Psr Update And Hold On Patch.
#381
Posted 11 June 2020 - 04:25 AM
#382
Posted 11 June 2020 - 04:35 AM
Dauntless Blint, on 11 June 2020 - 03:30 AM, said:
This = inflated PSR true or false?
What? I don't see how that's relevant. You complained about nascar so I told you how to break nascar.
FYI in case you have missed it I have been consistently against the group queue/solo queue mash up. While it's here I'm going to drop with a group, but I don't think it's good for the queue overall.
#383
Posted 11 June 2020 - 04:56 AM
#384
Posted 11 June 2020 - 05:34 AM
Thorn Hallis, on 11 June 2020 - 04:56 AM, said:
PSR reset and changes to how PSR is calculated is not about making sure top players only play with top players, and very low performing players only play with other very low performing players. It's about getting a better balance of skill on each side of a match. Currently the matchmaker, in my opinion, basically just grabs 24 random pilots and throws them into a match. The current matchmaker is essentially blind to player skill.
#385
Posted 11 June 2020 - 05:48 AM
Brauer, on 11 June 2020 - 05:34 AM, said:
i'd say the solo matchmaker actually created properly balanced drops if you were to go by tiers only. the randomness in player skill you saw in your drops was due to PSR being fubar
group queue MM (which is used in soup queue) is a blind chicken, tho.
#386
Posted 11 June 2020 - 06:15 AM
w4ldO, on 11 June 2020 - 05:48 AM, said:
group queue MM (which is used in soup queue) is a blind chicken, tho.
The MM is blind because PSR is not based on skill, it is based on Match Score*Matches Played. The #matches you played should not have any impact on your skill rating. All the current MS proposals will include Matches Played as a main component, which is why they won't do anything to improve match making.
David Sumner, on 11 June 2020 - 02:58 AM, said:
Except.
Groups > Solo for wins. Yes?
So there will always be more Solos on the losing side, simply because more of the winning side will be in a group.
Take 4 solo players with 400 avgMS and 2 WLR, and 4 group players with 400avgMS and 10 WLR. Would you prefer the MM say, they are equal and therefore putting them on both sides will create a balanced match? Or would you want the MM to say the 4 players grouped up is stronger, and therefore more good players need to be added to the team with 4 solo players to balance the teams?
That tells you whether MS or WLR is better. The MS system is designed to screw solo players over by being blind to the impact of groups (high WLR).
MrMadguy, on 11 June 2020 - 12:50 AM, said:
1) Do we want PSR to be W/L dependent?
Yes - Make values different for win and loss.
No - Make them the same.
2) Do we need true zero-sum or not?
Yes - change PSR according to relative performance vs other players.
No - change PSR according to relative performance vs fixed value.
3) How fast do we want player to climb/drop?
Set PSR change values accordingly.
Everything else - is purely about MS calculation. What actions mean more skill and what don't. At the end precise MS calculation isn't that important now. Except may be reducing DMG component. Because it can never be perfect anyway.
So we just need to answer to this 3 questions above and make changes as soon, as possible.
Another thing, that would be nice - per 'Mech PSR modifier. But such system is way too hard to implement.
1/2/3 doesn't change the fact that a match score based system will set PSR = <PSR change> x <#matches played>. Just how is the # matches played an important factor in your skill? It's not. It's why the current system doesn't work either since it is also <PSR change> x <#matches played> just with a different PSR change formula.
Just pointing out that W/L doesn't have this problem, so integrating W/L into MS will not incorporate any of the benefits of W/L.
Edited by Nightbird, 11 June 2020 - 06:44 AM.
#387
Posted 11 June 2020 - 06:29 AM
David Sumner, on 11 June 2020 - 03:38 AM, said:
I'm already lunch for them.
It's impossible, even with max turn boosts, to actually keep them in sight long enough to hit with more than a glancing laser shot.
When (not if) the rest of the side runs off and leaves you on your own (which is what actually happens, rather than my "standing at the back" by choice), you're toast for the first light that comes along.
