Elo Worthless
#141
Posted 14 November 2013 - 11:14 AM
#142
Posted 14 November 2013 - 11:25 AM
Kunae, on 14 November 2013 - 11:13 AM, said:
And I am not in error. Almost every one of your posts on this subject are all about "if I am having fun, nothing else matters about the performance of my team" and "as long as I take 1 or 2 with me, I am happy with the match". That's textbook selfish behavior.
So if doing my part to help win is selfish, I am selfish! Wanting to win and expecting to win is two different goal. I cannot see MM stopping an experienced player (who can actually type fast) from typing "Form up at E7, Scouts, go left side and find the enemy." I you are such a great commander, you should lead not Rambo to win. From my experience over the last few months to many PUGs don't want to follow a leader, They don't actually lead, and that is why they cannot win. I am not a leader, I know my limitations. So I do my thing, I follow Like Mechs, I provide massed fire, and an armor buffer. If nobody takes the lead, I cannot o more than I already do. So I will continue to be a selfish bastord and enjoy the efforts I put in towards the teams success.
Appogee, on 14 November 2013 - 11:14 AM, said:
Good it is a good playing to win Mech!
#143
Posted 14 November 2013 - 12:13 PM
Joseph Mallan, on 14 November 2013 - 11:00 AM, said:
It's exactly what the MM is expecting from you : to compensate for the lack of skills of the new players he paired you with. The MM expects that putting on one hand, a few good players with a flock of new players and on the other hand, a set of average players, it will result in a balanced match. Guess what ? It is not working and the situation is the same since the introduction of ELO ratings back in march. Check the forums and you will find nearly a hundred threads talking about the exact same situation.
And the best part is that the ELO ratings were introduced to eliminate PUG stomping : before it was random, now it's the MM that purposely brings this kind of match. I don't see any kind of improvement.
Plus, I don't think it is a good experience for a new player too : how can he have fun or learn something in these unforgiving match where his first mistake is usually his last too ? And it also creates a negative mood/trend : instead of being helped, new players are used as meat-shield if not worse.
So stop playing the white knight : you don't know how it works and you don't care about teamplay.
Edited by SgtKinCaiD, 14 November 2013 - 12:16 PM.
#144
Posted 14 November 2013 - 12:19 PM
Joseph Mallan, on 14 November 2013 - 11:25 AM, said:
Just for the Lulz ...
Edited by SgtKinCaiD, 14 November 2013 - 12:20 PM.
#145
Posted 14 November 2013 - 01:16 PM
#146
Posted 14 November 2013 - 01:45 PM
Mudhutwarrior, on 14 November 2013 - 01:16 PM, said:
From what I understand, the Elo ranges the MM uses has been at like 600-900pts since August.
I did similar to what you describe, last weekend. 2nd match had 2 founders on the other team. Most matches had some, plus at least half of the mechs in any match were general meta mechs. eg. DDC's, 2ppc + Ac20 highlanders, AC40 jaegers, etc.
The skill level was noticeably less, primarily in situational awareness and weapon accuracy.
The new player experience in this game currently sucks, and it has nothing to do with "premades".
#147
Posted 14 November 2013 - 02:06 PM
Kunae, on 14 November 2013 - 01:45 PM, said:
I did similar to what you describe, last weekend. 2nd match had 2 founders on the other team. Most matches had some, plus at least half of the mechs in any match were general meta mechs. eg. DDC's, 2ppc + Ac20 highlanders, AC40 jaegers, etc.
The skill level was noticeably less, primarily in situational awareness and weapon accuracy.
The new player experience in this game currently sucks, and it has nothing to do with "premades".
I think it all has to do with a lack of ingame voice. Noobs in the dark with only a keyboard while against premades susing voice comamnd and control which is vastly superior in the fight. They have no differnt experince than casuals and pugs do. Its the same for everyone who does not or cannot team up.
#148
Posted 14 November 2013 - 02:08 PM
#149
Posted 14 November 2013 - 02:10 PM
#150
Posted 14 November 2013 - 04:19 PM
It is just so awesome when you win... but that only happens when they are real banged up or real inept or both.
Of course, you have to be very good yourself to pull it off, which shows MM/Elo is FUBAR to begin with.
But to often it is 10 mechs to your 1. or your 10 to their 1 and you scramble among yourselves to get an assist so it isn't such a wasted match.
I don't enjoy being on either side of a ROFLstomp.
