The way players recieve or lose rating points makes them want to play less.

[lightstalker.1498]

The match could be any combination really but for a visual

 

1700	1300	1500	1600	1400
1600	1600	1700	1300	1300

 And if you do the math, each horizontal line is going to average out at 1500, but win or lose, the rewards won't be the same for each player on each team, and I do not believe this is fair, equitable, or "right" considering the state of the game. I believe it actually causes players to play less. It may be a hidden culprit to population decline, disregarding the lack of content updates.

 

I won't say I understand the reasoning why you get an  greater |absolute| amount of points for losses than wins. Because it is apparent that with an even positive or negative reward system, if you lose more, you will go down, and if you win more, you will go up, and if you win and lose the same, you will stay where you originally placed (50% WR). So, I will just say that I can "accept it".  That aside, it doesn't make sense for players on the same team to receive or lose different points from one another. Perhaps the match would award a winner with 12, and the same match would take away 14 for a loser. This package should be the same for everyone on the same team. Not have one player who might lose 8 and gain 18, while another will lose 20 and gain 3.

 

Everyone was subjected to the same terms and conditions in the match. Now relative to their rating, they had to either carry hard as (expletive), or they were pitted against a guild lord of a player. Both are struggles on their own, and should be awarded equally, and not with a skewed "equity". Because of this equitable point system, that doesn't make it fair for everyone, players are inclined and persuaded to play less, and it is apparent when most of the top top players are 120-ish games played a season, with the few SPVP addicts that sit close to 500 ensuring we have those outliers.

 

I suggest keeping the cost for a loss, but change it so everyone on the same team receives the same rating reward or loss. It is not like players can form a full team, and decide who their team mates are, in ranked. If one player is getting +12 or -14, then that should be the reward package for all players on the same team.

 

 

MY MAIN POINT IS TO ONLY REWARD THE POINTS BASED OFF THE TEAM AVERAGE VERSUS THE ENEMY TEAM AVERAGE, AND NOT BASED ON INDIVIDUAL PLAYER AVERAGE VERSUS ENEMY TEAM AVERAGE.  (this is not even remotely fair, because you have to consider the effort as a whole, and when considering the player as part of that whole, and not as an individual when rating)

 

final note: copypasta from excel is weird

Edited December 2, 2021 by lightstalker.1498

[Multicolorhipster.9751]

You right.

It's all RNG, it's annoying. I don't think basing it off teams would be much better though.

I mean; if the teams were still imbalanced, wouldn't you get or lose pretty much the same as you always would? And it would still be completely outside of anyone's control.

 

I'd rather see something that isn't RNG going into it. AKA an actual measurement of player skill. AAKA how ranked works in any other game.

 

Just throw in end of match stats and also win/loss streaks into the mix so that the person currently sitting in our spawn saying gg after the first midfight doesn't lose less points than the rest of us.

[Genesis.5169]

I side with this, for another reason which is I'd like to queue with people much lower rank then me and not lose 26 points and only gain 5. Queue is built to force you to play with your own kind if not you have to win 5 times for every lose you take which if you care about ranked you will either just refuse to queue, deal with it and carry the team (not all classes to can do this i don't care what rank you are).

I recommend if there's a massive rating mismatch for one or 2 players on the team have them lose 1 rating and gain 1 or lose1 rating and gain nothing but this 26 point nuke you guys do legit made me not want to queue for a couple of days i took a 100 rating loss in 4 games as much as i like pvp i do not like grinding through g2 and the hit or miss kitten show that bracket is.

Sidenode : Alot of players complain about plat players and high ranking players ruining there game's im pretty sure this is exactly why good players are punished so hard in queue. I'd be perfectly happy to see rating mismatch to high to gain points and not counting those ranked games to the leader board too btw.

Edited December 3, 2021 by Genesis.5169

[Bazsi.2734]

10 hours ago, lightstalker.1498 said:

I suggest keeping the cost for a loss, but change it so everyone on the same team receives the same rating reward or loss. It is not like players can form a full team, and decide who their team mates are, in ranked. If one player is getting +12 or -14, then that should be the reward package for all players on the same team.

 

 

MY MAIN POINT IS TO ONLY REWARD THE POINTS BASED OFF THE TEAM AVERAGE VERSUS THE ENEMY TEAM AVERAGE, AND NOT BASED ON INDIVIDUAL PLAYER AVERAGE VERSUS ENEMY TEAM AVERAGE.  (this is not even remotely fair, because you have to consider the effort as a whole, and when considering the player as part of that whole, and not as an individual when rating)

This is a bad idea. The team averages are usually really close to each other, so this would just standardise rating gain/loss to 13 points.

In the current system your winrate will not let you climb above a certain point, if you let players climb infinitely with positive winrates rating would lose its meaning.

