24 replies to this topic

[Shredhead]

I think we all can agree that this tournament was fundamentally flawed. It was not about personal skill, but who could sacrifice the most time in a gaming marathon and/or share his account with others. So we should take another approach to see where we stand based only on the data the tournament provided us.
We take our tourney score, divide it through the number of games played and multiply the result x 100.
We should have played at least 25 games in that bracket to have viable data, proving our position with a screenshot. I'll start, I only played with my Hunch in the medium bracket:



Doing the maths, I get 52 : 37 = 1,405... x100 = 140,5 or 1,405 points per match.
Comparing this with the number one player in that bracket. He gets 607 : 371 = 1,636... x 100 = 163,6 or 1,636 points per match. Kudos to him, he's clearly better than me!
Number two on the list gets 599 : 532 = 1,125... x 100 = 112,5 or 1,126 (rounded) points per match.
We could make our own leaderboard with this data, if someone would be so kind to make a simple excel table with all the provided numbers and maybe the top 25 players of every bracket.

[Exoth3rmic]

The higher your Elo becomes, the better your team mates are meant to be and the more evenly matched to your opponents and team mates you will be.

If we consider that each match can only contain a maximum number of points, and that these points are available to all players on your team and that not all points will necessarily be earned or paid out - you would expect that the higher your Elo is - the less points you will be able to earn as your teammates will be earning proportionately more, and your opposition will be denying more.

So points per match avg doesn't guarantee an accurate scaling at the 'low' end as at the top with a low game sample regardless of what multiple you wish to apply.

Edited by Exoth3rmic, 25 February 2013 - 03:49 AM.

[Accused]

It's obvious PGI wants ELO data. It's okay, people did honestly spend a lot of time to get up there and they deserve some credit for their dedication. However this wasn't a really turny, if you realize that then it will be easier to sleep.

[Wispsy]

Exoth3rmic, on 25 February 2013 - 03:48 AM, said:

The higher your Elo becomes, the better your team mates are meant to be and the more evenly matched to your opponents and team mates you will be.

If we consider that each match can only contain a maximum number of points, and that these points are available to all players on your team and that not all points will necessarily be earned or paid out - you would expect that the higher your Elo is - the less points you will be able to earn as your teammates will be earning proportionately more, and your opposition will be denying more.

So points per match avg doesn't guarantee an accurate scaling at the 'low' end as at the top with a low game sample regardless of what multiple you wish to apply.

From what I have seen most of the time the higher your Elo the worse your teammates become.

[Exoth3rmic]

Wispsy, on 25 February 2013 - 03:55 AM, said:

From what I have seen most of the time the higher your Elo the worse your teammates become.

;-)

[Voidsinger]

Wispsy, on 25 February 2013 - 03:55 AM, said:

From what I have seen most of the time the higher your Elo the worse your teammates become.

Yeah Wispsy, I've noticed you're getting stuck with me on your team more and more.

[Chemie]

Wispsy, on 25 February 2013 - 03:55 AM, said:

From what I have seen most of the time the higher your Elo the worse your teammates become.

Correct. They do not match equal ELO players...they match the total team ELO. This means if you are crazy high, you are more likely to get a crazy low ELO on your team to even it out. Stupid but PGI logic for you.

Also, the match maker does not try to balance weight so you are also just as likely to get 5 lights while the other team gets 5 assaults so ELO is only partof the game outcome.

Edited by Chemie, 25 February 2013 - 04:05 AM.

[ciller]

Thanks for calling me a bad player man. Kinda rude..

I have no idea why I got so few points per match. I got crazy amount of kills usually, I win almost all my duels, I get assists and what not - but my points per match is terrible. Maybe it was all the horrendously bad matches I had where my team wouldn't do anything. Or maybe it was the build I had. Or maybe it was something else.. but please try to be less rude.

I worked hard, I had lots of wins, I had great games often. I'm sure I had terribad matches too but yeah..

I would like if the tourney next time just took your top matches and used that. Like if we could drop as much as we wanted and our 20 best matches of the weekend would determine our score, and keep trying to get our averaged metric up on those 20 by getting a killer game in - this would be much better. That way people can still drop tons, dont need to drop a crazy amount, and still gives us the best players via their best matches.
Winner would be the best accumulated Match Score of your top 20 matches during the weekend.

__________________________

Chemie, on 25 February 2013 - 04:03 AM, said:

Correct. They do not match equal ELO players...they match the total team ELO. This means if you are crazy high, you are more likely to get a crazy low ELO on your team to even it out. Stupid but PGI logic for you.

Also, the match maker does not try to balance weight so you are also just as likely to get 5 lights while the other team gets 5 assaults so ELO is only partof the game outcome.

