Ken I never squashed your point of view I think what you have to say is valid. I just thinking I’d would say a majority of people have stuck feeling with these 50/50 split from what I’m seeing. Truthfully I don’t have a perfect way to fix it. I honestly have a rank in mind I feel I deserve that isn’t onyx. I don’t feel I’m an onyx player. still when I rank up to D2 and then take three or four wins and then one rough loss. With little gain for the wins I might add, and fall back to half D1 yeah I get a little bit flustered. Those wins should mean something. I have hard time having fun with the fear of losing my diamond without knowing fully why at times.
1 Like
I agree 100%. Keep it simple. No one wants to feel they are stuck, and it should be very clear to anyone playing ranked why they are rising or falling.
1 Like
Its tough bro. The context of the match, average mmr between both teams is what’s really looked at. Let’s take this example though in terms of winning more, losing less.
A player can be low platinum. This player got gold rating after 10 tough matches, and then afterwards started winning games. Each time this player was about to rank up, would lose against a team of players above their teams and individual skill level. Let’s say this was a repeating factor. The player could win against peers of gold and platinum, but once any daimonds came into the picture would lose, even with 1 on his team.
The match maker sees the repeating factor, OK. It’s decided this player skill ceiling is platinum tier as of right now. There’s some options.
- give this player as many fair matches as possible within the respected skill spectrum they are at? Only challenge the lobby when this player has a hot streak should it happen? This would measure if the player has surpassed the current ceiling.
Resulting in winning lots of matches, but against against equal to or slightly below their level with expected results and statistics. No crazy kd, accuracy, just average performance. Not getting 30 kills 5 deaths and good obj plays. This results in very little csr gain, and the prevention of moving up is to have that 1 tough match erase any build up of csr should that player lose that 1 single match. (Sounds frustrating, but does it make sense? Is it too harsh? This is what is under scrutiny). This is a means to keep the player in their actual skill range since this is a ranking system and not a progr system.
Or
- push challenging games onto player where the player is expected to lose, but jump the rank up should the player win. Resulting in a far greater loss amount, bad stats from nothing but hard matches, but a higher rank and faster increases should the player win. After sustaining many losses though, does it not make sense to scale matches down? This is where the issue is. Scale matches down…win a lot of matches against peers…but to move up you need to win matches that matter. Which is against people better than yourself, and to a degree.
I believe the ratio of csr gain and loss against peers should be tightened. This is what’s most frustrating right now. That in a scenario where you may baaarely have even 2% advantage over the other team in terms of beating them…you’ll get 1csr for the win, upwards of -10 for a loss at times. When the win % is far above the other team, then and only then does it make sense for a harsher derank. Would be easier for us tell if we had some form of graph detailing how it all works so when we lose a lot we can sit back and say damn…we had a bad game against players we should have beat…or…those players are on the upswing and below where they currently are if they won by such a big margin.
We can all try our best to make sense of it…with many users having great points…but there are also many variables at play that factor so much of our matches, the outcomes, the csr gain or loss,etc. I definitely understand the frustration though.
1 Like
The last bit really sells it for me man. I’m even more stressed now that I realized that my buddy that is D6 is almost statically that same as me except deaths…i would say fair enough if my obj stats weren’t beast mode above his. Even if that obj stats aren’t worth as much our rank should be closer especially if I have more wins. D1 vs D6 from deaths’ mainly, this is what inspires shellfish play. Don’t reward support players and why should they bother supporting anymore…
I’ve seen onyx matches barely anyone touches the ball. I’ve got teammates that have reach onyx and wont play now. They got sick of trying to keep it all the time.
same experience, its insanity. you go negative in slayer and get rewarded. have best k/d but last place or 3rd you go up one. win 20/25 and get objective games 3 vs. 4 no chance to win and you lose all the progress… im fine having harsh losses but reward wins correctly and dont let uneven bad servers affect players
That only happens when you reach the peak of your skill level. For instance if you were Gold 5 but playing like a Platinum 6. You would lose almost no rank for a loss and gain huge amounts of rank per win.
