227 replies to this topic

[Mystere]

Ruar, on 24 September 2017 - 02:35 PM, said:

The problem with the match score is it is primarily based on damage. A mech who leads a push, and dies early because of it, could be the key to the team winning, but that player will have a low score due to low damage.

There are quite a number of things that a player can do in this game to secure wins for the team but which are at best next to impossible to measure.

A drop caller who stays away from the fight (or intentionally dies early) to give orders and consistently wins is in all likelihood a great commander who has horrible stats other than W/L.

Heck even someone who just serves as the team mascot must be doing something right if the W/L ratio is significantly above 1.

Edited by Mystere, 24 September 2017 - 09:07 PM.

[KeeningBanshee]

Mystere, on 24 September 2017 - 08:51 PM, said:

There are quite a number of things that a player can do in this game to secure wins for the team but which are at best next to impossible to measure.

A drop caller who stays away from the fight (or intentionally dies early) to give orders and consistently wins is in all likelihood a great commander who has horrible stats other than W/L.

Heck even someone who just serves as the team mascot must be doing something right if the W/L ratio is significantly above 1.

Dude... I got a story about that. I was dropcalling a match like I always do, this Kingcrab pushed the corner of an entrenched enemy pushed first like a CHAMPION not even firing just torso twisting to soak as much damage as possible. As a result our whole followed behind and the enemy was completely broken backing up and scurrying away, we ended up wiping them out 12-1. I went on to praise that Kingcrab pilot calling him the MVP of the match cause he was in making sure that strategy was successful and it was wildly successful. I wanted him to know that he did a great job & to make sure he got some credit/validation for what he accomplished (since MWO sure was not going to in XP/CBills/Stats.) Some id1ot who barely did 300 damage himself started badmouthing the Kingcrab cause "he only did 27 damage" totally ignoring the sacrifice that pilot had made to push and die first then watch us get all the glory.

That is the behavior that gets stomped down by the time people have played enough to reach Tier 1... the willingness to sacrifice to achieve Victory and only the dedicated peek-snipers concerned with KDR or some other generic stat measurement (that they are usually not even that great at) because that is the only thing MWO shows/rewards. People can do amazing things which are 100% resulting in Victory for the Team yet by MWO judging/rewards standards they are actively penalized for NOT hiding in the back and living the longest being carried to Victory of dragging their Team down in Defeat.

Edited by KeeningBanshee, 24 September 2017 - 09:42 PM.

[MechWarrior5367362]

KeeningBanshee, on 24 September 2017 - 09:39 PM, said:

I was dropcalling a match like I always do

Stop dropcalling pug matches.

[KeeningBanshee]

The Weka, on 24 September 2017 - 09:47 PM, said:

Stop dropcalling pug matches.

Where is the fun in that??? That is actually the most exciting & enjoyable part of the game for me... giving my brain something to process by having it run endless predictive pattern models in real time, quiets my brain down so I can have some peace and quiet up there.

[justcallme A S H]

Cloves, on 24 September 2017 - 08:43 PM, said:

Arcana, you where not seeing a tier 4 player, you where seeing either a smurf account (someone cheating the MM system)

There is no cheating if you're allowed to do it freely. And PGI allows this.

Plus if you make a new account, perform really well in matches you'll be T3 unless than 10 games in every instance. PGI at least got that part right.

[H I A S]

El Bandito, on 24 September 2017 - 06:16 PM, said:

Boy come on over and get curb stomped by me at Super Smash Bros Melee.

Nah, i know in suck at such games
But at least in would not represent the average.

[Black Ivan]

Since Tiers don't represent how good a player is, just how much he played the problem starts here.

PGI would have to start measuring the actual skill of players

[MechWarrior5367362]

If you actually want to smurf in T4/T5, as opposed to T3, you would need to deliberately throw your cadet matches.

[Cloves]

So how does the PSR multiplier work? Is the magnitude of the PSR change dependent on the number of games played? Is that why I can barely see my bar move these days? ...and good lord I spent months in tier five and four, I guess that comes from learning on locusts...

[justcallme A S H]

Cloves, on 24 September 2017 - 11:13 PM, said:

So how does the PSR multiplier work? Is the magnitude of the PSR change dependent on the number of games played? Is that why I can barely see my bar move these days? ...and good lord I spent months in tier five and four, I guess that comes from learning on locusts...

http://mwomercs.com/...-as-best-i-can/

http://mwomercs.com/...-tiers-and-psr/

[Shifty McSwift]

People who actually care about their leaderboard scores, tend to play like the biggest ********. But who can blame them when you can't even trust the guy next to you, not to have dumped all his armor and boated LRMs when you are a brawldog, you can't even trust them to engage in simple tactics like covering fire or support some 60% of the time.

So it's hard, there's a point where being too laidback means you are just a detriment with a **** build or style, and if you go too far into competitive mode, you turn into a heartless Rambo playing for solely his score, which is just as often a detriment to the team.

People seem to forget the in a 24 man deathmatch, a good game will result in 20 or more of you dying. If you are consistently one of those last ones left alive in "good games" it doesn't actually mean you are a good player (a good leader maybe), it more indicates that you are the most willing to use your team like pawns to achieve your goal, which is never dying and always conserving your leaderboard score.

Seems like there are a lot of people out there who are living in some fantasy land where they expect to never die and always get good score, and at the same time, expect "good games" to happen. Would you like the moon too?

