This works for stacked teams. It doesnt work for solo players. I cant win alone a match. I can try to carry the team, but if i get matched with player that cant keep up, then its a lost case.
Personal performance is not weighted, only W/L, and that doesnt work for a solo player in ranked.
1 Like
Because it’s more concise and easier to control the variables. It’s very possible to make up for one teammate playing badly. It’s impossible to make up for 3. Add to that the inability for most people to even remotely understand how objectives work and you realize how much better slayer is in a ranked setting.
What biases
Probably
Definitely not. For starters there isn’t an oddball playlist or any way to play oddball outside of ranked. And when your oddball experience is playing solo against teams when your teammates are anchors or leave or only playing slayer and ignoring the oddball, etc, then you can see where it would be a skewed average.
Empathy for who? Who do you want me to feel sorry for?
Completely illogical outlook
What I said you were lying about and what you quoted were unrelated. Maybe you just misunderstood
People always say that when they’re losing an argument online.
Name one
Name one
You’re the one that somehow removed the block in the first place, because I didn’t. Well you know you don’t have to constantly pester people if you don’t want to, that’s a choice you’re making.
2 Likes
Which kind of defines empathy… or the lack of it.
The speed of ranking. Almost exponential compared to simple fixed W/L sorting systems.
Accuracy is comparable.
They do have their own foibles, like rank degredation - but that’s part of why you have Seasons.
And the speed of ranking allows them to adjust much faster to fluctuations in form etc.
You would be pretty silly to go back to a more primitive system from the dark ages.
First off they are different ratios. One is K per minute. The other K is per death.
Secondly, and most importantly, one affects your ranking (MMR). The other one doesn’t.
Thirdly, one is a good reflection of your skills in a 1v1. The other potentially reflects your ability to camp.
I still can’t fathom why that nuance is relevant when the goal is to get as many kills as possible either way. You’ve gotten 15 kills and 6 deaths either way, the game doesn’t treat it differently.
Why would camping even matter if it was how you got the kills
Both those statements show a fundamental lack of understanding of how kills happen in Infinite.
You should never even have a 1v1 unless you have no choice. Having a high KD does not mean you’re camping, it means you’re good at killing people without dying.
what I still don’t understand is how players that have less than 200 matches played, are unranked, or Platinum 6, and finish 2-16 are playing in High Diamond - Onyx matches. Can’t understand that
1 Like
Time is a major factor in determining skill. Capability to perform effectively in an efficient manner.
2 Likes
Doesn’t mean they can’t be well versed / skilled in FPS… or have played Halo before… or have multiple accounts.
Unranked players still have a MMR. And it’s seeded from last Season’s.
Could still be mid to high Diamond is post placement and their MMR is well above their CSR.
Or they could be buddied up with another player.
And, of course, the match maker may have no choice (low population).
That’s the frustrating bit.
And that’s great advice - get as many kills as you can. Stick to that.
From a ranking point of view though - KPM and KD are two different things. Completely different beasts.
You can have a high KPM but a low KD.
And, of course, you can have a very high KD but a low KPM. And this is where the distinction is important. Playing for KD can lead to some toxic team behaviours. And slow your ranking journey down.
And in terms of reviewing your own performance and seeing where you need to improve, then 16 kills is kind of irrelevant if you don’t take the time of the game into account.
So it’s very important to not tell people they are same. They are not.
It’s an illustrative point.
But if it makes you feel better - KPM is a better reflection of your ability to engage the enemy.
And that’s why it’s a metric that is actually useful to weight into the ranking system. A Diamond player will have a very predictable KPM as you progress through Platinum, Diamond, and Onyx lobbies.
It doesn’t - but it can promote such play (it’s a legitimate strategy).
KD can be manipulated. I can camp the -yoink- out of an Onyx lobby and come away with a high KD. It would be particularly easy in some of the objective games.
But there is no way I could maintain an Onyx level KPM without actually being Onyx level.
KD is a rubbish stat.
It’s a bit hard to know for sure. Unfortunately services such as Halo Data Hive have been shut down for now.
But it appears to be a delta of 15 for everyone.
And quitters cop the full -15.
Everyone else on the team got -3.
Very interesting. Thank you. I’ll plug it into the spread sheet and have a play with some Infinite data.
No, they haven’t.
The inherent problem is that personal skill influences your MMR.
And including it the immediate post game algorithm effectively double dips it’s influence on the CSR.
Not at all. If your KD is very low, especially below 1.0, outside of playing in an actual peak setting like a LAN tournament with the best players on earth, then its proof that someone is not good at the game.
The only reason someone should ever have a KD close to 1 is because they’re playing in a setting where they’re completely evenly matched with those they’re playing with and against, consistently.
It’s useful on an individual basis. Game for game. In the context of the role you were trying to play in that particular game.
As a career stat it’s much less useful.
Generally true.
But the problem is in the other direction. A high KD isn’t guaranteed proof that you are particularly good at it. Other than protecting your KD.
Ideally we would have better access to our KPM. A graph of KPM vs opposition MMR would be a fantastic insight into a player’s rank and skill.
Not at all
I don’t get why you’re so against someone having a high kd
its no different than KD in that regards
KD is not a great indication of a good player.
Lot’s of people play in a way to farm their KD.
Or they spend too much time playing in team set ups that attract lower ranked players.
etc.