You've never dropped with me then, I have a habit of trying to make sure the stragglers are not left behind, when I am playing a faster mech. I also know how to move my Assault mechs (my primary weight class) in order to keep enemy lights from going all around me, making the job of clearing them easier... I do this with all of my Assaults.... Am I successful every time, nope, not on your life, but I can usually get enough damage on them to make them think twice. I have an Atlas with stock engine and zero of the mobility skills unlocked and have no issues..... Maybe the problem isn't the team leaving the Assaults behind, but learning how to move the mech you've chosen in a more impactful manner...
#388
Posted 11 June 2020 - 07:04 AM
Paul Inouye, on 08 June 2020 - 02:52 PM, said:
Please keep in mind, we're going to need buy-in from the majority of players.
After having gone through all the posts in this thread so far, I do not believe the community will agree on something specific. I do hope Paul/Russ/Engineers have at least looked at Nightbird's idea which doesn't use MS at all, I personally believe it to be the best way to go after having reviewed his data.
https://mwomercs.com...thread-we-need/
Only thing I have to add to this thread is echo something LordNothing has stated a few times: Don't wait a few months hoping the community comes to a consensus... do something, even if it is implementing what you had originally planned.
#389
Posted 11 June 2020 - 07:38 AM
First is that our player base has shrunk to the point where just about anything we do with PSR and MM won't really matter much.
This leads me to ask, what is the goal of these changes? Is it to appease the current player base to try to stave off more people leaving, or is it meant to encourage more people to try and play? The answers to those to questions are very different.
First, I'll explain why I ended up leaving in the first place. It was at the time that faction play become simply Inner Sphere vs Clan, as a Smoke Jaguar, we had a pretty tight knit group of units and some reliable unitless people. Once that change was implemented, the quality of play in invasion mode dropped significantly. Players would throw the match if the planet fight wasn't for their clan. I realize at this point that the player base is too small now to go back to faction specific, but that was one reason I left initially. The second and far more relevant to this loss of player base, is the choices we now get before drop. Giving those choices has lead to many maps almost never getting played. This is significant because prior to that, you had to build your mech to be able to handle any environment. Now you have players building super high heat builds or just plain missile boats because unless you are terribly unlucky, you won't have to play Terra Therma or other maps that have high static temp and plenty of cover to duck behind from missiles. Also, the missile boat influx has increased because the maps that are more open are also usually picked. Those choices at the beginning of the match may appear to be a good idea at first, but it has resulted in lopsided builds being standard.
For a new player, seeing these maps that never get picked, trying to compete with lopsided builds, and figuring out the mechanics (like asking why a Clan ER Med laser has a longer cooldown than the large version) of the game all at the same time can be overwhelming. All of you hardcore vets complaining about "carrying" noobs should be trying to help them understand the tactics and why's instead ripping into them on coms and bitching about your team not pulling weight. Not all people on coms are that way, but there is plenty of bitching.
I am focusing on the things that had me leave the game in the first place, because I am relatively certain this is the reason the player base ended up shrinking to this point. Everything we have in place for PSR and MM are probably fine. What we need to focus on are ways to increase the player base to let those systems actually function as intended. I'm not saying that reverting the changes that made me leave will help, but coming back, seeing these things as they are now and the general state of the game and community, is saddening. I really liked this game back when it came out. Even in my absence, I would log in to see if things had changed back for the better... and every time it ended up being the same or worse. And now, we're here. I have a long friend list in game, and it seems many are doing the same thing I was. Logging in every now and then to see if the game brought back the things it broke.
You all can try changing the PSR all you want to cater to the elites that are still here. They'll be here to the end and the player base will keep shrinking. Honestly, I do think removing the pre-drop selections for map/mode and putting faction play back to keeping Houses and Clans seperate will bring old players back. Maybe not, maybe the current player base won't be able to handle the faction portion, maybe the choices pre-drop are a good thing... all I know, is that the biggest issue isn't with the PSR (Something your avg player doesn't care about, and honestly shouldn't even be shown) or even with the way the MM works. The problem is the player base it has work with. The other issue is how the top tier players don't try to help the noobs get better, explain the tactics, have patience with the noobs.
