It doesn’t matter what levels your teammates are. Your win rate is 50% ish which would indicate the algorithm is working as expected. Your individual performance is what changes your position,
You were onyx 1506 in season 1, which is before Feb.
343i also said they were making some changes which would mean players would likely place a whole division below where they were in season 1.
So in your case it would’ve been around Diamond 1 (1206 CSR). You’ve since managed to return to near enough the same CSR level though.
I’d be willing to guess that the 50CSR difference is probably from you getting stressed out, losing your temper, playing worse as a result.
Like I said (edit) once you have reached your actual level there should be very little movement without a conscious effort to improve your skill. In fighting games you watch back videos of yourself to determine what mistakes you make out of habit and work on improving them, with shooters you should be looking at starting strats for each map, which mistakes you make most and try to consciously work on improving them. Playing over and over is a much slower way to improve,
well I finally got my D6 back so none of you can say I didn’t do it
and I’ve been watching videos and streams about strats, jumps, movements tricks, when and how to rotate objectives, etc. I’m not new to video games. I’ve been making videos to show my practice progress.
This system is broken garbage, and it needs an overhaul.
I understand getting beaten by a better team/player. But the number of games you get with teammates who go double negative is way too often. Doesn’t matter how hard you frag, you are not winning obj game modes with teammates who can’t win individual gunfights.
If the metric is how well you slay, then it should be slayer only. Making most of the game modes objective then not giving the objectives ANY weighting in how you rank up is the dumbest decision of all.
Every match there’s an albatross to drag you down, but at least in obj they can get some points that way. The problem is they don’t get anything and contribute nothing but a headache.
How is it supposed to be fun for them either?
Edit: It took me 22 freaking games to go up 19 CSR.
Based on my experience playing ranked the past 2 seasons, the game doesn’t reward playing the objective at all. Higher ranked players know this and would rather lose with a positive k/d than try to break a hill/grab the ball and go negative. It rewards some of the most selfish playstyles rather than team play.
Thing is, there is only 50CSR between Diamond 6 and onyx 1500; with 300 CSR between each rank group. Considering the max rank you have ever made is onyx 1506 you’re hovering around your current maximum so you’re not going to see much improvement without considerable skill improvements.
You’re right about the system needing an overhaul. Problem is, once you reach your max rank there isn’t much point in playing until you can guarantee you’ve improved.
The other problem is that if the game is targeting a 50% win/loss (it is) and the objectives carry no weighting (they don’t) then it may as well be a team slayer game because you’re going to win half of the matches (you do)…which results in rank stagnation…unless one focusses on TS metrics for rank increase (kills, ttk, deaths, time played in game). Again though…MMR will adjust. It’ll show that you potentially lose more objective matches whilst having a high KD, which means you’ll end up in the same place.
The only way they can sort this out is to bite the bullet and to track a wider selection of performance indicators, otherwise you may as well treat objective games as team slayer variants.
You may see some difference in team skill variance tonight though, they dropped a pod in the last 24 hours that severely reduces party rank disparity. I believe they made some other “undisclosed” changes also which should make rank calculation more accurate. You’re still only going to win 50% of the time, but the pool of players should be more finely selected and the onyx Vs gold should stop entirely.
Keep in mind it’s not the ranking system doing this - it’s the match maker - struggling with the lower population.
You definitely need to adjust your play.
And accept that a lower ranked team mate can go negative and still contribute.
Probably not double figures though. To which end there seem to be two major problems with the wider ranked teams;
It makes playing the objective hard if you don’t have mics. You can definitely work with lower ranked players - but you need to be on the same page. Sadly this doesn’t seem to happen a lot. But even if they aren’t talking - they are often listening.
When your team doesn’t gel, for whatever reason, it falls apart hard. So you end up with big losses and players with double figure drops. Those same players may do better in other teams. And the final numbers don’t necessarily represent the actual skill level of the players involved.
Your MMR shouldn’t take a huge hit as long as you’ve met your expected number of kills and deaths. It kind of buffers you a bit.
If I was 343 I would have tell the matchmaker to at least make the game slayer if the teams are widely spread. That is the game mode easiest to work with in these circumstances.
I think the important bit is to view all these games as slayer games with an objective.
Certainly watching a few high level games seems to back up the general advice re: kill first and then go objective.
I’m getting better at Oddball. And to a certain extent CTF. But I still struggle with zone games - I’m ok with getting into the zone - just not very good at defending it.
Pretty much all the usual suspects have proved fruitless.
If you rank players up on them (assists, damage, objective score) you are just exposing the player to games they can’t win and you have to wait for them to lose and rank back down (while ruining the game for their team mates in the process).
I would like them to look more closely at damage. Particularly combinations or ratios of effective damage (leading to kills) and damage escaped (where you didn’t die). Those may be a way of rewarding assist type plays.
And for objective games look at partitioning time with the objective from your KPM. So if you get 16 kills in a 10 minute game and you held the oddball for 2 minutes - then your KPM for the game is 16 / (10-2) = 2.0 KPM (instead of 1.6).
Step 1 is the XP rank.
Step 2 is some sort of specific reward for winning a ranked game, even if it wasn’t a win that advanced your MMR.
Thing is - the team CSR score is an indication of how the matchmaker works. It takes all players, generates the mean, then finds another team with a similar mean score. People who reside outside of the normal curve, ie LUCID, skew the mean by being the overall average up. He is what, Onyx 2400? That means if he is matched with a team of platinum 1’s (900CSR) his rank brings the team average up to 1275 (4x diamond 2 players).
