Yeah I play solo duo when ever I can instead basically
1 Like
Developers have explicitly stated that being in a fireteam, no matter how well you know each other or how well you complement each other’s playstyles, causes you to lose a bit more CSR on a loss and gain a bit less on a win than you would if you queued solo.
3 Likes
You guys are all Onyx playing mostly against other Onyx players. You’re already in the upper echelons of the ranking. Ranks are graded on a curve. You can’t expect to keep climbing continuously, even if you’re winning more than you’re losing. At some point you’ve reached the appropriate rank and you’ll fluctuate around there.
wrong, If my win % is constantly increasing the system is failing me by not getting me to a rank where I’m closer to a 50% win rate.
3 Likes
What I’m saying is that if you queue as a stack that stack should have its own separate rank.
So I’ve made a new account and what’s hilarious is that I match with high Onyx players consistently, because that’s where my hidden MMR is at I guess, but because of that my rank is moving very slowly in the low Diamonds.
It’s a very strange system they have. They system knows my skill is Onyx 1700 because it matches me there, but it also knows my rank is Diamond 1 there (versus Onyx on my main account) and so I get little CSR for beating Diamonds, but lose nothing when I lose against Onyx 1800s etc.
I really think a more productive conversation can be had around tweaking the ranking system so that it continues to give people fair matches, but also rewards people for their good matches.
What I really don’t like is that a string of just bad games can ruin rank, but the system knows the skill of the account is high. Like wtf is my second account matching with Onyx players consistently when I solo queue, but the system doesn’t seem to want to change my rank to the level it matches me at.
So strange.
I also stand by the idea of ranking stacks at a unit, together and showing it somewhere. Then having another rank for a player as a individual outside of their stacks.
2 Likes
Just want to add in my two cents. I submitted a bug report about the ranking system because it felt like it was unintended behavior. The Support team led me to the forums since I guess it is intended behavior. I’ll just copy some excerpts from my support ticket so that hopefully some 343 devs can see them:
Just for some reference, here is my halotracker: https://halotracker.com/halo-infinite/profile/xbl/hogiyogi597/overview
A quick real world example. I played a session with 3 of my Onyx friends. We won 5 games in a row. I went positive or even every single game and I went up a total of 4 points. We lost the next game of Oddball and I went 22/29/14 which in my opinion is not that negative for losing an Oddball match, and I lost 9 points. 1 game set me back “progress” I gained after 5 games.
Summarizing this session, we won 9/14 games. Every time I went positive or even. I gained a total of 13 points from those 9 wins. From the 5 losses, I lost 48 points giving me a net -35 points for a win percentage of ~65% and an average positive KDA. What a fun system. I would also note that we played against full teams of Onyx and even higher Onyx than us ~1994.
You can verify on my stats about my claims to be H3/H5 top rank by looking at my Service Record.
I just want a ranking system that feels fair and fun. So far, this ranking system gives me neither of these.
Best method to understand is to use predictions on Halotracker. 50/50 games show nothing, if one team is favoured you’ll get a % of winning. Essentially doing well in 50/50 or predicted loss games consistently is what will help you rank up faster. Performing average will result in little rank movement. If you are slightly below expectations your rank will drop faster than it rises. Let’s look at your last 10 games:
50% - loss - 0.83
50% - win - 0.64
50% - win - 1.18
48% - loss - 0.82
50% - loss - 0.71
100% - win - 1.17
22% - win - 3.00
72% - win - 0.77
71% - loss - 1.07
54% - win - 1.25
Apart from the 22% win game being a really weird anomaly (you won 50-15) despite having overwhelming prediction against you, your stats are pretty standard for the level. You are almost winning games expected to win, and losing game expected to lose, draw games are going either way with average performance. As this is all happening around the 1800 rank it would suggest where you are currently is pretty accurate given your steady/below average performances in 50/50 games for this level. Having a good performance in a heavily predicted win is par for the course and won’t help you rank up quicker.
Do more analysis yourself, but based on my knowledge of the system and using your last 10 games as a reference point I wouldn’t say anything strokes me as too out of the ordinary here.
2 Likes
quite interesting that the only thing you used for performance was KD. So I really can’t take your “analysis” seriously lmfao. Wins are simply not rewarding enough and losses are just dumping CSR.
