[Locked] MATCHMAKING FEEDBACK UPDATE – August 31

Mr khuRAM wrote:

eLantern wrote:

Spartan21097 wrote:

Can someone here help me understand why I lost 15 CSR in this game??

Mr khuRAM wrote:

...

Here is a breakdown of Halo 5’s matchmaking system. I suggest giving the CSR and MMR sections a careful read.

OK I know all that smart -Yoink-, you try being in diamond 6 spending days with a small margin of error to lose games you literally have to win every game at diamond 6, there's a difference in reading the material and actually witnessing it first hand.

The difference is in the personal experience itself. Nobody is saying that pushing one's skill beyond where they've historically existed is easy or achieved without oscillation. The thing is that more often than not the system is working as intended and the intention is reasonable when fully comprehended. Often knowing why what's happening is helpful but, of course, not always; nevertheless, this is why I shared the breakdown. In case you didn't fully grasp all the reasoning behind the results that are frustrating you I hoped that the breakdown would bring additional awareness and a sense of acceptance to the difficult reality that you face in achieving rank progression once your CSR has been generally converged to your MMR.

Mr khuRAM wrote:

WHAT I DONT UNDERSTAND! You have a perfect template to fix the BR just use the template for the H3 throwback playlist, better yet why dont you just leave that playlist their permanently because some people are sick and tired of going into ever match with pistols and you know the population increases every time you release that playlist, If i have to play this game for another 2 years before H6 then at least make the BR equal or better than the pistol because right now its absolutely useless in multi player

Pop does not increase when we run H3 throwback. It's one of our better rotationals, but still behaves like a rotational --- it gets a decent crowd the first week, and then quickly peters off to being less popular than, e.g. BTB.

eLantern

• “If your Pre-Match CSR is less than your Pre-Match MMR: you will gain +15 CSR for a WIN or you'll lose -1 CSR for a LOSS.
• If your Pre-Match CSR is greater than your Pre-Match MMR: you will gain +1 CSR for a WIN or you'll lose -15 CSR for a LOSS.
• Comparing Pre-Match CSR to Pre-Match MMR means that an individual’s in-game performance won’t have an immediate effect on CSR. It’ll alwayslag by at least one match; in other words, an aspect of the Post-Match CSR adjustment will be reflective of a potential MMR skill change that occurred from your previous match or previous matches. This often leads to some confusion as players are unable to relate their most recent individual performance to the adjusted amount.”

Thanks for this, I was not aware there was a lag in the CSR update.

ZaedynFel wrote:

LovelyJoey21605 wrote:

QX wrote:

LovelyJoey21605 wrote:

I'm really enjoying these changes, especially with how well matchmaking is working!

6/7 games results in defeat, with my last game going 50-6 in RANKED SLAYER.

Why bother having the ranks in effect if they are not going to be used anyway?

Can we PLEASE have more maps in quick-play so I don't have to TOUCH ranked. It is infuriating to know peoples rank immediatly (and get an inkling of their MMR), and even more so because you then KNOW that the game has set you up.

Halo 5:
"Yeah, you went 50-6 in Ranked Slayer, but hey, we always knew you were going to loose that so we wont lower your CSR by much."

Am I the only one wanting more maps in quickplay, so I don't have to touch ranked for the variety?

Josh could you look at this players recent games on Monday? He has two games with very blowout scores which seems weird to get considering he is ranked platinum.

Joey they do use the ranks when matchmaking. Besides those two very lopsided games the rest of your losing streak very pretty close. Losing 6/7 just means you we're unlucky. I'm sure you've also had winning streaks too they just are not as memorable.

I will say though once one of your teammates leaves the game you can quit without an additionally penalty given to you. Basically the game will be treated as a loss for you while the first quitter will be given the quitting penalty. I recommend doing that if you lose a player and know you can't 3v4. Especially in that game 50-6 game where you were 2v4. All you are doing is wasting your time, the enemy's time, and your own sanity by playing that game to completion.

Quick play will probably never get more maps since it is meant as a landing point for new players to learn the game. That's why it's limited to slayer (the easiest gamemode to understand) and also has 3 of the most iconic H5 maps that are in pretty much every gametype. Adding more maps will only add more confusion to new players as they have to learn more maps at the same time as learning mechanics.