[MechWarrior5367362]

Cloves, on 24 September 2017 - 11:13 PM, said:

So how does the PSR multiplier work? Is the magnitude of the PSR change dependent on the number of games played? Is that why I can barely see my bar move these days? ...and good lord I spent months in tier five and four, I guess that comes from learning on locusts...

It is definitely the case that there is a sizable PSR multiplier in your early matches. Not sure what it is, or for how many matches.

It may also be the case (it seem so to me, but perceptions are subjective) that moving from T2 to T1, for example, takes more PSR points than moving from T4 to T3.

[Cloves]

I am making a distinction between leader-board scores and PSR as I believe they are not the same thing and do not reflect one another. Paul seems to say there is a fixed point change for the different match score results, and I have observed my own change varying amounts. With this last event playing odd and horrible builds with many a match with less than 200 damage, it still climbed...

-edit
Not saying he is wrong or that it goes up on a lousy score loss, just that I have memories of large movements of the bar and am not seeing them now that I am in lower tier 2. Maybe it was the LRM boating ...

Edited by Cloves, 24 September 2017 - 11:56 PM.

[The Basilisk]

Main problem with matchmaking in MWO (QP)is, was and always will be that the mech you take into game isn't weighted properly.
A T1 in a stock blackjack will not win against an other T1 in a tweaked shadowcat.

In Group QP there is additionaly the completely unfathomable thing with group organisation level.
Groups that drop with lots of undertonnage, groups that just drop together without any form of organisation or dropcaller just wild chatter, groups that speak loadout and strategies prior to drop...this isn't manageable by any matchmaker.

And last but not least the matchmaker works with false data. I got several players in my unit that are only T1 because the teams they play with carried them there.
Others are so extremely good that they tend to cut a path of devastation throught the enemy team before the rest of the team is able to fire a single shot.

At this point with that large skill and equipment disparities I do not think there is any point in matchmaking except the difference between beginner, experienced and veteran level of players and the mech they drive.
The rest is intangible for any MM.

[MechWarrior5367362]

There's no mystery about the thresholds.

Lose with less than 250 match score? PSR goes down.

Lose with 250-400 match score? PSR remains equal.

Lose with 400+ match score? PSR goes up.

Read Paul's thread. It's in there.

[Cloves]

The Weka, on 24 September 2017 - 11:36 PM, said:

It is definitely the case that there is a sizable PSR multiplier in your early matches. Not sure what it is, or for how many matches.

It may also be the case (it seem so to me, but perceptions are subjective) that moving from T2 to T1, for example, takes more PSR points than moving from T4 to T3.

Yes! this is what I have been experiencing as well! Like they shrunk tier 3!

[Cloves]

The Weka, on 24 September 2017 - 11:51 PM, said:

There's no mystery about the thresholds.

Lose with less than 250 match score? PSR goes down.

Lose with 250-400 match score? PSR remains equal.

Lose with 400+ match score? PSR goes up.

Read Paul's thread. It's in there.

Yes, the thresholds are there, but not the magnitude of the change.

what if you made tier 2 contain twice as many "steps" as tier 3? The bar would appear to move slower up or down for the same value loss or gain.

[Lily from animove]

H I A S, on 24 September 2017 - 02:23 PM, said:

I mean look at all the sub 250 MS sub 1,5 KD guys at Tier1. How can a MM create good games for the low percentage of decent Players with so much terribadness around?
From my experience the average MWO Player is unskilled at videogames. Maybe because they simply dont play any other competitive PvP games and they are here for their BT fix?
A Part of the Com says "We only want hard competitive matches and play the best of the best". In the meantime they curbstomp Tier5 Pugs at CW and refuse to participate in ranked Mode.
It's so sad.

Maybe just make the MM random.

No, they are not, a palyerbase is never too bad. players are distributed rather equal amongst all games.

bad MM comes form three things:

1. inproper rating (which definately is the case in MWO since neither the selected score number, nor the way how score is defined are indicating good players)
2. insufficient playercount. - no matter how good rating and MM is, if the playercount is too low theres not enough people to put together.
3. inproper MM algorithm(the one we have given looks fine if we didn't had issues no 1 and 2)

Edited by Lily from animove, 25 September 2017 - 12:10 AM.

[adamts01]

Shifty McSwift, on 24 September 2017 - 11:20 PM, said:

People seem to forget the in a 24 man deathmatch, a good game will result in 20 or more of you dying. If you are consistently one of those last ones left alive in "good games" it doesn't actually mean you are a good player (a good leader maybe), it more indicates that you are the most willing to use your team like pawns to achieve your goal,
Seems like there are a lot of people out there who are living in some fantasy land where they expect to never die and always get good score, and at the same time, expect "good games" to happen. Would you like the moon too?

It's normally 2 or 3 good players on each side, and if winning is a priority, then it's better for those players to stay alive longer to deal more damage. It's admirable to be in the front and aggressive, and personally more fun, but it doesn't net as many victories as hanging back and using potatoes as bait. So which is the "good" player, the one who dies early or the one who wins the match for the team?

[Shifty McSwift]

adamts01, on 25 September 2017 - 12:15 AM, said:

It's normally 2 or 3 good players on each side, and if winning is a priority, then it's better for those players to stay alive longer to deal more damage. It's admirable to be in the front and aggressive, and personally more fun, but it doesn't net as many victories as hanging back and using potatoes as bait. So which is the "good" player, the one who dies early or the one who wins the match for the team?

Well see you have already positioned your argument from the context of a "bad game", if only 2 or 3 out of 12 actually have the skill/equipment to make a difference then you have a game with 9 or 10 potatoes on one team...