I don’t think it was like this right at the beginning. At that stage the games were tight.
It didn’t take long though for players to realise they could manipulate the squads - so you would get those teams with people grinding Onyx teaming up with smurfs and sandbaggers.
And it only got worse into S2 - hence 343’s move to put limits on the squads.
But before they could the population fell to the point that the match maker was compromised.
Fingers crossed we get back there soon.
This should be good… 
What’s your evidence for this?
It’s not how it’s described in TrueSkill2. It’s not how it worked in Halo 5. And it’s not how it’s described by 343 in their blog;
I think people are overestimating the effect of using a current MMR to seed a new playlist (eg. the infamous bots into ranked post on Reddit) - or the effect of the global MMR on an immature MMR (while it wide and volatile).
Each playlist should have it’s own MMR offset curve.
And we saw recently with FFA playlists such as LSS you could look up your MMR in those playlists - and they were clearly different from your Ranked MMR.
And we’ll soon have even more evidence when we get a second ranked playlist next week (touch wood).
But anyway. If you want to end this topic I’ll add the following quote from 343, also from their ranking blog;
But this would involve players;
I see at least another 100 posts in this thread’s future.
Games should be, but every once in awhile teams just pop off and there’s nothing you can truly do about it.
Let me just ask this one question: You believed that the current SBMM system was designed for player retention because, in your words,
So then if this is the case, why is the population stretched so thin?
So this quote literally just proved the post. Sure, 343 isn’t “intentionally enforcing it,” no one said they were. And i really dont appreciate your strawmanning in every single post you write. I said from the very beginning that their system is gatekeeping certain challenges, that theyve created a social system that favors results over enjoyment. Youre the only one peddling conspiracy crap, in a pretty disrespectful way in an attempt to undermine everyone who disagress with you. Youre literally just invalidating this discussion by claiming to be the end all be all of this topic, and strawmanning everone else out of the way by telling them to take off their tinfoil hat.
Anyway, 343 admitted to designing a system that results in an average of a 50% win rate. No, 343 doesn’t give us unfair matches to enforce that win rate, the system they designed does that job for them. They say it’s “not a goal,” so then why did they design a system with an expected average value of 50%? This isnt conspiracy talk. This is calling 343 out on their BS. They say one thing, but the writing is (in this case literally) on the wall.
I’m just saying whatever logic they said or claim to be isn’t working… people are creating alt accounts and going straight into ranked and its taking them forever to rank up even if they go 20+ and win… since the logic cant calculate your social MMR it handles Ranked differently… these people play with high onyx players to get lower plats and low diamonds… go 60-2 in btb for 500 games and you will face high onyx in ranked… its directly correlated… go ahead and test it you seem to know more about these systems than Max Hoberman.
Skill ceiling equals 50 percent win rate once you’ve hit that ceiling. Before the ceiling there is/was positive reinforcement with leveling up, thus more motivation to win that final battle. Once I’ve hit my ceiling the 50 percent win rate is only if my play at my hardest every match. Sometimes I’m buzzin and it’s hard to catch up to these young whippersnappers
That is what i experienced too
The forecast of expected Kills and Death does not take in account:
I want just hear a logical explanation why the forecast of expected death/kill should be correct.
Why does the forecast not take in account the objective? Like flags returned, carrier stopped, stronghold captured/defended?
Because it does not take in account the points listed above
Im not telling you that 343 is evil. Im just telling you that this system is broken, it never worked correct. And now with the low population that remained , how should this system work?
Desync, lack of content, no XP progression, et al.
Many more obvious reasons.
The evidence for SBMM and player retention would actually suggest that the population would be stretched even further without it.
I don’t set out with this in mind. In fact I tend to structure my posts to answer what has been posted.
You did. Based on Waypoint’s “expected kills”. But I’ve asked a couple of times for you, or anyone, to tell me how the stat proves 343 is manipulating anything?
Ok?
Let’s do a quick run through the thread…
I know that specifically taking bites of quotes can take people out of context. But I think we can see which side of the fence you are sitting on in regards to 343 forcing WL to manipulate your rank.
Because that is the basic goal of a ranking - match making system.
If your ranking system is accurate then putting players of the same rank together will be a 50:50 result.
Just like the population is spread across a skill curve. Each division (gold to onyx) is one standard division from the mean. And if you play someone a division above you then you will have a 25:75 win rate.
I’m unaware of ranking being slower than usual?
The system specifically needs 46 games to rank you (from a 4v4 playlist).
And that’s if you play consistently.
Going 20+ may rank you up faster - it just depends if that game expected you to go 20+ or not.
Not quite sure what you are getting at here?
But each playlist has it’s own offset MMR. Again you need enough games in each mode to stabilise them to be more independent of your global stats.
I have. Recorded average ranked MMRs after spending a few hours each week in social mucking around in friends.
There did seem to be a dip in opponent MMR for a few games - but it corrected very quickly. And I got the same, and even worse, dips in opponent MMR wiith drops in form playing Ranked.
This is consistent with TrueSkill2 having parameters (or at least global stats) representing “form”.
I have never purported to be an expert. I have said multiple times we need someone from 343 to step in and provide Q&A (like Menke used to).
