87 replies to this topic

[Rkshz]

First of all, I do not want to offend anyone. Therefore, if something in the below written offend you, it is not intentional.

And so about the balance in QP: I do not know the "quick play" balance formula. On this I will reflect on the basis of my guesses.

I'm sure everyone has come across a situation where there are a lot of skillful players in one team, and there are no-names in the other - as a result, we have another stomp 0-12
You can often see in the chat how the team writes "we already dead" (or something like that)... Did you think why stomps happen so often?

Everything is very simple!

How I think the matchmaker will distribute the players by TIER or AMS (average match score).
But the players TIER doesn't matter - and never had it, because the devs did not do it correctly. Which does not mean anything, because TIER-1 can even take a lrmboat (I am not lrm-racist, but lrm is a trash in our reality).

The matchmaker does not take into account the indicators of WLR and KDR - although these are shown to be the main!

The problem of stomping in the game is most often due to a stupid balance - everyone who has encountered stomping knows this.

If you tell me that the matchmaker selects teams for AMS (average match score), I will say that this is still not correct, because AMS is not working correctly.

Why AMS is not working correctly?
Because the main indicators for high AMS are the amount of damage done, KMDD and kill assists or kills (kills and kill assists equal on scores). And this is a terrible mistake.

Why?
The fact is that if the kills was not done, then none of the players will receive a "kills assistant" and "KMDD".
That is, the player who kills the enemy does TWO necessary things - he gives the team a "kill assists" + "KMDD" and helps his team to win (and often brings victory).
No kills - no kill assistance, no KMDD, no scores, no victory! It's very simple, but not for devs - because they wanted to make a team game, in a shooter, really? Everyone wants to kill and win, nobody cares about team play! Team play happens situationally, and not everyone wants it.

We need to change the scoring formula, and after which we need to change the matchmaking formula.

Need to improve:
Solo kill - is the most underestimated indicator. If the player made a solo kill, he helped his team a lot, because the enemy did not have time to inflict damage or kill someone from your team.
Kills - the one who made the kill gave his team a "kill assist" + "KMDD" and helps his team to win.

Need to lower:
Kill assists - does not have the right to be equal to the kill, because you need to deal 1 damage to the enemy's mech to get kill assists.
KMDD - is a controversial indicator that does NOT indicate a player's efficiency.
and all other indicators too...

The scoring formula needs to be changed!
Because those who do more kills and solo kills should get more points than the one who did more damage or kill assists etc..
After that, we need to change the matchmaker formula!
Divide players into TWO groups - those who have AMS above 250 (for example) and those who have below 250 = these groups should not overlap in games (it also prevents twinks from messing around with newbies).
Everyone will be happy!
Beginners will play with others like them, and skilled players will play with skilled ones.

P.S.
You can change the balance of weapons and mechs... you can improve the maps, but if you do not change the balance of the players in the QP formula, then the stomps continue - then give me at least one reason why this game should be attractive to the players?

I brought over 40 friends to the MWO and none of them stayed because nobody likes stomping.
As they say: You need 5 minutes to find the game, you need 5 minutes to load into the game and reach the enemy, and you need 30 seconds to die... this is how an MWO beginner sees!

Edited by Rkshz, 31 August 2021 - 01:38 AM.

[justcallme A S H]

Stim & Poly (2+) on one side. Some JGx (3) on the other.

If you were going to pick something to use as an example you aren't doing yourself any favours with that screenshot...


Anyway we know the MM does not do any detailed matching aside from Tier. No secondary balancing no nothing. Feels like first come first serve which in higher tiers is definitely a problem.

Maybe it gets addressed at some point.

[Aidan Crenshaw]

One Question: What makes you think the Opfor had no premades? I know from experience that Stimraug often plays together with Poly Cat and LightningKats.
I would advise against making conclusions based on things you can't possibly know for sure.

[Bud Crue]

Show us the end of match summary screen that supports your argument. The show us "lots" of such screen shots. This single pre-match screen means nothing.

