Jump to content

No More Xp Bar


65 replies to this topic

#21 Dingo Battler

    Member

  • PipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 357 posts

Posted 04 February 2016 - 04:23 PM

Its funny how people say its an XP bar, when we had that topic a few days ago about tier 1s seal clubbing their way through tiers 4&5

Even if its an XP bar, the grind is so hard that after thousands of matches, some people don't even move a tier.

View PostTier 1 Smurf, on 04 February 2016 - 04:13 PM, said:

n the last town hall, right after all the ban wave stuff, Russ was talking about how he can see what were the last 30 matches to kick off, see the tier and average deviation. It seemed to me there were far fewer T1 matches going on than T3 and 4. He ever said there was a T3 vs T4 happening iirc. So I imagine T1s being rarer have the worst end of the MM.


If you're right, making T1 the 20% would make T1 matches more common, not less common, since it looks like less than 20% of players are in T1 now, given the longer wait times compared to other tiers. If anything, T1s should match more easily, given that T1 players practice more often, but this is nothing like that.

#22 5LeafClover

    Member

  • PipPipPipPipPipPip
  • The 1 Percent
  • The 1 Percent
  • 317 posts

Posted 04 February 2016 - 04:25 PM

View PostSQW, on 04 February 2016 - 04:15 PM, said:


Seriously?

And just how will you rate the players as GOOD? K/D rate? DMG per game? Time logged in? Number of mastered mechs? Mix of everything? Your method automatically assumes player skill is a knowable quantity which if that is so, PGI's Tier system would work far better than what you are proposing. Also, if Tier 1 will only ever have 20% of total population, who on earth would want to advance to a rank where you have to queue for 15min?

And if you know how percentages work, dropping 5 tiers to 4 tiers wouldn't make a difference. lol

Go back to the drawing board OP.


What utter garbage.

OP hasn't suggested any change to the mechanism PGI use to raise or lower PSR. Neither did he suggest that match making shouldn't still make up games from players in neighbouring tiers.

The point was (as I understand it), to adjust the up/down mechanism to be neutral, rather than having a bias toward moving up which causes the XP bar effect (THE TITLE OF THE THREAD BTW). And then secondly to distribute players into 5 equal size tiers. At the moment, it's quite possible that 40% plus (for example) of the player base exist in a single tier. How can that be a good thing?

#23 Lyoto Machida

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Overlord
  • Overlord
  • 5,082 posts

Posted 04 February 2016 - 04:29 PM

View PostSQW, on 04 February 2016 - 04:15 PM, said:


Seriously?

And just how will you rate the players as GOOD? K/D rate? DMG per game? Time logged in? Number of mastered mechs? Mix of everything? Your method automatically assumes player skill is a knowable quantity which if that is so, PGI's Tier system would work far better than what you are proposing. Also, if Tier 1 will only ever have 20% of total population, who on earth would want to advance to a rank where you have to queue for 15min?

And if you know how percentages work, dropping 5 tiers to 4 tiers wouldn't make a difference. lol

Go back to the drawing board OP.


Posted Image

#24 Mister Blastman

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Survivor
  • Survivor
  • 8,444 posts
  • LocationIn my Mech (Atlanta, GA)

Posted 04 February 2016 - 04:34 PM

View PostMonkey Lover, on 04 February 2016 - 01:51 PM, said:

I know lots of people who play all the time and are in the same tier they started off with. It seems to be working to help some of the lower skilled players.


Russ did say i think during the townhall but maybe tweeter they're going to adjust the Tier system.


There's not enough bads falling through the strainer...

#25 White Bear 84

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 3,857 posts

Posted 04 February 2016 - 04:49 PM

View PostMystere, on 04 February 2016 - 02:02 PM, said:

Back to Elo then? Posted Image


Wash your mouth out with soap young man!

Posted Image



#26 Mystere

    Member

  • PipPipPipPipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 22,783 posts
  • LocationClassified

Posted 04 February 2016 - 04:58 PM

View PostAccused, on 04 February 2016 - 02:21 PM, said:

When will they implement Tier 0? My win/loss - k:d:r isn't changing.


I prefer Level 6. Posted Image

Edited by Mystere, 04 February 2016 - 05:01 PM.


#27 DAYLEET

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 4,316 posts
  • LocationLinoleum.

Posted 04 February 2016 - 05:20 PM

Whats the point when 2-3 tier are always mixed together? For a long time ive been experimenting by timing waiting time and cancelling my wait time and ive been saying that it didnt work because i knew for a fact that t3 t1 forumers where in my match. Well last townhall Russ mentioned how it really works, a T1 can wait 1minutes and be put with 23 other T2 who all waited 2 seconds because the waiting and releasing is per person not per tier. So in conclusion, who cares about psr? Its most important function is to let the new players learn together rather than be canon fodder.

