849 replies to this topic

[Five by Five]

Will people on the winning team potentially still go down in rating? On a zero sum, I would hope to see all 24 players in the match grouped and graded on a bell curve. That might be more complicated to do, since an heavy has greater damage potential than a light, but with enough data, a balance should be achievable.

As far a resetting, . . . I'm a recent returning player, so I haven't really "invested" much time into my tier rating, so it's not very important to me. But the idea of the tier is to get better games, so sure go ahead reset it.

[GaelicWolf]

From a perpetual Tier 3 who is probably lower than that.
.
.
.
RESET PLEASE

[Impp]

I think the comment that your teir should move up or down based on personal performance and not solely victory is well taken. I think the fact that victory gives you some level of score bonus is enough. I think folks that don't support the victory should move down and folks that do well on a loss should move up. It might even discourage the folks that join a match and afk. I don't think win or loss alone is fair in quickplay as you don't have control of your team. If there is a small score plus to upward moment with a victory that should by enough.

[East Indy]

General comment to concern about the seeding matches: since players of all skills have been matched together for months and months, if those 20 are somehow the worst you ever play, you won't be stuck in "ELO hell," and in fact you'll be an advantage to your team for as long as the matchmaker thinks you're not as good as you actually are. If rating changes are adjusted, you'll average out per your leaderboard/Jarl scores and get to the correct tier even quicker. The end result being, ideally, better matchups.

Lockheed_, on 26 May 2020 - 10:45 AM, said:

Just to get some clarity for myself, the new PSR system will be Zero-Sum if implemented right?
So does that mean, a lost match is always a drop in SR no matter how my personal performance was? even if we played 9:12 and I did all 9 kills and twice the amount of dmg than the next best team mate?

Short answer: not necessarily. Even if Paul doesn't adjust his matchscore-to-rating-change system, matchscores of 250+ from the losing side will either retain or increase your rating and while some players will pass you others will drop. But, the better that system reflects player performance -- for example, the guy who puts up a fight even in a stomp -- the less match result will color the matchmaker's perception of you as a player.

[Three Cats In A Trench Coat]

DO IT!

just make tiers good dude, everyone will profit

[Ignatius Audene]

God. Pls do it asap.

[Star Dust]

1) Performing well or great on a losing team should not result in a large PSR loss.

2) Not performing well on a winning team should not result in large PSR gain. If anything it should be neutral or a loss in PSR for someone that performs exceptionally poorly.

Will the above 2 statements be true for the PSR change?

Edited by raarz, 26 May 2020 - 11:19 AM.

[Exhall]

This is some excellent deflection. The main point of this announcement is that mixed queue is staying, but sweetening the deal with a PSR improvement is a nice way to get us to accept this awful move. Well, I don't. Tried the game again after a long break and it's just awful. Don't think for a moment that a tier reset will keep elite groups from harvesting pugs. As long as mixed queue is here, I won't be.

[Mangeras]

Reset dat Tier thing. Gimme Tier 4, lemme rank up by earning it and I am happy <3

[Ult9876]

East Indy, on 26 May 2020 - 11:04 AM, said:

But, the better that system reflects player performance -- for example, the guy who puts up a fight even in a stomp -- the less match result will color the matchmaker's perception of you as a player.

That's more based around how matchscore is calculated though, which could easily be tweaked too to improve certain behaviour in matches.

[Brauer]

For those players arguing "damage/kills shouldn't give a lot of points, progress to objective should be more heavily weighted." Think about it this way. By killing or crippling a mech you are both 1) eliminating or decreasing that mech's ability to sit on a square and turn a capture point, and 2) eliminating or decreasing that mech's ability to damage or kill your own mechs.

Also, consider which takes more skill. Sitting on a square on a map to turn a conquest point, or engaging the enemy.

There are situations where flipping caps is more impactful than contributions to the fight, but even in competitive settings where teams are organized and spend far more time considering how to control the map and caps than in quick play, doing damage to and killing the enemy is of extreme importance. If you can't take your opponent's pieces off the field you are unlikely to win.

Biasing rewards towards sitting on cap points or rushing inanimate buildings will not change that getting kills and doing damage is a key way to win games.

Edited by Brauer, 26 May 2020 - 11:50 AM.

[Nearly Dead]

I saw it again today. Playing my Vapor Eagle wATMs, winning game where I screwed, got to the yellow circle and ended up boxed and dead in a three way cross fire. Rest of the team won, I got a skill rating increase for one or two assists and maybe 100? damage if that. Another match I got one of only two kills on our team, did the most damage but we lost and my skill rating dropped.

That is a bit absurd.

We need an objective measurement of how much a player contributes to each match.

K.I.S.S. Top 3 TS per team go up, bottom 3 go down, middle 6 stay the same.

My personal dead horse, Average damage divided by mech weight for the most recent 100 matches.

