227 réponses à ce sujet

[archangelV]

FirroSeranel wrote...

You know, I keep hearing people say things, on the forums, on IRC, in voice chat... things like, "Score doesn't matter.  The best player could be playing support and not score high at all!"  Or, "N7 rating doesn't mean anything at all."

I beg to disagree.

Okay, I'll grant, score and N7 rating aren't a 100% guaranteed predictor of player skill.  And I'll grant, score shouldn't be the -goal- by any means.

I have soloed gold cerberus and had team matches where I topped the scoreboard. I have had games where I was at the bottom. There was no difference in my effort or skill. The only difference was over powered teammates slaying all my debuffs or in the case of sliver under powered enemies.

Modifié par archangelV, 18 juillet 2012 - 11:37 .

[Legbiter]

N7 ranking is a pretty good indicator of how well you'll play more often than not. Sure, once in a blue moon you'll get crappy high N7's and good lowbies but most of the time it's a solid indicator

[GGW KillerTiger]

I have never earned points for revives and I'm a walking difibulator in this game because I always get randoms that can't stay up. Hell none of my friends can even stay up ...... So my score is always low because I'm too bussy reviving and not getting points. Also once I read you thought we get points for revives I stopped reading due to you talking about stuff you don't know about.

[loungeshep]

I like to think I"m between crappy and good as most my N7 rating is me learning the MP.

[Computron2000]

Eh someone tied their ego to their n7. Some classes are built for AoE. Some aren't. When i run a QMI, human soldier or any grenade class, i bomb the spawn spots which means i get a lot of points. When i run my human engineer, i cannot do that and have to kill enemies much slower.

If i'm in a match with a QMI as a human engineer, topping the board means the guy playing poorly (which is also why i LOL at "yada yada because i top the scoreboard"), not because i'm suddenly much much better at playing.

There are also those with the mobility and durability to reach a mob and kill them (while blocking your shots). They also die often but they can top the scoreboard. Then theres the sniper that can top the scoreboard but staying in position while everyone is doing a hack circle or the guy that primes everything for people to bursting/explosion and gets the points (did this with the fury while testing). Plenty more ways to get points which ignores the contributions of others.

Personally, i don't rate people by the scoreboard. I rate if they did objectives and the number of times they dropped and the number of revives they did when they were in position to help. People who did well in all 3, i tip my hat off to them

[FirroSeranel]

GGW KillerTiger wrote...

I have never earned points for revives and I'm a walking difibulator in this game because I always get randoms that can't stay up. Hell none of my friends can even stay up ...... So my score is always low because I'm too bussy reviving and not getting points. Also once I read you thought we get points for revives I stopped reading due to you talking about stuff you don't know about.

Yes you have.  They're medals at the end of the round, for 5, 15, and 25 revives, I believe.

[FirroSeranel]

Computron2000 wrote...

Eh someone tied their ego to their n7. Some classes are built for AoE. Some aren't. When i run a QMI, human soldier or any grenade class, i bomb the spawn spots which means i get a lot of points. When i run my human engineer, i cannot do that and have to kill enemies much slower.

If i'm in a match with a QMI as a human engineer, topping the board means the guy playing poorly (which is also why i LOL at "yada yada because i top the scoreboard"), not because i'm suddenly much much better at playing.

There are also those with the mobility and durability to reach a mob and kill them (while blocking your shots). They also die often but they can top the scoreboard. Then theres the sniper that can top the scoreboard but staying in position while everyone is doing a hack circle or the guy that primes everything for people to bursting/explosion and gets the points (did this with the fury while testing). Plenty more ways to get points which ignores the contributions of others.

Personally, i don't rate people by the scoreboard. I rate if they did objectives and the number of times they dropped and the number of revives they did when they were in position to help. People who did well in all 3, i tip my hat off to them

AoE, huh?  I won't say (unlike some people who are totally willing to sling around ridiculous generalizations) that AoE -doesn't- help generate score.  But on the other hand, it's hard to score any higher than a Geth Infiltrator with a Javelin, with good aim.  Which involves no AoE at all.

