all under the mindset of raids that dont necessarily help in pvp combat.
This is a long road to develop a good matching system, we are still "early on" in that journey. Let's give them a chance before we just start calling it a poor design because it doesn't match the exact parameters we think it should, especially since we dont know if that's even true.
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
And that’s probably where the algorithm is badly designed. Players set the defense teams, players lead the offenses, not algorithms - and the parameters we point to, are the ones determining the outcome. If the algorithm looks at others, it’s missing its intended utility.
Sorry, I may be over simplifying, I should have said are not as high a priority. I'm sure they are taken into account.
The problem is players feel that there should be very strick limitations on certain parameters, and in all honesty I have no idea if that's the right approach either. But to say it's just wrong, is not right either. We dont know enough to make that call.
His point still holds, the algorithm is clearly poorly designed if it's taking into account all of the parameters but ones that are clearly not impacting the outcomes as much are getting higher priority than other parameters.
From what we have been told they are trying to match on more of a player by player basis, which could be more accurate then a simple numbers match of GP and # of g12 toons.
Not that this is the case here but there is a lot of strategy that goes into TW and to always say "it's not an even match" is not always fair on the system either, I'm sure many guilds could overcome a deficit that they didn't realize. There is a lot to be said for farming strategy also, people can have a bunch of g12 toons all under the mindset of raids that dont necessarily help in pvp combat.
This is a long road to develop a good matching system, we are still "early on" in that journey. Let's give them a chance before we just start calling it a poor design because it doesn't match the exact parameters we think it should, especially since we dont know if that's even true.
I understand that, I'm just saying you can't discount his point because of simplifying it, either way there is still definitely something wrong with the algorithm that is leading to huge mismatches in some wars, but then still able to make for fairly even wars in other guilds.
Also we've had 20 TWs now, I'm not really sure how much longer it will count as "early on" in the matchmaking system. The wars I've had recently have been fairly even, but I know if I were in a guild that kept getting mismatches I would be really discouraged from playing in TW and would want this to be altered asap or at least have some sort of acknowledgement that the algorithm is clearly working well for some guilds and not as well for others and that they're looking into that.
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
And that’s probably where the algorithm is badly designed. Players set the defense teams, players lead the offenses, not algorithms - and the parameters we point to, are the ones determining the outcome. If the algorithm looks at others, it’s missing its intended utility.
Sorry, I may be over simplifying, I should have said are not as high a priority. I'm sure they are taken into account.
The problem is players feel that there should be very strick limitations on certain parameters, and in all honesty I have no idea if that's the right approach either. But to say it's just wrong, is not right either. We dont know enough to make that call.
His point still holds, the algorithm is clearly poorly designed if it's taking into account all of the parameters but ones that are clearly not impacting the outcomes as much are getting higher priority than other parameters.
From what we have been told they are trying to match on more of a player by player basis, which could be more accurate then a simple numbers match of GP and # of g12 toons.
Not that this is the case here but there is a lot of strategy that goes into TW and to always say "it's not an even match" is not always fair on the system either, I'm sure many guilds could overcome a deficit that they didn't realize. There is a lot to be said for farming strategy also, people can have a bunch of g12 toons all under the mindset of raids that dont necessarily help in pvp combat.
This is a long road to develop a good matching system, we are still "early on" in that journey. Let's give them a chance before we just start calling it a poor design because it doesn't match the exact parameters we think it should, especially since we dont know if that's even true.
I understand that, I'm just saying you can't discount his point because of simplifying it, either way there is still definitely something wrong with the algorithm that is leading to huge mismatches in some wars, but then still able to make for fairly even wars in other guilds.
Also we've had 20 TWs now, I'm not really sure how much longer it will count as "early on" in the matchmaking system. The wars I've had recently have been fairly even, but I know if I were in a guild that kept getting mismatches I would be really discouraged from playing in TW and would want this to be altered asap or at least have some sort of acknowledgement that the algorithm is clearly working well for some guilds and not as well for others and that they're looking into that.
Ok..... How many of your guild's players are participating in the current TW, and how many defence spots are available per territory?