[justcallme A S H]

Aidan Crenshaw, on 27 August 2021 - 03:51 AM, said:

One Question: What makes you think the Opfor had no premades? I know from experience that Stimraug often plays together with Poly Cat and LightningKats.
I would advise against making conclusions based on things you can't possibly know for sure.

Yep Verti/LK and Stim/Poly would have been in group of 4 no doubt there so its a skilled 3man vs a skilled 4man. As I said, this is a super bad example.

[Davegt27]

Rkshz, on 27 August 2021 - 03:33 AM, said:

First of all, I do not want to offend anyone. Therefore, if something in the below written offend you, it is not intentional.

And so about the balance in QP: I do not know the "quick play" balance formula. On this I will reflect on the basis of my guesses.

As you can see in the screenshot (I have a lot of such screenshots), premades of JGx and MASC are in one team. It is not normal - this is an imbalance. With what score the opposing team lost, you can guess!

Why didn't the matchmaker assign us to different teams?



Everything is very simple!

How I think the matchmaker will distribute the players by TIER or AMS (average match score).
But the players TIER doesn't matter - and never had it, because the devs did not do it correctly. Which does not mean anything, because TIER-1 can even take a lrmboat (I am not lrm-racist, but lrm is a trash in our reality).

The matchmaker does not take into account the indicators of WLR and KDR - although these are shown to be the main!

The problem of stomping in the game is most often due to a stupid balance - everyone who has encountered stomping knows this.

If you tell me that the matchmaker selects teams for AMS (average match score), I will say that this is still not correct, because AMS is not working correctly.

Why AMS is not working correctly?
Because the main indicators for high AMS are the amount of damage done, KMDD and kill assists or kills (kills and kill assists equal on scores). And this is a terrible mistake.

Why?
The fact is that if the kills was not done, then none of the players will receive a "kills assistant".
That is, the player who kills the enemy does TWO necessary things - he gives the team a "kill assists" and helps his team to win (and often brings victory).
No kills - no kill assistance, no scores, no victory! It's very simple, but not for devs - because they wanted to make a team game, in a shooter, really? Everyone wants to kill and win, nobody cares about team play! Team play happens situationally, and not everyone wants it.

We need to change the scoring formula, and after which we need to change the matchmaking formula.

Need to improve:
Solo kill - is the most underestimated indicator. If the player made a solo kill, he helped his team a lot, because the enemy did not have time to inflict damage or kill someone from your team.
Kills - the one who made the kill gave his team a kill assist and helps his team to win.

Need to lower:
Kill assists - does not have the right to be equal to the kill, because you need to deal 1 damage to the enemy's mech to get kill assists.
KMDD - is a controversial indicator that does NOT indicate a player's efficiency.
and all other indicators too...

I think that players are divided into two types: killers and everyone else.
Killers - these are skill players, these players spent a lot of time to play well. It is these players who must have high scores and it is to these players that the matchmaker must distribute teams.
If this is not changed then the stomps will continue.

P.S.
You can change the balance of weapons and mechs... you can improve the maps, but if you do not change the balance of the players in the QP formula, then the stomps continue - then give me at least one reason why this game should be attractive to the players?

I brought over 40 friends to the MWO and none of them stayed because nobody likes stomping.
As they say: You need 5 minutes to find the game, you need 5 minutes to load into the game and reach the enemy, and you need 30 seconds to die... this is how an MWO beginner sees!

like what was said your screen shot is not good example

but I understand where your coming from

some of the answers can be found in past pod cast/town halls (if you listen carefully)
for example the past lead programmer said in a POD cast when he tightened up the ELO
stomps accrued more often (that means when they put more high ELO players on the field they had more stomps)
he said they did not know why that happened

another thing mentioned in past pod casts was what PGI did with new players
(you see in bygone era players would choose where they wanted to play, from beginner to advanced) well PGI said
they give new players a boost, the 2nd time i heard it mentioned the new player boost he stated how many matches they gave new players a boost

the important thing is they are not only capable of boosting players but are willing to do it
(they don't realize once you boost a player for any reason you just ruined your game)

the game does not work the way we think it works
last year when Paul was working on the game he asked the players about modifiers

he got silence no one said anything
very interesting



I think Paul could quickly talk over every ones heads about this game

[pattonesque]

