Jump to content

My Thoughts On Elo (Yep, Another One)


156 replies to this topic

#101 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 26 February 2013 - 08:13 AM

View PostJoseph Mallan, on 26 February 2013 - 08:09 AM, said:

No the best players don't have that right. The best players in the game should expect to be in the top 3 places at the end of every match. Win or lose. Individual performance does not insure team victory.



Nope.

There have been numerous times that I have ended up with the highest match score, both before and after ELO. I don't think I'm that good.

I'm certainly competent, but I jump mechs (I have 43 now) like every two games, so I rarely get to the point if the dedicated jump sniper (I hate sniping) or expert marksman.

I can wing ballistics on the run with the the best of them though, ACs are pretty constant on nearly all of my mechs that can support them.

#102 PhoenixFire55

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 5,725 posts
  • LocationSt.Petersburg / Outreach

Posted 26 February 2013 - 08:14 AM

View PostJoseph Mallan, on 26 February 2013 - 08:09 AM, said:

No the best players don't have that right. The best players in the game should expect to be in the top 3 places at the end of every match. Win or lose. Individual performance does not insure team victory.


It does not of course. But it tips the balance in favor of the team that has a better player. IMO the best player in game can tip it enough to win 80% of his games if he is dropped with 15 random people. As for the top 3 places that is exactly what I am saying as well.

#103 armyof1

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • 1,770 posts

Posted 26 February 2013 - 08:16 AM

View PostPhoenixFire55, on 26 February 2013 - 05:32 AM, said:


Just as you are tied of 8:0, 0:8 stomps now you will eventually grow tired of 8:7, 7:8 nailbiters. ELO has good motivations behind it I can't argue with that, but it is very wrong in some of its core principles.

Well I'll bet I won't get tired of even games compared to all-out stomps. Really what excitement would I get out melting virtually defenseless guys that have hardly grasped the concept of torso twists? Because it'll make me feel better than average and net me more cbills? It's the same thing as if I'd go and beat a little kid in a sport I've played a lot longer than the kid has, would that give me any real satisfaction? Of course not.

#104 Steven Dixon

    Member

  • PipPipPipPipPipPipPip
  • Knight Errant
  • 621 posts

Posted 26 February 2013 - 08:24 AM

As a solo player I've been having far more fun after Elo. Sometimes their is a pugstomp one way or another (nothing is perfect), but the majority of the time they are epic, close matches. I've had many matches that are right down to the wire. It does make me wonder in cases like this if its simply a case of some people having bugs with the system (perhaps their Elo is in a strange place) of if they are really just mad they can't pugstomp as easily and just don't want to admit it.

#105 PhoenixFire55

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 5,725 posts
  • LocationSt.Petersburg / Outreach

Posted 26 February 2013 - 08:26 AM

View PostRocketDog, on 26 February 2013 - 08:13 AM, said:


Whoa there! That's a big assumption.

A far better way to ensure the longevity of the game would be to make sure that even new and/or not-very-good players (I was the former, I'm now the latter) can still have fun with the game and don't get turned away by crushing defeat after crushing defeat.


In a completely random drop a new player has an equal chance to end up on winning team and losing team. He will win 50% of his games and he will lose 50% of his games. If he is good enough to tip the balance in favor of his team he will win more and visa versa. It is as it should be.

Just as I said I believe that haveing 50% wins all the time no matter what you do (thats what ELO is) is far less fun for everybody. IMO the only reason to defend current ELO system is when you know you will be winning far less matches without it then you will with it.

From what I've seen its usually really good players who can't have fun when they are losing. Smth to do with being overly competitive I think. If you are just like myself willing to admit that you aren't anywhere near really skilled players in this game then you should only worry about haveing fun. From my experience ELO is less fun because you always know what will happen before it happens.

There is nothing wrong with occasionaly haveing a really bad player or a really good player on your team as well.

#106 Thirdstar

    Member

  • PipPipPipPipPipPipPipPipPip
  • 2,728 posts
  • LocationIndia

Posted 26 February 2013 - 08:36 AM

View PostPhoenixFire55, on 26 February 2013 - 08:26 AM, said:

Just as I said I believe that haveing 50% wins all the time no matter what you do (thats what ELO is) is far less fun for everybody.


You're still saying stuff like this with a straight face. Who died and made you FUN emperor?

#107 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 26 February 2013 - 08:45 AM

View PostYokaiko, on 26 February 2013 - 08:13 AM, said:



Nope.

