Skip to main content

Forums / Community / Matchmaking Feedback & Discussion

Matchmaking Update 3/17/2020

OP ZaedynFel

  1. 1
  2. ...
  3. 7
  4. 8
  5. 9
  6. 10
  7. 11
  8. ...
  9. 39
Sorry guys - but matching sucks. I am a low to
mid diamond player. I am sick of getting wrecked by 2 champions and an onyx dude vs diamond players.
I can’t believe that even happens.
Which playlist? Never happens to me in slayer. I'm D1 currently and almost never see onyx in my matches. Always high plat to mid diamond at the most. If I ever do see one, there's always a gold player on their team as well.
Chimera30 wrote:
Flipfadez wrote:
Kashyy wrote:
Chimera30 wrote:
I thought boosting was no longer a thing
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/9d45e440-4e4e-4bac-a06f-bbb96327260b/players/i%20800%20get%20fived?gameHistoryMatchIndex=1&gameHistoryGameModeFilter=All
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/254b5b23-841b-4f6b-a323-23e4526cbe0d/players/i%20800%20get%20fived?gameHistoryMatchIndex=0&gameHistoryGameModeFilter=All
I see Onyx/Diamonds vs Onyx/Diamonds.

Can you clarify what the issue here is?
My Onyx teammate underperforming not once, but twice and not getting punished because he's in a to2 with a lower ranked player
I'm not really seeing "boosting" here. People have bad games/days. It happens.
I've had bad games and games, so of course I know the feeling, but even CSR doesn't forgive or care. It's a machine/system, but even then, how would it know?
The system looks at your entire play history, so it can calculate a mean value for your skill across time. Negative outliers from this mean would be what the system sees as having a very bad match. Let's say you have once bad match, and just do really poorly. Your MMR might take a big hit behind the scenes, but luckily CSR can't shift as much as MMR, so one bad game won't send you from Diamond to Silver. Assuming you play normally the next match, your MMR will update back to around your normal skill value, meaning any CSR lost from the bad match should be made back quickly. Even a few bad matches in a row wouldn't lead to a permanent dip in rank. It would only be permanent if you kept having "bad" matches, because after a certain point it would be indicative of a loss of skill rather than bad luck.
Personally, I think that judging a players skill based on KPM and DPM is not an accurate measure. A lot of flaws into that system resembles the flaws reach had in his first season. If a player hogs power weapons/power ups, does that determine whether he's good or not? Why isn't assist not taken into consideration when it comes to measurement of MMR?

Bungie, realized that error and made their system strictly wins and losses. This is a TEAM game. A player that averages 6 kills, but 10+ assist a game deserves just as much credit as the player that gets 17 kills and 5 assists with 5-6 deaths. It's an insult to players that are more supportive.
I agree with this statement.

I do not feel the current Halo 5 ranking system is very good or accurate.
All factors that contribute to a win by each individual player should count as positive “points” in the equation.

Contributory points should be: Total Kills, KPM, Assists, DPM, Protector, Bodyguard, Distraction, Winning. Of course the intangibles can’t be quantified unfortunately such as communication, positioning but using all variables would give a much more accurate rank for a skill based system.

Otherwise, the full picture of the positive impact is not truly being captured and therefore simply not mathematically accurate.
If adding assists and all that other crap into the model makes it predict the winner less often, then it's less mathematically accurate. And that's exactly what happens. KPM predicts the winner of a match correctly like 69% of the time. KPM+assists predicts the winner correctly less often. Josh and the matchmaking team have the ability to actually test all the factors you think should go into the model and see how they work. And they did test them, and it turns out they don't work. Now, if you want to design a skill system that doesn't use winning as the predominant factor, then maybe you could find a way to incorporate all those other factors you feel are so important to determining skill.
Points taken.

and yes I agree winning should be the most heavily weighted variable. Then KPM, and so forth. Each variable should carry a weight within the equation however in determining skill. It just just not accurate if assists for example are not factored in. Or a distraction as this contributes to the outcome of a match in a positive way and should be counted as a positive contribution.

it would be like in basketball doing away with tripe doubles if suddenly “assists” were no longer counted. It doesn’t make sense .
Flipfadez wrote:
Chimera30 wrote:
Flipfadez wrote:
Kashyy wrote:
Chimera30 wrote:
I thought boosting was no longer a thing
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/9d45e440-4e4e-4bac-a06f-bbb96327260b/players/i%20800%20get%20fived?gameHistoryMatchIndex=1&gameHistoryGameModeFilter=All
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/254b5b23-841b-4f6b-a323-23e4526cbe0d/players/i%20800%20get%20fived?gameHistoryMatchIndex=0&gameHistoryGameModeFilter=All
I see Onyx/Diamonds vs Onyx/Diamonds.