#28 Kyynele

    Member

  • PipPipPipPipPipPipPip
  • The Vicious
  • The Vicious
  • 973 posts

Posted 04 February 2016 - 07:32 PM

View PostAlistair Winter, on 04 February 2016 - 02:30 PM, said:

Do pure Tier 1 matches ever happen? Any Tier 1 players who can confirm or deny?


Well, at least in group queue pure T1 matches do happen.

Seems to be very rare in solo queue. Or, there are so many not-very-goods in T1 already that it's impossible to tell.

#29 Tier 1 Smurf

    Member

  • PipPipPip
  • Death Star
  • 61 posts

Posted 04 February 2016 - 08:03 PM

View PostKBurn85, on 04 February 2016 - 04:23 PM, said:

If you're right, making T1 the 20% would make T1 matches more common, not less common, since it looks like less than 20% of players are in T1 now, given the longer wait times compared to other tiers. If anything, T1s should match more easily, given that T1 players practice more often, but this is nothing like that.

By further diluting the competition in T1 it may make shorter queues yeah, but really we haven't a clue how the MM goes about releasing values for cross tier matching, the damn thing might wait 5secs before widing the player delta for crossing tiers.

#30 adamts01

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 3,417 posts
  • LocationPhilippines

Posted 05 February 2016 - 12:17 AM

View PostDAYLEET, on 04 February 2016 - 05:20 PM, said:

Whats the point when 2-3 tier are always mixed together?

View PostTier 1 Smurf, on 04 February 2016 - 08:03 PM, said:

By further diluting the competition in T1 it may make shorter queues yeah,

I think this would tighten things up as well as provide faster matches. Pretend there are only 3 tiers,as there are 3 types of mixtures right now (Tiers 1-3, 2-4, 3-5) I would assume the population could handle that. Right now MM reaches all the way down to the middle men to fill tier 1 matches, and down to absolutely new players in tier 5 to fill tier 3 matches. My proposal would avoid that. I think this would provide better matches at all levels. And if it turns out there are more than enough players for 3 tiers, PGI could consider having 4, for even tighter matching.

View PostSQW, on 04 February 2016 - 04:15 PM, said:


Seriously?

And just how will you rate the players as GOOD? K/D rate? DMG per game? Time logged in? Number of mastered mechs? Mix of everything? Your method automatically assumes player skill is a knowable quantity which if that is so, PGI's Tier system would work far better than what you are proposing. Also, if Tier 1 will only ever have 20% of total population, who on earth would want to advance to a rank where you have to queue for 15min?

And if you know how percentages work, dropping 5 tiers to 4 tiers wouldn't make a difference. lol

Go back to the drawing board OP.

I never said PSR should change, that's how PGI determines who's good. That's an even messier subject than this. I fully realize it has flaws.

If 20% of the population was tier 1 then the wait time would be exactly the same as any other tier. And if there aren't enough players online for that, then make it 4 tiers with 25% of total players in each, or 3 tiers with 33% of players in each. If tiers aren't mixed, that makes a bigger pool so quicker matches.

This proposal isn't entirely selfish. There are plenty of tier 3 players better than me, this would keep brand new tier 5's from fighting them. Even if there are only 3 tiers, the middle guys never have to go against the supper l33t or absolute noobs.

#31 adamts01

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 3,417 posts
  • LocationPhilippines

Posted 05 February 2016 - 12:22 AM

View PostTier 1 Smurf, on 04 February 2016 - 08:03 PM, said:


By further diluting the competition in T1 it may make shorter queues yeah, but really we haven't a clue how the MM goes about releasing values for cross tier matching, the damn thing might wait 5secs before widing the player delta for crossing tiers.
I've had many instant drops against very new players. PGI said there's a 2 minute wait till all gates are open, but that 2 minutes could be for 1 player and everyone else is instantly grabbed in to his game. I have a feeling that's what happens more often than not.

#32 MrMadguy

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,300 posts

Posted 05 February 2016 - 12:29 AM

Currently system works the way, that your average match should be the following in order to stay at the same PSR rating:
Posted Image
Posted Image
And the best match, you can have in this case, is following:
Posted Image
Posted Image
I still can't make a screenshot of ~325MS match - I couldn't even realize, that it would be that hard to earn such match score. I'm currently at too high PSR level and can't drop due to bubble rule - I have to earn <100MS avg match score in order to drop - so I'm currently at endless lose/stomp streak.

Here is the real fix, that should be applied to current PSR system. We should get rid of PSR bias towards increasing: win MS thresholds should be risen, so average match score, you have to earn in order to stay at the same level, will rise too.
Posted Image

Edited by MrMadguy, 05 February 2016 - 12:31 AM.


#33 Alistair Winter

    Member

  • PipPipPipPipPipPipPipPipPipPipPip
  • Storm
  • Storm
  • 10,823 posts
  • LocationBergen, Norway, FRR

Posted 05 February 2016 - 12:30 AM

View PostKyynele, on 04 February 2016 - 07:32 PM, said:


Well, at least in group queue pure T1 matches do happen.

Seems to be very rare in solo queue. Or, there are so many not-very-goods in T1 already that it's impossible to tell.