There have been numerous times that I have ended up with the highest match score, both before and after ELO. I don't think I'm that good.

I'm certainly competent, but I jump mechs (I have 43 now) like every two games, so I rarely get to the point if the dedicated jump sniper (I hate sniping) or expert marksman.

I can wing ballistics on the run with the the best of them though, ACs are pretty constant on nearly all of my mechs that can support them.

Same here, a hot match. But we are not those top ten players. I myself will get a 1,000-1,400 point game or I lead the board in damage and xp from the losing side with 700+ point games. Elo should track K/D not W/L. Winning is dependent on the teams performance not mine specifically. I get 15% more wins with a dedicated team, but K/D is of my doing. Maybe add Assists into the W/L and you would get a better individual performance for Elo.

Edited by Joseph Mallan, 26 February 2013 - 08:47 AM.


#108 PhoenixFire55

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 5,725 posts
  • LocationSt.Petersburg / Outreach

Posted 26 February 2013 - 08:48 AM

View Postarmyof1, on 26 February 2013 - 08:16 AM, said:

Well I'll bet I won't get tired of even games compared to all-out stomps. Really what excitement would I get out melting virtually defenseless guys that have hardly grasped the concept of torso twists? Because it'll make me feel better than average and net me more cbills? It's the same thing as if I'd go and beat a little kid in a sport I've played a lot longer than the kid has, would that give me any real satisfaction? Of course not.


Chances that you stomp somebody are same as chances that you will get stomped. The need for better interactive in-game tutorial for MWO is obvious, but it is a different matter. What I'm saying is that without ELO you will have games that end 8:0, 8:2, 8:4, 8:6, 6:8, 4:8, 2:8 and 0:8 with nearly same probability, while with ELO you will always have 6:8 and 8:6 games. I'm not sure about you, but to me it seems that no-ELO is way more fun than ELO.

True satisfaction is when you have a team where nearly everyone is a new player, you talk to them, make them communicate, help each other, and even if you lose that game being stomped you know that those guys learn from it and will strife to become better and help others do same.

#109 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 26 February 2013 - 08:49 AM

View PostJoseph Mallan, on 26 February 2013 - 08:45 AM, said:

I get 15% more wins with a dedicated team, but K/D is of my doing. Maybe add Assists into the W/L and you would get a better individual performance for Elo.


I dunno.

I had a 2.43:1 W/L before the tourney, and I pulled 2:1 throughout solo dropping the entire weekend (I sync'd like the first three with a couple corpies for kicks...and ended up shooting at them). That is about a 50/50 split solo and 2/3/4man dropping.


I haven't watched the numbers enough since OB to really tell you the difference, but I don't think its that great (at least in my experience)

....some of that is just having people to ***** at in a game without having to type it, I know I can be an *** at times.

#110 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 26 February 2013 - 08:53 AM

View PostYokaiko, on 26 February 2013 - 08:49 AM, said:


I dunno.

I had a 2.43:1 W/L before the tourney, and I pulled 2:1 throughout solo dropping the entire weekend (I sync'd like the first three with a couple corpies for kicks...and ended up shooting at them). That is about a 50/50 split solo and 2/3/4man dropping.


I haven't watched the numbers enough since OB to really tell you the difference, but I don't think its that great (at least in my experience)

....some of that is just having people to ***** at in a game without having to type it, I know I can be an *** at times.

You had an 18% drop in wins over the weekend. :)

#111 PhoenixFire55

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 5,725 posts
  • LocationSt.Petersburg / Outreach

Posted 26 February 2013 - 08:54 AM

View PostThirdstar, on 26 February 2013 - 08:36 AM, said:

You're still saying stuff like this with a straight face. Who died and made you FUN emperor?


View PostPhoenixFire55, on 26 February 2013 - 08:42 AM, said:

You keep twisting what I have said and keep making it personal. You are not making your point any stronger doing that.


#112 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 26 February 2013 - 08:58 AM

View PostJoseph Mallan, on 26 February 2013 - 08:53 AM, said:

You had an 18% drop in wins over the weekend. :)



True, but that was also grinding off games.

If it looked bad charge in kill something, and die, exit and do the next drop for another weight class (I spent a LOT of the weekend in the top 50, up to 6th at one point).

So pulling 2:1 was surprising when I saw the individual class stats. Mediums where my worst with a 50/50, W/L However, I pulled over 5/1 in in assaults and I never drove a D-DC


Hold on I'll get the numbers.