Rkshz, on 27 August 2021 - 06:42 AM, said:

if I delete the screenshot can you understand the point of the post?
or will you talk about the chances (which did not exist) of the other team?
yes, the screenshot is not the best, should I look for others among hundreds? will it change something?

they're just saying that the mm put one of the better groups in the game on the other team. So it's not like you have JGx/MASC against a bunch of randos, it's JGx/MASC against Stimraug, Poly Cat, and LightningKats

[justcallme A S H]

Rkshz, on 27 August 2021 - 06:42 AM, said:

if I delete the screenshot can you understand the point of the post?
or will you talk about the chances (which did not exist) of the other team?
yes, the screenshot is not the best, should I look for others among hundreds? will it change something?

Of course and it is unfortunately the same discussion we've had time and time again.

I wish it was better, we all do!

[D V Devnull]

justcallme A S H, on 27 August 2021 - 07:03 AM, said:

I wish it was better, we all do!

In order to help this discussion not die as a slug-fest here, I shall interject one question...

Why is it that Tonnage Balance was not re-implemented alongside PSR Balance for making a Match to play?

~D. V. "either I am crazy, or Tonnage needed to be factored with PSR in MatchMaking still" Devnull

[Rkshz]

justcallme A S H, on 27 August 2021 - 07:03 AM, said:

Aidan Crenshaw, on 27 August 2021 - 03:51 AM, said:

Bud Crue, on 27 August 2021 - 03:55 AM, said:

yes you are right, the screenshot is incorrect, I deleted it, and the post was updated

[Saved By The Bell]

I have many bad fights lately. I dont know why they place me in teams who do 100-200 damage. I dont play much as before.

[Saved By The Bell]

I have 3 stomps today. 0-12, 2-12, 0-12. Time to stop play this game.

[John Bronco]

Rkshz, on 27 August 2021 - 03:33 AM, said:

As they say: You need 5 minutes to find the game, you need 5 minutes to load into the game and reach the enemy, and you need 30 seconds to die... this is how an MWO beginner sees!

I certainly agree this game has way too much time not playing the game while trying to play the game. Especially if you are higher tier. Not sure what the solution is, but letting people use the mechlab while searching would be nice (and is a feature that currently exists but only as a bug).

D V Devnull, on 27 August 2021 - 08:10 AM, said:

Why is it that Tonnage Balance was not re-implemented alongside PSR Balance for making a Match to play?

Because people want to play matches and the wait times are too long already.

Edited by John Bronco, 27 August 2021 - 08:41 AM.

[Nightbird]

The current PSR, and by extension, the matches made by the match maker, is designed by the community. In regards to its pattern of making unbalanced teams in upper tier matches, it is caused by the PSR ceiling of 5000 which can be reached with a meager WLR of 1.2, and blinds the MM to great players that have a WLR of 2, 3, 5, 10, etc. This is working as intended as a reward to allow skilled players an advantage in the MM.

[Remover of Obstacles]

Nightbird, on 27 August 2021 - 08:57 AM, said:

The current PSR, and by extension, the matches made by the match maker, is designed by the community. In regards to its pattern of making unbalanced teams in upper tier matches, it is caused by the PSR ceiling of 5000 which can be reached with a meager WLR of 1.2, and blinds the MM to great players that have a WLR of 2, 3, 5, 10, etc. This is working as intended as a reward to allow skilled players an advantage in the MM.

I agree the matchmaker could be tuned to for less upward progression (zero sum was the plan, right?). There is a big difference in skill level in Tier 1.
There is even a bigger skill difference between a solo Tier 3 player and a JGX pre-form (substitute in teams of players with WLR of 2+) .
This results in a lot of players getting killed quickly (not even reaching 100 damage in a match) and matches snowballing (12-4 in less than five minutes of play).

