Tier Ranking Reset
#61
Posted 06 June 2018 - 03:12 PM
#62
Posted 06 June 2018 - 03:33 PM
Wouldn't it be better to divide the success by the total number of matches played making more an average? I mean if it took you 250,000 matches to get the same score as El Bandito,Shane or BrunoSS in 10,000 matches you would still never be in their tier.
and/or
Create a timer that slowly degrades match score if you stop playing for more than two weeks?
Were did I get this paint brush? Hey!! not nice!
Edited by Popcat, 06 June 2018 - 03:34 PM.
#63
Posted 06 June 2018 - 03:36 PM
Popcat, on 06 June 2018 - 03:33 PM, said:
Wouldn't it be better to divide the success by the total number of matches played making more an average? I mean if it took you 250,000 matches to get the same score as El Bandito,Shane or BrunoSS in 10,000 matches you would still never be in their tier.
and/or
Create a timer that slowly degrades match score if you stop playing for more than two weeks?
Were did I get this paint brush? Hey!! not nice!
#64
Posted 06 June 2018 - 03:43 PM
Grus, on 06 June 2018 - 03:12 PM, said:
It would depend on several factors. Would PGI keep the current PSR setup or modify it, and would PSR have a consistent reseeding of Tiers?
Example.
Take myself as an example. After 23 seasons, I have an average MS 240, average 1.12 W/L with almost all matches in Solo queue except for maybe 10 in Group (most drops as a group/unit have been in FP). Heavy on meds then heavies lights then assaults.
- Played 3386 games since Season scoreboard started July 2016. Remember PSR started August 2015 so 11 months before Seasons implemented
- Seeded as a high Tier 3 (3.7ish) from Elo stats when PSR went live Aug 2015.
- No record or such of when I hit Tier 2.
- Hit Tier 1 in Nov 2016 (Season 5). I had added a remark to a post I was working on cause the forums said Tier 2 but the game had me at Tier 1 after a few drops.
- Max Tier 1 April 2017 (Season 10)
- So 14 months to move from approx Tier 3.7 to Tier 1 then another 5 months to max out T1.
And if PGI did not separate the Solo/Group PSR, then that would give an incentive for players who really want to move up to actually GROUP UP and drop in GP.
Edited by Tarl Cabot, 06 June 2018 - 03:46 PM.
#65
Posted 06 June 2018 - 05:46 PM
#66
Posted 06 June 2018 - 06:25 PM
Internal Obedience XIII-omega, on 06 June 2018 - 05:46 PM, said:
Great idea. I agree.
#67
Posted 06 June 2018 - 06:57 PM
El Bandito, on 06 June 2018 - 06:55 AM, said:
It does make a difference, as every three months, I would be able to play with people that are competent, with good builds, for at least a month, during NA prime time. It would be just like September 2015 all over again. And baddies will stay at T4-5 most of the time, and have less chance of grouping with me. Although MWO's population is relatively small, it is still large enough to make tier system work to a degree.
PGI (or Russ himself) is dead set on trying to let PSR to have upwards bias as a rule. I doubt we can contest that--which is why tier reset is the next best (or easiest) thing to do. Although I wouldn't mind having both.
I think upward bias makes sense up to T4 or T3, but T1 should be for top performers.
#68
Posted 06 June 2018 - 07:06 PM
#70
Posted 06 June 2018 - 07:59 PM
My PSR increased the same amount I lost on a loss while doing nearly 200 damage. Now, I could see if I touched every cap point. But seriously, it went up????
#71
Posted 06 June 2018 - 08:30 PM
#72
Posted 06 June 2018 - 08:45 PM
KodiakGW, on 06 June 2018 - 07:59 PM, said:
Yep.
The proof right there PSR weights a win far, far too highly compared to individual performance - Inv Perf being the metric it should be using. And it a dynamic variable too based on the altering numbers. So just a "set and forget" system like it essentially has been.
Especially as Average Match Score rose inline with Skill Maze introduction (and the survival tree). PSR wasn't touched at all so suddenly people started climbing even faster with rubbish performances.
#73
Posted 06 June 2018 - 09:11 PM
Kaeb Odellas, on 06 June 2018 - 06:57 PM, said:
I think upward bias makes sense up to T4 or T3, but T1 should be for top performers.
maybe some kind of curve. bubble up in lower tiers so those are reserved for noobs like they should be, curve levels off around t2 and t1 is strictly players who maintain consistently good scores, possibly having a slight bubble down to prevent it from flooding with taters.
also the rate of change needs to go up as well. so if i get bogged down trying to level a trashmech, it drops me fast enough where i can have good games while still leveling said trash mech and recovery time would be short when i switch back to metabots.
Edited by LordNothing, 06 June 2018 - 09:12 PM.
#74
Posted 06 June 2018 - 09:16 PM
justcallme A S H, on 06 June 2018 - 08:45 PM, said:
Yep.
The proof right there PSR weights a win far, far too highly compared to individual performance - Inv Perf being the metric it should be using. And it a dynamic variable too based on the altering numbers. So just a "set and forget" system like it essentially has been.
Especially as Average Match Score rose inline with Skill Maze introduction (and the survival tree). PSR wasn't touched at all so suddenly people started climbing even faster with rubbish performances.