So

Lights rank 505 score 39 w24/l10 so 2.2:1 w/l
Mediums rank 467 score 47 w20/l20 (note that was while xping my trebs)
Heavier rank 373 score 61 w36/l23 w/l 1.53:1
Assaults rank 426 score 61 w34/l17 w/l 2:1

So ok, more like 1.7:1....and that was enough for 66th overall.

Like I said I actually pushed my KDR up, which is no small matter from 2,700 kills.

Also like I said I never really watched the delta for win rates between grouped and ungrouped.

Edited by Yokaiko, 26 February 2013 - 09:06 AM.


#113 armyof1

    Member

  • PipPipPipPipPipPipPipPip
  • Bad Company
  • 1,770 posts

Posted 26 February 2013 - 08:58 AM

View PostPhoenixFire55, on 26 February 2013 - 08:48 AM, said:


Chances that you stomp somebody are same as chances that you will get stomped. The need for better interactive in-game tutorial for MWO is obvious, but it is a different matter. What I'm saying is that without ELO you will have games that end 8:0, 8:2, 8:4, 8:6, 6:8, 4:8, 2:8 and 0:8 with nearly same probability, while with ELO you will always have 6:8 and 8:6 games. I'm not sure about you, but to me it seems that no-ELO is way more fun than ELO.

True satisfaction is when you have a team where nearly everyone is a new player, you talk to them, make them communicate, help each other, and even if you lose that game being stomped you know that those guys learn from it and will strife to become better and help others do same.


Well then we have totally opposite ideas of what's fun. The reason ELO was needed is simply because the matches were way too uneven. I used to see more 8:0 matches than I saw 8:4 which is just ridiculous if you are looking for competitive matches. How much fun is it for a new player to just get stomped within 2 minutes or even when they win hardly do anything because the other guys did all the killing before you even had time to get into it? Honestly the only people that had the most fun without ELO was guys that loved to feel superior and could wreck a lot of newbies to feel that way while getting all the exp and cbills while doing it. That's not fun, that's just sad. I really don't understand how you could think it would be fun for new players to get stomped on a regular basis, more likely they'll try a few games, get tired of it and never play again.

#114 Mister Blastman

    Member

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

Posted 26 February 2013 - 09:03 AM

I'll admit there is a problem with ELO--not sure what it is or how it decides who is on what team... but there is one.

Take one match last night my group of four had. I played for several hours and rotated out a few teammates periodically in our 4-man (so our ELO was never constant) but one thing was constant... the 0-100 damage players on our team. It wasn't every drop, but, it was happening frequently enough that I was scratching my head wondering what these players were doing grouped with us. So on one insightful match...

My PUG players, well, 3 of the 4 PUG players decided... Hey, lets go cave! ... On Conquest. On Frozen City. Yeah. I knew we had some winners because one guy kept spamming "OOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOOO" into chat repeatedly. They proceeded to go through the cave all the way to the enemy starting point it looked like... I think. Not sure, I lost track of them.

The rest of us went to my designated point and then we /were/ going to go to the next point... however... We were outnumbered 8 to 5 given our PUGs wandering off. So we waited. We had a missile heavy boat on our team so the rest of us spotted for him to soften the enemy up. We picked off stragglers as best we could to even the odds.

Eventually, the five of us prevailed and smashed the enemy to pieces. At the end of the match... Our PUGs were dead. They managed to do... hmm... 2 and 11 damage for two of them... The other was less than 100. Yeah. I don't know what they did or where they went but they died without doing anything.

Long story short we adapted. Had the other team been a co-ordinated sync drop it might not have gone as well. I liken these PUG players as playing with a handicap. Maybe that is what the matchmaker is doing? Using an average ELO score and balancing out the very highs with very lows? Who knows. It is very frustrating to say the least sometimes. There were other drops last night where we didn't fare so well due to our PUG players.

#115 80Bit

    Member

  • PipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 555 posts
  • LocationIdaho

Posted 26 February 2013 - 09:06 AM

You seem to have come down with a nasty case of elitism.

First, you did not actually point out how Elo has changed your experience Based on your complaints however, what I can infer is that before Elo you and 1-3 friends would drop in groups and PUG stomp and feel like you were God's gift to MWO, and that was fun for you. Now you are getting your butt kicked at least 50% of the time because Elo has forced you to drop against equally qualified players, and you don't like that.