Team score for the most recent 100 matches, or a month, or almost any period, as long as it is a rolling average with the oldest dropping off.

If they want "tiers", make a large number of them, easily divisible, so they can combine them into groups to meet the needs of the match maker. 24, 36, 48 etc. Anything divisible by 3, 4, 6, 8 etc.

[VonBruinwald]

Brauer, on 26 May 2020 - 11:49 AM, said:

For those players arguing "damage/kills shouldn't give a lot of points, progress to objective should be more heavily weighted." Think about it this way. By killing or crippling a mech you are both 1) eliminating or decreasing that mech's ability to sit on a square and turn a capture point, and 2) eliminating or decreasing that mech's ability to damage or kill your own mechs.

Also, consider which takes more skill. Sitting on a square on a map to turn a conquest point, or engaging the enemy.

There are situations where flipping caps is more impactful than contributions to the fight, but even in competitive settings where teams are organized and spend far more time considering how to control the map and caps than in quick play, doing damage to and killing the enemy is of extreme importance. If you can't take your opponent's pieces off the field you are unlikely to win.

Biasing rewards towards sitting on cap points or rushing inanimate buildings will not change that getting kills and doing damage is a key way to win games.

From the sounds of things Win/Loss is going to be the biggest contribute to PSR Gain/Loss.

If so it is of benefit to increase the contribution of objectives over that of simple kills as it expands on the tactical level of play.

• Playing the objective is the most efficient way of increasing match score,
  
• But going for kills gives you a higher turnover of wins which can be more effective in the long run.

Remember "Thinking man's shooter".

[Pz_DC]

1. Reset PSR will work ONLY if You will make really better calculation of it after => before reset PLEASE post calculations somewhere around so we can not only say "yes/no" but see why it's "yes" or "no".
2. Tighter Tier balancing IMHO is not needed - even with, as You noticed, increased "steamroll rate" as for me it's going well now.
3. Maybe those "best of the best" should have their own "hiden" mark in queue so MM will split them between both teams?..
4. Still we have a moments like - "single pushing AS7 pull allied team into fight, died with like 150-200 dmg but give a allied team initial "impulse" so they was able to get ot of cover and start killng" - that AS7 died with low in-game stats but if we will look on match strategy he made over 50% of win alone... Or single sniper who pulled 1-2 000 "spread" damage with no kills while other player do 300 directly into enemy CT and get 3 solokill... Those things need to be noted, sad I can't say You how I see it

P.S. know it's wrong place but...
5. Maybe we will have some gameplay changes to make MWO look clothier to mission-based game instead as it is "FFA based game" ?...

[Beadhanger]

Do eeeet!

[Brauer]

VonBruinwald, on 26 May 2020 - 12:04 PM, said:

From the sounds of things Win/Loss is going to be the biggest contribute to PSR Gain/Loss.

If so it is of benefit to increase the contribution of objectives over that of simple kills as it expands on the tactical level of play.

• Playing the objective is the most efficient way of increasing match score,
  
• But going for kills gives you a higher turnover of wins which can be more effective in the long run.

Remember "Thinking man's shooter".

No. It flatly does not make sense to boost the rewards for objectives by all that much because 1) sitting on cap points or shooting stationary buildings is not challenging (rewarding players more for actions that require zero effort or skill is a bad idea), and 2) sitting on cap points is largely not as important as getting kills in QP.

If you check out players with high WLRs you'll see that generally they all are pretty damn good at doing damage and securing kills.

[DevinMace]

I really feel like there needs to be a 2 prong elo system in place, 1 for if you win or lose, and 1 how you do in those matches

1. You gain or lose elo based on who your playing in a match some sort of combined elo score vs the other.
2. Elo changed based on how you do vs your averages in matches

The result would be if someone decided to smurf for example they would fly up the standings really fast because of how they perform in matches, where as someone who gets carried by a team would not go up as much, but also if they don't play with the team they will get smashed down to were they belong.

[OmgKllL]

Imho the PSR management of the current game isn't that bad, whan I play decently I never get the "decrease".
There's always something you can do to get the most out of every situation in the game, even in the dumbest match with the most crippled team.
That's what I've learned during this year of playing MWO and that's what I love of this game!
It's supposed that you know the maps and the team's mechs, otherwise you are just running blindly into your loss...

[Rustyplate]

The tier system in my opinion is garbage. Been playing a couple years now. Losing rank because your team loses. really fair......not. Biggest problem with this game has always been wait time. Been suggesting to use the bot system like world of warships. I like this game a great deal. But when i spend more time running practice matches than i ever do in actual matches. I mean really? I like the latest changes. Far better than it was. I would wait 2-5 days sometimes to get a match. But really have some more work to do.

[Dusty Mayalith]

I'm all for a Reset

Honestly, I don't really see them as being needed at all, give it a personal time played value, or additional achievement type value instead of any sort of skill gate system.

Let everyone fight everyone, because **** happens and we like fast que's!