1 Like
Where do you find the estimated win percentage for each game? I would love to be able to analyze myself but the system seems pretty closed to me.
And to generally address what you brought up. Regardless of how the system works, if it results in me questioning why I am not going up then imo it is not a good system. If I beat someone that is ~2000 rating and I am at ~1800 and still go up 1 point, that to me does not seem to work well. In your example, I should have been expected to lose against someone almost 200 CSR higher than me but I won… but instead of going up any substantial amount I went up 1. I just don’t see how that makes any sense.
I think I’d have to agree with @illeston3 in the sense that at the Onyx level, KDA should matter less (unless you are going incredibly negative). At this point, wins should matter more than anything if the teams are roughly evenly matched. If I beat a team that is of equal or better skill, I should go up. If I lose against a team that is equal or slightly less skill, I should go down.
Regardless of how the system works, it is a **** feeling to go up 1 every win and then lose 1 game and lose hours worth of progress… So either the system needs to be more aggressive with who I am being matched against to give me more chances to prove that I should be a higher rank or it should be less punishing on a loss. Everyone is entitled to a bad game here and there. But losing 10 points on a loss and only gaining 1 on a win seems a bit ridiculous. I shouldn’t be expected to “top frag” in order to go up. I contributed to my team as much as the next guy. I should go up on a win.
Even more frustrating. A few nights ago, I was tied with my friend for CSR and we were playing together. We lost, I went positive and went down like 6 and he went negative and went down like 2… So what is the ranking system doing there?
1 Like
This is one of the biggest things for me. Every other match it feels like I’m trying to figure out what happened on rank gains and losses. Maybe the system isn’t that bad, but the transparency of what’s going on could be vastly improved.
How I feel as well. Nothing worse than ending a night further back than where I started from. It just ends up feeling like I wasted my time.
Agreed
This is one example of dozens that I’ve experienced.
gyazo.com/82220741c15520af4671c7b1e1ce514f
6 wins in a row and I lose 1 match and I’m lower than where I started. My win % keeps going up I just broke 60% but my CSR is stagnant. Doesn’t the system strive to put you around a 50% WR? The system is failing us.
this is my stats from 100 matches. I was about 1550 and I’m now 1562 with a 71% win rate. PLEASE tell me the system is working well, I DARE YOU.
gyazo.com/d4cc45f22379e1f8603dae7c6abe19b2
2 Likes
One thing I have noticed with friends is that only the person that places first on the team gets a big boost… Everyone else usually gets like +1. You should see if that is the case for you.
If that is the case, that is really naïve. There can be more than one good player on a team.
2 Likes
Most of the time for games where theres very little points earned its across the board for everyone.
I have had a few games where I’ve only gained a couple and someone gains 10+
But to continue on how the system is functioning, I sort of think there is a problem with matching players incorrectly. I just played 2 solo games in crossplay and the odds for my team were very favored for me. The problem is, the games were way too close for how bad I was destroying the other team. If I picked up the ball my team just could not defend, so I was forced to only slay and I slayed them and it was a pretty close game even so.
Give me longer ques to put me in a match where my teammates are on the same level as me. This seems to be the problem with the “prediction” system is because it’s just not placing the correct skilled (CSR) players together and there is too big of a variance… Especially when you are solo queing.
3 Likes
Here is an example of my last match where we were the underdogs (even though the other team was mostly high diamond, but we had a diamond 1 and 2). Me and the other guy had to completely pop off in order to make the game even close… Even after the fact of us having 41% chance to win and I completely popped off, I get 3 points… Incredible.
tracker.gg/halo-infinite/match/226ed5e8-10e4-44e0-8c10-f70990143ff8
Where are you seeing your percentage to win? I’d like to be able to analyze my own games more.
you have to click into the match and then look top right, it only shows the % on maps where you can see updated CSR. I’m not sure why it doesn’t show it for every match but yeah.
On Halotracker you can click on individual games and it will say top right the % chance for each team to win. If it doesn’t say anything then the game can’t determine a winner aka 50%.