The question has to be asked though, are these lower rated players only winning because of their higher ranking friends?
These same players can’t rank up when they play alone, yet their friends still manage to. So it would suggest the answer is yes.
I believe you may be overlooking a lot of what people have been expressing, and the really frustrating experiences that are pretty much detailed and listed out many have experienced. It’s not coincidence that there exists so many cases of it.
There are some instances where you can clearly see someone’s own performance is holding them back, perfect example of why they’re stuck where they’re at. Then there’s instances where the lowest ranked player in a high ranked lobby performs on par with his team mates but gains 1 point for a win and loses more than them during a loss.
The team make up affects the rankings drastically. This is where its just difficult to figure what to do. On one hand you have a player that belongs in those lobbies and doesn’t have the rank that shows it. The losses can be very close games. Not losing by land slides , and not always losing. The team mates can have that close to 50% win loss you’re speaking of. How could the lower ranked player ever get a higher ranking if the match outcomes usually consist of that win loss ratio, and given the evidence that the wins amount to much less than the losses, resulting in negative overall csr each time they play together.
I’ve seen the cases before where it’s really play time that has created the gap between them with 1 person not on as often. That player gets more gains playing alone or with others, but wants to play with their main friends. So it results in a frustrating experience most of the time.
Its like taking a team of friends that all play together. Let’s say they’re all upper onyx. 1 player decides to make a new account. Other 3 are at their peak already in terms of highest csr. How do you see the experience of the new account going, as they play placements with those team mates and playce into D1. How will the rank up journey go? Will the matches cater to the new account and suddenly they’re winning every game so that player gains rank faster? Or will it go pretty much just like its been described here?
The higher the players are in csr, the more drastically affected. If its 3 friends in 1500s, what we’d normally see is a small inflation of their rank and their team mate would attain onyx fairly quickly. If those 3 friends were 1700s+ it becomes a very different experience. The statistics change drastically and you see less crazy high performances with large positive kdr. Less opportunities for the system to get flags on that new account that would suggest they’re under ranked. The reason for this is just the sheer caliber of the players within the lobby. I’m sure you get that without me needing to explain it.
Perhaps this is why other games put a lock on who you can play with. You can’t play with players that are so many tiers above you. To my knowledge we’ve never seen this in a halo game, nor heard of it ever being thought to attempt. They probably don’t want to place things that wouldn’t allow us to play with our friends.
I think the match maker still tries to make fair match making based on average mmr across 2 teams…but it pulls in very high csr players still. Should that new account play without those other high ranked players, the player faces no players of that caliber so Is able to exceed expectations and have the larger statistic margins in their wins by being that 1 player who influences the match to a high degree each match.
Its not only a case for high onyx examples since skill gaps exist across all levels, but I used the onyx example since it’s easier to get, and the general level of player can be seen as showing an extreme case vs lower tier players having this problem may not feel it AS much, yet still probably deal with it.
I may not be replying about the exact reference you’re pointing towards in terms of the player not being able to win alone. You may need to quote what I said so i can remember. I’ve replied a lot and can’t remember, maybe I used a different example or you’re referring to someone in this thread dealing with the problem theyre trying to express.
The worst player (who gains 1 CSR for a win) is offset by the enemy team having an equally worse player, or by the full team comprising of players lower skilled than their own teams best player(s). The system is aiming to match players on a 50/50 basis so their low skill is offset by equal low skill. Winning that game doesn’t prove they are of higher skill, it only proves they are capable of performing at the minimum expected level for their current rating.
Everyone who has cited they should be a higher rank has not provided evidence that supports this in any way.
Ultimately if they play better they will win more games the system expects them to lose, perform better against their expectations and rise in rank. People are gaining minimal CSR for wins when their rank is either equal to or above the MMR expectation. Just like someone who is climbing due to an increase in skill will see minimal CSR losses when losing and large CSR gains when they are winning.