Seriously though Josh/ZaedynFel, I agree with QX, can you take a look at what happened with my ranked-matchmaking 10/13/2018? I'm kinda curious because that loosing streak was ridiculous, and peppered with games I should not have been put in.

I'm especially curious as to what happened with matchmaking here:

https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/cc14ee31-38da-4583-9e27-d9e4889e6701/players/lovelyjoey21605?gameHistoryMatchIndex=1&gameHistoryGameModeFilter=All

Just glossing over the CSR's of both teams I'm clearly put into a game I'm EXPECTED to loose, where the teams are undoubtedly unbalanced. Yet my drop in CSR is HUGE, indicating that my MMR might even be below Platinum 6, which further begs the question: Why was I put into that game at all?

Pre-game MMR shows 3 Diamonds, 1 Plat vs. 3 Diamonds, 1 Plat for a very small overall MMR gap.

But, one of the best Diamonds on your team played Borderline Silver that match. Maybe AFK?

Going into the match you had a 43% chance of winning (within the 60/40 we shoot for), but the system updated its estimate after that bottom player quit to 10%, meaning that bottom player played much worse than expected.

1° problem: the enemy team was a diamond team (looking the stat not mmr) so Ts2 judged one diamond without good reason platinum.
2° problem: the team of this poor guy was a noob platinum team (only the diamond was a diamond) so TS2 for a match where this 2 platinum played at diamond level considered them diamond.
This a problem you can't ignore but you did it because in last thread there was a platinum who jumped at onyx for losing 2 match against diamond

This +/- 1 or +/- 15 CSR system is brutal for enjoyment of the game. My main account (LVL 149) is "level-locked" in doubles at diamond 4/diamond 5. This past weekend I played 9 hardly budged, lost 1 game from a 3-16 teamate and lost rank to diamond 4.

I decided to start a fresh account and got to Champion 44 in about 30 games, with extremely similar stats (KDA and win %) as my main.

Now I know the reasons and logic behind this system. And it is amazing for search times currently given the halo playlist populations. But in no past Halo have I had to win 15 games just to go even rank for the 1 game I lost. It simply isn't fun. And it's heartbreaking to simply make a new account and jet rocket past my main rank within 20-30 games when I am the exact same player so my skill didn't really change, only my account.

Josh I am not trying to make another "this ranking system sucks" thread. Like I said it is amazing for search times and games. But for ranking up and down? Something needs to change there is so much discrepancy and volatility which makes a really focused 10 game win streak feel like a complete waste of time and not rewarding in the slightest.

Thank you for your time
WWorkage

Why as a Platinum player I can win 3 games in a row, with positive KDA spreads, against a team with a diamond player and gain less rank then i would lose with 1 match for a defeat under the same conditions? Why conspire against me?

QX wrote:

LovelyJoey21605 wrote:

ZaedynFel wrote:

LovelyJoey21605 wrote:

QX wrote:

LovelyJoey21605 wrote:

I'm really enjoying these changes, especially with how well matchmaking is working!

6/7 games results in defeat, with my last game going 50-6 in RANKED SLAYER.

Why bother having the ranks in effect if they are not going to be used anyway?

Can we PLEASE have more maps in quick-play so I don't have to TOUCH ranked. It is infuriating to know peoples rank immediatly (and get an inkling of their MMR), and even more so because you then KNOW that the game has set you up.

Halo 5:
"Yeah, you went 50-6 in Ranked Slayer, but hey, we always knew you were going to loose that so we wont lower your CSR by much."

Am I the only one wanting more maps in quickplay, so I don't have to touch ranked for the variety?

Josh could you look at this players recent games on Monday? He has two games with very blowout scores which seems weird to get considering he is ranked platinum.

Joey they do use the ranks when matchmaking. Besides those two very lopsided games the rest of your losing streak very pretty close. Losing 6/7 just means you we're unlucky. I'm sure you've also had winning streaks too they just are not as memorable.

I will say though once one of your teammates leaves the game you can quit without an additionally penalty given to you. Basically the game will be treated as a loss for you while the first quitter will be given the quitting penalty. I recommend doing that if you lose a player and know you can't 3v4. Especially in that game 50-6 game where you were 2v4. All you are doing is wasting your time, the enemy's time, and your own sanity by playing that game to completion.