Is that it, or could you be more specific on how Elo has made it impossible for you and your friends to have an impact on a match?

#116 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 26 February 2013 - 09:10 AM

View Post80Bit, on 26 February 2013 - 09:06 AM, said:



Is that it, or could you be more specific on how Elo has made it impossible for you and your friends to have an impact on a match?



No way.

Even if all 16 players in a match are of the same ELO bracket ( and there is no evidence that is happening like ever) one 4 man that is working well together will trump 8 guys typing.....if they are even trying to.

#117 PhoenixFire55

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 5,725 posts
  • LocationSt.Petersburg / Outreach

Posted 26 February 2013 - 09:12 AM

View Postarmyof1, on 26 February 2013 - 08:58 AM, said:


Well then we have totally opposite ideas of what's fun. The reason ELO was needed is simply because the matches were way too uneven. I used to see more 8:0 matches than I saw 8:4 which is just ridiculous if you are looking for competitive matches. How much fun is it for a new player to just get stomped within 2 minutes or even when they win hardly do anything because the other guys did all the killing before you even had time to get into it? Honestly the only people that had the most fun without ELO was guys that loved to feel superior and could wreck a lot of newbies to feel that way while getting all the exp and cbills while doing it. That's not fun, that's just sad. I really don't understand how you could think it would be fun for new players to get stomped on a regular basis, more likely they'll try a few games, get tired of it and never play again.


I don't really understand why ALL matches should be even. Are all football games you watch on TV even? Football is a team game. A superior team will nearly always win. An inferior team should work to become better and win or accept that it can not become better and not be bothered with losing. Would you like to see your favorite football team to be 'balanced by ELO', just so it loses its edge and begins to win 50% and lose 50% matches at random, when a match outcome will be decided pretty much by a flip of a coin rather then their skill they worked so hard to improve?

The other thing is that ELO made games even worse balanced. I can't say about your games, but in my games all I see is terrible balance in terms of mech types as well as players skills. The fact that somebody uses a 3L Raven rather then a Jenner to make his ELO sky high doesn't mean he is a better player then a guy in a Jenner.

A new player will get stomped by ELO = 1300 guys when he starts anyway, but if after that he'll be stuck in ELO hellhole with griefers, TK'ers and afk'ers I believe he'll abandom this game much sooner then he would have had with just random matchmaker.

Also, if you were looking for competitive matches why aren't you doing 8 vs 8?

#118 FerretGR

    Member

  • PipPipPipPipPipPipPipPip
  • Legendary Founder
  • Legendary Founder
  • 1,445 posts

Posted 26 February 2013 - 09:17 AM

View PostPhoenixFire55, on 26 February 2013 - 06:43 AM, said:

Best players in game have all rights to expect that 8 out of 10 games will end up in their favor, worst players in game have no rights to expect that that less then 8 out of 10 games will end up not in their favor. With ELO both types will have 5 out of 10 games ending up in their favor. I fail to see how ELO is fair. Just as I fail to see how people who bring up arguments to show that ELO isn't fair are bad.

What ELO does is encouraging noobiness, no matter how bad you play you will still be winning 50% of your games and you will still be getting same c-bills and exp as everybody else.


Your entire argument, as far as I can tell, is based on this premise: that "bad" players shouldn't be earning as much as "good" players. I want you to tell us WHY that's the case. WHY should good players be winning 80% of their matches? WHY should bad players be losing 80%? There's nothing inherent that would necessitate this. The developers inserted a system into the game that actively fights this, so we can assume given this that the developers prioritize even matches over increased earnings for pros. The system is working as intended. You'd have the system changed.

You are prioritizing earnings over competition. Simply put, I want you to explain why increased c-bill earnings take priority for you over competitive matches. You're hung up on money, a minor part of this game, when you should be concentrating on gameplay and in-game enjoyment, which is everything.

This is the way competition works, IMHO. It's not competition if you're constantly facing opposition that can't compete. Consider major league baseball: last year, the top team in the league had a .6 W/L. Why do you suppose that's the case? Well, the long and the short of it is that the league system is set up in much the same way as Elo: if you're a good enough player, you're in the major leagues, facing other major leaguers. If you can't hack it, you're playing triple A ball in Florida or something. But you're playing with other AAA-level players and the competition is still very even.

The system you'd like to have would have the Washington Nationals playing against the Butler, PA little leaguers. That's no fun for the little leaguers and it's no fun for the Nationals.


