The rate within the match doesn’t matter, it’s still the same kd at the end. The match didn’t change lengths.
You’re not listening.
Performance over time matters. And that’s what Darwi is trying to make clear to you. He referenced two performance scenarios with the same KD ratio, but different KPMs & DPMs.
1 Like
You’re not listening.
It’s the same number of kills either way. 15 kills in a 10 minute game don’t change anything if you get them all at once or spread out, it’s still 15 kills in the same time frame.
You’re all comparing apples to oranges. Get a high kd and win the match, nothing else matters.
The match’s total time length is the factor. I’m not referring to just moments in a match.
15 kills in a match that lasted 10 minutes vs 15 kills in a match that lasted 5 minutes. Obviously, the KPM is better in the 5 minute match.
It doesn’t matter if in that 10 minute match the player was able to also achieve 15 kills in 5 minutes. If no additional kills were achieved the final 5 minutes that players KPM is still going to calculate the total kills by the time it took to complete the match, so it’s going to be worse than the player who achieved it in a match that ended in 5 minutes.
1 Like
The higher your kills at the end of the match, the higher your CSR change.
The time between doesn’t matter if you were getting a high kd either way. 10 kills is 10 kills, whether you got them all early in the match or throughout the match.
Jeez guys, just kill people, it’s simple.
You can still search for them here on Waypoint. The database is still intact.
Zaedynfel is a good start. I was reading the last Ranking update. You were posting in the last bit!
And thanks for the last couple of posts.
1 Like
Not necessarily.
Clearly this is a difficult matter for you to comprehend, so I’ll try explaining again.
First of all, for any possible positive change to CSR a player must win the match. This is because CSR can only go up on a win and down on a loss. Wins are rewarded. Losses are punished.
Secondly, the degree of positive change to your CSR will be influenced by where your MMR exists in relation to your pre-match CSR. If your MMR is above your pre-match CSR you’ll earn a better positive adjustment than if your MMR is below your pre-match CSR.
MMR adjustments can go either positive or negative in relation to the match outcome, but the match outcome remains the primary factor based on expectation. Were you expected to win or not. Winning unexpectedly can net the biggest positive adjustment to one’s MMR.
Then there’s the individual performance expectations. These are based on KPM (Kills Per Minute) & DPM (Deaths Per Minute), but you have to understand that the minute aspect is calculated by the length of the match. Not independent moments throughout a match.
This is why KD isn’t the be all, end all in determining a CSR adjustment. And why having 20 kills in a 10 minute match isn’t better than 12 kills in a 5 minute match.
2 Likes
I was always trying to get a sound understanding of the systems in place and how they worked together, I’d request improvements where they were needed, plus help others to better understand the systems too — like the good work you’re doing now.
1 Like
So long story short, win the game and get a lot of kills while dying less, ie a high kd.
Glad you all finally see I’m right.
You’re all overthinking it. You can’t just calculate how good you’re supposed to do, because that number is arbitrarily formed by the matchmaker.
So slay out and win, ignore everything else, and you rank up more.
That’s of course a lot easier when you’re not saddled with a Diamond 2 in a mostly Onyx lobby.
It’s frustrating.
It is a complicated system under the hood. For sure. But instead of debating the nuances and suggesting improvements on how we can improve - we have this caveman attitude - “I don’t understand how the CSR changes. I don’t want to understand it. The WHOLE system is -yoink-”.
Add in a bit of warped nostalgia for the “good ol’ days” (spoiler alert, they weren’t), and it’s definitely a baby in the bathwater scenario.
I’ve definitely noticed that the biggest anti-ranking posters come from the D5/6 ranking. And psychologically this is interesting. In Halo 2 they were probably relatively satisfied treading water at Rank 42. And in Halo 3 they would have been happy (and oblivious) perched with their Rank 50. But the Diamond to Onyx barrier is creating an artificial tension.
