Jump to content

First Few Matches Extreme Points? What?


43 replies to this topic

#21 RickySpanish

    Member

  • PipPipPipPipPipPipPipPipPip
  • Veteran Founder
  • Veteran Founder
  • 3,510 posts
  • LocationWubbing your comrades

Posted 21 July 2018 - 01:22 AM

View PostSurn, on 21 July 2018 - 01:10 AM, said:

Ash, white knighting...

the real issue here is that Elo is crap as a ranking system because MechWarrior is not chess! Players win because they have superior builds just as much as superior skills.


One ought to consider their activity in the Mech-bay as part of the match, surely? Elo works exceptionally well in a duelist situation where the outcome of each match is entirely dependent on the actions of the two competitors. It breaks down a little in team based games, but I think it's a pretty good system for Solaris 7.

#22 Mechwarrior8985372

    Member

  • PipPip
  • 45 posts

Posted 21 July 2018 - 04:16 AM

You get bonus SSR for your 1st 10 games, you have to play 25 minimum to qualify for rewards, if someone wins 10 games who is not supposed to have won those 10 games, they are going to lose the next 15 and drop back down, or get no more SSR for wins vs bad players, and good players will have risen above the SSR for having won 1st 10 games. The season is 3 months, it will balance out.

This helps distribute the playerbase to their correct SSR fast, which is great for lower skilled players to not get demoralized by constant losses and is very good when there is 20 people in a Queue, which happens when RJBASS or nuttyrat or other streamers stream solaris 1v1, people get good, close matchups.

You do not get bonus SSR from people still in their 1st 10 games for beating them, you will get your normal +3 or whatever while they get -188 or whatever it is, For PGI, what they have implemented is decent.

It would be nice if we had more people playing, or a "multiqueue system" where you tick what divisions your willing to queue into, so we get more matches more easily with more close fights, at the moment 7 queues is too many, hopefully MW5 will bring in more people to MWO.

It would be nice if PGI had communicated this new Season 2 10 placement matches system to us, would have helped ease some of this concern by clarification.

It would be nice if after the Tutorial there was a mechlab / queue types tutorial in game that you had to watch b4 you could play.

Edited by ReichKaiser, 21 July 2018 - 04:17 AM.


#23 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 21 July 2018 - 05:18 AM

View PostSurn, on 21 July 2018 - 01:10 AM, said:

Ash, white knighting...

the real issue here is that Elo is crap as a ranking system because MechWarrior is not chess! Players win because they have superior builds just as much as superior skills.


Me, white knight? Please. Spare me the woeful song.

So in all your wisdom - what would be a better system exactly? Spell it out. I mean you complain about people legging and what not, yet you let yourself. You are hardly the source of consistency.

View PostReichKaiser, on 21 July 2018 - 04:16 AM, said:

You get bonus SSR for your 1st 10 games, you have to play 25 minimum to qualify for rewards, if someone wins 10 games who is not supposed to have won those 10 games, they are going to lose the next 15 and drop back down


I just played 2 games. I was 8-0 in a particular Div.

I earnt +7 ELO from those 2 games.

So in furtherance to my earlier statement of 10 games. It appears that it is 'hard capped' to 'X' ELO. Now I don't know what that ELO is exactly but I would say based on my ELO that is is hard capped at 2100 ELO.

So either 10 games or 2100 ELO... Whichever comes first.

This then - is an even MORE reasonable system than I first thought. The fact is hard a cap limit.

Anyway the tin-foil-hat-crew will still tin-foil-hat. So let them be I guess.

#24 Edustaja

    Member

  • PipPipPipPipPipPipPip
  • 730 posts
  • LocationFinland

Posted 21 July 2018 - 05:40 AM

View Postjustcallme A S H, on 21 July 2018 - 05:18 AM, said:


Me, white knight? Please. Spare me the woeful song.

So in all your wisdom - what would be a better system exactly? Spell it out. I mean you complain about people legging and what not, yet you let yourself. You are hardly the source of consistency.