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
And that’s probably where the algorithm is badly designed. Players set the defense teams, players lead the offenses, not algorithms - and the parameters we point to, are the ones determining the outcome. If the algorithm looks at others, it’s missing its intended utility.
Sorry, I may be over simplifying, I should have said are not as high a priority. I'm sure they are taken into account.
The problem is players feel that there should be very strick limitations on certain parameters, and in all honesty I have no idea if that's the right approach either. But to say it's just wrong, is not right either. We dont know enough to make that call.
His point still holds, the algorithm is clearly poorly designed if it's taking into account all of the parameters but ones that are clearly not impacting the outcomes as much are getting higher priority than other parameters.
From what we have been told they are trying to match on more of a player by player basis, which could be more accurate then a simple numbers match of GP and # of g12 toons.
Not that this is the case here but there is a lot of strategy that goes into TW and to always say "it's not an even match" is not always fair on the system either, I'm sure many guilds could overcome a deficit that they didn't realize. There is a lot to be said for farming strategy also, people can have a bunch of g12 toons all under the mindset of raids that dont necessarily help in pvp combat.
This is a long road to develop a good matching system, we are still "early on" in that journey. Let's give them a chance before we just start calling it a poor design because it doesn't match the exact parameters we think it should, especially since we dont know if that's even true.
I understand that, I'm just saying you can't discount his point because of simplifying it, either way there is still definitely something wrong with the algorithm that is leading to huge mismatches in some wars, but then still able to make for fairly even wars in other guilds.
Also we've had 20 TWs now, I'm not really sure how much longer it will count as "early on" in the matchmaking system. The wars I've had recently have been fairly even, but I know if I were in a guild that kept getting mismatches I would be really discouraged from playing in TW and would want this to be altered asap or at least have some sort of acknowledgement that the algorithm is clearly working well for some guilds and not as well for others and that they're looking into that.
I was not trying to dismiss what he said, I just wanted to not over simplify anything, there is a big difference between comparing things and not looking at others vs. Comapring all factors and properly weighing them. I miss spoke and didn't want to over simplify this whole concept.
From what we have been told they are trying to match on more of a player by player basis, which could be more accurate then a simple numbers match of GP and # of g12 toons.
Then it’s failing too. Because we have had a lot of matchmaking results where it’s not just a couple of rosters with 144 g12, but overal superiority on a player-by-player comparison. Average GP, mean GP, median GP, average/mean/median number of g12, same about zetas... when you’re systematically on the lower branch of the comparison, a pattern appears. And that pattern is ugly.
It’s so at least since the change in matchmaking (which led to a fix). The mismatches are way more pronounced now than earlier (where, when facing guilds with 6-8M GP, we could beat when playing smart - but it was a early in the TW life, and many guilds approached them in an inefficient way).
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
And that’s probably where the algorithm is badly designed. Players set the defense teams, players lead the offenses, not algorithms - and the parameters we point to, are the ones determining the outcome. If the algorithm looks at others, it’s missing its intended utility.
Sorry, I may be over simplifying, I should have said are not as high a priority. I'm sure they are taken into account.
The problem is players feel that there should be very strick limitations on certain parameters, and in all honesty I have no idea if that's the right approach either. But to say it's just wrong, is not right either. We dont know enough to make that call.
His point still holds, the algorithm is clearly poorly designed if it's taking into account all of the parameters but ones that are clearly not impacting the outcomes as much are getting higher priority than other parameters.
From what we have been told they are trying to match on more of a player by player basis, which could be more accurate then a simple numbers match of GP and # of g12 toons.
Not that this is the case here but there is a lot of strategy that goes into TW and to always say "it's not an even match" is not always fair on the system either, I'm sure many guilds could overcome a deficit that they didn't realize. There is a lot to be said for farming strategy also, people can have a bunch of g12 toons all under the mindset of raids that dont necessarily help in pvp combat.
This is a long road to develop a good matching system, we are still "early on" in that journey. Let's give them a chance before we just start calling it a poor design because it doesn't match the exact parameters we think it should, especially since we dont know if that's even true.
I understand that, I'm just saying you can't discount his point because of simplifying it, either way there is still definitely something wrong with the algorithm that is leading to huge mismatches in some wars, but then still able to make for fairly even wars in other guilds.