I find the whole concept of ranking systems interesting. My maths isn’t quite up to the level to go into a super deep dive - but I enjoy the mechanics.
I feel bad when people are getting frustrated or not enjoying the game because of some misunderstanding with the ranking system. So I’m happy to try and help. What I need to stop doing is getting into inane discussions with people who don’t want to be helped.
2 Likes
Because its designed to gatekeep
1 Like
Nobody truly knows how the current ranking system works. The guy who designed it left 343 about a month after launch. Looks like they’re only now figuring out how to make changes to it.
What we know for sure, is that it’s designed to force you to lose half of your games. Regardless of skill level. It was their way of getting you to play more games to complete your challenges.
1 Like
Really can’t see the connection.
But hey, you keep doing you.
There is plenty of information available about TrueSkill2.
We know a fair bit from Halo 5 - and we’ve seen very little to think it would have changed that much.
Tinkerings around CSR for sure.
It’s an evolving process. They should always be tinkering to make it better.
You don’t know that.
But if you do know something - I would be very interested to see your sources. As long it’s not just a bunch of anecdotal complaints from people who have hit their skill ceiling.
Really?
1 Like
Is this not exactly what they’ve said time and again? That the goal is to have your win/loss be as close to 50% as possible?
That means they’re setting you up to lose 50% of the time
2 Likes
There is a subtle difference.
Matching teams equally based on rank/skill so the results naturally fall into a 50:50 pattern. Combined with a super accurate ranking system that can detect fluctuations in form and rank players up and down quickly - so that your games get harder or softer as needed. Your rank evolves with your skill level.
As opposed to what you are proposing - a system which deliberately sabotages select games to lock you into your rank.
Actually. I take it back. There is no subtly. It’s chalk and cheese.
Yes there’s info on trueskill. I was referring to the technical build itself and how it’s a complicated system to edit. Especially when the original programmer is no longer there.
Agree
We do know that. 343 has repeatedly stated that you should be winning 50% of your games. Not even pros can get a win % much higher than 60%
you ever get a challenge to win 4 matches? if you get 4 straight wins you’re done and maybe get offline. But if you lose more than a couple, now you’re playing for an extra hour 
1 Like
everything you think would happen with a 1-50 system is what’s happening with the current system
One of the principles of TrueSkill2 is that it’s developer friendly.
When you read the paper it’s pretty much set out to make it plug and play. There are only a few tuneable parameters. They specifically acknowledge that developers don’t want to spend resources on this.
You would hope he had an understudy or two.
And that’s the principle of TrueSkill2 - and the accompanying match maker.
If it has you ranked correctly and the teams matched evenly (on those ranks) then the game could go either way. Hence the evolution of a 50% win rate.
The top end of town is harder to control. It’s harder for the match maker to find them equal opposition. So many do tend to drift above 50%. But because it’s a team game and one player can’t win every match off their own back - it shouldn’t drift too high.
I have.
Why would you “get offline”?
If you’ve just won four games against similar or lower ranked opponents - your MMR won’t have changed. The next match should be just another evenly matched contest at your current skill level. It could go either way.
If you’ve on a bit of a streak and have improved - your MMR will go up a bit and you’ll get some harder opposition. But because you are in the zone you should easily be able to compete. Again - hopefully an evenly matched contest at your new skill level. It could go either way.
Just move onto the next challenge.
After losing a couple - your MMR will dip. And you’ll get softer matches. Letting you salvage your 50% rate.
Acknowledging your wink as a tongue in cheek statement.
I wouldn’t consider trueskill super accurate, at least not for objective ranked modes considering it values kill/death and kpm performance and nearly ignores any player who sacrifices their life for the objective often. The system feels like it promotes ignoring the objective and fragging out over objective play, which means it favors one “type” of skill over another
1 Like
I have a feeling you believe faults in systems are more to do with the end users reaction rather than any fault with the system. Which doesn’t make any sense, considering a systems primary objective should be to not require its end user to adapt to all of its flaws