I think i was running into this. Every match I would wish to be in the "elite" groups but got groups where I didnt recognize anyone, they would scatter every where, and die piecemeal despite my efforts in herding them. Then I would get angry, Id say stuff in a tired fury, oh sigh!! What a terrible weekend it was, haha.

Edited by ciller, 25 February 2013 - 04:59 AM.

[Psychobunny]

Accused, on 25 February 2013 - 03:52 AM, said:

It's obvious PGI wants ELO data. It's okay, people did honestly spend a lot of time to get up there and they deserve some credit for their dedication. However this wasn't a really turny, if you realize that then it will be easier to sleep.

i wish they didnt call it a tournament than and be honest about it, and reward active players than, but whatever..
usualy a tournament is fair, and to make it fair is a bit hard with ppl ending up with random players, and therefor they should not base it on wins / loses, maby damage done and received, but its hard to get into detail, cause there is plenty of things to take into account.. weight class, team mate skills, lag, capping, aiming...
it has to have some luck based games, cus in the very start of the game, you picked your weapons, and now you have to deal with what you got..

i would like a tournament where we do 100 games, solo, with weight class control, damage done and taken, kills, assists and deaths, caps (maby), and if it has to be that detailed, players team mates skill rate taken into account.
maby they did this already i dunno :X

..forgot to mention disconnections, maby those games shud be reseted if early, or disqualified, meh..

Edited by Psychobunny, 25 February 2013 - 05:00 AM.

[ciller]

Shredhead, on 25 February 2013 - 03:43 AM, said:

Doing the maths, I get 52 : 37 = 1,405... x100 = 140,5 or 1,405 points per match.
Comparing this with the number one player in that bracket. He gets 607 : 371 = 1,636... x 100 = 163,6 or 1,636 points per match. Kudos to him, he's clearly better than me!
Number two on the list gets 599 : 532 = 1,125... x 100 = 112,5 or 1,126 (rounded) points per match.
We could make our own leaderboard with this data, if someone would be so kind to make a simple excel table with all the provided numbers and maybe the top 25 players of every bracket.

i don't know how accurate that is either mostly because it is using the metric that PGI started with (the tourney score). This value is derived from numbers that include matches played then you divide by matches played again? I don't like it. You'd want to track stats in a different way right from the beginning to determine the best player, something like an average or accumulated score of your top 20 matches during the weekend would work nicely.
Fiddling with the numbers that PGI produced is like playing with a Cennobite puzzle box, you wont accomplish anything useful aside from summoning demons (forum demons in this case)

[Exoth3rmic]

ciller, on 25 February 2013 - 04:59 AM, said:

Thanks for calling me a bad player man. Kinda rude..

If this is directed at me, then it wasn't my intention to imply that you were 'low' end and therefore bad, it was my intention to show that regardless of the number of games you've played (and by definition you can't be high MM with low sample of games at this time) you can't conflate numbers to compare with people that have now played a larger sample of games as without knowing their actual Elo you can't take in to account the variables I then went on to highlight - like your team, the opposition etc

[Chrithu]

I happen to have played exactly 100 matches. My score is at 830 or 0.83, which made me number 204 in the "Heavy"bracket.

Have to add though that while my winrate of close to 50% stayed the same my K/D ratio for those 100 matches is way balow my overall of 1.6. I think I didn't even get above 1 in those macthes. Why? Because a great number of the matches went down totally different from what I was used to during the last two months.

I also saw people during the tournament I haven't seen a lot in that time before either, because they are known to drop in teams mostly, which kind of proves that there is not really much pugstomping going on or I would see those names a lot more. It also proves that those playing in teams obviously are used to different tactics that you do not see in the PUG matches which gave them an advantage. In a way you could say that those Team players are more skilled because their situational awareness and their tactical decision making is just a lot more trained and purposeful. That is related to my lower K/D during the tournament. Or so I think.

Bottom line: I earned my place mostly by doing worse than I used to before during the tournament.

Edited by Jason Parker, 25 February 2013 - 05:35 AM.

[ciller]

Exoth3rmic, on 25 February 2013 - 05:26 AM, said:

If this is directed at me, then it wasn't my intention to imply that you were 'low' end and therefore bad, it was my intention to show that regardless of the number of games you've played (and by definition you can't be high MM with low sample of games at this time) you can't conflate numbers to compare with people that have now played a larger sample of games as without knowing their actual Elo you can't take in to account the variables I then went on to highlight - like your team, the opposition etc

More at OP really. He called me bad specifically elsewhere.

And yes, I wish more variables would be taken into account, you are totally correct. I really really wish the next tournament is quite different and not based on accumulated number of matches played. This is coming from 2nd place medium with most matches played. IT IS HORRIBLE.