Also we've had 20 TWs now, I'm not really sure how much longer it will count as "early on" in the matchmaking system. The wars I've had recently have been fairly even, but I know if I were in a guild that kept getting mismatches I would be really discouraged from playing in TW and would want this to be altered asap or at least have some sort of acknowledgement that the algorithm is clearly working well for some guilds and not as well for others and that they're looking into that.
Ok..... How many of your guild's players are participating in the current TW, and how many defence spots are available per territory?
We get all 50 to participate and theres 25 spots per territory.
I exactly want them to tell me that the matchup we have is WAI OR that they recognize this is a problem and they are (again) working on it. I could care less how the algorithm is actually working and what variables they use.
I also want them to know that half our guild isn't even playing this TW, mostly out of a complete lack of care about the game mode they designed and feel is WAI.
and for the record...our opposing guild had 49 signup and we had 50. So let's lose the active GP argument once and for all.
One of the issues we are facing is while we might have a similar match up on GP, their teams totally out match ours.
Example is we have quite a few FO, Sith and Empire teams but these do not easily match the standard defence of CLS and JTR teams. We have 25 CLS teams, 8 JTR teams and 1 Kenobi at 7*.. Yet we are matched with 2 front walls of Full G12 CLS/ 7* GK or multiple JTR teams more than likely because of our other squads match that 90-100k range.
Removing our issue of having a few eclectic squads, I personally think this is the main issue with the matchmaking. Squads are so different in defence or offence power that if it is a straight match on squad power there is always a very strong chance of a mis match.
I don't have an answer other than matching on specific toons which would be pretty complex..
One of the issues we are facing is while we might have a similar match up on GP, their teams totally out match ours.
Example is we have quite a few FO, Sith and Empire teams but these do not easily match the standard defence of CLS and JTR teams. We have 25 CLS teams, 8 JTR teams and 1 Kenobi at 7*.. Yet we are matched with 2 front walls of Full G12 CLS/ 7* GK or multiple JTR teams more than likely because of our other squads match that 90-100k range.
Removing our issue of having a few eclectic squads, I personally think this is the main issue with the matchmaking. Squads are so different in defence or offence power that if it is a straight match on squad power there is always a very strong chance of a mis match.
I don't have an answer other than matching on specific toons which would be pretty complex..
I actually 100% disagree with you. If 2 guilds are evenly match on GP, then it's all strategy (farming, guild make-up, and map placement). Yes if you dont have key toons for pvp combat then that's on you not the dev team to match you to someone who made the same farming decisions. IMHO
The only situations where this starts to become less true is when it's the number of guild members starts to get skewed, like 50 vs 30 or something like that. Then the 30 could have an advantage of a wider range of toons.
I wish my zeta drop rate was as good as the drop rate of horrible tw matchups. That’s all I’m saying lol
LMAO don't we all!
In all seriousness though. It's fair to say the MM is not perfect. In aggregate it probably serves it's purpose, but there are a lot of glaring holes in the system. Comparing player by player is probably not as good as comparing guild to guild imo. Obviously that's an opinion since we can't base that on fact. I can see how matching 2 guilds that have 100m but one with 30 players and one with 50 would be a mismatch, but putting 2 guilds that have an equivalent number G12 players shouldn't be. In the lower tiers, I can understand the need to balance around guilds that aren't all at 50 members. But in the upper tiers, it probably should be based on 50 members (note that many guilds purposely run 48-49 to get an easier matchup). I'm sure they know a lot more about this than we do, but they also don't necessarily play this game every day either. We do. And I agree with the comments that they are just wasting everyone's time when they put these gross mismatches together. Many guilds like ours put almost as much (maybe more) time into TW preparation than they do actually playing the TW. I've been on both sides of mismatches, and I can tell you it is a huge let down on both sides. Not fun for anyone and a huge waste of the time and effort put into preparing for it.
Hey all, look here now. If you want a match where you are the clear favorite, here is what you do: lose.
Lose your TW matches again and again and again and again. You will (probably) find that your opponents' get less and less powerful. If you keep beating the opponents that out-power you, do you really expect that the algorithm should give you easier matches?
