**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.

Flawed Match Making

12346»

Comments

  • Darksun987Darksun987 Posts: 83
    DNA3000 said:

    Just bc people say something is actually a problem bc it makes them feel better saying they were wronged instead of they just weren't good enough, doesn't mean it's actually a problem.

    I say it is a problem because it is mathematically provable that such tactics destroy the validity of the rating system, which then makes match making systems which rely on the integrity of the rating system also invalid.
    Bruh, boosts, revives, team healing potions... what validity are you on about?

    I'd be down for them to make a legit tournament as the last season of every year and in that tournament there's no consumables of any kind and straight ELO and that would be valid and confer actual bragging rights.

    As it is today it's just another resource farming mechanism.

  • Gregdagr8Gregdagr8 Posts: 380 ★★★
    Based on zero response from Kabam (shame on you), and off season matchups I've seen, it's pretty clear this matchmaking system is going to match based on prestige for season 16 as well. Enjoy another season of just fighting allies with the same prestige as you, not the allies around you on the leaderboard regardless of prestige to ACTUALLY see who's better in war. What a joke. RIP war for another season.
  • Patchie93Patchie93 Posts: 1,898 ★★★★
    If they want to Match based on prestige/war rating then a compromise would be that your season points are calculated based on war rating + Prestige. Not sure if I can explain it right but imagine
    Base points for a war x (war tier based on rating) x (y% of prestige of)

    Then at least if your stuck in tight wars with other high prestige alliances you'll at least get bonus season points for your hardship.
    Could also make it
    % of total prestige between the 2 alliances
    Or % of opponent's prestige

    Just tossing ideas out since obviously they aren't gonna go back to just war rating matchmaking
  • Matching by Prestige on top of War Rating would be fine (actually desired) for any Tiers EXCEPT for the top several tiers.

    In lower tiers, you easily have a mix of newer or weaker ally's looking to progress, alongside experienced and strong ally's who no longer take things as seriously. Pitting a 1000-1500 WR ally with Prestige of 7000 would definitely not be fair to match against a similar 1000-1500 WR but only 3000 Prestige ally. The 7000 Prestige ally which is probably just a casual (or not caring about AW) one would step all over the 3000 Prestige ally who might actually be trying to win and advance their ally.

    But once in upper Tiers, matchups should strictly be done by War Rating (and playing against those around you in the standings, so long as the matchups are not repeated).
  • Matching by Prestige on top of War Rating would be fine (actually desired) for any Tiers EXCEPT for the top several tiers.

    In lower tiers, you easily have a mix of newer or weaker ally's looking to progress, alongside experienced and strong ally's who no longer take things as seriously. Pitting a 1000-1500 WR ally with Prestige of 7000 would definitely not be fair to match against a similar 1000-1500 WR but only 3000 Prestige ally. The 7000 Prestige ally which is probably just a casual (or not caring about AW) one would step all over the 3000 Prestige ally who might actually be trying to win and advance their ally.

    Actually, this doesn't even work in lower tiers, although it is complicated to explain why. Let's say the average prestige of a 1500 war rating alliance is about 6000. In actuality we're going to have a spread of alliances at that rating, let's say most of them are between 3000 and 9000. Most will probably cluster near the average, and so most of them will get matched against most of them most of the time. The problem comes at the extremes. If you start using prestige as the secondary match criteria then the alliances closer to 3000 will tend to get matched *only* against the other alliances around that prestige. And by definition, half of them will win (outside of the rare tie). You are in effect giving half of the low prestige alliances "automatic wins" in that war rating. What's more, as they win they will rise in rating. When they get to 1550, or 1600, or 1700, they will actually be *even more extreme outliers* in that war rating, because they will be even lower than the average. So they will continue to get preferentially matched against other outliers.

    Suppose I "inject" 50 "low prestige" alliances into the 1500 war rating area. 25 of them will automatically become 1560-ish alliances after one war, 12 will become 1650 alliances, and at least one of them will make it to 1800+ *automatically* because they won't get matched against anyone else besides each other or other low prestige alliances that happen to be already there.