View PostPhoenixFire55, on 26 February 2013 - 09:12 AM, said:

I don't really understand why ALL matches should be even. Are all football games you watch on TV even? Football is a team game. A superior team will nearly always win. An inferior team should work to become better and win or accept that it can not become better and not be bothered with losing. Would you like to see your favorite football team to be 'balanced by ELO', just so it loses its edge and begins to win 50% and lose 50% matches at random, when a match outcome will be decided pretty much by a flip of a coin rather then their skill they worked so hard to improve?


Awesome, you went to the sports analogy without my prompting! You say "A superior team will nearly always win", and yet, that's not the way it works in professional sports. Of course in any given season you'll have teams that go on hot streaks or what have you, but over the course of a season in any sport with enough games to start to settle out to nice averages (I use baseball because of their 162 game season), that simply doesn't happen. The best teams in the league end up with a .6 win percentage. Why? Because those leagues exist as a real-world Elo. The best all get grouped together to play against each other.

Edited by FerretGR, 26 February 2013 - 09:29 AM.


#119 Joseph Mallan

    ForumWarrior

  • PipPipPipPipPipPipPipPipPipPipPipPipPipPipPip
  • FP Veteran - Beta 1
  • FP Veteran - Beta 1
  • 35,216 posts
  • Google+: Link
  • Facebook: Link
  • LocationMallanhold, Furillo

Posted 26 February 2013 - 09:21 AM

View PostYokaiko, on 26 February 2013 - 08:58 AM, said:



True, but that was also grinding off games.

If it looked bad charge in kill something, and die, exit and do the next drop for another weight class (I spent a LOT of the weekend in the top 50, up to 6th at one point).

So pulling 2:1 was surprising when I saw the individual class stats. Mediums where my worst with a 50/50, W/L However, I pulled over 5/1 in in assaults and I never drove a D-DC


Hold on I'll get the numbers.

So

Lights rank 505 score 39 w24/l10 so 2.2:1 w/l
Mediums rank 467 score 47 w20/l20 (note that was while xping my trebs)
Heavier rank 373 score 61 w36/l23 w/l 1.53:1
Assaults rank 426 score 61 w34/l17 w/l 2:1

So ok, more like 1.7:1....and that was enough for 66th overall.

Like I said I actually pushed my KDR up, which is no small matter from 2,700 kills.

Also like I said I never really watched the delta for win rates between grouped and ungrouped.

You do know that doing that may or may not change the teams outcome right? i died early in a few matches and did the same, Didn't look to see if the team won or not.

So what Phoenix and I have shown is that being on a Premade team is worth a +15%-18% bonus to your W/L record.

Win loss has no impact on how good a player you are. I have died first in a match, delivered 600-700 damage 1 kill 5 assists. the team lost, & I have deliver 70 Damage 0 Kills 0 assists and the team rolls the opfor without me. Win/Loss is no metric for personal performance. Damage delivered, Kills/assists, components destroyed these tell how good YOU are.

#120 Yokaiko

    Member

  • PipPipPipPipPipPipPipPipPipPip
  • Bad Company
  • Bad Company
  • 6,775 posts

Posted 26 February 2013 - 09:23 AM

View PostFros7bite, on 26 February 2013 - 09:16 AM, said:

This. The fact that you're losing 50% of the time is a great sing of the system working PERFECTLY.



My W/L is still 1809/771

Even after hero weekend, a cumulative W/L of 2.35:1 spilt between 2/3/4 mans and solo dropping, of that my 8 man play is statistically insignificant.

View PostJoseph Mallan, on 26 February 2013 - 09:21 AM, said:

You do know that doing that may or may not change the teams outcome right? i died early in a few matches and did the same, Didn't look to see if the team won or not.

So what Phoenix and I have shown is that being on a Premade team is worth a +15%-18% bonus to your W/L record.

Win loss has no impact on how good a player you are. I have died first in a match, delivered 600-700 damage 1 kill 5 assists. the team lost, & I have deliver 70 Damage 0 Kills 0 assists and the team rolls the opfor without me. Win/Loss is no metric for personal performance. Damage delivered, Kills/assists, components destroyed these tell how good YOU are.


Absolutely, I said plain out I just don't know. I haven't tracked it, my numbers are positive, so I'm more or less happy.

Edited by Yokaiko, 26 February 2013 - 09:24 AM.






3 user(s) are reading this topic

0 members, 3 guests, 0 anonymous users