Think of it as a round robin tournament bracket. If you win, you move on to face tougher opponents. If you lose, you fall back into the second chance bracket. Only difference in TW is that you never get eliminated, it's just a never-ending tournament.
Yes, there will be matches where you are the underdog. Should you complain about it? NO! Did the number 16 seed University of Maryland-Baltimore County complain when they were matched up against number 1 seed Virginia in the 2018 NCAA basketball tournament? NO! They were happy about it, then went out and exceeded all expectations by beating Virginia on the court! Did the Hickory Huskers complain when they were matched up against the taller, stronger South Bend Bears in the 1951 Indiana High School Basketball State Finals? NO! They stuck to their strengths and pulled out the upset victory! Do I have an analogy that does not involve the sport of basketball? NO! But that's not the point!
The point is, if you are winning any TW matches at all, then you should expect your opponents to get harder to beat.
No sense complaining about it ... unless you have never won a TW. Then maybe you have a legit gripe.
One of the issues we are facing is while we might have a similar match up on GP, their teams totally out match ours.
Example is we have quite a few FO, Sith and Empire teams but these do not easily match the standard defence of CLS and JTR teams. We have 25 CLS teams, 8 JTR teams and 1 Kenobi at 7*.. Yet we are matched with 2 front walls of Full G12 CLS/ 7* GK or multiple JTR teams more than likely because of our other squads match that 90-100k range.
Removing our issue of having a few eclectic squads, I personally think this is the main issue with the matchmaking. Squads are so different in defence or offence power that if it is a straight match on squad power there is always a very strong chance of a mis match.
I don't have an answer other than matching on specific toons which would be pretty complex..
I actually 100% disagree with you. If 2 guilds are evenly match on GP, then it's all strategy (farming, guild make-up, and map placement). Yes if you dont have key toons for pvp combat then that's on you not the dev team to match you to someone who made the same farming decisions. IMHO
The only situations where this starts to become less true is when it's the number of guild members starts to get skewed, like 50 vs 30 or something like that. Then the 30 could have an advantage of a wider range of toons.
You misunderstand what I mean here as I agree its all about strategy, what you have farmed and the composition of your guild.
My point is if that taking the snippets of posts by the devs, the algorithm seems to match players by a mixture of overall gp, squad numbers over 6000 and top squad gp. People are going to perceive to see a mis match because all toons are not equal even if a squad has a similar GP lvl.
A lot of the comments you will see is of people saying... "I cant believe we been matched with so and so squad when they have X number of this squad yet we only have Y number."
I would hazard a guess if you saw the averages, they would have similar GP average for the squads between the two. Its the composition of those squads that may be different what I did not very eloquently state.
I don't know if that's accurate...we were 16-4, and this last one our 66m GP guild got a 92m GP guild. And we throttled them.
It doesn't seem like it's getting harder at all. We've obviously had tougher matches (see the 4 we lost and a few of our wins) but it's been seemingly random.
One of the issues we are facing is while we might have a similar match up on GP, their teams totally out match ours.
Example is we have quite a few FO, Sith and Empire teams but these do not easily match the standard defence of CLS and JTR teams. We have 25 CLS teams, 8 JTR teams and 1 Kenobi at 7*.. Yet we are matched with 2 front walls of Full G12 CLS/ 7* GK or multiple JTR teams more than likely because of our other squads match that 90-100k range.
Removing our issue of having a few eclectic squads, I personally think this is the main issue with the matchmaking. Squads are so different in defence or offence power that if it is a straight match on squad power there is always a very strong chance of a mis match.
I don't have an answer other than matching on specific toons which would be pretty complex..
I actually 100% disagree with you. If 2 guilds are evenly match on GP, then it's all strategy (farming, guild make-up, and map placement). Yes if you dont have key toons for pvp combat then that's on you not the dev team to match you to someone who made the same farming decisions. IMHO
The only situations where this starts to become less true is when it's the number of guild members starts to get skewed, like 50 vs 30 or something like that. Then the 30 could have an advantage of a wider range of toons.
You misunderstood what I mean as I agree its all about strategy, what you have farmed and the composition of the guild.
