Teams have a weighting applied to their MMR to account for the fact they play better together.
Some squads seem to be manipulating the system to grind CSR. A couple of Onyx players with a mid-range and a low sandbag account.
They grind out a few points for each win and use their personal performance vs the lower ranked individual opponents to mitigate any losses.
If the majority of those wins were against similarly or lower ranked teams you arenāt going to rank up a lot.
Thatās pretty good for 20 odd games.
It gets slower as you get towards the pointy end.
I think your rank is compared to the opposition teamās average rank. So they could easily be ranked the same or lower than you.
I havenāt worked out out yet.
First person out is -15.
Second person out could be-15 or -3.
I think it depends on the timing.
Everyone else is-3
I do get the you must win vs high or similar ranked teams but its something out of my control. Like playing vs some onyx trying to rank boost a gold is just a pointless risk on my end when iām going to gain literally nothing but risk losing everything.
Then the people leaving straight away meaning any rank loss is just stupid.
I get there are alot of variables at play but what they have atm just doesnāt feel like its addressing the current state of play. Iāll probably reach onyx eventually by ramming my head against the wall but its such a chore trying to navigate having pointless matches where if me or someone on my team is having a bad match for whatever reason then I could lose about 8 games worths of grind
Itās important not to get hung up over any single game. You need to give it time. You will get good games and bad games. It all comes out in the wash.
Shouldnāt be the case.
Even though they have a wider range of skills in the squad, the team MMRs will hopefully be the same. What you gain for a win and lose for a loss in terms of MMR is the same.
Yes. I know the system wants to bring your CSR and MMR back together, but people are already -yoinked- off with the quitter. Itās not the time for them to pinch back more than everyoneās fair share of CSR (which is -3).
And games that start 3v4 shouldnāt be abandoned.
Two things.
Skill shouldnāt be a grind. You are what you are.
You can improve, obviously, but it usually takes weeks to show discernible change. Itās not something you can, or should, grind in a day.
And secondly, the loss is only taking away what youāve ground out in CSR. Over those eight wins you didnāt do enough to convince the system that your MMR should change - so your CSR drifts up on itās own. And then on the very next loss it will bring them back if it can (max of -15 per loss).
Itās frustrating. But itās just your CSR oscillating around your rank. By less than 1% of the total ranking scale. As far as your MMR is concerned, you CSR isnāt changing at all.
If you want to āgrindā your CSR you first need to āgrindā your MMR. You have to convince the system you are a better player than you are now. Once you have done that your CSR will follow (and then oscillate around itās new baseline).
Iāve been absolutely stomping people in btb.
Finally made it to plat 6 on a decent streak, but still dealing w the do-nothings.
Two of my losses, I absolutely murdered the low onyx playersā¦as inā¦it was a massacre for them.
The more I play, the more I get the feeling I deserve onyx being as Iām pretty consistently the best or second best player of the 8.
Itās just such a grindā¦as posted aboveā¦to lose a ton but win half of what you lose.
Although my last two wins, I gained a lot of CSR
Doesnāt help with ranked.
Well done.
Remember itās pretty much about the win.
The best player on the losing team is still a loser.
And the only personal performance metrics of value are kpm and dpm.
The key is to do what you need to do to boost your MMR. Just grinding your CSR is always going to end in losing it on the next loss.
Which is good news. It suggests your MMR has moved ahead of your current CSR.
I went 15-5 in a slayer earlier against a team with a higher average MMR and deranked because my teammates were outclassed. Iāve just learnt to accept that this ranking system is straight up trash and moved on. Now I just play it when I want those weapon set ups, rather than to actually rank up
My mmr has BEEN ahead of my car apparently., Because when I DO win, I get the max CSR
The only ranked game on tracking I could find for the last 10 days for you going 15-5;
You didnāt de-rank. Your CSR change was 0 points.
You had 15 killsā¦ but otherās in the game had around the same number. So your kill/minute wasnāt out of the box.
Regardless of that. Your CSR change has nothing to do with personal performance. Itās about W/L, who you beat, and your MMR.
You were actually in the favourite team (1418 to 1281). Projected to win 53%.
Your team-mates did struggle. Particularly the D6 (-3) and the D4 (-7). But you were essentially beaten by their 1650 Onyx player. But could you have done anything to help? Could you have co-ordinated more. As the in-form/best player could you have been more aggressive? Less defensive? I just ask because your team-mates had 23 assists between them to your 1.
Apologies if this is the wrong game or you are talking about an alt account.
Lol 1 assist?? No wonder he left that part out. 15-5 but cleaned up all his teammates kills. You should derank lolz 
Wrong game. The enemy had 83% chance of winning with an 1815 peak player in their team. It was slayer on catalyst yesterday. Although for some reason halo tracker is now showing I didnāt lose CSR in the game. That is definitely not what was shown in game, or yesterday when I checked halo tracker right after the game.
Of course I may have been hallucinating. But halo tracker doesnāt show every CSR movement between every game reliably so I donāt trust it.
Heās literally talking about the wrong game. I went 15-5 against a team with a peak 1815 onyx on it.
I have not had any other 15-5 games in the last 10 days, so I donāt know whoās stats he is looking at
https://halodatahive.com/Infinite/Match/4323acc3-986a-46a2-ba9c-e40ec533ffd6?gamertag=Stoatbag&page=1
Yes, the game where we had 83% chance of losing. You say we were favoured to win?
Iām not āsayingā anything.
The link to the Halo Data Hive says 53% in your favour.
The fault here appears to be the way the various sites calculate the %'s. And that isnāt a small discrepancy.
I wish we had a public API. We should be able to check this out ourselves!
Your MMR team average was 137 higher. So it makes sense that you should have been favourites.
Statistically, a gap of 300 should be 75:25 head to head. How that translates to 137 I donāt know (it wouldnāt be a linear relationship).
Point number 4 above. Thatās where you said it. I now understand why, and thatās fair enough. If the two sites have massively different balances then isnāt that indicative of how confused and obtuse this matchmaking system is?
It must be to do with massive differences between MMR and CSR. To me that is the only way that sort of discrepancy can happen. I just donāt think CSR is helpful in any way. Show MMR, and have a separate progression based system for people to grind
Nobody actually knows how it works and it seems really inconsistent both in matchmaking and CSR outcomes.
Not really. Itās just confusing how two different sites are using non-API sanctioned data.
TS2 averages three MMR curves of each team and then subtracts the curves from each other. The result, also a curve, tells you the expected win rates.
These two sites are taking the two raw averages and applying their own made up algorithm.
Unlikely. Again the two sites are just using two numbers and making -yoink- up.
Most of the players on the field probably had quite close CSR vs MMR. The ranked players only had CSR changes of 1 to 4 points.
The two unranked players could well have been wide apart. But we donāt see their CSR.
I think the way 343 are doing it is confusing. Iām not convinced the MMR is necessarily the way to go. It would be messy early on.
But they could easily simplify the CSR.
Someone suggest 1 to 117. I think thatās a superb idea.
I still canāt believe they thought dropping that was a good idea.
If they continue on this path they need better communication. Both in and out of the game. Players should be able to work out exactly what happened each game.