Quick play will probably never get more maps since it is meant as a landing point for new players to learn the game. That's why it's limited to slayer (the easiest gamemode to understand) and also has 3 of the most iconic H5 maps that are in pretty much every gametype. Adding more maps will only add more confusion to new players as they have to learn more maps at the same time as learning mechanics.

Seriously though Josh/ZaedynFel, I agree with QX, can you take a look at what happened with my ranked-matchmaking 10/13/2018? I'm kinda curious because that loosing streak was ridiculous, and peppered with games I should not have been put in.

I'm especially curious as to what happened with matchmaking here:

https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/cc14ee31-38da-4583-9e27-d9e4889e6701/players/lovelyjoey21605?gameHistoryMatchIndex=1&gameHistoryGameModeFilter=All

Just glossing over the CSR's of both teams I'm clearly put into a game I'm EXPECTED to loose, where the teams are undoubtedly unbalanced. Yet my drop in CSR is HUGE, indicating that my MMR might even be below Platinum 6, which further begs the question: Why was I put into that game at all?

Pre-game MMR shows 3 Diamonds, 1 Plat vs. 3 Diamonds, 1 Plat for a very small overall MMR gap.

But, one of the best Diamonds on your team played Borderline Silver that match. Maybe AFK?

Going into the match you had a 43% chance of winning (within the 60/40 we shoot for), but the system updated its estimate after that bottom player quit to 10%, meaning that bottom player played much worse than expected.

First off, thank you for answering, I really appreciate that!!

No, I'm pretty sure there weren't any AFK that match. Can you see steps traveled by individual Spartans, or is that not logged?

What do you mean "best Diamonds on your team played Borderline Silver that match"?
Was there anyone on that loosing team actually playing above Silver in that match?

What was the post-match MMRs?

I had this game almost right before;
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/dd382dc2-8fde-436e-bde4-b51fa0e52a18/players/lovelyjoey21605?gameHistoryMatchIndex=3&gameHistoryGameModeFilter=All

Does the game not keep track of win/loss streaks?

What he is saying is that those diamond level players historically have MMRs and performed around the diamond level. However in the particular game you played they played at the equivalent level of how a silver level player would have performed if he was put in your game.

Now this could happen for a variety of reasons. The player had to go AFK for a decent amount of the match, was drunk/high, just let someone else play on their gametag such as their sibling or significant other. All of these are unpredictable to 343 so they can't know that the player who almost always plays at a diamond level will suddenly go to a silver level with no warning.

Also I'm pretty sure the game doesn't track win and loss streaks to "break" them. If you win enough you will go up in MMR and if you lose enough you will go down. That will eventually break any streaks since you will be facing harder/easier opponents.

Or because TS2 can't judge people skill. I had the same problem in the summer season where i was evaluated diamond 5 (onyx 1634) and my friends like this platinum 1 (diamond 2) platinum 5 (onyx 1550) diamond 6 (at least onyx 1550), almost all the game we won that season the enemy didn't reach 30 kill.
Then me and an other member in the last thread we talked about a player who jumped from mid plat to onyx for losing 2 games against diamond during the placement

NBK DarkWarlock wrote:

QX wrote:

LovelyJoey21605 wrote:

ZaedynFel wrote:

LovelyJoey21605 wrote:

QX wrote:

LovelyJoey21605 wrote:

I'm really enjoying these changes, especially with how well matchmaking is working!

6/7 games results in defeat, with my last game going 50-6 in RANKED SLAYER.

Why bother having the ranks in effect if they are not going to be used anyway?

Can we PLEASE have more maps in quick-play so I don't have to TOUCH ranked. It is infuriating to know peoples rank immediatly (and get an inkling of their MMR), and even more so because you then KNOW that the game has set you up.

Halo 5:
"Yeah, you went 50-6 in Ranked Slayer, but hey, we always knew you were going to loose that so we wont lower your CSR by much."

Am I the only one wanting more maps in quickplay, so I don't have to touch ranked for the variety?

Josh could you look at this players recent games on Monday? He has two games with very blowout scores which seems weird to get considering he is ranked platinum.