My point is, taking snippets of the devs posts it seems that part of the algorithms look at player gp, number squad numbers over 6000 and a match between the top gp of squads.
A number of comments you see is "I cant believe we have been matched with so and so guild with they have 'X' number of this squad when we only have 'Y' number. What i would hazard a guess at is if you looked at both guilds the numbers actually are reasonably close, its the composition of the squads that is different.
People are going to perceive this as a mis match as not all squads are equal even at a similar GP lvl
Hey all, look here now. If you want a match where you are the clear favorite, here is what you do: lose.
Lose your TW matches again and again and again and again. You will (probably) find that your opponents' get less and less powerful. If you keep beating the opponents that out-power you, do you really expect that the algorithm should give you easier matches?
Think of it as a round robin tournament bracket. If you win, you move on to face tougher opponents. If you lose, you fall back into the second chance bracket. Only difference in TW is that you never get eliminated, it's just a never-ending tournament.
Yes, there will be matches where you are the underdog. Should you complain about it? NO! Did the number 16 seed University of Maryland-Baltimore County complain when they were matched up against number 1 seed Virginia in the 2018 NCAA basketball tournament? NO! They were happy about it, then went out and exceeded all expectations by beating Virginia on the court! Did the Hickory Huskers complain when they were matched up against the taller, stronger South Bend Bears in the 1951 Indiana High School Basketball State Finals? NO! They stuck to their strengths and pulled out the upset victory! Do I have an analogy that does not involve the sport of basketball? NO! But that's not the point!
The point is, if you are winning any TW matches at all, then you should expect your opponents to get harder to beat.
No sense complaining about it ... unless you have never won a TW. Then maybe you have a legit gripe.
I agree with TVF : I don’t feel it’s like that. For the last few months, it’s become worse and worse, while we lose more. We achieved big successes in November and December, so I too considered it possible, when we faced greater odds later, that it was due to our overachievement earlier - but mismatches continued.
I would have to check, but I’m pretty sure the last four TW have been increasingly worse - despite consecutive losses. The last was 19M GP, the previous one around 12 or 14M - we lost it, yet we get a worst match.
Replies
my guild got matched with alliance alpha, then wallet warriors 2 weeks in a row, both are like 16m GP over us (alliance pi)
I'm not attacking you, but how exactly did you come up with this number?
Swgoh.gg
yep
Argument about lower guilds is not really relevant.
There is so much differences in what ppl have.
Last tw we faced a guild 29mil GP higher (80-109), guess what, they wiped the floor with us but we put up a decent effort.
This time the difference is 6/7mil.
As others, every single tw we have faced a bigger guild since forming in Jan this year.
We won quite a few but also a few have been totally unfair matchup's.
I'm still struggling to see how the matchmaking is done.
For instance on this one they have more jtr's, more zeta talzins, more zeers, more 7* gk's.
So we used to think as we got a few zeta talzins might be why we getting higher guilds but not too sure now.
We only have 22 profiles of theirs to go off but seems a majority of the bigger members are on swgoh.gg.
Our 50 profiles shows 191 G12 toons, their 22 profiles shows 141.
So although cant be confirmed, they probably have more gear 12 than us.
We have 350 zetas, their 22 profiles shows 275, so they probably have more zetas.
So why this matchup in terms of GP seems fairer, they seem to have more of most of the meta than us, more g12 and more zetas.
We however have faced a guild with 15/20mil GP more than us that had TB GP and had a lot less high end meta teams.
May the force with us.
I understand that you want them to reassure you of things, but they have stated that.
https://forums.galaxy-of-heroes.starwars.ea.com/discussion/162893/dev-post-matchmaking-clarification-post-3-15/p1
What we want is for them to explain whether or not they think these matchups are WAI, or do they concur that its broken and they are working on it. Or at least acknowledge that these matchups shouldn't be happening.
The only communication we got so far is "hey look everyone we fixed it (but not really), but we will continue to monitor".
The purpose of this thread was 2 fold.
1.) Receive some communication from CG that these matchup are indeed NOT what they wanted.
3.) and to let CG know that our guild is now using this TW as a practice/sandbox mode and that our guild no longer cares about this aspect of the game (which certainly can't be what they wanted), because of the same algorithm that is supposed to keep things evenly matched.