See, the group queue is another reason why the current PSR system is problematic. Because PSR, like elo, is hugely dependent on winning. And if you're with a decent group and you have fairly average skills and a great drop commander, then you can get crazy good results in the group queue. People end up with 3.0 or 5.0 WLR or whatever. That's basically impossible in the solo queue, unless you're in the top 0.5 percentile or something like that.

Of course, the point of the PSR is to ensure good matches and I guess pure T1 matches in the group queue are pretty good anyway during NA prime time, due to the presence of organized groups with drop commanders.

#34 DAYLEET

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 4,316 posts
  • LocationLinoleum.

Posted 05 February 2016 - 12:34 AM

View PostTier 1 Smurf, on 04 February 2016 - 08:03 PM, said:

By further diluting the competition in T1 it may make shorter queues yeah, but really we haven't a clue how the MM goes about releasing values for cross tier matching, the damn thing might wait 5secs before widing the player delta for crossing tiers.

We DO know and i said it already in this thread.


View PostDAYLEET, on 04 February 2016 - 05:20 PM, said:

Last townhall Russ mentioned how it really works, a T1 can wait 1minutes and be put with 23 other T2 who all waited 2 seconds because the waiting and releasing is per person not per tier.

Tier releasing is per player, in under 60seconds the match maker wont put me in other tier but LOL it will put other tier who already waited with me... Townhall around 2h55.

Edited by DAYLEET, 05 February 2016 - 12:36 AM.


#35 Soultraxx

    Member

  • PipPipPipPipPipPip
  • 427 posts
  • LocationEngland

Posted 05 February 2016 - 12:37 AM

View PostMonkey Lover, on 04 February 2016 - 04:01 PM, said:


My understanding is you drop down to tier2 levels at 1 minute and tier3 after 2. I can say i often have to wait more than 3-4 minute.


I never knew this. Considering I rarely wait more than a minute (normally 30 secs or less) then Im dropping mostly with people of my Tier I guess?

#36 adamts01

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 3,417 posts
  • LocationPhilippines

Posted 05 February 2016 - 12:40 AM

View PostAlistair Winter, on 05 February 2016 - 12:30 AM, said:

See, the group queue is another reason why the current PSR system is problematic.
Very true. And I don't have a solution. I pretty much only play in group que now because solo que is such a mess of new players and YOLOs. It's the same problem of tonnage mattering in CW but only the weight class mattering in solo. You can't balance both ways.

View PostMrMadguy, on 05 February 2016 - 12:29 AM, said:

Here is the real fix, that should be applied to current PSR system.
I agree PSR has flaws, mostly that players can't move down, as in your case. But this bypasses all of that. If someone is doing poorly consistently in tier 1, they'll be pushed down by someone who does consistently well in tier 2 who moves up.

Edited by adamts01, 05 February 2016 - 12:40 AM.


#37 DAYLEET

    Member

  • PipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 4,316 posts
  • LocationLinoleum.

Posted 05 February 2016 - 12:42 AM

View PostSoultraxx, on 05 February 2016 - 12:37 AM, said:

I never knew this. Considering I rarely wait more than a minute (normally 30 secs or less) then Im dropping mostly with people of my Tier I guess?

Group get released earlier than solo though, solo was 60 and group was 30.

#38 adamts01

    Member

  • PipPipPipPipPipPipPipPipPip
  • Death Star
  • 3,417 posts
  • LocationPhilippines

Posted 05 February 2016 - 12:43 AM

View PostSoultraxx, on 05 February 2016 - 12:37 AM, said:



I never knew this. Considering I rarely wait more than a minute (normally 30 secs or less) then Im dropping mostly with people of my Tier I guess?
No. Dayleet and I both mentioned how it works. You can instant drop with someone 2 tiers lower than you who has been waiting 2 minutes and needs a slot filled. From what I observe, this is a constant cycle and all that happens to me.

#39 MrMadguy

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,300 posts

Posted 05 February 2016 - 01:10 AM

View Postadamts01, on 05 February 2016 - 12:40 AM, said:

I agree PSR has flaws, mostly that players can't move down, as in your case. But this bypasses all of that. If someone is doing poorly consistently in tier 1, they'll be pushed down by someone who does consistently well in tier 2 who moves up.

Tiers separate players by SKILL, not by number of players, because it's the purpose of MM - to separate players by skill.

About my fix. May be ~325MS average match - is a little bit too high. May be 250-275MS will be enough, but you see the idea - players should be allowed to drop at win. Win thresholds should be risen at least by 100.

Edited by MrMadguy, 05 February 2016 - 01:31 AM.


#40 XX Sulla XX

    Member

  • PipPipPipPipPipPipPipPipPip
  • Ace Of Spades
  • 3,094 posts

Posted 05 February 2016 - 01:32 AM

View PostAlistair Winter, on 04 February 2016 - 02:30 PM, said:

Do pure Tier 1 matches ever happen? Any Tier 1 players who can confirm or deny?
Yes





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users