This is what the problem is. The further you get away from the mean population CSR (platinum 3 - 1050 CSR) the fewer players there are of that skill level available at that specific time of playing. Therefore the matchmaker brings in the closest available rank, taking into account ping and time to play, then works out what it needs next to bring that CAR value up to a matching team CSR mean. If they don’t loosen some factors, they cannot match the player, therefore the higher a rank you are the more likely you are to have an unbalanced team….unless you enter matchmaking with a full team with matching (ish) CSR scores. A team of 4 onyx players (1800 average team CSR) will have to wait longer, but will eventually find a team to play against with a similar Team CSR. At least - they would if the player base is healthy.
You’ve touched upon the issue of not tracking objectives also. Again though, the system tracks if you’re getting lots of kills in objective, but if you’re not winning objective then you’re not ranking up in there either….plus if you’re getting loads of kills and losing them, your MMR is increasing, this meaning your next slayer game will be harder. You then could lose this, get less kills than expected, then derank back to where you were before owning objective carriers,
Like I’ve said earlier, if it didn’t work you wouldn’t have approx 50% games won.
Ultimately I feel there should be a slayer ranked playlist, it won’t sort the problem, which is that higher ranks over perform relative to team CSR, but that actually revolves around the limitation to player skill tracking based on the cost of data storage for algorithm analysis.
In short. If you can’t rank up any more then that is your skill ceiling. you need to improve and you can only do this by labbing out new processes, improving aim, or by playing with people better than you and learning from them.
People have a lot of complaints about the matchmaking system but they also have no idea how it works, then get angry and complain. in reality their improvement would still result in them winning approx 50% of the time and with wider skill variance likely because they’re solo queueing and don’t understand that the higher their rank, the more likely a team CSR variance actually is as a result,
But not sure how the match maker is supposed to deal with this?
It can only work with the players it has.
And to some extent we are all now just starting to experience the problem that player’s like LUCID has from the start! The only advantage is that his “crappy” team mates are more likely to be experienced players than ours.
This only seems to be a problem with mid-lower ranked teams.
The good teams share the objective. And prioritise killing and not dying as a strategy to get numbers before pushing the objective.
I think the key message is that if you want to rank up you need to recognise the “onyx meta” (I think that’s what the cool kids call it).
I don’t profess to be a good player by any means - but if you relying on other people to kill so that you can secure the objective and then dying every time when you do - you are doing it wrong.
This is an interesting point. I came across an old Menke post the other day which suggested that each mode in a playlist has it’s own MMR. But the match maker uses the average of all the MMRs to find you a game (keeping in mind that the map and game mode are chosen AFTER the teams have been chosen).
So your performance in said objective game’s influence on your Slayer MMR may be a bit more nuanced than that.
And besides… the same thing could be said across the rainbow of modes. You might be awesome at KOH but really bad at Oddball. Or even better on one map over another. So a bit of luck with maps and modes could see you push your MMR up more in a session than you would expect - and it may not be pretty when you get your next game in a map / mode combination you aren’t so good at.
Personally I seem to find a groove until I get a couple of CTF games. It’s not that I don’t like CTF, I really do… but it’s the one mode that you are going to get steaked at if your team doesn’t work together.
Word.
Improving skill in any endeavour is a laborious process. Whether it be Halo, playing the piano, or crochet.
It’s not going to happen overnight.
And this is where 343 need to step in. Provide the support or tools we need to make the system more approachable.
The issue there is, the game shouldn’t be FORCING a 50% win/loss on anyone. It shouldn’t be actively looking to sabotage people who won a game too many.
It’s matching on when the algorithm thinks people need a loss to bring their wins down to 50%.
I have proof. On halotracker it tells you in the match details what chance each team had to win. The system keeps forcing me to have matches where I have only a 30-40% chance of winning, because my win ratio is about 52%.
It does this by putting worse and worse teammates on your team, against higher ranked players and full stacked teams. Almost every game I had yesterday was my team of 4 solo players against teams of 4.
I have proof the system is actively sabotaging me. I know for a fact it’s doing it to everyone else in my situation.
If you have one really skilled player matched with below-average players to balance out the match vs a full team of average/above-average players, they can focus on the skilled player the entire game and make them perform as if they are average. With the time it takes to recharge shields, having a team that can hold map pressure is extremely important to winning games.
Exactly. It skews everything. You can’t just hard carry people who aren’t helpful at all, it’s like playing a 3v4 or worse. It’s so frustrating to just constantly be getting teamshot with no support, or having to do the objective yourself when you need to be slaying. Especially when they don’t use mics, don’t speak English so you can’t communicate, don’t ping, don’t listen, etc etc
It’s insurmountable. The game is FORCING losses by doing this on purpose, and anyone that disagrees is being disingenuous.
It’s objectively worse. Above/slightly above average players would almost 100% be better off fighting 2v4 without the dead weight dragging us down. At least in 3v4/2v4 you don’t have people feeding kills to the enemy.
No. No you don’t have any proof; just an assumption and lack of understanding regarding basic probability.
The game tries to match 50:50 each time. It cannot always do it because of the population size; this is ever more evident when you are a higher rank because the population size is smaller
The system 343 uses to calibrate peoples rank is working just as they intended, they have you chasing it, it’s aggressive tactics to keep you playing this game, and only this game, you’re behaving just as they intended. Every time I play this game, and knowing all the slimy manipulative tactics they use, I feel like I’m betraying my own code of ethics … oh well, it’s only a game, is that a valid excuse?