Also matchmaking would need to be adjusted(basicly remade from the ground up) to account for this. Currently the players are kept in close proximity to each other, so rating is meaningful for the matchmaker. If you play 120 matches with a 90% winrate with standard +13/-13, you end up as 2390. But you can arrive at the very same rating if you played 400 games with a 60% winrate. So as you can see, rating no longer indicates skill, there is a weak correlation at best.  So what do you base matchmaking on at this point? If you treat these 2 players as equals, you're setting up a curbstomp.

The current system is designed to make you settle into a rating after a while. Doesn't matter how much the 60% winrate guy grinds, he'll never ever get close to the 90% winrate veteran. There are so many things needing a change/fix in sPvP imo, but the rating and matchmaking system is not one of them.

And if seeing those +5-s and -20-s is really that annoying that the players leave solely because of that, the whole system needs a redesign from the ground up, you cannot work infinite climb into the current one.

Edited December 3, 2021 by Bazsi.2734

[Morwath.9817]

3 hours ago, Bazsi.2734 said:

This is a bad idea. The team averages are usually really close to each other, so this would just standardise rating gain/loss to 13 points.

In the current system your winrate will not let you climb above a certain point, if you let players climb infinitely with positive winrates rating would lose its meaning.

Also matchmaking would need to be adjusted(basicly remade from the ground up) to account for this. Currently the players are kept in close proximity to each other, so rating is meaningful for the matchmaker. If you play 120 matches with a 90% winrate with standard +13/-13, you end up as 2390. But you can arrive at the very same rating if you played 400 games with a 60% winrate. So as you can see, rating no longer indicates skill, there is a weak correlation at best.  So what do you base matchmaking on at this point? If you treat these 2 players as equals, you're setting up a curbstomp.

The current system is designed to make you settle into a rating after a while. Doesn't matter how much the 60% winrate guy grinds, he'll never ever get close to the 90% winrate veteran. There are so many things needing a change/fix in sPvP imo, but the rating and matchmaking system is not one of them.

And if seeing those +5-s and -20-s is really that annoying that the players leave solely because of that, the whole system needs a redesign from the ground up, you cannot work infinite climb into the current one.

I disagree and my main argument will be: the best players are playing the most, as they tend to be the most dedicated.

I would go even further and give players extra +1 points for each "top of something in team". Sooo, if you're top damage, kill, offfense, defense, revives and healing on your team - that would be +6. So instead -13 you lose only 7. 

Probably I'd also reduce gains and loses based on how close were games, because cmon, why would you lose so many points on game that was 498-500... Even if you would know it's 100% lost as people on your team are either clueless, afk or just trolling, you would have more motivation to keep playing.

So you would have more factors like : a) average rating between teams, b) personal performance, c) final score. This way we could see more accurate reflection of player skill (as performance) than just win/lose.

Edit: Now, even if leaderboard would feel grindy with those factors you've two groups of players. Players who are good and have high impact, and despite "average rating between teams" being same in their matches, they win more per game as: 13 (base value on equal rating) + 1-6 (number of top stats) + extra points for stomping. Meanwhile weaker player who has lesser impact would get same 13 (base value) + problably 0-1 (bad players don't top stats usualy) and + probably barely anything (as they don't stomp). As well better players would lose less for loss than worse players for same reasons (e.g. they are more likely more often stomped 500-50 than better players).

It could lead to situations where much better players get like +20 per win and -5 per lose, while bad players would probably be around +13 per win and -10-12 per lose. Ofc, assuming matchmaker could build "equal teams based on average MMR".

Edit2: I'd love to see on leaderboard average match score.

Edited December 3, 2021 by Morwath.9817

[Arheundel.6451]

12 hours ago, Multicolorhipster.9751 said:

You right.

It's all RNG, it's annoying. I don't think basing it off teams would be much better though.

I mean; if the teams were still imbalanced, wouldn't you get or lose pretty much the same as you always would? And it would still be completely outside of anyone's control.

 

I'd rather see something that isn't RNG going into it. AKA an actual measurement of player skill. AAKA how ranked works in any other game.

 

Just throw in end of match stats and also win/loss streaks into the mix so that the person currently sitting in our spawn saying gg after the first midfight doesn't lose less points than the rest of us.

Players have been saying this for years now and it' s the main reason why PvP was abandoded by the vast majority

[Genesis.5169]

3 hours ago, Arheundel.6451 said:

Players have been saying this for years now and it' s the main reason why PvP was abandoded by the vast majority

No players been complaining about getting stomped by higher ranking players for years hence the set up matchmaking to make sure if your stomping people you get punished for being in the wrong queue and you log off or do unranked. Many of you are misinterpreting why this is how it is. It not to make it easy to climb its to make it easy to fall so there is no stratification and/or the best player only play with the best players.