I just played 2 games. I was 8-0 in a particular Div.

I earnt +7 ELO from those 2 games.

So in furtherance to my earlier statement of 10 games. It appears that it is 'hard capped' to 'X' ELO. Now I don't know what that ELO is exactly but I would say based on my ELO that is is hard capped at 2100 ELO.


I'm curious, what was the the SSR score for your opponent for your last two matches?

#25 Surn

    Member

  • PipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 1,073 posts
  • Twitter: Link
  • LocationSan Diego

Posted 21 July 2018 - 06:41 AM

View PostRickySpanish, on 21 July 2018 - 01:22 AM, said:

One ought to consider their activity in the Mech-bay as part of the match, surely? Elo works exceptionally well in a duelist situation where the outcome of each match is entirely dependent on the actions of the two competitors. It breaks down a little in team based games, but I think it's a pretty good system for Solaris 7.


I completely disagree, the entire math basis of Elo is predicated on 1v1 with a normal distribution of opponent skill levels.

It is far too simplistic for MechWarrior, or any team game.

#26 Surn

    Member

  • PipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 1,073 posts
  • Twitter: Link
  • LocationSan Diego

Posted 21 July 2018 - 06:55 AM

View Postjustcallme A S H, on 21 July 2018 - 05:18 AM, said:


Me, white knight? Please. Spare me the woeful song.

So in all your wisdom - what would be a better system exactly? Spell it out. I mean you complain about people legging and what not, yet you let yourself. You are hardly the source of consistency.




I created a fully automated MechWarrior Registry with both planetary resource based combat and a bounty system.

Because pilot skill is a key component in Battletech, I worked with a mathematician and we created the Ranked Percentage Index based on Strength of Schedule. It is now used in college basketball as the starting point for created the March Madness bracket in the NCAA.

It was created for this game, MechWarrior!

RPI is far superior for any game with a team or with a variable like equipment used (mech).

The calculations are still posted on mechreg.org and if they need sql stored procedures that calculate team and match, I have them...

because we used to have a server, when you played on that server, it would automatically update your stats, planet stats, essentially Faction or quickplay or Solaris.


I do sometimes leg, because sometimes it is required...however I do not enter matches with a legging optimized mech and strategy.

#27 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 21 July 2018 - 07:32 AM

View PostEdustaja, on 21 July 2018 - 05:40 AM, said:


I'm curious, what was the the SSR score for your opponent for your last two matches?


+4 and +7 ELO after 2100 ELO reached, for a win.

So for the TOP TIER players, it is working, as intended.

Edited by justcallme A S H, 21 July 2018 - 07:35 AM.


#28 Scout Derek

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Infernal
  • The Infernal
  • 8,015 posts
  • Google+: Link
  • Twitch: Link
  • LocationSomewhere where you'll probably never go to

Posted 22 July 2018 - 07:57 PM

Whoever thought it was a good idea to make elo gain be so high for the first few initial matches needs to stop. It's pretty bad, a guy with 10 wins and 1 loss being able to get top 10 so easily?

That's just sad man, real sad. I remember hearing some talk about elo decay and wouldn't mind it.

Play one match, daily. or even two for that matter to keep position.

#29 Marius Evander

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,113 posts

Posted 22 July 2018 - 09:50 PM

View PostScout Derek, on 22 July 2018 - 07:57 PM, said:

Whoever thought it was a good idea to make elo gain be so high for the first few initial matches needs to stop. It's pretty bad, a guy with 10 wins and 1 loss being able to get top 10 so easily?

That's just sad man, real sad. I remember hearing some talk about elo decay and wouldn't mind it.

Play one match, daily. or even two for that matter to keep position.


Have to play 25 to get rewards

#30 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 22 July 2018 - 09:57 PM

Yeah once people play 15-20, have a few big losses, it does even things up.

While I agree the system, or rather bonuses, are a big high early. The premise behind it is actually quite good, just lacking in execution department a little.

Over the course of 25 games though it will even out IMO.