There are only so many “examples” of it because players feel they deserve to be a higher rank than they are when that is in an extremely large majority of cases, just their own warped interpretation of their own skill level.
The big measurement is performances Vs expectations so if a player was playing with 3 1700+ then you could determine win rate Vs opponents. This would identify based on win % when that player is used as a balancing tool to determine what their effective MMR is. If the MMR is underestimated then they will disproportionately win games even ones expected to lose. If it is overestimated they will lose significantly more. This will help identify the true MMR of the player. The fact the data takes into account all games ever played, it is very hard to imagine the system not generating a strong picture of a players skill level when compared to millions of other players and being able to match games so effectively and consistently as a result. It’s evidence that the system is working to a degree. Can it be fine tuned more? Probably. Would it result in these players gaining a higher rank? Unlikely.
Doesn’t matter once the game caps you it will drag you back when solo.
Say you have a team that is one Onyx,one diamond five, a diamond three, and a diamond one like my team often is. You will be pulling in the worst possible teams that play against… The MMR system will bring in a team of onyx squids or cheaters because it doesn’t know how to pair up. If we do get a legit squad that is the appropriate real ranks they, the match usually goes pretty even or we will win one or squeak it out and then you get one point towards rank. Or if the team’s a squid team were likely barely gonna win or squeak it out or lose. All of which hurt you or you barely get ranked because it isn’t considered not a good matchup. These players aren’t honestly ranked. The system corrects fast but making new accounts is so simple. Not to mention getting deranked for lag outs and cheaters…
Early on in halo infinite release I played very objectively and focused on being a support player I basically have to wait for the rank to reset now to flip the games understanding of what it believes me to be as a player. Other playlist obviously I have a positive KDA. In rank objective I focused on the on objective and being a support player, not reckless tunnel vision but support. I won a lot of games because of that. Now is the arduous journey of protecting my KDA for a few months now so the game MMR system can collect enough stats to release me from my wall.
Is all really a huge letdown right now because I was having so much fun with my friends ranking up but now that we’ve all placed it is totally ruined our experience playing together. I’m talk 20+ year halo guys, we do care. My cousin asked me if I’m even have fun because he can tell how tied up in the stats I am now.
1 Like
Exactly, the game should be exciting because your winning. So many games I am scrambling around just for kills instead of enjoying victory
I’m not sure you understand. The best way to rank up right now is to play for stats. When you play for stats, your losses hurt less and your wins reward more. That’s the problem.
I get and agree with this. However, I think the system should have a better concept of team history.
If I’m playing with ransoms I met on discord and just wanted to party up with, it shouldn’t be the same expectation for a win as playing with a squad that runs together all the time.
I don’t want them to punish trying to make new friends. Honestly, I have concluded the best way to rank up is to solo cue, because it’s easier.
I just go into the games, bait my team mates, stat and sorta try to win the game. The losses don’t really hurt because I’ll have the best stat line in the game and the wins are decent rewards.
What does this do: it makes me not want to make new friends or play with existing friends. It creates bad habits that wouldn’t actually work in a competitive setting (so in this sense, your rank has little to do with your competitive skill or ability). Like at a certain point your movement is good enough, your shot is good enough, and you have your clutch moments. Great.
Do you actually know what’s going on in the game and how to control the map with your teammates? Is that a skill we want to sus out in ranked or is ranked just about being a selfish player?
That’s not true. The system determines that players in a squad perform better than solo players for the same MMR. Therefore it will register the teams average MMR, let’s pretend for argument’s sake, your team is rated 1400 MMR. The enemy team comprised of just solo players could be rated 1500 MMR average. Numbers are made up, I don’t know the real values. They are not a team, but the game will still be close on paper. How does the system know? Millions of games for reference point and an expected win % that can determine if the squad is effective or not. Some teams of four are basically the same as four solo players as their teamwork is lacking which means they will disproportionately lose games Vs expectation and therefore alter the MMR scores of the team until they are no longer losing disproportionately.