Basically they are ring fencing the top 2% of the game to only play with the top 2% or suffer. This game has never ever taking into consideration the top players in any game mode and in all honesty i think anet staff spites them more times then not with every release.

Edited December 3, 2021 by Genesis.5169

[Lacdanon.1483]

29 minutes ago, Genesis.5169 said:

No players been complaining about getting stomped by higher ranking players for years hence the set up matchmaking to make sure if your stomping people you get punished for being in the wrong queue and you log off or do unranked. Many of you are misinterpreting why this is how it is. It not to make it easy to climb its to make it easy to fall so there is no stratification and/or the best player only play with the best players.

Basically they are ring fencing the top 2% of the game to only play with the top 2% or suffer. This game has never ever taking into consideration the top players in any game mode and in all honesty i think anet staff spites them more times then not with every release.

This doesnt seem to work the way you have described at least not in my case, i constantly get matched against top 20 as g3 and if i win i get 4-10 points and if i lose i get negative 18-23

[Genesis.5169]

12 minutes ago, Lacdanon.1483 said:

This doesnt seem to work the way you have described at least not in my case, i constantly get matched against top 20 as g3 and if i win i get 4-10 points and if i lose i get negative 18-23

Thats untrue. 😕
Only time i had a negative gain on win/lose was when i was playing with silvers in g3 and plat when i was there and i played vs other top 10 players i only lost 3 to 5 rating why lie about this?

Edited December 3, 2021 by Genesis.5169
Absolutely no idea why this guy is making kitten up seriously wtf?

[Lacdanon.1483]

16 minutes ago, Genesis.5169 said:

Thats untrue. 😕
Only time i had a negative gain on win/lose was when i was playing with silvers in g3 and plat when i was there and i played vs other top 10 players i only lost 3 to 5 rating why lie about this?

It literally happen 30 mins ago, I even posted a screenshot of the team I played against, like you said why would I make this up?

You shouldnt be so quick to dismiss others players experiences because they don't match up with your own it's very close minded.

[Multicolorhipster.9751]

8 hours ago, Genesis.5169 said:

No players been complaining about getting stomped by higher ranking players for years hence the set up matchmaking to make sure if your stomping people you get punished for being in the wrong queue and you log off or do unranked. Many of you are misinterpreting why this is how it is. It not to make it easy to climb its to make it easy to fall so there is no stratification and/or the best player only play with the best players.

Basically they are ring fencing the top 2% of the game to only play with the top 2% or suffer. This game has never ever taking into consideration the top players in any game mode and in all honesty i think anet staff spites them more times then not with every release.

Having your actual skill and performance impact your rating gain/loss would be a huge benefit to top players who can consistently do well in games. To everyone really, so long as they try and do well for their rank.

 

I have to disagree too because Arenanet has always spoiled the top % of PvP players. They gave them back DuoQ when they cried about it, it took them YEARS to take any long-term action against their match manipulation that they'd been caught multiple times for(and they still didn't even ban a good chunk of them), and they even brought a former top player in to be solely responsible for PvP/PvP balance. 

 

I think you're mistaking Arenanet for the playerbase here, because a lot of the playerbase dislikes the top 1% because a small % of that 1% are arrogant egomaniacs willing to do everything and anything besides play like everyone else for virtual clout.

[lightstalker.1498]

I don't know what to do, but the current system isn't right either.

Edited December 4, 2021 by lightstalker.1498
I'm an idiot.

[JusticeRetroHunter.7684]

The problem is that this is fundamentally how Elo works. Studying the algorithm for a bit you realize that it doesn’t work without the mechanism of forced wins or losses. (Gaining more for wins where you are fated to lose or vice versa where you lose more if you are fated to win)

 

 this harkens back to my discontent with statistical algorithms like Elo…modified Glicko, and so on, which you can find on the thread “Why is ranked Pvp so bad.”

 

The reason the system hasn’t changed this way of rewarding points (or much at all really) is because it can not be changed…not in a way that makes any real sense. 
 

It’s a hard problem for sure and my personal belief is like you point out in your post, a move away from statistical algorithms into systems that reward some kind of merits based on a large number of parameters. Fully relying on Elo/Glicko  to delineate a rating leads to statistical oddities (Elo Hell).

Edited December 4, 2021 by JusticeRetroHunter.7684

[Morwath.9817]

52 minutes ago, JusticeRetroHunter.7684 said:

The problem is that this is fundamentally how Elo works. Studying the algorithm for a bit you realize that it doesn’t work without the mechanism of forced wins or losses. (Gaining more for wins where you are fated to lose or vice versa where you lose more if you are fated to win)

 

 this harkens back to my discontent with statistical algorithms like Elo…modified Glicko, and so on, which you can find on the thread “Why is ranked Pvp so bad.”

