**Mastery Loadouts**
Due to issues related to the release of Mastery Loadouts, the "free swap" period will be extended.
The new end date will be May 1st.
Due to issues related to the release of Mastery Loadouts, the "free swap" period will be extended.
The new end date will be May 1st.
Matchmaking Discussion [Merged Threads]
This discussion has been closed.
Comments
Its taking fun away and giving sadness opposite of what it meants to be.
War seasons pit your alliances roster and skills TOGETHER to put EVERYONE in order from best to worst, therefore it must be possible for anyone to math anyone else so long as war rating is equal, as the war rating is a numerical value given to your alliance that signifies its combined roster and skills.
Prestige is NOT the truest and best factor for determining ability, we are not rewarding prestige now, we are ignoring it. Those whose alliance grew up being over rewarded for the focus one that one factor: low prestige. Have become used to the rewards they received by avoiding fair competition on an even playing field. This season is a shambles but I don’t think any solution will appease those in the long run who have become used to rewards above their game progression level, and think they were performing well due to being shielded from fair competition by a systemically unfair system. If gw is correct with his claims that prestige was brought in due to his suggestions then he and kabam are to blame for leaving that system in place for so long, prestige is not a good factor for determining ability for progression. I also point out once again that the true new bronze and low silver alliances were also being shafted by having to fight over and over again the alliances 6 tiers above them and gold 2 reward bracket. And those of you arguing for the broken system are actually supporting a system that shafted them for the manipulators too.
Systemic inequality causes problems, this season is purely an indicator of how broken it was and how badly it needed fixing
Better late than never
We *want* to match approximately comparable alliances most of the time. But how do we do that? We can't match based on alliance rating or prestige: those only measure roster. They don't measure playing strength. We want to match roughly equal *competitors* and that has to take both into account.
But there's no way to just look at an alliance and know this. The only way to know if two alliances are roughly equal is to have them fight each other over and over, and see if they win and lose about half the time. That's impractical, so that's out.
An *approximate* way to figure this out is to have a bunch of alliances face each other collectively, and if they win against each other collectively about half the time, then we can say very roughly they are probably all equally matched. It is the best we have.
So we invent this thing called "war rating" which is our best guess as to how strong an alliance is. Every time they win, we increase this. Every time they lose, we decrease this. We assume that if their actual war rating is lower than their actual strength they will keep winning until they face equal competition, and then they'll start losing. When they win and lose about half the time, their war rating matches their strength. We assume this is happening to everyone.
No system is perfect, so there will be glitches. Alliances will form, disband, players will progress. An alliance's true fighting strength is never going to be exactly equal to their numerical rating. But if we keep matching them against everyone else at their numerical rating, they will only *stay* at that rating if they can win about half the time. If they can't, if they start losing more than winning they will drop in rating. If they win more than they lose they will rise. The system corrects itself over time, because you cannot have the "wrong rating" for long: actual competition self-corrects.
This is true if everyone has roughly the same chance to face everyone else of the same rating. If this is true, then everyone at that rating belongs there. Anyone who doesn't belong there will drop lower or rise higher. Only the alliances that win 50/50 (or so) will remain there. Everyone else will leave. This ensures something very very important:
EVERY ALLIANCE OF THE SAME WAR RATING HAS ROUGHLY THE SAME STRENGTH
This is the fundamental basis for how match making works. And if it was working correctly all along so called "bigger" alliances would ALWAYS have to face equally strong competition. They would *never* match against push overs, because push overs would never win often enough to rise in rating high enough to face them. There's a road from push over to top tier alliance, and you would have to go through that entire road to face them.
Why is this not happening now? Because the old system was broken. And this is the only way to fix it.
This doesn't "reward" bigger alliances. If you believe those alliances are getting more than they should this season, you have to believe they were getting less than they should in previous seasons, because rank rewards are a zero-sum game: there are only so many places at the top (and for that matter, basically everywhere). And if you believe alliances are getting less this season than they should, that can only be true if they were getting more than they should have in previous seasons for the same reason.
No one is "winning" or "losing" overall because of this problem with match making, at least not in general terms. I'm sure there are edge cases where someone is coming out ahead and someone else is coming out behind. That always happens when large changes like this occur. But *in general* the "winners" this season were "losers" last season and vice versa. And once every alliance that has the same rating also has the same approximate strength, which is how the system is supposed to work, matches will even out again.
And a year from now when someone posts a picture of them matching an alliance with double the alliance rating but equal war rating and claims that's unfair and the system needs to change, remember people like that were the cause of all of this in the first place. And remind them why they're wrong.
Also: war is just plain different when you're fighting with less than three BG. The competition is just a lot more random, because obviously the vast majority of alliances taking war seriously would be running three BGs. I'm in a similar position where I sometimes run two and sometimes one BG. The level of competition is wildly diverse. Sometimes we're running up against casual but very strong alliances that aren't full. Sometimes we are running up against up and comers that are still building up their alliance membership. Sometimes we match against people who give even less Fs than we do and look like they are trying to make a statement by placing their defenders in Bizarro arrangements.
War rating is probably going to stabilize relatively quickly among the 3xBG alliances. But in the one or two group alliances, things will probably always be wonky, because they seem to have always been kind of wonky.
Again these broken war ratings are a result of systemically focusing on a dynamic that doesn’t equate to ability.
Technically as I said at the start a numbers system to keep matchups within a big range like 25% of prestige would probably work but during this season it actually would have prolonged the levelling effect
Imagine there's twenty 9k prestige alliances and they are all roughly equal in strength. So they match against each other and win about half the time, so their war rating remains roughly stable. Let's say that's 2500.
Now imagine there's two hundred 5k prestige alliances. These are more moderate alliances, and there are more of them, and thus they have a distribution of strength: some stronger and some weaker. They all match against each other as well. They all start around 1500 rating, say. The stronger ones will increase in rating, and the weaker ones will decrease. But if the system keeps matching them against each other only, the stronger ones will keep drifting higher and higher in rating. And the higher they go, the less likely they will match against anyone else besides that same group of two hundred, because it gets increasingly less likely they will find a "suitable match" - the higher their rating gets, the fewer alliances that have that same rating *and* prestige. They become an outlier.
If they keep matching against similar prestige, their rating can just keep climbing without limit. Eventually it can exceed 2500. At that point, they've climbed above all of the 9k prestige alliances. They all continue to fight each other, so their ratings don't rise. But the strongest 5k prestige alliance (in this example) could have 2600 war rating. That means they are now the top alliance in this example, and every other 9k alliance drops down one.
This is a highly oversimplified example, but it illustrates how the older system could cause lower strength alliances to rise above higher strength ones, and then actually push them downward. Because ranking is zero-sum, the rewards that alliance gets come from the alliances they pushed downward.
Remember: with the current system doing this is probably suicide because when those guys jump into the new alliance they will have to face alliances with almost double their prestige and probably much higher war strength. But a prestige match window allows them to dodge that competition and ask the game servers to find them easier prey. And the more often they win, the harder the game servers will be forced to look to find competition, because they will become more and more extreme outliers (alliances with low prestige and high war rating are increasingly rare the more disparity between those values). Which will probably make their matches increasingly weird.
The Rewards had to be restructured. I don't argue otherwise. Saying people went down because others were going up isn't wholly accurate. Not when the Win/Loss ratio is so close to breaking even. You need at least 75% wins to make significant progress through the Tiers.
One thing I'm damn sure of is taking advantage of the momentum of coming up against Alliances with 3-4x less strength isn't any kind of accomplishment or solution.