Is this working as intended? Apparently yes if the above quote is correct and we assume that quick matches (release valve wide open) are being prioritized.
Is it helping make more competitive games? No. Why would PGI think that they could get better games by adding groups to solo queue? Oh wait, they did...

Is it making the game more fun?
For the seals. No.
For the other players that get a "free win", but don't really contribute to the outcome of the match. No.
I am not even sure the seal clubbers are enjoying being paired against or teamed with players that paraphrase in game voice coms, "aren't (expletive redacted) helping".


Unfortunately population is low and severely less than optimal choices need to be made.

......
Starting Tuesday, April 28/2020 we are going to start a 4 week test on the live servers. This test is going to help us monitor the game's public queues and the overall effects of combining solo and group queues.

This is not just a simple combination of queues and seeing what happens, as we've taken steps to help address both internal and community concerns over that type of implementation. It's important to keep the teams as balanced as possible and to have games kick off in a relatively quick manner.

How the combined public queue will work:
- All matches will remain 12v12.
- Max group size will be restricted to 4 players. All group sizes of 2,3 and 4 are valid group sizes.
- When the Match Maker builds a 12-player team, it will ensure that there is no more than one 4 player group or as close as possible (2 player + 2 player for example) per team and all remaining slots will be filled with solo players.

What we expect:
- Playing with a smaller group of friends will be much easier to accomplish.
- Matches should kick off drastically faster than previously.
- Matches will become more competitive.

What we are monitoring:
- Match Maker ability to maintain even team building.
- Tonnage restrictions for groups. (note: we can adjust these dynamically/real time)
- Group composition and any release valves that may kick in to make games kick off.

[justcallme A S H]

D V Devnull, on 27 August 2021 - 08:10 AM, said:

Why is it that Tonnage Balance was not re-implemented alongside PSR Balance for making a Match to play?

The pinned post in this section explains why 3/3/3/3 was removed so probably have a read of it.

Outside of that - it is only very, very rarely I see a screenshot where the tonnage variance is more than was possible with 3/3/3/3... Most of the time it isn't a lot different as in the old system you could have 190t variance or there about.

[D V Devnull]

And here I was thinking my question would have sparked far better discussion which I could have actually hung back and just simply read... how mistaken I was... catching up now...

---

John Bronco, on 27 August 2021 - 08:41 AM, said:

Because people want to play matches and the wait times are too long already.

justcallme A S H, on 27 August 2021 - 11:56 AM, said:

The pinned post in this section explains why 3/3/3/3 was removed so probably have a read of it.

Outside of that - it is only very, very rarely I see a screenshot where the tonnage variance is more than was possible with 3/3/3/3... Most of the time it isn't a lot different as in the old system you could have 190t variance or there about.

I see that the two of you have SEVERELY misread & misinterpreted me. Please read on to understand exactly why...

I very much do NOT wish to return to the days of those ugly "3/3/3/3" setups under any circumstances. Those were definitely draconian (NOT meaning "Draconis Combine", but as in the dictionary "Draconian" definition) in their very nature, and would most certainly slow the MatchMaker by far too many factorable degrees of speed. It would gravely hurt Game Balance in general to re-institute those ever within MWO's lifetime, seeing as we don't want those up-to-195 Ton difference style of stomps happening again.

Instead, what I would think is a better course comes along the lines of a Dynamic Tonnage Balance that instead adds up the Total Tonnage of each Team formed. It would then swap players in and out during Team Formation based on BOTH the player's PSR and their Mech's Tonnage to ensure that both Teams to be put on any given field came within 50 Tons difference of each other. This would avoid having problematic PSR and/or Tonnage disparities even if the makeup of Teams looked totally goofy in the aftermath of any Match played, and thoroughly stays away from the massive MatchMaker slowdowns of hard-locked ugly "3/3/3/3" junk setups occurring. Heck, my guess is that the delay would only be just a few extra seconds with this thought of Dynamic Tonnage Balancing being added in order to ensure prevention of easy stomps. But hey, I guess that is just a pipe dream for now, seeing as they would likely have to spend a lot of time coding?