 

The reason the system hasn’t changed this way of rewarding points (or much at all really) is because it can not be changed…not in a way that makes any real sense. 
 

It’s a hard problem for sure and my personal belief is like you point out in your post, a move away from statistical algorithms into systems that reward some kind of merits based on a large number of parameters. Fully relying on Elo/Glicko  to delineate a rating leads to statistical oddities (Elo Hell).

 

ELO is perfect for 1v1 games, where outcome is 100% based on player performance without extra factors like "will someone on your team be much weaker than enemies on average" or rather extreme "will someone on your team afk/DC" and so on.

 

This is why I believe looking at personal performance and performance difference between teams (points vs just win/lose) would give us much more accurate MMR leading to much better matchmaking itself.

[Vancho.8750]

4 minutes ago, JusticeRetroHunter.7684 said:

The problem is that this is fundamentally how Elo works. Studying the algorithm for a bit you realize that it doesn’t work without the mechanism of forced wins or losses.

 

again this harkens back to my discontent with statistical algorithms like Elo…modified Glicko, and so on, which you can find on the thread “Why is ranked Pvp so bad.”

 

The reason the system hasn’t changed this way of rewarding points is because it can not be changed…not in a way that makes any real sense. 

Statistically the Elo system works in the game, the issue is that is that it doesn't factor in abnormalities and with lower population abnormalities become the status quo, also win/loss based system based only on that is pretty kitten stupid if you ask me when you are 1/10 of the game. The matchmaker also has added variables that also throw off the system even more in low pop environment, the you have played more games so you have added MMR thing. Another throw off on the win/loss thing is that is not based on Class/Elites so rerolling is pain in the kitten. There is also the duo queue that throws of the balance of the system, making duo queuing for lower rated players a pain and making the game harder for them while increasing the win rate of high rate players. The problem I find with statistics is that on paper everything can look good/bad but the reality to be quite different since the statistic is missing a variable. 

There is allot of guff added to the system that is supposed to help but in reality it does the opposite and other options that would help are not used at the pretext that the system is kept clean even though it isn't.

I solo queue, and from my experience playing the absolute minimum per day helps you climb and queuing and grinding games one after another just drops you in unfair matches ( putting you against top end duo 3 times in a row in the same team that blew out 2 games in a row, does not promote playing the game much).

The rating system also feel really off since for some reason it expects you to have way higher win loss ration to keep on the same level since you lose 20 and gain 9 or something like that, just because you queued with random people that are lower rating, the whole thing is kind kitten since by default Elo is dragging you to 50% anyway, so the system is pulling you to G2, that is when people start gaming the system and not the game to keep higher elo. Now that I mentioned Gold this one is ridiculously volatile you have Really really experienced players that are there cause the system decided to chuck them in there and total newbs which for some reason are considered equal on skill, there is also the total newbs that get validated on high elo just out of luck. The system seems that expects you to solo carry games when you have higher elo, even though the game is not a solo game and it is actively punishing you for playing.
I hope the pop will bounce back when EoD launches, since when the Streamers were promoting the game few months back the game had an uptick of players and the Plat region filled up and the games were not that volatile, but the issue will be to keep those players after.

[JusticeRetroHunter.7684]

30 minutes ago, Morwath.9817 said:

 

ELO is perfect for 1v1 games, where outcome is 100% based on player performance without extra factors like "will someone on your team be much weaker than enemies on average" or rather extreme "will someone on your team afk/DC" and so on.

 

This is why I believe looking at personal performance and performance difference between teams (points vs just win/lose) would give us much more accurate MMR leading to much better matchmaking itself.

 

Yes, I agree.

 

30 minutes ago, Vancho.8750 said:

Statistically the Elo system works in the game, the issue is that is that it doesn't factor in abnormalities and with lower population abnormalities become the status quo, also win/loss based system based only on that is pretty kitten stupid if you ask me when you are 1/10 of the game. The matchmaker also has added variables that also throw off the system even more in low pop environment, the you have played more games so you have added MMR thing. Another throw off on the win/loss thing is that is not based on Class/Elites so rerolling is pain in the kitten. There is also the duo queue that throws of the balance of the system, making duo queuing for lower rated players a pain and making the game harder for them while increasing the win rate of high rate players. The problem I find with statistics is that on paper everything can look good/bad but the reality to be quite different since the statistic is missing a variable. 

There is allot of guff added to the system that is supposed to help but in reality it does the opposite and other options that would help are not used at the pretext that the system is kept clean even though it isn't.

I solo queue, and from my experience playing the absolute minimum per day helps you climb and queuing and grinding games one after another just drops you in unfair matches ( putting you against top end duo 3 times in a row in the same team that blew out 2 games in a row, does not promote playing the game much).