You shouldn’t be going up against that many cheaters, but it’s proportional to the playerbase. You play cheaters, so does everyone else. It’s factored in based on averages. Your rank is a distribution of the entire community. If you are that unlucky, then play more games. It averages out. You will play the same number of smurfs or cheaters as everyone else over a large timeframe, at at a comparable MMR rating to yourself.
Your MMR is affected across every playlist so technically your ranked rating is no different to how you perform in social. Don’t forget you are being paired loosely on skill in social so you may be performing well, but that doesn’t mean you are a high ranking player if you are doing it against lower rated players. MMR moves pretty damn quick which is why smurfing is essentially useless beyond a handful of games. You are behind an ability wall, not a stats wall. Say if you gave your account to a 1700 Onyx player, they would be seeing large CSR gains rapidly after a handful of games because they would perform significantly better than expectation repeatedly and against scaling opposition.
I’m sorry that your friends and yourself put so much stock in a rank that you can’t enjoy playing the game together due to it. Ranked is there to secure close and competitive games on a more consistent basis than social, not to constantly climb. Every player hits their skill wall at some point. They will then hover around a single rank until they improve. This doesn’t mean just stats, although obviously the two are directly linked together. Being so focused on the rank will take away the enjoyment, I’ve had some great losses in Halo ranked that were super intense. Why be frustrated about losing a bit of CSR when you’ve had a nail biting game? I guess that’s just difference of opinions. I enjoy a good game more than anything and I don’t care if I’m Gold 6 of that means I get 50-49 games more frequently.
Playing the objective does not equal a lower rank. That mentality will only hold you back. You need to learn to have a greater impact on each game you play and how to adapt to fulfill the role the team needs. That’s how you will get past your wall.
I think you need to look past the paper’s theory and into the actually implementation.
People in a fire team don’t necessarily play better together. Play styles can be mismatched. However, when they are matched, the game doesn’t reward the lowest ranked player in the fire team. I’ve gone on 20 game winning streaks against higher level Onyx player than me, just to have it all lost by one lag out or loss. I literally won’t play with friends anymore because trying to be organized and the win the game doesn’t rank me up.
For instance, over the winter break I played mostly by myself and was easily progressing to 1588 Onyx, my goal was to hit 1700 or so because that’s where I can compete in 8s (custom games with Onyx’s on discord). However, the second the 8’s lobby dies and we decide to go on ranked we just have a really bad time.
We’ll all sit there and play together as a team and not really gain anything after win streaks. Which is strange because we play the game efficiently so the stats might not be that high. Compare this with me literally letting a setup fall apart when playing open queue by myself, just so I’ll have an opportunity to break it again and get more kills. You might call it playing your life, but at a certain time your life should be played to defend an objective or so your team can capture the flag, or so your team can secure a power weapon or trade your life for two kills.
Like I get it bro, you read the TruSkill2 doc, so did I. The system still sucks. Wins and losses are all that should matter. If you can carry the ball for 2 minutes against Ogre2, Frosty, Snakebite etc with 0 kills, I don’t really care. Your rotations and movement must be sick.
2 Likes
Trueskill2 factors in skill level prior to and following the game, so can determine the parties effectiveness by teaming up which would mean after some games it should determine if the squad is performing greater than expected or worse which should in turn provide better match ups after some games, not ideal but otherwise it would favour teams too much. It is a known variable that greater squad numbers produces greater win %, consistently in online games. It may be hard to start out and find new players especially when the players are rank driven and not looking to develop an actual team with some initial losses if the team isn’t gelling as expected right out the gate.
My entire playstyle in almost every game is revolved around filling any gaps in the team. I’m pretty high ranked in most games, but I’ve been the top rank in every Halo across most playlists. You don’t have to be selfish, you need to be the opposite. You will win more games and do better as a team if you help your teammates thrive. If the enemy gets rockets, more potential deaths for you, less teammates to teamshot when they die to rockets, less map control, less ability to move and complete the objective. I don’t care if I get the rockets, I just need the enemy team to not get rockets. My team will be better for it. Teammates are important to how a player can perform. A good team elevates a player and a bad team will make a good player struggle at times. People who think they need to be more selfish to improve are taking backward steps to their own learning curve.