Maybe try something like faction play drop decks to a quick play mode with the ability to mix clan and IS mechs.
Edit: Also, the majority of avg players don't usually visit forums. Unless it's an MMORPG, this is a FPS game. "It's a FPS, no quests to research, login and go." Granted, Units had/have their own sites and such, but your avg player just wants to play, doesn't care about PSR, MS (Only because new players need C-Bills for new mechs and equipment), but generally, this type of thing isn't something the avg player cares about.
Edit 2: Congrats on making it through all way to the end (both of my 2 cents, and the current end of the posts.)
Edited by Ghoja, 11 June 2020 - 08:02 AM.
#390
Posted 11 June 2020 - 08:03 AM
w4ldO, on 11 June 2020 - 05:48 AM, said:
group queue MM (which is used in soup queue) is a blind chicken, tho.
Yes, PSR is the main culprit. That's why the matchmaker basically grabs players at random right now, because it simply cannot tell the difference between someone with a sub-1.00 WLR and a multiple world champ with a 2.00+ WLR. The matchmaker overall is blind to player skill right now because of how bad PSR is, and group queue is a separate and significant issue as it is even more blind to the impact of groups.
#391
Posted 11 June 2020 - 08:04 AM
Brauer, on 11 June 2020 - 05:34 AM, said:
PSR reset and changes to how PSR is calculated is not about making sure top players only play with top players, and very low performing players only play with other very low performing players. It's about getting a better balance of skill on each side of a match. Currently the matchmaker, in my opinion, basically just grabs 24 random pilots and throws them into a match. The current matchmaker is essentially blind to player skill.
After re reading the original PGI posts, I think we are talking about a fine tuning adjustment of PSR that will do virtually nothing to improve the customer experience because matching players by tier is turned off in due to the "Quest for the Holy Wait Time".
If I am getting what they meant from what they said, there aren't enough players in line to sort players by tier. Certainly not enough premade groups in line to match 4 man vs 4 man. If this is the case this is all kabuki theater intended to make us feel better; and the matches are already, and will remain, "[color=#00FFFF]as good of matches as we can possibly see based on the player base numbers[/color]".
Maybe I am misinterpreting what was posted, I find it hard to believe they would to to this much angst and work to get people into accurate skill tiers and then ignore tier when matchmaking.
#392
Posted 11 June 2020 - 08:17 AM
#393
Posted 11 June 2020 - 08:21 AM
Nearly Dead, on 11 June 2020 - 08:04 AM, said:
After re reading the original PGI posts, I think we are talking about a fine tuning adjustment of PSR that will do virtually nothing to improve the customer experience because matching players by tier is turned off in due to the "Quest for the Holy Wait Time".
If I am getting what they meant from what they said, there aren't enough players in line to sort players by tier. Certainly not enough premade groups in line to match 4 man vs 4 man. If this is the case this is all kabuki theater intended to make us feel better; and the matches are already, and will remain, "[color=#00FFFF]as good of matches as we can possibly see based on the player base numbers[/color]".
Maybe I am misinterpreting what was posted, I find it hard to believe they would to to this much angst and work to get people into accurate skill tiers and then ignore tier when matchmaking.
Well we'll see. I can't remember if the matchmaker balances the number of Tier 1s on each side of a match, for example. If it does that then a PSR reset and change can provide some benefit. If it does not then the benefit may be quite limited. Since making any changes to the matchmaker itself is off the table according to PGI (last I knew) that limits the available options.
Cipher2012, on 11 June 2020 - 08:17 AM, said:
Drop in a group.
Take a power position.
Gun down everything that shows.
This is how you break nascar. It is extremely easy to punish if you have some backup.