#151
Posted 14 November 2013 - 05:41 PM
Kunae, on 14 November 2013 - 10:40 AM, said:
If the higher Elo players happen to die on one side, really early for whatever reason, a 3-0 lead can quickly turn into a 3-12 loss. I've seen it happen way too many times.(Edit: This usually happens because the low Elo folks tend to run away and hide at the first sign of enemy fire.)
Also, if the higher Elo players happen to just be screwing around in mechs, leveling odd variants/loadouts, etc, then there's a pretty reasonable chance it's going to be a losing match.
Track it Kunae. Anecdotal evidence is notoriously untrustworthy. Keep a running tally of win/loss over 50 matches and what you scored in them and were you 'fit' in scoring for your team. Top 5, top half, bottom half, bottom 5.
You're going to find that you won plenty of games where you walked around the wrong corner early on and scored <100 points for the match. Statistically you will have more wins where you scored relatively low than you have matches where you lost when you scored very high.
That's not how the human brain works though. Your brain keeps and weighs memories of being wronged or denied what you felt was your just reward far better than it keeps memories of you being carried. That's fine, it's how we all work but don't mistake that for statistical accuracy.
It doesn't matter if you come here and post any of it. If you find that actually, statistically, you almost never tank and the team wins but you often rock the field and still lose than you've got a statistical basis from which we can examine a flaw in the matchmaker. Currently though there's no actual evidence or data to back up that position other than the opinions of some people on the forum.
Please don't take the following personally - it's not directed at you but at the routine position on the forums that peoples opinions trump actual data.
Sam Harris -
"Water is two parts hydrogen and one part oxygen. What if someone says, “Well, that’s not how I choose to think about water.”? All we can do is appeal to scientific values. And if he doesn’t share those values, the conversation is over. If someone doesn’t value evidence, what evidence are you going to provide to prove they should value it? If someone doesn’t value logic, what logical argument could you provide to show the importance of logic?"
You seem like a smart guy. If you're right and there's statistical evidence to support it, I'm your guy. I'll be all over it on white on rice, we'll dig into the metrics and identify exactly what has gone wrong and cry bloody tears at PGI to fix it. The problem is that quite a few people (myself included) tend to track our matches in statistical spurts just to fact-check our opinions of our own performance. You've heard me bitching about BMs, right? Well, when I check my statistical peformance what I find is that I had a string of losses early on that have left me at a win/loss deficit. I'm actually a tad over 1.0 win/loss for the last 50 matches, it just doesn't *feel* that way because I'm playing it as a sniper and not a brawler. I'm not in the middle of the fight blowing people to pieces so I don't connect to the victories as much. My opinion of my performance doesn't match the data, so I know my opinion is wrong.
The 1G still needs more torso twist though - 60 degrees simply isn't enough with a wide torso, it makes it impossible to protect yourself. It needs fixed but since the 1G is (P) only right now it'll probably stay gimped until it's a public model to avoid cries of 'P2W!'
How you think you're doing is different from how you're actually doing. I'd bet dollars to donuts you're comfortably over a 1.0 win/loss when pugging. This means that regardless of your performance (nobody can be 'on' 100% of the time) you're dropping with competent enough people in premades or pugs to consistently win more than you're losing. You're dropping most the time with good teams. Sometimes even good players can pull a dumb, sometimes bad players can have a good game. Don't mistake individual statistical extremes for 'every game'.
Track some yourself. If you see a problem, bring it here. I love dicing sh*t like that up. If I'm wrong I'll be the first to say so. I don't think I am though. I think you'll look at it and see that you're doing well and as a given rule your team does well. Staying over a 1.0 win/loss means that normally it's the other team failing their luck roll, not you.
#152
Posted 14 November 2013 - 07:12 PM
ELO isn't worthles its just that pre-made groups skew the ELO usefulness. If you are a pre-made group, your ELO doesn't reflect the increased chances of winning that come with being a pre-made and the whole damn point of ELO -- to determine your odds of winning a match. That means the pre-made group gets placed with players who, would be a competitive mix if everyone were just a bunch of puggers playing puggers, but end up geting repeatedly stomped becuse of the pre-made skew!
Then, when a person who usually groups, starts pugging, they end up being put in a MUCH higher ELO bracket then they should be because of the pre-made skew and fail to win match after match.
Quickest fix would be to add a pre-made size factor to the ELO calculations to balance ELO and group size on the fly.