I’m trying to help people to realise that but everyone is too busy thinking the system is wrong to even consider assessing their own ability. Selfish play won’t get you far.
1 Like
Literally not true. To rank up in Onyx, you have to play selfishly.
TruSkill does give me harder games when teaming with a good team, and I appreciate that. It can get really sweaty and you can meet some pros. Dope. However, why the hell don’t I get rewarded for winning and why do I get punished for losing 1 out of 20. l
Watch a high level match making stream versus a high level 8s stream. When pros play match making they are literally trolling and just going for kills. That’s how they get these high ranks. When pros play 8s, they are trying to win the game. The gameplay looks world apart.
If the ranked system were working effectively, high level ranked play should actually play like medium level competitive play. It doesn’t.
Every player I know in two separate discords is an Onyx. Very few of us can bear playing ranked together because to get higher level in Onyx, we all know, we have to stat. It’s hard to stat with players you know. I don’t feel bad doing that with randoms. I’ll make sure we still win, but it shouldn’t feel like I’m playing dumb easy games with Onyx’s in solo open cross play just to rank up.
1 Like
They are all that matter, realistically. You can’t lose CSR if you win and you can’t gain CSR if you lose.
Depends on the wins, if you’re beating better teams and performing against then you would go up faster. But you aren’t. Do you think you won 20 games in a row against much better teams? No realistically if the teams were even, in a perfect world you’d win 10/20 on average.
If they release head to head and I’m Onyx 1700 and I beat 100 Plat 4’s in 1v1, does that evidence I’m capable of being Onyx 2,000? Not at all. It just proves I’m good enough to beat people I’m expected to beat.
So why does the system pair you against opponents that aren’t good enough to prove you are deserving of a higher rank? It can’t. You’re 4 stacking at the top 1% so it’s finding you the best games possible, for your timezone and connectivity without making you wait too long. They’ve already extended the waiting time. Playing solo makes it easier to gain rank because it’s easier to prove you are overperforming Vs expectation as your expectation is lower when not in a stack and you are having to adapt around all different teammates. Plus it’s easier for the system to get more balanced games, especially at the top percentile.
Also it’s slower to climb in Onyx, that’s just a feature.
1 Like
Dude what are you talking about? The game isn’t going to give us 20 games against weaker opponents and it didn’t. We checked the Halotrackers after each game.
I’m telling you I know how to go up faster, the answer is to stat. I don’t want to do that because it’s trolling, but it’s what works. Trolling doesn’t make a better player or more likely to perform on the main stage, but it works to rank up in Onyx. That’s broken.
Why does a 20 game win streak get rewarded less than a 2 game losing streak gets punished? Do you really think the game gave us 20 easy to win games in a row as a troll, then we somehow got two more and bungled them. And if you think that, isn’t that broken? Why did we sit through 20 games we were expected to win?
Like i’m just trying to follow your appeal to authority logic here. You’re wrong, spoiler alert, the system is broken. We played opponent ranked higher than us and some were very clearly able to organize at a high level, and some we know were teaming. Still each win was worth 1-3 CSR and the losses hit us for 10-15.
Again, pro players will tell you this and discords full of Onyx players will tell you this. If you want to gain CSR in Onyx, winning is secondary. Your primary objective is to get as high KDA as possible. If my team has a setup, I’ll let the enemy team break it so I have kills to clean up. If my team need someone to push a setup and entry frag, I won’t do it, no way. I don’t want the death on my stat line.
Does this make me a more skillful player? No. Unless you appeal to authority and define TruSkill2 as the holy definition of skillful player. This trolly matchmaking play style that developed with Onyx players will win 0 actually competitive games. Why does high level ranked play look NOTHING like competitive play?
2 Likes