Can you clarify what the issue here is?
My Onyx teammate underperforming not once, but twice and not getting punished because he's in a to2 with a lower ranked player
I'm not really seeing "boosting" here. People have bad games/days. It happens.
I've had bad games and games, so of course I know the feeling, but even CSR doesn't forgive or care. It's a machine/system, but even then, how would it know?
The system looks at your entire play history, so it can calculate a mean value for your skill across time. Negative outliers from this mean would be what the system sees as having a very bad match. Let's say you have once bad match, and just do really poorly. Your MMR might take a big hit behind the scenes, but luckily CSR can't shift as much as MMR, so one bad game won't send you from Diamond to Silver. Assuming you play normally the next match, your MMR will update back to around your normal skill value, meaning any CSR lost from the bad match should be made back quickly. Even a few bad matches in a row wouldn't lead to a permanent dip in rank. It would only be permanent if you kept having "bad" matches, because after a certain point it would be indicative of a loss of skill rather than bad luck.
Personally, I think that judging a players skill based on KPM and DPM is not an accurate measure. A lot of flaws into that system resembles the flaws reach had in his first season. If a player hogs power weapons/power ups, does that determine whether he's good or not? Why isn't assist not taken into consideration when it comes to measurement of MMR?

Bungie, realized that error and made their system strictly wins and losses. This is a TEAM game. A player that averages 6 kills, but 10+ assist a game deserves just as much credit as the player that gets 17 kills and 5 assists with 5-6 deaths. It's an insult to players that are more supportive.
I agree with this statement.

I do not feel the current Halo 5 ranking system is very good or accurate.
All factors that contribute to a win by each individual player should count as positive “points” in the equation.

Contributory points should be: Total Kills, KPM, Assists, DPM, Protector, Bodyguard, Distraction, Winning. Of course the intangibles can’t be quantified unfortunately such as communication, positioning but using all variables would give a much more accurate rank for a skill based system.

Otherwise, the full picture of the positive impact is not truly being captured and therefore simply not mathematically accurate.
If adding assists and all that other crap into the model makes it predict the winner less often, then it's less mathematically accurate. And that's exactly what happens. KPM predicts the winner of a match correctly like 69% of the time. KPM+assists predicts the winner correctly less often. Josh and the matchmaking team have the ability to actually test all the factors you think should go into the model and see how they work. And they did test them, and it turns out they don't work. Now, if you want to design a skill system that doesn't use winning as the predominant factor, then maybe you could find a way to incorporate all those other factors you feel are so important to determining skill.
Points taken.

and yes I agree winning should be the most heavily weighted variable. Then KPM, and so forth. Each variable should carry a weight within the equation however in determining skill. It just just not accurate if assists for example are not factored in. Or a distraction as this contributes to the outcome of a match in a positive way and should be counted as a positive contribution.

it would be like in basketball doing away with tripe doubles if suddenly “assists” were no longer counted. It doesn’t make sense .
There are few players and around the world it is now easier for me to have assistance ... then I leave one shot and someone who earns Kills for me .....
Chimera30 wrote:
Flipfadez wrote:
Kashyy wrote:
Chimera30 wrote:
I thought boosting was no longer a thing
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/9d45e440-4e4e-4bac-a06f-bbb96327260b/players/i%20800%20get%20fived?gameHistoryMatchIndex=1&gameHistoryGameModeFilter=All
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/254b5b23-841b-4f6b-a323-23e4526cbe0d/players/i%20800%20get%20fived?gameHistoryMatchIndex=0&gameHistoryGameModeFilter=All
I see Onyx/Diamonds vs Onyx/Diamonds.