The rating system also feel really off since for some reason it expects you to have way higher win loss ration to keep on the same level since you lose 20 and gain 9 or something like that, just because you queued with random people that are lower rating, the whole thing is kind kitten since by default Elo is dragging you to 50% anyway, so the system is pulling you to G2, that is when people start gaming the system and not the game to keep higher elo. Now that I mentioned Gold this one is ridiculously volatile you have Really really experienced players that are there cause the system decided to chuck them in there and total newbs which for some reason are considered equal on skill, there is also the total newbs that get validated on high elo just out of luck. The system seems that expects you to solo carry games when you have higher elo, even though the game is not a solo game and it is actively punishing you for playing.
I hope the pop will bounce back when EoD launches, since when the Streamers were promoting the game few months back the game had an uptick of players and the Plat region filled up and the games were not that volatile, but the issue will be to keep those players after.

 

Yes i also agree. However, technically the algorithm works the same way, no matter what the population is...it's just that larger population makes the negative behavior of the algo negligible.. and in concurrence those negative effects are amplified the lower the population. In my research of the algo, this fault is simply a fault with what these algo's are really doing when they are calculating rating for people, especially in 5v5 rando que environments. 

 

In addition having more population is not a real solution to this problem...because getting higher population is not a solvable thing really. If people were robots and anet could program them to play guild wars 2 spvp, then it would be a solution, but folks aren't robots and we can't make them do anything. There's also an uncertainty in variety of actions anet can take...since there's no bonifide...certified way to increase population...maybe there is... but it comes from a social engineering perspective rather then an implementable, empirical one.

 

I also believe heavily...not 100% sure on...but pretty confident in, that the issues with the algorithm come from the fact that it's a 5v5 multiplayer environment where the team decides on an individuals rating...and that this is something that makes the issue pervasive and persistent no matter what the population is.

Edited December 4, 2021 by JusticeRetroHunter.7684

[Vancho.8750]

1 minute ago, JusticeRetroHunter.7684 said:

In addition having more population is not a real solution to this problem...because getting higher population is not a solvable thing really. If people were robots and anet could program them to play guild wars 2 spvp, then it would be a solution, but folks aren't robots and we can't make them do anything. There's also an uncertainty in variety of actions anet can take...since there's no bonifide...certified way to increase population...maybe there is... but it comes from a social engineering perspective rather then an implementable, empirical one.

Basically what Riot games is doing with League of Legends, in game social engineering and promotional social engineering.
Basically we can reach the conclusion that PVP sucks cause Arenanet suck at marketing their game as a whole, not being able to sell the game spoils the whole meal.

[JusticeRetroHunter.7684]

28 minutes ago, Vancho.8750 said:

Basically what Riot games is doing with League of Legends, in game social engineering and promotional social engineering.
Basically we can reach the conclusion that PVP sucks cause Arenanet suck at marketing their game as a whole, not being able to sell the game spoils the whole meal.

 

Ya pretty much. Riot is a great case study for how important good marketing is, not to mention Riot also is now open to multiple revenue streams now rather then just games...they basically set up their own path to making movies, and a cinematic universe too.

 

Here's the other thing...I think to do anything close to what Riot can do, requires a complete paradigm shift in thought...a new level of intellect that I don't think Anet has right now. Not insulting them by any means tho...what i'm saying is that there's different kinds of "smart." there's the mathy-smart guys sure...but then there is also those that are conceptually...creatively smart. People that think out of the box and can see that bigger picture. 

 

Riot indeed sees the big picture...they see where entertainment is headed into the future, and they got the right people on it, no funny business. Can we say the same thing for Anet? I think they've improved some of their decisions over time... but i don't think anyone could forget the very cringe live action commercials they had. That cringe usually comes from the lack of just being able to see a bigger picture...no conceptualizing what the final product is gonna look like (and how audiences would react to it) and really felt like at most...a bunch of out of touch market department that think they know "what's hip."

Edited December 4, 2021 by JusticeRetroHunter.7684

[Reikou.7068]

On 12/3/2021 at 5:01 AM, lightstalker.1498 said:

The match could be any combination really but for a visual

 

1700	1300	1500	1600	1400
1600	1600	1700	1300	1300

 And if you do the math, each horizontal line is going to average out at 1500, but win or lose, the rewards won't be the same for each player on each team, and I do not believe this is fair, equitable, or "right" considering the state of the game. I believe it actually causes players to play less. It may be a hidden culprit to population decline, disregarding the lack of content updates.

 

I won't say I understand the reasoning why you get an  greater |absolute| amount of points for losses than wins. Because it is apparent that with an even positive or negative reward system, if you lose more, you will go down, and if you win more, you will go up, and if you win and lose the same, you will stay where you originally placed (50% WR). So, I will just say that I can "accept it".  That aside, it doesn't make sense for players on the same team to receive or lose different points from one another. Perhaps the match would award a winner with 12, and the same match would take away 14 for a loser. This package should be the same for everyone on the same team. Not have one player who might lose 8 and gain 18, while another will lose 20 and gain 3.