Mostly I just try to get people on track to what we should be realistically expecting. Like at least reading the TrueSkill 2 papers, watching some of the GDC talks. Or at least looking to Halo 5 to expect what has been carried over to Infinite.
Having said that, somewhat cheekily, Max probably hasn’t had a lot to do with TS2. So maybe…
And that was a joke people.
It’s just a stat calculated from your past history.
Your KPM vs that level. Multiplied by the game duration (in retrospect) to give a value.
It is affected by all the game quality factors; pings, team dynamics, squads, and other things. But so is your KD, KDA, assists, and all the other stats we hold so dear.
It shouldn’t be looked at any differently to any of those.
It’s a value calculated from your past game history.
It’s an average representation of how you’ve played against that level of opponent before.
If you’ve been playing consistently - it should be very accurate.
Two things;
Objective games are essentially Slayer with a job to do. We know that the better teams concentrate on Slaying first and then push the objective when they have the numbers advantage.
We know from TrueSkill2 research that the only stat that has extra ranking prediction is KPM (and to a lesser extent DPM). That’s not to say that Objective play isn’t important - it is - but it’s value is in the win. If you rank someone up harder on the basis of objective score then they have a miserable time losing until they de-rank again.
The system should work to the point that you can predict the result. Not each individual game per se - but overall.
Matching players of equal skill should be 50:50.
Players a division apart should be 75:25.
You can validate your system by looking back across the results and seeing that various skill gaps have the appropriate W/L ratios. If you have the right results across a rank then you know the players are equal skill. And if they have the right results with players above and below them then you know you have them in the right order.
I think where it breaks down is when people start using the ranking system as a career grind. What’s broken is the lack of an XP progression system.
Again, you already cleared the air here. Stop strawmanning this conspiracy bullcrap. This is straight from the horses mouth. “We don’t give you unfair matches just to enforce a 50% win rate. But the system will still do that anyway. But it’s not our fault that the system that we designed is enforcing a 50% win rate. Oh, but its not actually enforcing a 50% win rate, we just expect that to happen.”
Like you copied and pasted a bunch of my past quotes, but, ummm none of them are really incorrect. A conspiracy comes from the supposition that the information being offered is a lie and is meant to propogate a lie. I havent lied here. And you yourself added more information from 343 to reinforce that.
Please forgive me for my poor semantics. It’s not 343’s fault. It’s the algorithm. The algorithm that 343 made. My bad. Again though, exactly where did i say there were agents in 343 manipulating anything? It seems like youre still strawmanning here…
All of that is pure conjecture, because it completely ignores a variety of variables in each individual players experience. Especially when you consider that the way 343 has set up matchmaking, there is no clean 4v4 all golds. It’s more like a diamond mixed with 3 golds vs. a team of 4 platimum players. I wouldnt depend too much on the probabilities in a game like that, because as I’ve already said, 1 person cant be expected to win an objective based game. And, going off your math there, if im in plat and i outperform someone in diamond, shouodnt i receive a 25% boost to my rank xp at the end of the match? Doesnt make much sense to bog me down if my displayed more skill than someone who is supposed to be better than me
You have a theory that 343 is conspiring against you by manipulating team mates and setting expected kill targets that can’t be reached.
Is that not a conspiracy theory?
And all this is based on a bunch of anecdotal experiences.
And now you are using a quote from 343 specifically denying your theory as proof that you are right?
You do realise that the same endpoint doesn’t mean the same method.
And you are still dodging the question.
The first statement in your OP was…
So what specifically is “disturbing” about it.
You then link this statement with;
I’m just asking what is so egregious about “expected kills”. And how does it favour results.
Pretty sure this has already been covered ad nauseum. Because the algorithm then places a higher skilled person on a team with bad teammates who will underperform and basically guaruntee a loss. When they have a monetized challenge system in place that is based upon RNG playlists that also require you to win said RNG matches, it is an egregious conconction.
Would love to see the quote where i said any of this.
The quote says:
Do you really think that just because 343 says the words “we don’t enforce it,” that it automatically means that they dont enforce it? People can bend the truth ya know. Just because someone says something, it doesn’t mean its true. Especially if that something that was said was literally just contradicted in the sentence before it was said. Like honestly, are you just naive, or are you being willfully ignorant for the sake of Halo’s legacy?
Lol it also explicitly says “fair matches” but Mint’s experiences dont see to be fair if his objective is to win a game.
Conspiracy theorist 101.
If you outperform a Diamond team then your MMR will go up. If you do it consistently it will go up to Diamond and your CSR will follow as quickly as it can.
If you outperform a Platinum team that just happens to have a Diamond player in it - not so much.
You wouldn’t want to rank up in this circumstance. If you are only 50:50 vs Platinum teams you will struggle with the better teams that a personal Diamond rank will attract. But hey, at least you’d get your wish - your W/L won’t be 50% anymore
And where, pray tell, are you getting your metrics from?
Mint’s objective isn’t to win the game. In the case of his kill streak videos it’s to smurf his way into low level lobbies.
He’s literally been banned for forcing his way into games with opponents going at 8 kills a games vs his 42.