Can you clarify what the issue here is?
My Onyx teammate underperforming not once, but twice and not getting punished because he's in a to2 with a lower ranked player
I'm not really seeing "boosting" here. People have bad games/days. It happens.
I've had bad games and games, so of course I know the feeling, but even CSR doesn't forgive or care. It's a machine/system, but even then, how would it know?
The system looks at your entire play history, so it can calculate a mean value for your skill across time. Negative outliers from this mean would be what the system sees as having a very bad match. Let's say you have once bad match, and just do really poorly. Your MMR might take a big hit behind the scenes, but luckily CSR can't shift as much as MMR, so one bad game won't send you from Diamond to Silver. Assuming you play normally the next match, your MMR will update back to around your normal skill value, meaning any CSR lost from the bad match should be made back quickly. Even a few bad matches in a row wouldn't lead to a permanent dip in rank. It would only be permanent if you kept having "bad" matches, because after a certain point it would be indicative of a loss of skill rather than bad luck.
Personally, I think that judging a players skill based on KPM and DPM is not an accurate measure. A lot of flaws into that system resembles the flaws reach had in his first season. If a player hogs power weapons/power ups, does that determine whether he's good or not? Why isn't assist not taken into consideration when it comes to measurement of MMR?

Bungie, realized that error and made their system strictly wins and losses. This is a TEAM game. A player that averages 6 kills, but 10+ assist a game deserves just as much credit as the player that gets 17 kills and 5 assists with 5-6 deaths. It's an insult to players that are more supportive.
I agree with this statement.

I do not feel the current Halo 5 ranking system is very good or accurate.
All factors that contribute to a win by each individual player should count as positive “points” in the equation.

Contributory points should be: Total Kills, KPM, Assists, DPM, Protector, Bodyguard, Distraction, Winning. Of course the intangibles can’t be quantified unfortunately such as communication, positioning but using all variables would give a much more accurate rank for a skill based system.

Otherwise, the full picture of the positive impact is not truly being captured and therefore simply not mathematically accurate.
If adding assists and all that other crap into the model makes it predict the winner less often, then it's less mathematically accurate. And that's exactly what happens. KPM predicts the winner of a match correctly like 69% of the time. KPM+assists predicts the winner correctly less often. Josh and the matchmaking team have the ability to actually test all the factors you think should go into the model and see how they work. And they did test them, and it turns out they don't work. Now, if you want to design a skill system that doesn't use winning as the predominant factor, then maybe you could find a way to incorporate all those other factors you feel are so important to determining skill.
You can not determine who is going to win. It’s nearly improbable. Enhancers such as power ups and power weapons throws everything off. Am I skillful if I get camo/OS or Rockets/Snipers 70% of the time on maps and my KPM is high and my DPM is low? If those components weren’t in matches, then we can have an argument. Wins and losses should be the only factor in a game like this. Rarely have I heard people say that the system has paired them with people of their “skill” level.

Thats like if in basketball, a money ball was thrown in a game every three minutes where every basket is 4
4 points. It throws the balance off. Enough with trying to explain to people the system because what you explain doesn’t equate to what’s happening in matchmaking.

And if wins are the main focus, why do teams for the most part, go up 1 CSR with the exception of maybe one of two players in the fireteam?
Kashyy wrote:
Chimera30 wrote:
Flipfadez wrote:
Kashyy wrote:
Chimera30 wrote:
I thought boosting was no longer a thing
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/9d45e440-4e4e-4bac-a06f-bbb96327260b/players/i%20800%20get%20fived?gameHistoryMatchIndex=1&gameHistoryGameModeFilter=All
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/254b5b23-841b-4f6b-a323-23e4526cbe0d/players/i%20800%20get%20fived?gameHistoryMatchIndex=0&gameHistoryGameModeFilter=All
I see Onyx/Diamonds vs Onyx/Diamonds.

Can you clarify what the issue here is?
My Onyx teammate underperforming not once, but twice and not getting punished because he's in a to2 with a lower ranked player
I'm not really seeing "boosting" here. People have bad games/days. It happens.
I've had bad games and games, so of course I know the feeling, but even CSR doesn't forgive or care. It's a machine/system, but even then, how would it know?
The system looks at your entire play history, so it can calculate a mean value for your skill across time. Negative outliers from this mean would be what the system sees as having a very bad match. Let's say you have once bad match, and just do really poorly. Your MMR might take a big hit behind the scenes, but luckily CSR can't shift as much as MMR, so one bad game won't send you from Diamond to Silver. Assuming you play normally the next match, your MMR will update back to around your normal skill value, meaning any CSR lost from the bad match should be made back quickly. Even a few bad matches in a row wouldn't lead to a permanent dip in rank. It would only be permanent if you kept having "bad" matches, because after a certain point it would be indicative of a loss of skill rather than bad luck.
Personally, I think that judging a players skill based on KPM and DPM is not an accurate measure. A lot of flaws into that system resembles the flaws reach had in his first season. If a player hogs power weapons/power ups, does that determine whether he's good or not? Why isn't assist not taken into consideration when it comes to measurement of MMR?