 

Everyone was subjected to the same terms and conditions in the match. Now relative to their rating, they had to either carry hard as (expletive), or they were pitted against a guild lord of a player. Both are struggles on their own, and should be awarded equally, and not with a skewed "equity". Because of this equitable point system, that doesn't make it fair for everyone, players are inclined and persuaded to play less, and it is apparent when most of the top top players are 120-ish games played a season, with the few SPVP addicts that sit close to 500 ensuring we have those outliers.

 

I suggest keeping the cost for a loss, but change it so everyone on the same team receives the same rating reward or loss. It is not like players can form a full team, and decide who their team mates are, in ranked. If one player is getting +12 or -14, then that should be the reward package for all players on the same team.

 

 

MY MAIN POINT IS TO ONLY REWARD THE POINTS BASED OFF THE TEAM AVERAGE VERSUS THE ENEMY TEAM AVERAGE, AND NOT BASED ON INDIVIDUAL PLAYER AVERAGE VERSUS ENEMY TEAM AVERAGE.  (this is not even remotely fair, because you have to consider the effort as a whole, and when considering the player as part of that whole, and not as an individual when rating)

 

final note: copypasta from excel is weird

Been saying this exact thing for a while now.

I do not understand why rank point gains and losses are given based on individual rating, not team average.  It just seems like very strange logic.  

Edited December 4, 2021 by Reikou.7068

[Bazsi.2734]

On 12/3/2021 at 10:13 AM, Morwath.9817 said:

I would go even further and give players extra +1 points for each "top of something in team". Sooo, if you're top damage, kill, offfense, defense, revives and healing on your team - that would be +6. So instead -13 you lose only 7. 

Probably I'd also reduce gains and loses based on how close were games, because cmon, why would you lose so many points on game that was 498-500... Even if you would know it's 100% lost as people on your team are either clueless, afk or just trolling, you would have more motivation to keep playing.

So you would have more factors like : a) average rating between teams, b) personal performance, c) final score. This way we could see more accurate reflection of player skill (as performance) than just win/lose.

Edit: Now, even if leaderboard would feel grindy with those factors you've two groups of players. Players who are good and have high impact, and despite "average rating between teams" being same in their matches, they win more per game as: 13 (base value on equal rating) + 1-6 (number of top stats) + extra points for stomping. Meanwhile weaker player who has lesser impact would get same 13 (base value) + problably 0-1 (bad players don't top stats usualy) and + probably barely anything (as they don't stomp). As well better players would lose less for loss than worse players for same reasons (e.g. they are more likely more often stomped 500-50 than better players).

It could lead to situations where much better players get like +20 per win and -5 per lose, while bad players would probably be around +13 per win and -10-12 per lose. Ofc, assuming matchmaker could build "equal teams based on average MMR".

Edit2: I'd love to see on leaderboard average match score.

Okay so adding all that fluff to scoring just makes my initial concern worse. The current matchmaking requires player ratings to be close to each other, and the scoring reflects that. If top players would get 20-s instead of +4-s for beating people 2-3 rungs below them, their rating would skyrocket to the upper limit.
Now if you do not remove this upper limit, several people will be sitting on 2100 when the season ends, so eh... who won? Or you can remove it and watch these players farm absurd ratings in time, the numbers would break matchmaking after a while. So you need to rewrite matchmaking or people above a certain rating would not get matches.
And even if you rewrote matchmaking to let these players get matches, once you're above average in terms of skill, it just takes time to grind to get the highest rating, as I explained in my previous comment.

 

On 12/3/2021 at 10:13 AM, Morwath.9817 said:

I disagree and my main argument will be: the best players are playing the most, as they tend to be the most dedicated.

The OP noted how many on the top of the ladder only have the bare minimum of required games. We disagree with the reasons why, but the observation is objectively true. From time to time some players are having fun trying to place as many accounts into the top 50 as possible, but most of them have work&families(game is almost a decade old and most top players are of the oldguard) by now, and they only cling to this game because of their former investment and enjoying being good at the game.
So no, this statement is just not true. In a populated, competitive and everchanging game it would be, but Guild Wars 2 is not that.

[lightstalker.1498]

1 hour ago, Bazsi.2734 said:

Okay so adding all that fluff to scoring just makes my initial concern worse. The current matchmaking requires player ratings to be close to each other, and the scoring reflects that. If top players would get 20-s instead of +4-s for beating people 2-3 rungs below them, their rating would skyrocket to the upper limit.
Now if you do not remove this upper limit, several people will be sitting on 2100 when the season ends, so eh... who won? Or you can remove it and watch these players farm absurd ratings in time, the numbers would break matchmaking after a while. So you need to rewrite matchmaking or people above a certain rating would not get matches.
And even if you rewrote matchmaking to let these players get matches, once you're above average in terms of skill, it just takes time to grind to get the highest rating, as I explained in my previous comment.