#394
Posted 11 June 2020 - 08:29 AM
Andrzej Lechrenski, on 08 June 2020 - 07:38 PM, said:
Solution 2
Use match score only in a relative (within the same match) sense and give priority to W/L ratio. Winners gain PSR, losers lose PSR. Here is an idea of what that might look like using Russ' PSR numbers:
Winning team:
player 1 +24
player 2 +22
...
player 11 +4
player 12 +2
Losing Team
player 13 -2
player 14 -4
...
player 23 -22
player 24 -24
For sorting players in Tiers, based on their competitiveness, aka ability to affect the outcome of a match, this is the way to go.
First, the main questions.
Should win/loss effect player movement?
Absolutely !
All negative PSR for the losing team and All positive only PSR for the winning team is one of the ways to promote teamplay. Along with increasing score for team objectives.
You want the groups to help their team? This should do it. Player skill in MWO is a mix of aiming, mech-building and ability to play as a team.
What is the most common cause of a stomp? One team coordinating much better than the other one. That's why it must apply to Tiers.
Also, over the course of 20 games, if your W/L is about 1, it won't matter as long as the sorting is zero-sum. You'll get a true Tier movement relative to your actual skill. If you win often, you'll go towards the higher tiers, if you lose often you'll go down. Effectively sorting Tiers based on Ability to teamplay + General skill level.
Ok, one or two players might stand out on the losing team. Good for them, but the team still lost. These players will still go up when they win, probably a lot.
Someone did terrible on the winning team? That driver is the one who needs a little motivation, probably has no experience. That's the player base MWO needs. He's going to do way worse when he loses and will hopefully have an easier game next time because of the change.
Does it need to be zero-sum?
YES !
This is the only way to have a good distribution of players within the tiers over time. Also being zero sum per game is the easiest way to ensure the end result is. Anything non-zero sum risks a distribution skewed towards Tier 5 or Tier 1.
=============================
SUGGESTION
My suggestion is less aggressive than the one Andrzej put as solution 2 and more in line with others I've seen here.
Winning team
player 1 +16
player 2 +14
player 3 +12
player 4 +10
player 5 +8
player 6 +6
player 7 +4
player 8 +3
player 9 +2
player 10 +1
player 11 =0
player 12 =0
Losing Team
player 13 =0
player 14 =0
player 15 -1
player 16 -2
player 17 -3
player 18 -4
player 19 -6
player 20 -8
player 21 -10
player 22 -12
player 23 -14
player 24 -16
This give a little break from PSR change to the Players who carry a loss and Players who do nothing for a win. It's also more gentle to middle of the pack and shifts the top-win and bottom-loss players more heavily.
How will this cater to the different player styles:
The damage dealer : Win/loss doesn't matter. I can deal damage from any tier and get my c-bills, I still move up a bit from my big numbers so I'm happy.
The team player : I get ranked against TEAMS of a similar skill level. Those matches can be a challenge.
The role player : You don't finish the mission, you don't get paid. Seems fair my reputation is affected by the outcome.
The new pug : I won't go up until I learn to play, so matches won't get too rough against coordinated teams too soon.
Edited by Cluster Fox, 11 June 2020 - 08:35 AM.
#395
Posted 11 June 2020 - 08:45 AM
Cluster Fox, on 11 June 2020 - 08:29 AM, said:
For sorting players in Tiers, based on their competitiveness, aka ability to affect the outcome of a match, this is the way to go.
First, the main questions.
Should win/loss effect player movement?
Absolutely !
All negative PSR for the losing team and All positive only PSR for the winning team is one of the ways to promote teamplay. Along with increasing score for team objectives.
You want the groups to help their team? This should do it. Player skill in MWO is a mix of aiming, mech-building and ability to play as a team.
What is the most common cause of a stomp? One team coordinating much better than the other one. That's why it must apply to Tiers.
Also, over the course of 20 games, if your W/L is about 1, it won't matter as long as the sorting is zero-sum. You'll get a true Tier movement relative to your actual skill. If you win often, you'll go towards the higher tiers, if you lose often you'll go down. Effectively sorting Tiers based on Ability to teamplay + General skill level.