The issue is you aren’t constantly playing someone rated 200 above you, you’re playing a full team rated the same as you. So in the 50/50 games your teams are equally rated. This is why it’s important to consistently win and do well in the 50/50 games and games predicted to lose. I see the anomaly game where you were predicted to lose but your team won in a landslide. That’s great but doesn’t change the history of your last 100+ games. The system knows what you’re about so the only way to change its mind is to be consistent with this. If you are constantly doing well and beating people in these types of matchups then your rank will go up faster. I’ve only looked at 10 games and used k/d because I’m here to provide you with some clarity on the subject. It’s up to you to scrutinize the data yourself, look at your damage, average life, k/d in games expected to lose etc etc. K/d is less weighted in objective games, but still relevant to an extent.
The issue with giving players like yourself and illeston free rank for winning games is a pointless one, as the system knows you will fall back down if it did do that. If a platinum player won 20 games in a row by pure chance, should they now be in Onyx despite their playstyle remaining the same? If you searched matchmaking would you want that player on your team? Of course not.
Wins are the most important thing, but players acting like bad stats can be negated solely by wins are misguided. A player deserving of a high rank will be able to win and have good stats at their lesser rank, consistently. Stats shouldn’t be your focus though, wins should. But if you’ve hit a plateau it’s probably because games that matter the most aren’t going as they should. Why doesn’t the game give you more opportunities? Well the matchmaking is based on whoever is playing at that time and trying to find them close games. It may be that there aren’t enough people of similar rank to you to get the right games, however even in your last 10 you had a fair few 50% games and you didn’t bring the heat so I guess the rank may just be warranted.
Your friend likely has a better game history than you and consistently performs at a higher level purely based on measurable variables. Unless you’ve played every single game together, the waters will be muddied so it’s impossible to say who is the better player, and even stats can’t fully describe that but on paper your friend is better because they have a better body of evidence to support this.
People are so obsessed with their rank. It’s unhealthy. Play, enjoy your close and competitive games and ignore the rank. If you play for 5 hours and your rank barely budges after all those games, congrats you’re at your current skill ceiling. It may improve or decrease over time depending on how you evolve with the community. Both options are likely to happen, or you may have plateaued and will dance around that rank forever.
Seriously though if you want to discuss it any further then please leave me out of it. I wanted to provide some context for you, but I don’t care about if we think the system is accurate or not. You can look over every single game you and your friend have played if you want and really break down the differences. I have explained how it works, but I’m not interested in continuing a discussion about how it shouldn’t work like that and how it’s wrong. That’s just how it works.
1 Like
literally going 70%+ in my last 100 games, but I’m not performing good enough to increase more than 20 CSR in 100 matches. I’m not advocating for the system to work differently I’m advocating for the system to function well and have incentive. Currently the system sucks and there is no drive to play it as I’m literally just dropping W’s all day and not increasing.
CSR is important, I want to play better people on a regular basis outside of scrims/tournaments. This system simply just isn’t doing that.
You say the system can’t find players in my range… Bro, I’m getting INSTANT matches… the system isn’t even trying to find equal matches, it’s almost as if it’s just throwing me into lobbies like it does in casual lmao.
But w.e you haven’t really understood where we are coming from because you just go back to the usual “well the system works this way” blah blah blah. The system DOESNT WORK.
And yes, if someone gets on a winstreak of 20 wins, absolutely that person should be ranking up fast. If he doesn’t belong there he will derank. Did you never play halo 2? halo 3? those were the days where you could actually rank up and play people on a similar level.
I do honestly appreciate your replies @Ken_CTF . It has helped me understand how the current system is implemented.
I won’t argue with the way that it is implemented even tho I see examples that are contradicting to what we are discussing pretty much constantly. I guess the point I am making is, regardless of whether or not this is accurate or not, it is not a fun system. Game developers change things all the time to be less correct but more fun. Classic examples are the “coyote time” jumps and other things. The point of games is to be fun and encourage people to play.
I want to just relax and just have fun and enjoy close and competitive games. But honestly, the fun of a ranked playlist is earning Rank. Otherwise we would have a BR start no radar playlist that you can just play for fun and SBMM will pair you with the appropriate people anyway. Rank is literally the point.
The system should change to be more fun. More “fair” (to the people). More consistent. If I have to come ask on the forums to have the ranking system explained to me, it is a sh*t system. Just sayin’.
1 Like