    If you think that a 7000 prestige alliance is going to probably "step all over" a 3000 prestige alliance, then with random matching (at the same war rating) a 3000 prestige alliance will eventually by chance face one, and they will only get past them if they are very good. But with prestige matching, those 3000 prestige alliances will never face a 7000 alliance ever, so the only thing they have to do is beat other 3000 prestige alliances. If there are enough of them, some of them will get to very high rating because they can't all lose against each other. It is a kind of positive feedback that a two-criteria match system contains: it takes outliers and creates even more extreme outliers.

    To put it another way, if you're saying that on average a 3000 prestige alliance is much weaker than a 7000 prestige alliance, then by definition matching based on prestige (for a give war rating) means all the 3000 prestige alliance have statistically easier wars than the 7000 prestige alliances. Any one 3000 prestige alliance might be stronger than any one 7000 prestige alliance, but you're certainly having an easier time overall if all your matches get pulled from the set of all 3000 prestige alliances than the set of all 7000 prestige alliances. Which means matching by prestige offers an unfair advantage to the lower prestige alliances.
  • Yes, the lower Prestige teams, if they keep winnng against other similarly low prestige teams (“IF” being the key) will be able to rise up against more casual Higher Prestige Teams.

    But as you say, let’s start out neutral, and half of the teams win and half lose. Then of those winners (who had all climbed a little in WR), now those “winners” play against each other. And then only the 25% who are double winners continue to move up, then they are playing against 12.5% of their similar Prestige ally's. And gets narrower and narrower.

    ...(acknowledging that occasionally there may be some odd case wars where they are being match against a same Prestige + WR ally that due to shakeups, etc, is on a downward fall, and not part of the narrowing field of rising alliances)...

    So by the time lower Prestige ally's start to reach the “Tournament Caliber” Upper Tiers, there are actually not that many of them. But to continue on from there they are now starting to enter the HIGH LEVELS, where as I propose they now would have to actually start playing much higher Prestige teams.

    And that is where basically they will be kept, at the edge of that Tier Cutoff (where matching algorithm ignores Prestige). They will most likely lose matches when they are just above that threshold (matched against a high alliance), and when they fall back under then they might again match against a Same-Prestige team (but any of those low-Prestige teams up there have had to at least keep winning against “all other” low-Prestige teams.

    But they WILL NOT realistically be able to climb up WITHIN those upper Tiers. It will keep them just outside of that (so long as they are the “best of the best” of the lower-Prestige ally's).
  • ... and to further look at it from the Lower-Tier example. Having lower/newer teams face off against much higher Prestige teams (just because they have the same WR only), may essentially keep most of them from even advancing up past some of the very low Tiers at all (because there are enough High-Prestige Casual ally's sitting down there that would all essentially “Block” newer ally's from being able to Win enough wars to climb past them.
  • Yes, the lower Prestige teams, if they keep winnng against other similarly low prestige teams (“IF” being the key) will be able to rise up against more casual Higher Prestige Teams.

    But as you say, let’s start out neutral, and half of the teams win and half lose. Then of those winners (who had all climbed a little in WR), now those “winners” play against each other. And then only the 25% who are double winners continue to move up, then they are playing against 12.5% of their similar Prestige ally's. And gets narrower and narrower.

    ...(acknowledging that occasionally there may be some odd case wars where they are being match against a same Prestige + WR ally that due to shakeups, etc, is on a downward fall, and not part of the narrowing field of rising alliances)...

    So by the time lower Prestige ally's start to reach the “Tournament Caliber” Upper Tiers, there are actually not that many of them. But to continue on from there they are now starting to enter the HIGH LEVELS, where as I propose they now would have to actually start playing much higher Prestige teams.