What I would like to see is some type of response for the reports of unfair matchmaking. For those matches and reporting an unfair match it would be reassuring to get a response of “We have investigated your complaint and have concluded you were correct that the system did not give you a fair matchup”. I know that is a long shot but hey, I can dream.
From some of the previous information CG_Rydiggs has put out, I dont think they look at the data each time matching happens, they look at the data after the set has run. I have asked for a "bench mark" before so that we could have an understanding that you are looking for, to say, "yes this seems reasonable" and "no this doesn't", but the issue seems to lie in the fact that there are many parameters and that the ones the players focus on, are not what the algorithm looks at, so it's not a straight answer.
This also leads down the rabbit hole of each time he answers it gives insight into the algorithm, but let's avoid that discussion for now.
I have flagged him and I know he keeps an eye on these topic, but I dont fully expect you to get a direct answer.
Sorry, I may be over simplifying, I should have said are not as high a priority. I'm sure they are taken into account.
The problem is players feel that there should be very strick limitations on certain parameters, and in all honesty I have no idea if that's the right approach either. But to say it's just wrong, is not right either. We dont know enough to make that call.
His point still holds, the algorithm is clearly poorly designed if it's taking into account all of the parameters but ones that are clearly not impacting the outcomes as much are getting higher priority than other parameters.
From what we have been told they are trying to match on more of a player by player basis, which could be more accurate then a simple numbers match of GP and # of g12 toons.
Not that this is the case here but there is a lot of strategy that goes into TW and to always say "it's not an even match" is not always fair on the system either, I'm sure many guilds could overcome a deficit that they didn't realize. There is a lot to be said for farming strategy also, people can have a bunch of g12 toons all under the mindset of raids that dont necessarily help in pvp combat.
This is a long road to develop a good matching system, we are still "early on" in that journey. Let's give them a chance before we just start calling it a poor design because it doesn't match the exact parameters we think it should, especially since we dont know if that's even true.
I understand that, I'm just saying you can't discount his point because of simplifying it, either way there is still definitely something wrong with the algorithm that is leading to huge mismatches in some wars, but then still able to make for fairly even wars in other guilds.
Also we've had 20 TWs now, I'm not really sure how much longer it will count as "early on" in the matchmaking system. The wars I've had recently have been fairly even, but I know if I were in a guild that kept getting mismatches I would be really discouraged from playing in TW and would want this to be altered asap or at least have some sort of acknowledgement that the algorithm is clearly working well for some guilds and not as well for others and that they're looking into that.
Ok..... How many of your guild's players are participating in the current TW, and how many defence spots are available per territory?
I was not trying to dismiss what he said, I just wanted to not over simplify anything, there is a big difference between comparing things and not looking at others vs. Comapring all factors and properly weighing them. I miss spoke and didn't want to over simplify this whole concept.
It’s so at least since the change in matchmaking (which led to a fix). The mismatches are way more pronounced now than earlier (where, when facing guilds with 6-8M GP, we could beat when playing smart - but it was a early in the TW life, and many guilds approached them in an inefficient way).
We get all 50 to participate and theres 25 spots per territory.
I also want them to know that half our guild isn't even playing this TW, mostly out of a complete lack of care about the game mode they designed and feel is WAI.
and for the record...our opposing guild had 49 signup and we had 50. So let's lose the active GP argument once and for all.
Example is we have quite a few FO, Sith and Empire teams but these do not easily match the standard defence of CLS and JTR teams. We have 25 CLS teams, 8 JTR teams and 1 Kenobi at 7*.. Yet we are matched with 2 front walls of Full G12 CLS/ 7* GK or multiple JTR teams more than likely because of our other squads match that 90-100k range.
Removing our issue of having a few eclectic squads, I personally think this is the main issue with the matchmaking. Squads are so different in defence or offence power that if it is a straight match on squad power there is always a very strong chance of a mis match.
I don't have an answer other than matching on specific toons which would be pretty complex..
