With what I’m suggesting you would never get a negative win-lose ratio. Worst outcome is 3-3 which is a tie.
With what we have currently can potentially end up with 0-3.
It’s not a tie. A shuffle is not just one match. It’s 6 matches, one round is one match, and that’s how the rating system counts it.
Also, you’ve played those rounds. Don’t conveniently forget that. What you’re suggesting would make people win and lose rounds they didn’t even get to play. Which contradicts your earlier statement:
and also why
is wrong.
It’s a tie because in even MMR scenario it will result in 0 rating change.
Doesn’t matter how the other games went if there is a leaver. System should never punish you for not having your chance to beat him. So my soultion is the only reasonable one.
So instead of making me look up you could’ve said the constant is τ in glicko. I mean no one knows their implementation and talking about semantics is meaningless.
That’s why I told you it’s the instability and expanding parameters in the matchmaking causing the issues. I just told you it isn’t the k-factor to blame, because it’s not that simple, and then told you about the glicko systems if you were curious about the math of it, but that’s not really material to the convo.
It’s not punishing you. It’s just counting the rounds played, good and bad. You can’t have the good without the bad. And without both, it made people distrust the system and hate the game mode entirely.
But ok, if you want to talk about “punishments”, so you want to punish the people who won their rounds, by forcing them to forfeit the rest of their rounds?
There must be trust in the rounds played to count. That’s essential.
Sorry I made my point clear enough and I’m not going to bother replying with same stuff again and again. I don’t have the energy to argue with someone lacking basic reading comprehension skills, have a nice day.
What the hell are you talking about. It’s not magic, it’s just math :d Where do you get these ideas. You can just pull up the paper and take a look.
Because Blizzard did their own adaption of it, and so you can’t even base it on that.
But I also linked you to a way to fix that issue, which you didn’t seem to check.
I just don’t see what’s your issue. I don’t think there is anything to fix from blizzard’s perspective. Personally I’d like to see the rating fluctuation to be a bit smaller from single shuffle but as I said earlier, there is no way to win from blizzard’s perspective, someone will be unhappy anyway.
And what I linked to is a way to accomplish that. Because it’s a bigger problem than you seem to give it credit for.
The deviation value just isn’t compatible with the game mode the way it works now. So the gains & losses are too large, which makes the feeling of losing more pronounced.
Hence why putting in maximum numbers and minimum numbers brings it more within reason.
It is actually a huge problem and you are suggesting “simple way to accomplish that” - that is a true red flag you should pick up instantly yourself.
That’s why you have certain people doing it, not some random guys giving suggestion on forums on a system they can’t know how it actually works.
I’m just using generic terms that means lowering the volatility of the rating fluctuation, you are trying to give them advice how to fix something you cant’ know.
Like what do you think this even means?
The MMR system would remain unaffected because messing with that really does have more consequences than one can imagine without knowing their exact systems in place, but the way it grants the rating gains and losses would move in a more predetermined fashion, with smaller variations based on those MMR differences.
What do you think it means?
Right now, the way it is based solely on the MMRs relative to your own CR, is causing problems like not gaining rating for 4 (or more) won rounds in a shuffle after everything is over, which is causing bad optics in the eyes of the player base.
And can cause bigger-than-necessary losses for some rounds won. Which is also causing bad optics.
So simply putting in those maximums and minimums in the CR changes, would help mitigate a lot of those problems, although also slowing it down. But it’s needed.
I don’t think you quite understand how rating systems are supposed to work. You have some kind of expectation in a game. Like let’s say you are higher rated than the average lobby. You are expected to win 4 games. If you are expected to win 4 games, why would you ever gain rating for winning 4 games? You are already at your expected rating.
Now the thing MMR and CR being different is an interesting question, but in general case it’s kinda needed in WoW, but that probably shouldn’t be true for solo shuffle.
That’s not really what’s happening per se. It’s not expecting you to win a certain amount of rounds, to be specific. To start with, it’s a team game, so it’s “expecting” you to win based on the team MMR vs. the opponents’ team MMR. Your personal MMR “expects” you to either carry or be carried or provide an equal part in the round, based on the differences.
Then the personal MMR relative to the team MMR is designed to bring your personal MMR closer to your team MMR, by making you win less when you win against opponents lower than your CR/personal MMR, while your teammates wins more if they’re lower, and vice versa when you’re beneath the opponents’ MMR.
Simply put, the rating system is designed to eventually make your personal MMR reach your team’s MMR. Which is a function that doesn’t fit with the shuffle game mode. (And then it makes your CR reach your personal MMR, but anyway, it’s all connected.
The more games you play, the more its accuracy increases, as is the ultimate goal of rating systems overall.)
No, I think it’s probably needed for the spread of rating ranges. Redesigning that would be a huge undertaking.
Which is why I don’t like messing with the MMR system.
But the CR system is easier to mess with, with fewer consequences.
There is nothing to redesign. You just lower the volatility of the MMR change and call it CR in context of solo shuffle. The reason why it’s separated is because of in normal ladder you queue with different comps and people of different skill levels and they want to make as good games as possible as fast as possible without making your CR change too fast.
I just don’t know if this is true at all. It’s so hard to figure out but I think you are just wrong but no way I’m wasting my time thinking this through. There is someone who is paid good money to figure it out, but for some reason I don’t think it’s you.
You have a bit of misconception what MMR is. You think it’s some points you are supposed to grind, but it’s not. It’s there to put you in a game with similar skilled people and produce fair games. If you win more than the system thinks you should win, then it pumps you up and places you on a tougher game and so on.
I will do that.
But man you sound like a massive toxic di-ck, putting a terrible attitude on.
Would be fun if someone leaves to continue the game as 2v2.
One has to watch and the observer gets on one random team, if he loses it will also count as a lose.
I will take that as a compliment.
2 Likes
I would to.
This topic was automatically closed 30 days after the last reply. New replies are no longer allowed.