#31 Scout Derek

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • The Infernal
  • The Infernal
  • 8,015 posts
  • Google+: Link
  • Twitch: Link
  • LocationSomewhere where you'll probably never go to

Posted 22 July 2018 - 10:05 PM

View PostCadoazreal, on 22 July 2018 - 09:50 PM, said:


Have to play 25 to get rewards








View Postjustcallme A S H, on 22 July 2018 - 09:57 PM, said:

Yeah once people play 15-20, have a few big losses, it does even things up.

While I agree the system, or rather bonuses, are a big high early. The premise behind it is actually quite good, just lacking in execution department a little.

Over the course of 25 games though it will even out IMO.


20-25 is a good number, and I would most certainly have it at that instead of what it currently is, hell, anything is better than how it is right now.

(grinding up that +1 elo man. woooooooo Posted Image )

EDIT; the quoting system did something funky.

Edited by Scout Derek, 22 July 2018 - 10:07 PM.


#32 Dee Eight

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Big Daddy
  • Big Daddy
  • 6,271 posts

Posted 22 July 2018 - 10:56 PM

A few minutes ago I was 147th, 26 wins 8 losses. I started in the division with a dragon 1c design I wanted to try and did..average...8 wins 7 deaths and dropped into the thousand rank. Then I switched to a zeus I like. 9 win streak and then died. Then as I was writing this out, my second 9 win streak again was ended when my mech blew itself up without the override being on. Thanks PGI. So 26 & 9 now and dropped to 164th. My SSR is 1745. It took 18 wins, many against higher SSR opponents to claw my way to that. Meanwhile folks who put up that initial burst of wins and quit before a loss get to sit top spot on the leaderboard till the end of the season probably because of how PGI has altered the scoring formula to benefit such a strategy, and quite frankly that's not going to keep players in the queue once the event ends. And I know people keep repeating the "have to play 25 to get rewards" but that's assuming PGI doesn't alter that figure, just as they altered the scoring formula for the ELO change already.

Edit PS : I just won a match against a 100 elo lower rank where I removed the arms and leg from that dufus "meta" jager build with the xl and then popped the guy's head and my elo went down 7 pts... great new system PGI.

Edited by Dee Eight, 22 July 2018 - 11:03 PM.


#33 justcallme A S H

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • CS 2020 Referee
  • CS 2020 Referee
  • 8,987 posts
  • LocationMelbourne, AU

Posted 22 July 2018 - 11:24 PM

View PostDee Eight, on 22 July 2018 - 10:56 PM, said:

sit top spot on the leaderboard till the end of the season probably because of how PGI has altered the scoring formula to benefit such a strategy,


Except the top ELO from last season was around 2600-2700 ELO or around that IIRC. So the reality is, in a months time, that won't be the case at all of these people being at the top.

#34 Marius Evander

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,113 posts

Posted 23 July 2018 - 12:30 AM

It could certainly have been better but its allright, I just dont get how I went 7 wins 3 losses in one division and ended up at 1570 after 10 games, thats 4 more wins than losses with the bonus's ........ go figure....

#35 Dee Eight

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Big Daddy
  • Big Daddy
  • 6,271 posts

Posted 23 July 2018 - 12:46 AM

Probably same logic as my defeating a lower SSR cost me ELO points for some PGI reason.

#36 UnofficialOperator

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,493 posts
  • LocationIn your head

Posted 23 July 2018 - 08:31 PM

View PostDee Eight, on 22 July 2018 - 10:56 PM, said:

Edit PS : I just won a match against a 100 elo lower rank where I removed the arms and leg from that dufus "meta" jager build with the xl and then popped the guy's head and my elo went down 7 pts... great new system PGI.


Wow wtf?