Edited by nehebkau, 14 November 2013 - 07:17 PM.
#153
Posted 14 November 2013 - 07:39 PM
#154
Posted 14 November 2013 - 07:47 PM
nehebkau, on 14 November 2013 - 07:12 PM, said:
ELO isn't worthles its just that pre-made groups skew the ELO usefulness. If you are a pre-made group, your ELO doesn't reflect the increased chances of winning that come with being a pre-made and the whole damn point of ELO -- to determine your odds of winning a match. That means the pre-made group gets placed with players who, would be a competitive mix if everyone were just a bunch of puggers playing puggers, but end up geting repeatedly stomped becuse of the pre-made skew!
Then, when a person who usually groups, starts pugging, they end up being put in a MUCH higher ELO bracket then they should be because of the pre-made skew and fail to win match after match.
Quickest fix would be to add a pre-made size factor to the ELO calculations to balance ELO and group size on the fly.
Add allowing players to create groups of any size to your argument, and I'll back your play.
#155
Posted 14 November 2013 - 07:47 PM
Ghogiel, on 14 November 2013 - 07:39 PM, said:
But something must be amiss, or else the elo system would pretty quickly raise you up to the same level, since it is supposed to represent your individual skill.
If elo worked well, then your alt account elo would be the same as your main, unless you are just intentionally taking it down for some reason.
Don't get me wrong, I consistently see the same folks in the game I play too, but you also tend to see folks who clearly have no idea what they are doing.
#156
Posted 14 November 2013 - 07:51 PM
Ghogiel, on 14 November 2013 - 07:39 PM, said:
You must be mistaken. only anecdotal evidence that supports the original opinion is valid. All the rest is SHILLING FOR THE MAN.
You're part of the establishment now Ghogiel. Welcome to the Matrix.
#157
Posted 14 November 2013 - 08:02 PM
Roland, on 14 November 2013 - 07:47 PM, said:
If elo worked well, then your alt account elo would be the same as your main, unless you are just intentionally taking it down for some reason.
Don't get me wrong, I consistently see the same folks in the game I play too, but you also tend to see folks who clearly have no idea what they are doing.
Takes hundreds of games to seat for Elo at higher levels. Higher your Elo, lower your return for winning. Grits N Gravy had an awesome idea on how to address that, hopefully it got captured and passed along as feedback. reduce k-values and use a gaussian distribution for Elo and matching. Would fix a lot of things.
#158
Posted 14 November 2013 - 08:20 PM
Edited by Shanghai Shadow, 14 November 2013 - 08:26 PM.
#159
Posted 15 November 2013 - 04:37 AM
Shanghai Shadow, on 14 November 2013 - 08:20 PM, said:
Its like a stone skipping over the water now and the skips get shorter and shorter. Seen the cycles for near a year now and it seems were about to go under. Same defenders making excuses since I signed on and thousands of lone voices being shut down. Its obvious the player base is shrinking and only little spurts of new players come due to marketing which is almost non existant. I think the staunch defenders need to see the writing ont he wall. After starting a new acccount I am sure that how it is. 30 matches in I saw very few noobs while being one myself. In my real account I see them ever so rarely and my elo has not changed much from then to now. From a .65 to .74. over months.
I think PGI is looking for a way out of this game myself and they show very little interest besides money grabs. A year and not much to show for it past a couple of maps and tweaking. Maybe someone should approach them and see if a Kickstarter campaign could raise the coin to buy them out and find developers who actually care about the game and not just a paycheck.
#160
Posted 15 November 2013 - 05:01 AM
MischiefSC, on 14 November 2013 - 08:02 PM, said:
Takes hundreds of games to seat for Elo at higher levels. Higher your Elo, lower your return for winning. Grits N Gravy had an awesome idea on how to address that, hopefully it got captured and passed along as feedback. reduce k-values and use a gaussian distribution for Elo and matching. Would fix a lot of things.
I don't think this is actually true. It shouldn't take hundreds of games for elo to account for your skill. It doesn't in other games. If elo is working, then your initial games should boost you up fairly significantly.
I suspect that in two it does take hundreds of games, but this is merely illustrating that the system does not in fact work correctly. The reason it is taking so many matches is because your individual contribution is being heavily hidden by the noise generated by a 12v12 match.
8 user(s) are reading this topic
0 members, 8 guests, 0 anonymous users