[Chrithu]

ciller, on 25 February 2013 - 05:35 AM, said:

And yes, I wish more variables would be taken into account, you are totally correct. I really really wish the next tournament is quite different and not based on accumulated number of matches played. This is coming from 2nd place medium with most matches played. IT IS HORRIBLE.

I wish they'd make tournaments a team only thing, thus maybe giving the lone wolf solo drop PUG I am an incentive to look for a group before CW comes. Because as it is now I don't feel like grouping up just for the lulz of it and have people yelling "FOCUS GAMMA" at me.

[Wispsy]

Voidsinger, on 25 February 2013 - 04:00 AM, said:

Yeah Wispsy, I've noticed you're getting stuck with me on your team more and more.

Just incase...I meant relatively, compared to the other team. I still get placed with some awesome players but sometimes I end up against some pretty average teams with some truly unhelpful teammates. This is literally asking you to chew through a group of 8 people on your own if they stay together, faster then 8 people can chew through 7 solo stationary targets. Couple that with a number of them having weapons that cannot physically miss, well sometimes it can feel a little dissatisfying to do so much and lose to bad players! I do not really mind the challenge but quite a lot of people do. I also realise that when they first implemented matchmaking it took too long to find games and people had weirdly banded scores due to gathering data without acting on it to make matches. Then they made the matchmaking a lot wider and apparently it focuses most on making the scores average and not much about how close those scores might be to each other. With the solo endurance tournament they set up over the weekend a lot of the Elo scores will have spread out more. Hopefully this means that soon we will be seeing some tweaking and this will be less and less of an issue!

So in all, it does still match me with and against good players, but sometimes it really does not.

Edited by Wispsy, 25 February 2013 - 05:58 AM.

[MaddMaxx]

Thank God it is over. Well done to ALL. Now can we get back to QQ'ing about ECM or something else. All the "unfair tears" from the Tourney have worn out my Umbrella of ultimate coverage. LOL

P.S. Made it as far as #171 in the "Over-All" by Saturday and then went to bed. WGC (Golf) was on Sunday so no Mech'in. Ended the Tourney, in the Over-All at #323.

Lots of really good games played and it also showed how some folks just won't build an all-round, non-cheese, Mech design but would rather DC out, f*&^ over 7 others, then play out the Match on a Map they can't handle (cough Alpine cough). Damn shame really.

Totally looking forward to the next "Tourney".

Edited by MaddMaxx, 25 February 2013 - 06:11 AM.

[Joseph Mallan]

Shredhead, on 25 February 2013 - 03:43 AM, said:

I think we all can agree that this tournament was fundamentally flawed. It was not about personal skill, but who could sacrifice the most time in a gaming marathon and/or share his account with others. So we should take another approach to see where we stand based only on the data the tournament provided us.
We take our tourney score, divide it through the number of games played and multiply the result x 100.
We should have played at least 25 games in that bracket to have viable data, proving our position with a screenshot. I'll start, I only played with my Hunch in the medium bracket:

You ranked 401st in Mediums with 19W18L, I was 572nd with 22W/32L... Do your voodoo please, cause my brain is fried sir.

[Kdogg788]

Chemie, on 25 February 2013 - 04:03 AM, said:

Correct. They do not match equal ELO players...they match the total team ELO. This means if you are crazy high, you are more likely to get a crazy low ELO on your team to even it out. Stupid but PGI logic for you.

Also, the match maker does not try to balance weight so you are also just as likely to get 5 lights while the other team gets 5 assaults so ELO is only partof the game outcome.

Correct me if I'm wrong but isn't that how matchmaking is supposed to function? As in teams will be evened so that the match as a whole is even. It would not at all be even if the highest ELO players are all on one side. Having those lower scored teammates is just a result of having a high score and the game attempting to create more even sides. I think the weight classes are also biased against the higher ELO team. For example there were multiple times last night where my four man team faced 5-7 assaults, mostly Atlas' or sometimes all Atlas' with 1 or 2 of our own.

-k

[DocBach]

ciller, on 25 February 2013 - 04:59 AM, said:

I think i was running into this. Every match I would wish to be in the "elite" groups but got groups where I didnt recognize anyone, they would scatter every where, and die piecemeal despite my efforts in herding them. Then I would get angry, Id say stuff in a tired fury, oh sigh!! What a terrible weekend it was, haha.

But dude, the games where everyone was on the leaderboards were pretty freaking epic. They were like the games everyone promised me ELO would give us.

[Shredhead]

Joseph Mallan, on 25 February 2013 - 06:12 AM, said:

You ranked 401st in Mediums with 19W18L, I was 572nd with 22W/32L... Do your voodoo please, cause my brain is fried sir.

I needs your match score to work that little alchemical magic trick, good Ser, else I can't change that leaden bar into pure gold

Edited by Shredhead, 25 February 2013 - 06:45 AM.