Bungie, realized that error and made their system strictly wins and losses. This is a TEAM game. A player that averages 6 kills, but 10+ assist a game deserves just as much credit as the player that gets 17 kills and 5 assists with 5-6 deaths. It's an insult to players that are more supportive.
I agree with this statement.

I do not feel the current Halo 5 ranking system is very good or accurate.
All factors that contribute to a win by each individual player should count as positive “points” in the equation.

Contributory points should be: Total Kills, KPM, Assists, DPM, Protector, Bodyguard, Distraction, Winning. Of course the intangibles can’t be quantified unfortunately such as communication, positioning but using all variables would give a much more accurate rank for a skill based system.

Otherwise, the full picture of the positive impact is not truly being captured and therefore simply not mathematically accurate.
If adding assists and all that other crap into the model makes it predict the winner less often, then it's less mathematically accurate. And that's exactly what happens. KPM predicts the winner of a match correctly like 69% of the time. KPM+assists predicts the winner correctly less often. Josh and the matchmaking team have the ability to actually test all the factors you think should go into the model and see how they work. And they did test them, and it turns out they don't work. Now, if you want to design a skill system that doesn't use winning as the predominant factor, then maybe you could find a way to incorporate all those other factors you feel are so important to determining skill.
You can not determine who is going to win. It’s nearly improbable. Enhancers such as power ups and power weapons throws everything off. Am I skillful if I get camo/OS or Rockets/Snipers 70% of the time on maps and my KPM is high and my DPM is low? If those components weren’t in matches, then we can have an argument. Wins and losses should be the only factor in a game like this. Rarely have I heard people say that the system has paired them with people of their “skill” level.

Thats like if in basketball, a money ball was thrown in a game every three minutes where every basket is 4
4 points. It throws the balance off. Enough with trying to explain to people the system because what you explain doesn’t equate to what’s happening in matchmaking.

And if wins are the main focus, why do teams for the most part, go up 1 CSR with the exception of maybe one of two players in the fireteam?
Incorrect, the system is extremely accurate at not only predicting who will win each match (despite camo/OS, Rockets/Snipers), it is precise at predicting how many times you will win against a specific skill level.

The model already accounts for variation introduced by variations in individual player performances and any other factors like camo/OS.

You are also wrong about what happens in matchmaking. Players are winning and losing as predicted, and in fact, the number of kills you are getting per game is consistently predicted correctly. So, no, it most definitely does equate to what's happening in matchmaking.

Wins are the main focus, but each win has a different worth depending on how good your opponents are relative to you. Accounting for wins means the model that best predicts who wins each match gives the best estimate of each player's skill. The current model is great at that, so we know the MMRs are right.

Players getting less CSR are already close enough to their correct skill (MMR) and should not go higher because they are already winning as expected by their current CSR.

Players getting more CSR have a CSR below their actual recent performances, so will go up more to match if they earn it.

If your CSR is below your MMR, you will play teams above your CSR and, if you win, your CSR goes up quickly because you are beating teams better than your CSR.

If your CSR is above your MMR, you will play teams worse than your CSR, and if you beat them, you won't get much because you are beating teams worse than you. If you lose to them, then of course you go down fast because you are losing to people worse than you.
Flipfadez wrote:
Chimera30 wrote:
Flipfadez wrote:
Kashyy wrote:
Chimera30 wrote:
I thought boosting was no longer a thing
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/9d45e440-4e4e-4bac-a06f-bbb96327260b/players/i%20800%20get%20fived?gameHistoryMatchIndex=1&gameHistoryGameModeFilter=All
Quote:
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/254b5b23-841b-4f6b-a323-23e4526cbe0d/players/i%20800%20get%20fived?gameHistoryMatchIndex=0&gameHistoryGameModeFilter=All
I see Onyx/Diamonds vs Onyx/Diamonds.

