GA Algorithm - Worse Than Before??
June 25, 2019 9:21PM
edited June 2019
So I was excited that the GA matchup algorithm was going to be improved. I was also excited for GAC to start (correctly this time). The I go into my first round matchup...
The person I am against in round 1 has 220k more character GP than me (2.882 million to 2.661 million). Has a 7*, g12+5 Malak (I don't have Malak at all). Mod-wise he has:
68 15+speed mods to my 36.
57 100+offense mods to my 26.
This is not an improvement. This is far worse than any matchup I have ever faced in any previous GA. It's not better - it's worse. Please fix the game mode - especially if this is going to be the long term "dominant" mode to replace arena.
The person I am against in round 1 has 220k more character GP than me (2.882 million to 2.661 million). Has a 7*, g12+5 Malak (I don't have Malak at all). Mod-wise he has:
68 15+speed mods to my 36.
57 100+offense mods to my 26.
This is not an improvement. This is far worse than any matchup I have ever faced in any previous GA. It's not better - it's worse. Please fix the game mode - especially if this is going to be the long term "dominant" mode to replace arena.
In game name: Lucas Gregory FORMER PLAYER - - - -"Whale blah grump poooop." - Ouchie
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
1
- 60.8K General Discussion
- 9 Developer Q&A Archive
- 2.6K New Players
- 30.5K Arena and Character Strategy
- 4K Individual Events
- 4.6K Guild Events
- 4.6K PVP and Rivalry
- 50.1K Guild Recruitment
- 731 Art and Media
- 4.1K Off Topic
Howdy, Stranger!
Replies
So they did make it worse. Awesome.
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
I thought you were on a launch shard.
(I started in April 17 and I have exactly 68 myself)
Same. I never liked DR/Malak rosters versus rosters that had neither, but this is clearly worse than before.
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
Congrats on being lucky. Stay on topic please. If you want to "t", take it somewhere else.
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
So while the 4.5m GP account is still reasonable close to most in terms of character GP the 4.1m GP account isnt (500k difference with the highest char GP)
It certainly looks like something went wrong in matchmaking
*note the 4.1m GP is a very focused roster
You're the one that brought up mods. How on earth are they supposed to matchmake based on mods?
(and something tells me this isn't just a case of luck. since you brought that up too)
Seriously? Discord apps can check mods - how hard would it be to code them within certain ranges of each other?
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
It would be nice if they stated this is intentional and WAI. I sense a significant reallocation of my free time if that's the case.
I'm sure they can do it.
Why should they?
They indicated that the purpose of changing the matchup algorithm was so that players would "be matched up against someone who looks more like them." (from the GA matchmaking improvements thread. Whether or not you agree with that philosophy, that is their stated goal.
Big discrepancies in GP, mods or roster makeup (and I would consider one squad having DR/Malak maxed an another not even having them at all a fairly large discrepancy) do not reflect that stated goal. SOmething seem very broken regarding the current matchmaking. The same types of posts are proliferating Reddit - so hopefully the devs will see them there.
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
And if you give significant weight to the number of high speed mods as a determiner of how many effective units you can field, you've created an exploitable loophole in the system where units who don't care about speed mods- Baze, Droideka, Mission, Big Z, some Chewie builds, Resistance Trooper, Sith Trooper, Nest- become disproportionately beneficial in a way that the matchmaking algorithm would fail to account for.
This is not a failure of the matchmaking algorithm. You're just looking for an excuse to play the victim.
Apparently so are the several hundred people flooding the forums and Reddit with similar matchup concerns.
In the first (failed) launch of GAC, there was a guy in our bracket who had a 6* Traya, no Malak and no DR. Many of the rest of us had 2 (2 had all 3) maxed. And the differences in mods, g12, zetas, etc. were all way off compared to previous iterations of GA. I'm not "playing the victim". I legitimately think something is wrong with the matchup calculations.
The "new" algorithm has created far less even matchups within a given bracket than the previous one, as evidenced by the numerous screen shots (both in-game and from Discord) showing a much bigger discrepancy in matchups.
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
I hope they get to set the same number of defensive teams at least. The matchmaking is really in my favor this time, but it doesn't feel fair and I have a feeling it will be the other way around in 5v5 with fleet. While I'm glad that it's no longer purely based on raw GP, it still needs fine tuning.
If that's the metric we're going by, every single thing in the game is and has always been utter garbage and no one should spend one minute playing this garbage game.
You always set the same number of teams as your opponent. This was addressed in a dev post somewhere.
A lot of the people who actually show the full Discord readout- which is not the majority of those complaining- have pretty fair matchups but large toon GP disparities that appear to be caused by fluff.
We had one thread a few minutes ago where the OP was courteous enough to include links to rosters. They were complaining about a 400k difference in character GP, but I scroll down to the bottom of their opponent's roster and I'm seeing things like a 4* G8 level 85 Sith Marauder. The opponent had WAY more fluff.
These complaints are flooding, but are they really valid?
I'll let CG decide. That's why I am posting.
I realize my 8-player bracket is a small sample size - but the first GAC and this one are by far the worst GA matchups, in terms of balance, I have ever seen compared to any previous GA - no matter what metrics you want to look at - g12, zetas, character GP, mods, etc. There are much bigger discrepancies in all of those areas between the strongest and weakest players in these GAC brackets then there ever were before. I am sitting down dinner right now, but am more than happy to post full screen shots (and discord stats) to back up that claim tomorrow morning when I have more time.
But CG can check all that as well - and on a much larger scale. That's why I am bringing it to their attention.
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
2 of us have both revans. 2 have none. 3 have JR and 1 has DR. None have Malak.
Just stick to GP next time.
My first opponent has almost 600k (~50%) more character GP than me and is 2 divisions above me but I am deploying the same number of squads as I normally would.
The ironic part is that I am going to curb stomp him.
What's the problem exactly?
Mods are about the only thing I have in common with my opponent, so maybe it actually is. Otherwise I'm down 400k GP, 4 zetas, and 30 G11/12 characters.
I thought it was simple. Go by character GP when GA is just toons, and total GP when it's ships and toons.
People who had fluffed roster complained so cg change the algorithm to only account the top of you and your opponents rosters instead of all your roster. Thank the people who flooded the forum for this.
I'll type nice and slow, so you'll understand...
From the GAC match-up rework post from Crumb.... "they’ll be matched up against someone who looks more like them."
Now look at the bolded from the post you quoted. I'll wait. You'll see it eventually.
I wouldn't consider having the two characters that have been meta for the past 6+ months, while others have neither to be" looking like them. "
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*
I disagree with your fundamental complaint that the characters should be the same or similar. But even we accept that as a flaw, how does it make it worse than it was before? The matchmaking process has never factored that in. The only change we have identified is that it previously counted your entire character GP, and now it only counts the GP of the top X number of toons. If you want to argue it's worse, it has to be based around that change. I'd argue that they need to expand that number slightly (I'd suggest diminishing returns as you go past X), but I still can't see how this is worse.
Except that's not how it seems to be working. See the post from swgohfan29 for what I mean.
In game guild: TNR Uprising
I beat the REAL T7 Yoda (not the nerfed one) and did so before mods were there to help
*This space left intentionally blank*