There have been plenty of vigorous threads previous debating the utility of KD.
I’m not per se.
Just advocating that KPM is a much better metric.
A KD of 2.0 sounds impressive. But going 4 & 2 is not necessarily a great sign you were involving yourself in the overall team strategy.
Going 16 & 8 - same KD - and obviously a much better game.
Doing it in a 16 minute game (KPM of 1.0) isn’t as helpful as an 8 minute game (KPM of 2.0). Especially if your opponents are maintain a KPM of 15 or above vs you.
Obviously there are pros and cons of any metric. And context is important.
But KPM is much more useful.
Particularly because it is the ranking metric.
It would be so useful to see your KPM at your current rank and then just above it. That would be far more informative to your progression than raw K, KD, KDA, K-D, etc.
Except that KD is only useful if you add context.
Going 16 and 8 is a KD of 2.0. But so is going 4 & 2.
And then getting 4 & 2 in an 4 minute game is much more impressive than scoring the same in a 11 minute game.
All these have the same KD. But if add the context of raw kills and the time taken to get them you can make a value judgement on which were likely to have been the better performances.
You know what other metric includes the raw number of kills and the time taken to get them?
there is no fundamental difference
none
whatsoever
You still have to get high kills and low deaths for it to register as a high kd over time
Your belief that a high KD is primarily achievable by getting 3 kills and hiding every game makes no sense.
WHY ARE YOU ONLY GOING 4 AND 2, WHAT IS THIS OBSESSION WITH HAVING LOW KILLS BUT CALLING IT A HIGH KD, WHY IS THIS LIKE TALKING TO A BRICK WALL
2 Likes
WHY ARE YOU ONLY GOING 4 AND 2, WHAT IS THIS OBSESSION WITH HAVING LOW KILLS BUT CALLING IT A HIGH KD, WHY IS THIS LIKE TALKING TO A BRICK WALL
[/quote]
I think his obsession with low kills is trying say that players would play safe to guarantee a high KD rather than play for high kills. But doesn’t really make sense if you only get 4 kills in an entire game, I doubt you are winning the match unless your teammates are carrying the rest of the kills.
Regarding KPM, it might be not an accurate metric in modes like oddball, who would want to be the ball carrier if their KPM will be lower holding the ball most of the match. You see this especially in higher ranked play where people are allergic to holding the ball.
If KPM is weighted higher in your gains rather than objective points, many players who climb the ranks can be great at consistently holding a high KPM but not as strong as being an objective based player.
1 Like
Going 4:2 still gives you the win in a slayer match if everyone else goes +/- 0. 
One is a ratio per time. The other a ratio per deaths.
That is fundamental.
One is a metric that helps you to rank up. The other isn’t.
True. Or at least high kills relative to your deaths.
But you don’t have to do them quickly.
Which means you can have very high KD with a low KPM and that can work against your rank.
That’s the point.
It’s illustrative. To help make the point.
Math’s isn’t everyone’s strong suit - so it helps if you exaggerate the examples.
IS THIS WHERE I YELL BACK AT YOU?
IT WAS TO SHOW THAT YOU DON’T NEED HIGH KILLS TO HAVE A HIGH KD.
THAT’S HOW RATIOS WORK.
AND IT"S AN IMPORTANT POINT FOR PEOPLE WHO ARE WONDERING WHY THEY AREN’T RANKING UP.
The numbers were just to make the example easy to follow.
You can use more realistic examples if you like.
Going 12 and 6 is a KD of 2.0 - but someone who goes 13 and 12, which is a KD barely over 1, still ends up with a higher KPM.
It’s obviously less relevant in objective games.
But I believe Microsoft found it still added to the accuracy of the ranking overall. And it’s only a weighting.
The key is make sure you always involved in winning and protecting the oddball.
Yes, you won’t end up with as many kills - but you the idea is to not go low enough that it’s detrimental
Skilled players can still be bad team mates.
But a lot of this is on 343. Their poor communication allows mistruths like “KD = rank” to perpetuate.
It is. There is no weighting for objective points.
What I would do is adjust your KPM for time holding the oddball. So that your kills are calculated only against time with your gun out.
the game lasts the same amount of time either way
Yes you do, that’s what you fail to grasp
You’re not getting anywhere hiding in a corner with 4 kills while your teammates die. You’re proposing that people play extremely badly in ranked for no reason.
2 Likes
Josh explained to me that the limits of Halo 5’s design prevented them from addressing the skill-lag in the CSR formula.
Your[pre-match]CSR vs. OpposingTeam[pre-match]MMR + (Your[pre-match]MMR - YourTeam[pre-match]MMR))
So with Infinite perhaps they were able to finally design the game’s foundation to allow for a formula such as the one below to function.
Your[pre-match]CSR vs. OpposingTeam[post-match]MMR + (Your[post-match]MMR - YourTeam[post-match]MMR))
This formula captures the post-match performance in it’s entirety ensuring that the CSR adjustment isn’t reliant on the skill ratings from the previous match performance, but instead the most recent performance.
Don’t believe they did.
I couldn’t find any correlation between KPM and CSR changes. That was over 500+ data points.
I miss Halo Data Hive.
But because the CSR is changed by the MMR you can’t have KPM change both the CSR and the MMR. That is double dipping.