I actually 100% disagree with you. If 2 guilds are evenly match on GP, then it's all strategy (farming, guild make-up, and map placement). Yes if you dont have key toons for pvp combat then that's on you not the dev team to match you to someone who made the same farming decisions. IMHO
The only situations where this starts to become less true is when it's the number of guild members starts to get skewed, like 50 vs 30 or something like that. Then the 30 could have an advantage of a wider range of toons.
LMAO don't we all!
In all seriousness though. It's fair to say the MM is not perfect. In aggregate it probably serves it's purpose, but there are a lot of glaring holes in the system. Comparing player by player is probably not as good as comparing guild to guild imo. Obviously that's an opinion since we can't base that on fact. I can see how matching 2 guilds that have 100m but one with 30 players and one with 50 would be a mismatch, but putting 2 guilds that have an equivalent number G12 players shouldn't be. In the lower tiers, I can understand the need to balance around guilds that aren't all at 50 members. But in the upper tiers, it probably should be based on 50 members (note that many guilds purposely run 48-49 to get an easier matchup). I'm sure they know a lot more about this than we do, but they also don't necessarily play this game every day either. We do. And I agree with the comments that they are just wasting everyone's time when they put these gross mismatches together. Many guilds like ours put almost as much (maybe more) time into TW preparation than they do actually playing the TW. I've been on both sides of mismatches, and I can tell you it is a huge let down on both sides. Not fun for anyone and a huge waste of the time and effort put into preparing for it.
Lose your TW matches again and again and again and again. You will (probably) find that your opponents' get less and less powerful. If you keep beating the opponents that out-power you, do you really expect that the algorithm should give you easier matches?
Think of it as a round robin tournament bracket. If you win, you move on to face tougher opponents. If you lose, you fall back into the second chance bracket. Only difference in TW is that you never get eliminated, it's just a never-ending tournament.
Yes, there will be matches where you are the underdog. Should you complain about it? NO! Did the number 16 seed University of Maryland-Baltimore County complain when they were matched up against number 1 seed Virginia in the 2018 NCAA basketball tournament? NO! They were happy about it, then went out and exceeded all expectations by beating Virginia on the court! Did the Hickory Huskers complain when they were matched up against the taller, stronger South Bend Bears in the 1951 Indiana High School Basketball State Finals? NO! They stuck to their strengths and pulled out the upset victory! Do I have an analogy that does not involve the sport of basketball? NO! But that's not the point!
The point is, if you are winning any TW matches at all, then you should expect your opponents to get harder to beat.
No sense complaining about it ... unless you have never won a TW. Then maybe you have a legit gripe.
You misunderstand what I mean here as I agree its all about strategy, what you have farmed and the composition of your guild.
My point is if that taking the snippets of posts by the devs, the algorithm seems to match players by a mixture of overall gp, squad numbers over 6000 and top squad gp. People are going to perceive to see a mis match because all toons are not equal even if a squad has a similar GP lvl.
A lot of the comments you will see is of people saying... "I cant believe we been matched with so and so squad when they have X number of this squad yet we only have Y number."
I would hazard a guess if you saw the averages, they would have similar GP average for the squads between the two. Its the composition of those squads that may be different what I did not very eloquently state.
It doesn't seem like it's getting harder at all. We've obviously had tougher matches (see the 4 we lost and a few of our wins) but it's been seemingly random.
You misunderstood what I mean as I agree its all about strategy, what you have farmed and the composition of the guild.
My point is, taking snippets of the devs posts it seems that part of the algorithms look at player gp, number squad numbers over 6000 and a match between the top gp of squads.
A number of comments you see is "I cant believe we have been matched with so and so guild with they have 'X' number of this squad when we only have 'Y' number. What i would hazard a guess at is if you looked at both guilds the numbers actually are reasonably close, its the composition of the squads that is different.
People are going to perceive this as a mis match as not all squads are equal even at a similar GP lvl
I agree with TVF : I don’t feel it’s like that. For the last few months, it’s become worse and worse, while we lose more. We achieved big successes in November and December, so I too considered it possible, when we faced greater odds later, that it was due to our overachievement earlier - but mismatches continued.
I would have to check, but I’m pretty sure the last four TW have been increasingly worse - despite consecutive losses. The last was 19M GP, the previous one around 12 or 14M - we lost it, yet we get a worst match.