#37 Horseman

    Member

  • PipPipPipPipPipPipPipPipPip
  • The Mercenary
  • The Mercenary
  • 4,694 posts
  • LocationPoland

Posted 24 July 2018 - 02:55 AM

View PostDee Eight, on 22 July 2018 - 10:56 PM, said:

Edit PS : I just won a match against a 100 elo lower rank where I removed the arms and leg from that dufus "meta" jager build with the xl and then popped the guy's head and my elo went down 7 pts... great new system PGI.
Sounds like someone forgot to clamp the values.

View PostCadoazreal, on 23 July 2018 - 12:30 AM, said:

It could certainly have been better but its allright, I just dont get how I went 7 wins 3 losses in one division and ended up at 1570 after 10 games, thats 4 more wins than losses with the bonus's ........ go figure....

The change seems to be based on the SSR difference and outcome of the match. Losing to someone ranked much lower than yourself hits you harder than losing to someone equal, and losing to someone ranked far above you hits you much less than a loss against an equal ranked pilot..

#38 Rick T Dangerous

    Member

  • PipPipPipPipPipPip
  • Bad Company
  • 354 posts
  • LocationExactly above Earth's center

Posted 24 July 2018 - 03:12 PM

Would be nice if the ELO thingie could take into account what build a player is running and if the player is always running the same build or changes mechs and builds from time to time.

Player uses meta cheese and wins: no gain
Player uses meta cheese and looses: big loss
Player uses non meta build for the 1st time and wins: big gain
Player uses non meta build for the 1st time and looses: no loss
Player uses the same non meta build multiple times and wins: average gain
Player uses the same non meta build multiple times and looses: average loss

And maybe even the number of spectators and if they stay until the end should have an effect on the SSR gain/loss. When I spectate a match and see two Annis doing the turret thing I leave that boring parody of a fight. If the fight is not entartaining, the pilots don't deserve a reward. This is supposed to be Solaris, isn't it?

#39 Surn

    Member

  • PipPipPipPipPipPipPipPip
  • The God of Death
  • The God of Death
  • 1,073 posts
  • Twitter: Link
  • LocationSan Diego

Posted 24 July 2018 - 04:04 PM

View PostRick T Dangerous, on 24 July 2018 - 03:12 PM, said:

Would be nice if the ELO thingie could take into account what build a player is running and if the player is always running the same build or changes mechs and builds from time to time.

Player uses meta cheese and wins: no gain
Player uses meta cheese and looses: big loss
Player uses non meta build for the 1st time and wins: big gain
Player uses non meta build for the 1st time and looses: no loss
Player uses the same non meta build multiple times and wins: average gain
Player uses the same non meta build multiple times and looses: average loss

And maybe even the number of spectators and if they stay until the end should have an effect on the SSR gain/loss. When I spectate a match and see two Annis doing the turret thing I leave that boring parody of a fight. If the fight is not entartaining, the pilots don't deserve a reward. This is supposed to be Solaris, isn't it?


that is RPI in a nutshell, it takes into account more factors

#40 Marius Evander

    Member

  • PipPipPipPipPipPipPipPip
  • Ace Of Spades
  • Ace Of Spades
  • 1,113 posts

Posted 24 July 2018 - 05:01 PM

View PostRick T Dangerous, on 24 July 2018 - 03:12 PM, said:

Would be nice if the ELO thingie could take into account what build a player is running and if the player is always running the same build or changes mechs and builds from time to time.

Player uses meta cheese and wins: no gain
Player uses meta cheese and looses: big loss
Player uses non meta build for the 1st time and wins: big gain
Player uses non meta build for the 1st time and looses: no loss
Player uses the same non meta build multiple times and wins: average gain
Player uses the same non meta build multiple times and looses: average loss

And maybe even the number of spectators and if they stay until the end should have an effect on the SSR gain/loss. When I spectate a match and see two Annis doing the turret thing I leave that boring parody of a fight. If the fight is not entartaining, the pilots don't deserve a reward. This is supposed to be Solaris, isn't it?


Terrible idea, who decides what meta is, it takes PGI 6 months to identify a mech that's overperforming and you think they can do it for solaris weekly ?





1 user(s) are reading this topic

0 members, 1 guests, 0 anonymous users