Can you clarify what the issue here is?
My Onyx teammate underperforming not once, but twice and not getting punished because he's in a to2 with a lower ranked player
I'm not really seeing "boosting" here. People have bad games/days. It happens.
I've had bad games and games, so of course I know the feeling, but even CSR doesn't forgive or care. It's a machine/system, but even then, how would it know?
The system looks at your entire play history, so it can calculate a mean value for your skill across time. Negative outliers from this mean would be what the system sees as having a very bad match. Let's say you have once bad match, and just do really poorly. Your MMR might take a big hit behind the scenes, but luckily CSR can't shift as much as MMR, so one bad game won't send you from Diamond to Silver. Assuming you play normally the next match, your MMR will update back to around your normal skill value, meaning any CSR lost from the bad match should be made back quickly. Even a few bad matches in a row wouldn't lead to a permanent dip in rank. It would only be permanent if you kept having "bad" matches, because after a certain point it would be indicative of a loss of skill rather than bad luck.
Personally, I think that judging a players skill based on KPM and DPM is not an accurate measure. A lot of flaws into that system resembles the flaws reach had in his first season. If a player hogs power weapons/power ups, does that determine whether he's good or not? Why isn't assist not taken into consideration when it comes to measurement of MMR?

Bungie, realized that error and made their system strictly wins and losses. This is a TEAM game. A player that averages 6 kills, but 10+ assist a game deserves just as much credit as the player that gets 17 kills and 5 assists with 5-6 deaths. It's an insult to players that are more supportive.
I agree with this statement.

I do not feel the current Halo 5 ranking system is very good or accurate.
All factors that contribute to a win by each individual player should count as positive “points” in the equation.

Contributory points should be: Total Kills, KPM, Assists, DPM, Protector, Bodyguard, Distraction, Winning. Of course the intangibles can’t be quantified unfortunately such as communication, positioning but using all variables would give a much more accurate rank for a skill based system.

Otherwise, the full picture of the positive impact is not truly being captured and therefore simply not mathematically accurate.
If adding assists and all that other crap into the model makes it predict the winner less often, then it's less mathematically accurate. And that's exactly what happens. KPM predicts the winner of a match correctly like 69% of the time. KPM+assists predicts the winner correctly less often. Josh and the matchmaking team have the ability to actually test all the factors you think should go into the model and see how they work. And they did test them, and it turns out they don't work. Now, if you want to design a skill system that doesn't use winning as the predominant factor, then maybe you could find a way to incorporate all those other factors you feel are so important to determining skill.
Points taken.

and yes I agree winning should be the most heavily weighted variable. Then KPM, and so forth. Each variable should carry a weight within the equation however in determining skill. It just just not accurate if assists for example are not factored in. Or a distraction as this contributes to the outcome of a match in a positive way and should be counted as a positive contribution.

it would be like in basketball doing away with tripe doubles if suddenly “assists” were no longer counted. It doesn’t make sense .
It does make sense.

What you as a player think is contributing, does not necessarily contribute. We have tested each of those variables for whether or not they should be in the model at all, and they should not be there because they make the win prediction worse, not better.

If adding "distractions" to the model actually makes the system less accurate at predicting who will win the match, then "distractions" do not contribute more than just random noise.

In fact, players who are really good at getting distractions actually LOSE more than those who don't. So, no, that should NOT be weighted in the model in any positive form.

We do analyze these variables in depth and we will not add extra ones that actually don't improve predicting who WINS.
I think the above two points are really important for people to remember.

Ultimately, yes, there are a lot of factors and 'intangibles' that go into determining how 'good' or 'skillful' a player is.

However, I think the most objective measure is how often a player wins their match.

So, if we can agree that winning a match is the most important thing, then we can focus on what factors contribute to determining if a player is going to win their match.
That is what I think Josh is talking about. They are only going to include factors that increase the accuracy of the prediction that the model makes of whether or not the player wins their match.