 

The OP noted how many on the top of the ladder only have the bare minimum of required games. We disagree with the reasons why, but the observation is objectively true. From time to time some players are having fun trying to place as many accounts into the top 50 as possible, but most of them have work&families(game is almost a decade old and most top players are of the oldguard) by now, and they only cling to this game because of their former investment and enjoying being good at the game.
So no, this statement is just not true. In a populated, competitive and everchanging game it would be, but Guild Wars 2 is not that.

 

I thought about making the ends of the pvp rating spectrum less punishing, and the middle the most punishing. To push more players to play in prime time hours, or to play more games, and less dodging. With the average being the most punishing, because that is where the bulk will fall, and where many players who manipulate the board want their games to be.

 

Looks something like:

 

Tier	1	2	3		1	2	3		1	2	3		1	2	3		1	2	3
Division	BRONZE				SILVER				GOLD				PLATINUM				LEGENDARY
GAIN	15	14	13		13	12	11		10	9	8		8	9	10		9	8	7
LOSS	13	13	13		12	12	12		15	15	15		13	13	13		12	12	12

[lightstalker.1498]

Gold will be a trap for most, almost like now. But, plat players will want to play with plat players instead of duo-ing with lower rated players for easier games. Losses will be slightly less at the lower tiers to build motivation and desire, but will be lower from silver 3 and above, to punish failure. Failing in gold will be especially punishing, and making it into platinum and higher will feel safer than playing in lower leagues , when you don't belong there.

 

The rewards would be based on your team's division/tier. So playing with a Gold 3 team would reward or punish you with those rewards.

Edited December 5, 2021 by lightstalker.1498

[Morwath.9817]

3 hours ago, Bazsi.2734 said:

Okay so adding all that fluff to scoring just makes my initial concern worse. The current matchmaking requires player ratings to be close to each other, and the scoring reflects that. If top players would get 20-s instead of +4-s for beating people 2-3 rungs below them, their rating would skyrocket to the upper limit.
Now if you do not remove this upper limit, several people will be sitting on 2100 when the season ends, so eh... who won? Or you can remove it and watch these players farm absurd ratings in time, the numbers would break matchmaking after a while. So you need to rewrite matchmaking or people above a certain rating would not get matches.
And even if you rewrote matchmaking to let these players get matches, once you're above average in terms of skill, it just takes time to grind to get the highest rating, as I explained in my previous comment.

 

I think you miss serval consequences, like if average rating at the top is much higher, building teams that have same average rating would be much harder for matchmaking (+/-13 per win +/- modifers) especially when playing duoQ, thus you still couldn't farm too far away from average rating as it would stay the same for players at the top who have 2:1 winrates and better. What would change is that both rating and matchmaking would be more accurate and losing games would feel less punishing in case of better players than worse players as their personal performance could affect how many points they lose.

 

It could break current "tiers", but they should be changed to dynamic tiers rather than set in stone tiers.

 

You ask who would win season then? Well, Boyce would, like he should 🙃

Edit: About bare minimum of games most of players in top20 on EU got like 200-300 games played (only 8 got less than 200). Anyway you want your players to play, not have bare minimum of games while duoQ during off hours...

Anyway, I don't think caring too much for top 0,5% of players fighting over 1st place should be more important than making the best experience for average player, as top 0,5% playerbase isn't making gamemode alive, but average and a bit better than average players do.

 

Edited December 5, 2021 by Morwath.9817

[Bazsi.2734]

 

13 hours ago, Morwath.9817 said:

 

I think you miss serval consequences, like if average rating at the top is much higher, building teams that have same average rating would be much harder for matchmaking (+/-13 per win +/- modifers) especially when playing duoQ...

 

I did not miss this, this is one of my main concerns, also one of my major points against your suggestion. You keep getting +20 wins, so you end up several hundreds of ratings above the average (gold/low plat) players. This causes extremely long queue times for whoever is on top(or on the bottom if you're bad). As I said, matchmaking couldn't handle people being far apart in ratings.

 

13 hours ago, Morwath.9817 said:

...thus you still couldn't farm too far away from average rating as it would stay the same for players at the top who have 2:1 winrates and better. 

Ah yes, getting +20 wins, -6 losses, while maintaining a very strong winrate... it would be exactly the same as it is now, with +4 wins and -20 losses. How does this MATH?
 

13 hours ago, Morwath.9817 said:

What would change is that both rating and matchmaking would be more accurate and losing games would feel less punishing in case of better players than worse players as their personal performance could affect how many points they lose.