You'd be right if it was even close to possible to measure individual performance without encouraging statpadding instead of teamplay. The concept and delusion of "Personal" Skill Rating needs to go before we can move forward. Use the True Skill system or something similar, just get rid of individual performance tracking.
Anyone who has ever touched a book on game design knows individual performance tracking is both impossible and detrimental to gameplay and it's just another example that shows that PGI is sorely lacking in that department.
#75
Posted 06 June 2018 - 09:27 PM
If you are going to do some of the suggestions in here wont you end up with like no people in tier 1? Im an average 280-300 match score dude and I carry fairly regularly, there are a pretty high number of potatoes running around. so tier 1 may be a ghost town for solo Q particularly with all the seal clubing that goes on in Group Q inflating people stats.
#76
Posted 06 June 2018 - 09:40 PM
N0MAD, on 06 June 2018 - 07:06 PM, said:
Are you guys in Oceania? Do you play across all the servers? I wonder if we can apply the same arguments to the segments that play in far away timezones and remote severs. What is the total pop over there anyway?
#77
Posted 06 June 2018 - 10:02 PM
Savage Wolf, on 06 June 2018 - 09:16 PM, said:
Anyone who has ever touched a book on game design knows individual performance tracking is both impossible and detrimental to gameplay and it's just another example that shows that PGI is sorely lacking in that department.
I get accused of "stat padding" all the time. It makes me laugh when I win far more often than I lose. That isn't stat padding, that is over-performing(carrying).
Something many people don't understand, yourself included.
Stat padding is hiding 900m away with LRMs... Given they are ineffective there is a very simple way to fix that in PSR via proper calculations and weighting of in-game actions.
Blizzard have successfully had their match makers working off individual (peer-to-peer ELO) performance for over 20 years.
N0MAD, on 06 June 2018 - 07:06 PM, said:
That was adjusted, finally, a week ago.
T4 and T5 will no longer end up with T1. It has been finally hard locked that I won't see anyone under T3.
Problem is PSR is garbage in, garbage out... So it doesn't mean a hell of a lot cause eventually you climb.
#78
Posted 06 June 2018 - 10:32 PM
El Bandito, on 06 June 2018 - 06:55 AM, said:
It does make a difference, as every three months, I would be able to play with people that are competent, with good builds, for at least a month, during NA prime time. It would be just like September 2015 all over again. And baddies will stay at T4-5 most of the time, and have less chance of grouping with me. Although MWO's population is relatively small, it is still large enough to make tier system work to a degree.
PGI (or Russ himself) is dead set on trying to let PSR to have upwards bias as a rule. I doubt we can contest that--which is why tier reset is the next best (or easiest) thing to do. Although I wouldn't mind having both.
As someone who has played around with many alts (say hi to El Wizardo for me ) I would say without meaningful change to the system a reset will do nothing.
The current system gives the first ten matches a heavy bias, and you'd end up in lets say halfway through tier 3 in about fifty games? Even if you were to say spend 56% of those games in a trial Phoenix Hawk. Actually you personally would probably land in two.
PGI is seemingly convinced that the way to go is to separate by experience and not skill. I listened to every minute of the last podcast where Paul explained his view and why we have to just accept it.
If only we could convince them that they are wrong and that any effort to match people by their actual skill level would be more enjoyable for all involved. CW which someone told me is something you do from time to time as well is a great example of this. The skill disparity and group vs. pug argument is a long brewing issue.
Good players are often considered toxic for being good. Just today myself and the four other guys I was running with were accused of hacking cause we were bored and decided on a Grim Plexus defense to rush out the gate and push into them (and they happened to actually have a more organised in a group on the other side). A friend of ours happened to be on comms in the other group and came over to laugh about it a bit with us.
So I can't even imagine if there was a reset what it would look like when Proton, Twinky, or Writhenn are running around making 95% of 'tier 1' look like they just installed as it is....
#79
Posted 06 June 2018 - 10:45 PM
justcallme A S H, on 06 June 2018 - 10:02 PM, said:
I get accused of "stat padding" all the time. It makes me laugh when I win far more often than I lose. That isn't stat padding, that is over-performing(carrying).
Something many people don't understand, yourself included.
Stat padding is hiding 900m away with LRMs... Given they are ineffective there is a very simple way to fix that in PSR via proper calculations and weighting of in-game actions.
Blizzard have successfully had their match makers working off individual (peer-to-peer ELO) performance for over 20 years.
That was adjusted, finally, a week ago.
T4 and T5 will no longer end up with T1. It has been finally hard locked that I won't see anyone under T3.
Problem is PSR is garbage in, garbage out... So it doesn't mean a hell of a lot cause eventually you climb.
I'd add as someone who has dropped with you a bunch in CW while enjoying a Foster's, that one of the reasons you are good is that you are in fact very much so a team player. Of the guys in the top 30 on Jarl's, I have played regularly with 5 of them in CW and every single one of them is a team player. Might be said that it is part of what makes them a 'good' player.
#80
Posted 06 June 2018 - 10:47 PM
It's the PSR system that needs to change.. It needs to take into consideration more factors than it does now, and it needs to do this on a mech-by-mech basis..
I might be T1 in a Supernova-A, but T5 in a Pirate's bane..
21 user(s) are reading this topic
0 members, 21 guests, 0 anonymous users