~D. V. "I hate '3/3/3/3' and NEVER want extreme waits... so I think Dynamic Tonnage Balance may be a solution?" Devnull

[VeeOt Dragon]

i don't even think the MM takes tonnage into consideration anymore. i have been in multiple matches where one team has 4 or more Assaults and the has NONE or maybe 1 if they are lucky. this almost always results in a stomp unless the under-tonnage team has a lance of really good light pilots and the assault ream is completely incompetent.

[Rkshz]

Remover of Obstacles, on 27 August 2021 - 10:07 AM, said:

I agree the matchmaker could be tuned to for less upward progression (zero sum was the plan, right?). There is a big difference in skill level in Tier 1.

determining the balance of teams by TIER is a priori fail - and I explained it above
tonnage does not matter much (the difference of 200 tons doesn’t matter)

what is the difference between an experienced player and the rest?
An experienced player can:
1. Knows where to shoot (how to shoot effectively).
2. Knows how to dodge and spread damage.
3. Knows maps (where to run, where to hide).
4. Knows the behavior of the opponent (when you play a lot, you know who is playing how).
5. Has good heat control.
6. The most effective builds.

inexperienced players DO NOT know these aspects, so for experienced players they are just meat - this is the main reason for stomping

- AMS is a universal indicator of a player's skill, not a TIER
- a new player and an inexperienced player may NOT have a high AMS a priori
the only logical solution to how to balance in QP is to make a border so that players with low AMS do not fall into the groups of players with high AMS
any other solutions will not change ANYTHING, it is impossible, TIER is lying

[VeeOt Dragon]

Average Match Score isn't the indicator you might think it is because there are many times that no mater how well you do personally you get stuck with a team of utter incompetents. (i like to tell people that if I have the highest match score someone done ****** up. mostly because i am an average player most of the time.) also there are roles a player might build fore that don't lend themselves to a high Match score like lights carrying TAG and NARC. that NARC mech might have been the VIP of the match because thanks to them the LRM mech devastated then enemy. also the high damage numbers of LRM in itself could shift those numbers (i like LRM personally but it is a sandpaper weapon with a LOT of spread damage)

i also beg to differ that Tonnage does factor in. if you have two teams of equal skill but one has more damage potential than the other (say 2 more Assaults than the other team) then it is more likely than not the the higher tonnage team will take the win just on the fact that they have more armor and firepower on the field. tonnage still needs to be factored into the MM. right now the biggest problem is that the MM doesn't seem to look at the tonnage or PSR composition of groups when slotting them into a QP match. hence why there has been so much discrepancy in team balance since the consolidation of ques into just one option. (there is likely also the potential that you could game the system by have a group that is say 3 Tier 4 players and 1 Tier 1. even if the system averaged out the PSR of the group that Tier 1 would potentially be put into a Tier 4 match where they have no business being. sadly we don't have the player numbers to take groups out of QP right now)

also keep in mind that Match score is heavily effected by if you are on the winning or losing team where PSR is more just how you preformed compared to the rest of your team (hence why your PSR can go up even if you lose.) i think PSR is a much better way of gauging your performance than MS since there are many factors out of your control that can determine if you win or lose.

perhaps we might need 6 Tiers instead of 5 so the uber try hards (top 5%) don't get put into matches with lower skilled players. sadly we just don't have the player numbers for any further division.

Edit: also take into account people like me who are always trying out new build ideas (and couldn't give two shits about meta) so my match score tends to fluctuate all over the place. i don't honestly care but it is what it is. who knows i might be more skilled than my AMS makes it look.

Edited by VeeOt Dragon, 29 August 2021 - 06:23 AM.