Joey they do use the ranks when matchmaking. Besides those two very lopsided games the rest of your losing streak very pretty close. Losing 6/7 just means you we're unlucky. I'm sure you've also had winning streaks too they just are not as memorable.

I will say though once one of your teammates leaves the game you can quit without an additionally penalty given to you. Basically the game will be treated as a loss for you while the first quitter will be given the quitting penalty. I recommend doing that if you lose a player and know you can't 3v4. Especially in that game 50-6 game where you were 2v4. All you are doing is wasting your time, the enemy's time, and your own sanity by playing that game to completion.

Quick play will probably never get more maps since it is meant as a landing point for new players to learn the game. That's why it's limited to slayer (the easiest gamemode to understand) and also has 3 of the most iconic H5 maps that are in pretty much every gametype. Adding more maps will only add more confusion to new players as they have to learn more maps at the same time as learning mechanics.

Seriously though Josh/ZaedynFel, I agree with QX, can you take a look at what happened with my ranked-matchmaking 10/13/2018? I'm kinda curious because that loosing streak was ridiculous, and peppered with games I should not have been put in.

I'm especially curious as to what happened with matchmaking here:

https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/cc14ee31-38da-4583-9e27-d9e4889e6701/players/lovelyjoey21605?gameHistoryMatchIndex=1&gameHistoryGameModeFilter=All

Just glossing over the CSR's of both teams I'm clearly put into a game I'm EXPECTED to loose, where the teams are undoubtedly unbalanced. Yet my drop in CSR is HUGE, indicating that my MMR might even be below Platinum 6, which further begs the question: Why was I put into that game at all?

Pre-game MMR shows 3 Diamonds, 1 Plat vs. 3 Diamonds, 1 Plat for a very small overall MMR gap.

But, one of the best Diamonds on your team played Borderline Silver that match. Maybe AFK?

Going into the match you had a 43% chance of winning (within the 60/40 we shoot for), but the system updated its estimate after that bottom player quit to 10%, meaning that bottom player played much worse than expected.

First off, thank you for answering, I really appreciate that!!

No, I'm pretty sure there weren't any AFK that match. Can you see steps traveled by individual Spartans, or is that not logged?

What do you mean "best Diamonds on your team played Borderline Silver that match"?
Was there anyone on that loosing team actually playing above Silver in that match?

What was the post-match MMRs?

I had this game almost right before;
https://www.halowaypoint.com/en-us/games/halo-5-guardians/xbox-one/mode/arena/matches/dd382dc2-8fde-436e-bde4-b51fa0e52a18/players/lovelyjoey21605?gameHistoryMatchIndex=3&gameHistoryGameModeFilter=All

Does the game not keep track of win/loss streaks?

What he is saying is that those diamond level players historically have MMRs and performed around the diamond level. However in the particular game you played they played at the equivalent level of how a silver level player would have performed if he was put in your game.

Now this could happen for a variety of reasons. The player had to go AFK for a decent amount of the match, was drunk/high, just let someone else play on their gametag such as their sibling or significant other. All of these are unpredictable to 343 so they can't know that the player who almost always plays at a diamond level will suddenly go to a silver level with no warning.

Also I'm pretty sure the game doesn't track win and loss streaks to "break" them. If you win enough you will go up in MMR and if you lose enough you will go down. That will eventually break any streaks since you will be facing harder/easier opponents.

Or because TS2 can't judge people skill. I had the same problem in the summer season where i was evaluated diamond 5 (onyx 1634) and my friends like this platinum 1 (diamond 2) platinum 5 (onyx 1550) diamond 6 (at least onyx 1550), almost all the game we won that season the enemy didn't reach 30 kill.
Then me and an other member in the last thread we talked about a player who jumped from mid plat to onyx for losing 2 games against diamond during the placement

I mean sure if you want to disregard data based answers that T2.0 measure 343's definition of skill for H5 because you assume that your own anecdotal evidence is superior the I'm not going to argue with you. If you actually bring hard numbers and analysis then I will take your position of T2.0 is broken seriously.

Does T2.0 represent all skill? No it doesn't simply because 343 can't estimate communication and game sense. However the current system does a logically good job or deciding which players are better. It isn't perfect and it has a variance. However in terms of OPs question of why that player was regarded as silver I think my answer is more accurate than T2.0 being broken.