[edit]
It's important to remember that not all things actually contribute to a win. A distraction medal simply means someone else on your team saved your life. It doesn't necessarily mean you contributed to winning that match. There are lots of other factors, which when taken at their face value might seem to be good contributors, but when the data is analyzed, they might prove to not be factors in winning at all.
can you please explain this to me?
Played 8 games o slayer yesterday. Started at a Diamond 1.5, won 6 of those 8 games and ended up a Diamond 1.2
why if i only lost 2 out of 8 would I be ranked lower than i was. i get it is based to some degree off of player matched skill, but come on 6 wins 2 losses and i go down? and the last game we won, we only had 3 players and they had a full team! that is ridiculous! these kind of gaming session happen all the time. it just doesn't make sense. if you win more than you lose you should go up not down.
"If your CSR is below your MMR, you will play teams above your CSR and, if you win, your CSR goes up quickly because you are beating teams better than your CSR.

If your CSR is above your MMR, you will play teams worse than your CSR, and if you beat them, you won't get much because you are beating teams worse than you. If you lose to them, then of course you go down fast because you are losing to people worse than you."

This is why solo q became impossible. You loose game because you match bad teamate doing -15 against bad players even if you drop 30 kill. Why we match those players ? because there is not enough players in playlist (fair).

Worst part is they delete the only solo playlist "FFA".
Messi hihi wrote:
"If your CSR is below your MMR, you will play teams above your CSR and, if you win, your CSR goes up quickly because you are beating teams better than your CSR.

If your CSR is above your MMR, you will play teams worse than your CSR, and if you beat them, you won't get much because you are beating teams worse than you. If you lose to them, then of course you go down fast because you are losing to people worse than you."

This is why solo q became impossible. You loose game because you match bad teamate doing -15 against bad players even if you drop 30 kill. Why we match those players ? because there is not enough players in playlist (fair).

Worst part is they delete the only solo playlist "FFA".
that's great and all but if you win more than you lose, no matter what, you should go up not down. i know of no other sport, game, or business that you get penalized for winning more than you lose.
Messi hihi wrote:
"If your CSR is below your MMR, you will play teams above your CSR and, if you win, your CSR goes up quickly because you are beating teams better than your CSR.

If your CSR is above your MMR, you will play teams worse than your CSR, and if you beat them, you won't get much because you are beating teams worse than you. If you lose to them, then of course you go down fast because you are losing to people worse than you."

This is why solo q became impossible. You loose game because you match bad teamate doing -15 against bad players even if you drop 30 kill. Why we match those players ? because there is not enough players in playlist (fair).

Worst part is they delete the only solo playlist "FFA".
that's great and all but if you win more than you lose, no matter what, you should go up not down. i know of no other sport, game, or business that you get penalized for winning more than you lose.
The thing is that i win 90% time 1 csr and loose 8/9 so...
Messi hihi wrote:
"If your CSR is below your MMR, you will play teams above your CSR and, if you win, your CSR goes up quickly because you are beating teams better than your CSR.

If your CSR is above your MMR, you will play teams worse than your CSR, and if you beat them, you won't get much because you are beating teams worse than you. If you lose to them, then of course you go down fast because you are losing to people worse than you."

This is why solo q became impossible. You loose game because you match bad teamate doing -15 against bad players even if you drop 30 kill. Why we match those players ? because there is not enough players in playlist (fair).

Worst part is they delete the only solo playlist "FFA".
that's great and all but if you win more than you lose, no matter what, you should go up not down. i know of no other sport, game, or business that you get penalized for winning more than you lose.
You're in luck because that's exactly how it works. If you keep winning, you'll keep going up. The only difference is that, depending on the quality of opponent you face, you may only go up a little or you might go up a lot.
Messi hihi wrote:
"If your CSR is below your MMR, you will play teams above your CSR and, if you win, your CSR goes up quickly because you are beating teams better than your CSR.

If your CSR is above your MMR, you will play teams worse than your CSR, and if you beat them, you won't get much because you are beating teams worse than you. If you lose to them, then of course you go down fast because you are losing to people worse than you."

This is why solo q became impossible. You loose game because you match bad teamate doing -15 against bad players even if you drop 30 kill. Why we match those players ? because there is not enough players in playlist (fair).

Worst part is they delete the only solo playlist "FFA".
that's great and all but if you win more than you lose, no matter what, you should go up not down. i know of no other sport, game, or business that you get penalized for winning more than you lose.
"you get penalized for winning more than you lose."...I understand what you're trying to say here, but it's not just about winning or losing.