    And that is where basically they will be kept, at the edge of that Tier Cutoff (where matching algorithm ignores Prestige). They will most likely lose matches when they are just above that threshold (matched against a high alliance), and when they fall back under then they might again match against a Same-Prestige team (but any of those low-Prestige teams up there have had to at least keep winning against “all other” low-Prestige teams.

    But they WILL NOT realistically be able to climb up WITHIN those upper Tiers. It will keep them just outside of that (so long as they are the “best of the best” of the lower-Prestige ally's).

    I don't understand how you can acknowledge this is a problem, but just not care if it afflicts brackets below a certain level. If it is a problem in the high tiers, it is equally problematic in the lower tiers. Especially when it takes more work to solve it for just the high tiers than it would to simply eliminate the problem everywhere.
  • ... and to further look at it from the Lower-Tier example. Having lower/newer teams face off against much higher Prestige teams (just because they have the same WR only), may essentially keep most of them from even advancing up past some of the very low Tiers at all (because there are enough High-Prestige Casual ally's sitting down there that would all essentially “Block” newer ally's from being able to Win enough wars to climb past them.

    That's impossible. For those high prestige "casual" alliances to be in a particular rating at all, their average war performance has to be comparable to the average performance of all alliances in that rating. A prestige 9000 alliance at war rating 1500 CANNOT be harder for a low prestige alliance to beat than the average war rating 1500 alliance, because if they were harder to beat they wouldn't *be* war rating 1500. They'd win too often, and rise in rating.

    In general you're either going up, going down, or staying roughly the same in war rating. And eventually, alliances rise or fall to a rough equilibrium, at least to a first order approximation. If you face a 1500 war rating alliance it has to be just as good and just as bad as any other - or else you just happened to match against an alliance that is going up or going down. If they aren't going up or down that means by definition they are about as hard to beat as any other. If they are much harder to beat, that means they are going up and won't be there later. If they are much easier to beat that means they are going down and won't be there later.

    But it is literally impossible IF you match just by war rating, for there to be unusually strong alliance at a particular war rating that are much harder to beat than any other. Because in a match system where all alliances have a chance to face all other alliances of the same rating, you can't be stronger than your peers permanently. So blockades as you describe them are mathematically impossible.

    Ironically, they *are* possible in match systems that match by both war rating and prestige. Because in such a system, it is possible for an alliance to be much stronger than average or much weaker than average than all other alliances at the same rating, because they don't have to face all other alliances of the same rating. The problem you're describing only happens in the match system you're claiming attempts to address that problem.
  • QuikPikQuikPik Posts: 806 ★★★★
    edited February 2020
    Yep we have to endure another season with this match making system. For some unknown reason, the season leader board was down the entire off season.

    And not a single comment from Kabam even though KT1, Prof Hoff and RichTheMan made videos about this problem.
  • Gregdagr8Gregdagr8 Posts: 380 ★★★
    QuikPik said:

    Yep we have to endure another season with this match making system. For some unknown reason, the season leader board was down the entire off season.

    And not a single comment from Kabam even though KT1, Prof Hoff and RichTheMan made videos about this problem.

    They take down the ability to view the leaderboard like a week after the season ends. They've been doing that for a while now. No clue why.
  • HksBindraHksBindra Posts: 38
    They're probably fixing this now.
  • QuikPikQuikPik Posts: 806 ★★★★
    @HksBindra I doubt they'd change something like this in the middle of a season.

    For some further testing, I dropped my test alliance prestige from 4000 to 3200 and guess what...all of our matches also dropped down to other 3200 prestige alliances.
  • HksBindraHksBindra Posts: 38
    QuikPik said:

    @HksBindra I doubt they'd change something like this in the middle of a season.

    For some further testing, I dropped my test alliance prestige from 4000 to 3200 and guess what...all of our matches also dropped down to other 3200 prestige alliances.