I wanted to point out the result of this suggestion doesn't stabilize the PSR value of players at any skill level. All players will still have an <average PSR movement per match> and their final PSR value will be <average PSR movement per match> x <# of matches they play>. As they play more matches, people with a positive avgPSR movement will go to infinity, and with a minus avgPSR movement will go to minus infinity. Also since people play a different number of matches per week, it's very random where everyone ends up.
As long as you don't take # matches played out of the equation, the MM will essentially be ranking people based on # matches played instead of skill, something it does today and we can see the results.
Edited by Nightbird, 11 June 2020 - 09:16 AM.
#396
Posted 11 June 2020 - 09:31 AM
Nightbird, on 11 June 2020 - 08:45 AM, said:
I wanted to point out the result of this suggestion doesn't stabilize players at any skill level. All players will still have an <average PSR movement per match> and their final PSR values is <average PSR movement per match> x <number of matchs they play>. People with a positive WLR will go to infinity, and with a minus WLR will go to minus infinity.
As long as you don't take matches played out of the equation, the MM will essentially be ranking people based on matches played instead of skill, something it does today and we can see the results.
First of all, good thread on the WLR based MM.
To be honest, I still stand with my original suggestion of PSR movement relative to % of average team score. While not per-match zero-sum, it converges to it and does account for some sort of standard deviation.
I agree this will diverge for any W/L other than one. Honestly, without seeing the code, it's very hard to guess what is within the scope of PGI's PSR workable data. They already mentioned they can't really get player rolling averages. So I've been trying to suggest practical tweaks that, while imperfect, stick close to the originals.
A player movement could be weighted on matches played.
Option:
Multiplier = Vector * AmplitudeConstant / (matches (in season) + DampingConstant)
i.e.
- Ranked 12th on a loss = -16
- 70th match of the season
- Random constants for the example (20 matches break even)
PSR change = -7.1 = -16 * 40 / (70 + 20)
This stabilizes the player the more matches are played and amplifies the movement at the start of the season. Can PGI do that easily... no idea.
Edited by Cluster Fox, 11 June 2020 - 09:32 AM.
#397
Posted 11 June 2020 - 09:43 AM
Cluster Fox, on 11 June 2020 - 09:31 AM, said:
This stabilizes the player the more matches are played and amplifies the movement at the start of the season. Can PGI do that easily... no idea.
Like a simple avgMS calculation? It weighs early matches more than later matches thus dampens perfectly lol! The limitation PGI put on this thread is quite literally "you're not allowed to fix the problem"
Edited by Nightbird, 11 June 2020 - 09:58 AM.
#398
Posted 11 June 2020 - 10:07 AM
Mordenthral, on 08 June 2020 - 04:41 PM, said:
captureassist
capture - Gain for capturing a capture zone.
capturepulse - Gain for time you are capturing in a capture zone.
firstcapture - Gain for capturing the first capture zone in a match.
powercell_pickup - Gain for picking up a power cell.
powercell_dropoff - Gain for dropping off a power cell.
kill_powercell_carrier - Gain for killing a power cell carrier.
I absolutely agree with this. One of the biggest frustrations are teams that play every mode like skirmish and completely ignore objectives. These should have had higher weights all along. Nothing worse than doing your job on objectives when no one else is, losing because of it, and still losing rank. Otherwise, might as well just keep Skirmish and Domination, and drop the other modes altogether. I'd agree that it wouldn't be much fun, but as it stands now, that's pretty much what we're playing every match anyway.
As to the rest, I'm no engineer so I'm cool with whatevs.
#399
Posted 11 June 2020 - 10:15 AM
#400
Posted 11 June 2020 - 11:15 AM
Fix the game.
Than you will have 100 000 players and you will have a perfect MM even if the logic of building a match is bad.
4 user(s) are reading this topic
0 members, 4 guests, 0 anonymous users