In pretty much all sports, you actually have a proper schedule where you play each team a specific amount of times. It's a lot more fair. In H5 (and most video games) your opponents are not scheduled and controlled - the matches are generated based on the best available matchup.
Unfortunately, the best available matchup might continuously match you against significantly weaker opponents. Since that is the case, when you win against them, it stands to reason you should not be greatly rewarded for it, only marginally rewarded. And that's exactly what the current system does.

If the Patriots/Packers kept playing games against the Lions/Redskins (and similar teams) and kept beating them, does that make them superbowl champs? No. They should be rewarded, yes, but not with anything significant.
can you please explain this to me?
Played 8 games o slayer yesterday. Started at a Diamond 1.5, won 6 of those 8 games and ended up a Diamond 1.2
why if i only lost 2 out of 8 would I be ranked lower than i was. i get it is based to some degree off of player matched skill, but come on 6 wins 2 losses and i go down? and the last game we won, we only had 3 players and they had a full team! that is ridiculous! these kind of gaming session happen all the time. it just doesn't make sense. if you win more than you lose you should go up not down.
It has to depend on how good your opponents are and how good your teammates are.

Winning 6 of 8 if you are expected to win 8 of 8 means you should go down, not up.
AshamanND wrote:
Messi hihi wrote:
"If your CSR is below your MMR, you will play teams above your CSR and, if you win, your CSR goes up quickly because you are beating teams better than your CSR.

If your CSR is above your MMR, you will play teams worse than your CSR, and if you beat them, you won't get much because you are beating teams worse than you. If you lose to them, then of course you go down fast because you are losing to people worse than you."

This is why solo q became impossible. You loose game because you match bad teamate doing -15 against bad players even if you drop 30 kill. Why we match those players ? because there is not enough players in playlist (fair).

Worst part is they delete the only solo playlist "FFA".
that's great and all but if you win more than you lose, no matter what, you should go up not down. i know of no other sport, game, or business that you get penalized for winning more than you lose.
"you get penalized for winning more than you lose."...I understand what you're trying to say here, but it's not just about winning or losing.

In pretty much all sports, you actually have a proper schedule where you play each team a specific amount of times. It's a lot more fair. In H5 (and most video games) your opponents are not scheduled and controlled - the matches are generated based on the best available matchup.
Unfortunately, the best available matchup might continuously match you against significantly weaker opponents. Since that is the case, when you win against them, it stands to reason you should not be greatly rewarded for it, only marginally rewarded. And that's exactly what the current system does.

If the Patriots/Packers kept playing games against the Lions/Redskins (and similar teams) and kept beating them, does that make them superbowl champs? No. They should be rewarded, yes, but not with anything significant.
In the end, he's still getting penalized for winning more than he loses...

He doesn't control who he's playing against. If there was an option to have that his matchmaking filter is pairing him with only people of his rank/MMR or higher, then he wouldn't be expressing his concerns on the system. That's the problem that majority of halo players face. How is it fair that a person plays for hours, and constantly gets paired up with people that doesn't allow him to effectively rank up, just to lose one of those games and lose all of his progress? It's not his fault the population is low and it's not his fault he's matching up against those people. That's why the system isn't good for matchmaking. Because if the population dips, the majority of people that play are affected. That's why wins and losses should be the only thing that matters in relation to ranking up. You win? You go up a fair amount. You lose? You go down a fair amount. A lot of people don't play this game anymore, and majority of the people that do have to deal with this.
Kashyy wrote:
AshamanND wrote:
Messi hihi wrote:
"If your CSR is below your MMR, you will play teams above your CSR and, if you win, your CSR goes up quickly because you are beating teams better than your CSR.

If your CSR is above your MMR, you will play teams worse than your CSR, and if you beat them, you won't get much because you are beating teams worse than you. If you lose to them, then of course you go down fast because you are losing to people worse than you."

This is why solo q became impossible. You loose game because you match bad teamate doing -15 against bad players even if you drop 30 kill. Why we match those players ? because there is not enough players in playlist (fair).

Worst part is they delete the only solo playlist "FFA".
that's great and all but if you win more than you lose, no matter what, you should go up not down. i know of no other sport, game, or business that you get penalized for winning more than you lose.
"you get penalized for winning more than you lose."...I understand what you're trying to say here, but it's not just about winning or losing.