    @QuikPik they may have a reason now.
    Look at the alliance that's master rank 1 now. Overall master rank 4.
    They are the same losers I've been reporting here from the start.
    I've kicked many of them from my alliance back when my alliance was around 2500 rating.
    They can't even complete map 6 without items.
    They are a joke. Some of them aren't even level 60.
    And they're about to beat one of the top spending alliances in season rankings.
    I'm sure players from those top 3 alliances have emailed kabam.
    Kabam can't really afford them not playing.
    The only time they do something is when it starts affecting their pocket. Everyone knows this by now.
    I may be completely wrong but maybe they're finally fixing it.

    And you nailed it the first time - matchmaking is completely based on prestige.
    I offered a solution here but no one discussed that further.
    I think that solution would work well with everyone.
  • Gregdagr8Gregdagr8 Posts: 380 ★★★
    @QuikPik what are you seeing now? Our last two wars were vs 9200 and 9500 prestige allies that finished plat 1 and plat 2 last season. But we destroyed both of them (my ally is 10,455 and finished plat 3 last season). Our opponents died 173 and 154 times vs our defense. Also every opponent this season for us has been within us by 20 spots on the leader board as it should be. Last season we had matchups 464 and 600 spots away from us on the leaderboards. I think it's safe to say that last season was a joke and now these allies that got a free season of not matching with tough defenses are getting bashed back down the rankings. I think matchmaking is fixed now to what it should be. If a 9500 prestige ally can beat us, awesome, congrats. But at least this season we are now matching to give both a chance.
  • QuikPikQuikPik Posts: 806 ★★★★
    @Gregdagr8 not as far as I can tell. There's still lots of lower prestige alliances in the plat tiers.
  • RektorRektor Posts: 678 ★★★
    Picture it.. a 9000 prestige alliance with mostly r4 5* champs finishing 1st place in seasons above a couple dozen alliances that would beat them 300 to 0 if they could ever match in war.

    I wonder if that is what it will take for kabam to change matchmaking back to by war rating instead of by prestige.
  • GroundedWisdomGroundedWisdom Posts: 36,239 ★★★★★
    You don't Rank based on hypotheticals. You Rank based on the Points you put up from your own Wars.
  • QuikPikQuikPik Posts: 806 ★★★★
    Here's the updated list from 4-15-20. As you can see nothing has changed in their match making algorithm from last season.



    Green - highest prestige alliance in tier
    Orange - alliances where prestige is less than 20-30% of highest rated alliance
    Red - alliances where prestige is at least 30% less than highest rated alliance

    It's bad when in Plat 2 there's a 4300 prestige gap between the highest and lowest rated alliance. There's only 50 alliances in that tier. In Gold 2, the discrepancy is even worse. There's a 6800 prestige gap here. Yes there are a lot more alliances in Gold 2 but you should not have 4k prestige alliances in the same tier as an 11k alliance. Kabam is saying that an alliance fielding all R3 6* and R5 5* is as good at war as one that fields at best some R4 4*.
  • That 11k in Gold-2 is actually the outlier, and might not run a full 3 BG's, or at least plays casual without worrying about spending on Items, committing top champs to war, etc. So can’t really include them when saying how much of a Prestige spread there is in that tier.

    Wonder if the apparent change in matching that may have just happened here mid-season will have the full effect in standings by end of this season, or if maybe won’t truly be reflected until end of next season standings.
  • QuikPikQuikPik Posts: 806 ★★★★
    Remember, I've only done the top 50 in each tier. I'm sure there are plenty of 10k alliances below rank 50 in Gold 2.

    If Kabam has modified match making in the current season, it won't make much of a difference by the end of the season. The screen shots I saw in another post of 2 top rated alliances by war rating show something like +48/-8 adjustment. Alliances with inflated war ratings are still going to remain in the same tiers if they're only losing -8 in lop sided matches.
  • MoosetiptronicMoosetiptronic Posts: 2,106 ★★★★
    @QuikPik ah ha, it was your spreadsheet back at the start of the year!! Epic post, which laid to bare the ridiculousness of the prior system.
Sign In or Register to comment.