I explained why rating wouldn't be accurate. Even gave an example. Are you even reading my replies?
Here is my second and LAST attempt at explaining it. Currently whenever a match is played, the rating gained and lost by all participating players is basicly 0. What you lose is gained by someone else and vice versa.
You want to cushion losses and boost wins, that changes this 0 into a positive number. Meaning on the long run it's possible(and very likely) that everyone climbs. Playing a lot to compensate for a lower winrate could help surpass more skilled players in terms of rating, that's what my 120 games vs 400 games example was supposed to demonstrate. Rating at this point has more to do with how much you played, how good you are is secondary. Two pals flunking college could take turns to nonstop play on an account and end up above boyce and sindrenner at the end of the season. Even though they belong into plat 1.
 

What you said here is objectively false because MATH. If you just repeat again how your idea is all good without adressing this, I'm done with this conversation.

Also you mentioned that I "miss several consequences", which has some weird truth to it. My entry level analysis only shows how things would start spiraling out of control. As the season gets older, players rack up games at a different pace, of course there would be more and more distorting effects. Actually the more I think about it the worse your idea gets, but I don't feel like investigating why crashing schoolbus into a ravine is a bad idea, I'm just going to show how driving off the road is undesireable and not care about the rest.

 

13 hours ago, Morwath.9817 said:

Anyway, I don't think caring too much for top 0,5% of players fighting over 1st place should be more important than making the best experience for average player, as top 0,5% playerbase isn't making gamemode alive, but average and a bit better than average players do.

 

Ah yes, I'm no in the top 0.5%, who cares, let's ruin their game. Good mentality. 
When your average andy stops caring about integrity and afks a match or two in low gold, or tries a third party software just to  get banned shortly after, that hardly affects anyone.
When the top players get unhappy/salty/toxic-->stop caring, you get boosting, wintrading, selling monthly... the damage they can do the the gamemode is enormous. Do you really want to make them unhappy?

[Morwath.9817]

22 minutes ago, Bazsi.2734 said:

 

I did not miss this, this is one of my main concerns, also one of my major points against your suggestion. You keep getting +20 wins, so you end up several hundreds of ratings above the average (gold/low plat) players. This causes extremely long queue times for whoever is on top(or on the bottom if you're bad). As I said, matchmaking couldn't handle people being far apart in ratings.

 

Ah yes, getting +20 wins, -6 losses, while maintaining a very strong winrate... it would be exactly the same as it is now, with +4 wins and -20 losses. How does this MATH?
 

I explained why rating wouldn't be accurate. Even gave an example. Are you even reading my replies?
Here is my second and LAST attempt at explaining it. Currently whenever a match is played, the rating gained and lost by all participating players is basicly 0. What you lose is gained by someone else and vice versa.
You want to cushion losses and boost wins, that changes this 0 into a positive number. Meaning on the long run it's possible(and very likely) that everyone climbs. Playing a lot to compensate for a lower winrate could help surpass more skilled players in terms of rating, that's what my 120 games vs 400 games example was supposed to demonstrate. Rating at this point has more to do with how much you played, how good you are is secondary. Two pals flunking college could take turns to nonstop play on an account and end up above boyce and sindrenner at the end of the season. Even though they belong into plat 1.
 

What you said here is objectively false because MATH. If you just repeat again how your idea is all good without adressing this, I'm done with this conversation.

Also you mentioned that I "miss several consequences", which has some weird truth to it. My entry level analysis only shows how things would start spiraling out of control. As the season gets older, players rack up games at a different pace, of course there would be more and more distorting effects. Actually the more I think about it the worse your idea gets, but I don't feel like investigating why crashing schoolbus into a ravine is a bad idea, I'm just going to show how driving off the road is undesireable and not care about the rest.

 

Ah yes, I'm no in the top 0.5%, who cares, let's ruin their game. Good mentality. 
When your average andy stops caring about integrity and afks a match or two in low gold, or tries a third party software just to  get banned shortly after, that hardly affects anyone.
When the top players get unhappy/salty/toxic-->stop caring, you get boosting, wintrading, selling monthly... the damage they can do the the gamemode is enormous. Do you really want to make them unhappy?

 

You fail to see basic thing you can't fly too far away from average MMR, because if you do, the more rating you "farm" above it, the more unfair and unwinnable matches you will get: e.g. matchmaking will consider team made of you and mix of bronze-silver players is fair vs 5 gold-plat players.

 

Even if average MMR would be bigger as season is progressing, whats wrong with it? Do you think ELO isn't inflated over time in chess, or that it wouldn't if players could spam rating games like players can in GW2?

 

Do you know what you do with players who break ToS cause they got "unhappy/salty/toxic"? You ban them, at least in every other game, even if they are "espurtz" players.