In pretty much all sports, you actually have a proper schedule where you play each team a specific amount of times. It's a lot more fair. In H5 (and most video games) your opponents are not scheduled and controlled - the matches are generated based on the best available matchup.
Unfortunately, the best available matchup might continuously match you against significantly weaker opponents. Since that is the case, when you win against them, it stands to reason you should not be greatly rewarded for it, only marginally rewarded. And that's exactly what the current system does.

If the Patriots/Packers kept playing games against the Lions/Redskins (and similar teams) and kept beating them, does that make them superbowl champs? No. They should be rewarded, yes, but not with anything significant.
In the end, he's still getting penalized for winning more than he loses...

He doesn't control who he's playing against. If there was an option to have that his matchmaking filter is pairing him with only people of his rank/MMR or higher, then he wouldn't be expressing his concerns on the system. That's the problem that majority of halo players face. How is it fair that a person plays for hours, and constantly gets paired up with people that doesn't allow him to effectively rank up, just to lose one of those games and lose all of his progress? It's not his fault the population is low and it's not his fault he's matching up against those people. That's why the system isn't good for matchmaking. Because if the population dips, the majority of people that play are affected. That's why wins and losses should be the only thing that matters in relation to ranking up. You win? You go up a fair amount. You lose? You go down a fair amount. A lot of people don't play this game anymore, and majority of the people that do have to deal with this.
I've tested that approach and it's completely broken. Results in everyone just going up if they play more than others and have even just a 51% win %. So you have noobs passing pros as long as they play more, it's busted. Makes all of the ranks meaningless and everyone stops caring about them and trusting them.

Also, that system results in horrible accuracy, and does terrible at predicting who will win each match, which means the skill ranks it assigns are wrong and those players are in the wrong place.

The only reason you care about going up in ranks at all is because you know those ranks above you are better than you are. If we gave out points just for winning, that would no longer be the case and they wouldn't matter and it ruins the whole system. You wouldn't care about going up anymore since it wouldn't matter what your rank was. No one would trust it as a sign of skill.

The integrity and accuracy of the ranks are what make them worth going for.

You have to account for schedule difficulty. If you don't, then the ranks stop mattering.

You are wrong though about this happening to most players. This happens to a very tiny fraction of players because most are still matchmaking fairly most of the time.

We also aren't "penalizing" anyone. If you go down more than up over your matches, then you are playing worse than you should be, and deserve to go down, regardless of your win record. Otherwise, again, the ranks would be inaccurate and meaningless.

And the player in question DOES have some control over their matchmaking. They spend most of their time in parties. The larger the party, the less flexibility the system has for team balancing the match into a 50/50 situation. In fact, in the Slayer playlist this player is talking about, their solo queue matches have been 50/50.
So given how the system in H5 can predict accurately the winner of matches with high accuracy if in a tournament you knew everyone's CSR/MMR could you than accurately predict the winner of that tournament 9 out of 10 times let's say? (might be a dumb question but it just popped in my head and was bothering me)
SG FeLLoN wrote:
So given how the system in H5 can predict accurately the winner of matches with high accuracy if in a tournament you knew everyone's CSR/MMR could you than accurately predict the winner of that tournament 9 out of 10 times let's say? (might be a dumb question but it just popped in my head and was bothering me)
Maybe. Top pros do tend to have the highest MMRs as well.

But the ranks are based on how well people play in matchmaking, and the way they play. If there's a gap between what happens in matchmaking and what happens in tournaments, then the predictions would be off.

Also, if the pros are close enough in skill, it would say something like "Team A will win 65% of the time" so it wouldn't be a 9 of 10 thing. The best it could do if the matchups are close would be closer to random guessing.

Ironically, knockout tournaments are interesting because all else being equal, and under the average situations, the best team can have as low as a 40% chance of winning. Unless they are just way better.

That also goes up the longer the series are (best of 3 vs. 5 vs. 7 etc). More games, less random.
I think it'd be really cool if you released the system's predictions of each match in the major events, based on the players' matchmaking MMR.

It would be kind like how EA sports does their predictions of who will win the Stanley Cup sorta thing.
  • wow! I did not remember that there were so many additions in the update, or did you put all the ones you did not see? Now I got confused again, well the same thing happened to me recently when I went back in after a while and saw the amount of new things;) I felt so proud that I almost cried hahahaha
  1. 1
  2. ...
  3. 7
  4. 8
  5. 9
  6. 10
  7. 11
  8. ...
  9. 39