Also, saying that when you top the scoreboard is because everyone else played poorly differs from saying it's because you played well -how- exactly?  If you played poorly, -and- they played poorly, then you failed the match.  If all three of them played poorly, and you still won a Gold match, I dare say you played phenomenally.  And I dare say you'll have the score to prove it.

[FirroSeranel]

You know... this is actually rather hilarious. My premise, as -should- be completely obvious from the title of the thread, is simply that score and N7 rating matter. More precisely, that they don't "don't matter". That is a self-evident statement. Citing exceptions doesn't negate it. If you're arguing with my premise, then by definition you are saying that they don't matter AT ALL, EVER.

Quite aside from how absurd that is given the statement being evaluated, to state a generalization in the negative like that, is a logically unsupportable position.

For me to prove myself correct, all I have to do is cite one single reference where score or N7 rating were good predictors of skill. That's it. One reference, and I'm done. It matters.

For -you- to prove that they don't matter on the other hand? One reference doesn't do it. You'd have to systematically disprove every single instance in the history of the game in which a high-N7 played better than a low N7, and show that in every single instance in the history of the game, the top-scorer was actually the weakest link while the guy who scored 20k on Gold was actually the best player. Not once. Every single time, in every single group of players, in every single match, ever.

You can't. It's literally impossible. So stop trying!

Modifié par FirroSeranel, 18 juillet 2012 - 05:00 .

[Feauce]

Are you trying to justify something like booting people you arbitrarily judge as "dead weight" or a false sense of superiority? What's the purpose here? Neither score or N7 rating are accurate portrayals of how someone played during a match. There are a number of things that aren't scored, there's ways that someone will get points that should either be someone else's or split with them, and then there's the fact that the killshot is often times worth more than wearing an enemy down.

[Draining Dragon]

You get almost no points for objectives or revival.

[ltmarcos23]

FirroSeranel wrote...

Yes you have.  They're medals at the end of the round, for 5, 15, and 25 revives, I believe.

it is actually 5,10,15 just like melee kills. Just got it last night

Also I have played many a match with a friend where i set up combos and he detonates. I almost always top the scoreboard when playing one of those classes and he is usually near the bottom. but without him doing that it would be more difficult. Conversely if we run non biotics one is usually first the other 2nd. Dependent on class/weapons

[Zero132132]

No. First off, if you applied a linear regression with score and N7 ranking, you'd have nothing even close to 0.7 R^2 value. I guess I can't say this any more definitively than you can, since we have no data, but for this to be the case, an N7 4,000 would have to regularly outscore an N7 300 by a large amount. If it wasn't related to some baseline value instead, then an N7 4k would have to outscore an N7 300 by about 13x the lower score, around 5x an N7 800, etc. Even with a high baseline value, you're talking about a lot of variation in score. Might work a bit better if you used the natural log of N7 level instead, but I've played with N7 600s who are a lot better than some of the N7 4-6k people I've run into, so the portion of variability in score that could be accounted for by the variability in N7 probably wouldn't be as high as you're claiming.

This also says nothing about the t-value associated with the correlation between N7 ranking and score, and it's associated p-value (the probability that the correlation is genuine).

Sorry to be nitpicky, but don't put out arbitrary numbers if you don't want math people to get mad!

[Rick Lewis]

The N7 ranking tells you nothing of the players skill. People that are obsessed with it need to grow up.
I'm only 570 and I've played with lower N7's that were better and higher N7's that could barely play at all.

[Omega2079]

FirroSeranel wrote...

For me to prove myself correct, all I have to do is cite one single reference where score or N7 rating were good predictors of skill. That's it. One reference, and I'm done. It matters.

For -you- to prove that they don't matter on the other hand? One reference doesn't do it. You'd have to systematically disprove every single instance in the history of the game in which a high-N7 played better than a low N7, and show that in every single instance in the history of the game, the top-scorer was actually the weakest link while the guy who scored 20k on Gold was actually the best player. Not once. Every single time, in every single group of players, in every single match, ever.

You can't. It's literally impossible. So stop trying!

LOL, no. It doesn't work like that. Your statements show you don't understand what you're talking about.

You can't have it both ways. One example does not show correlation. Correlation does not imply causation.

Modifié par Omega2079, 18 juillet 2012 - 06:34 .

[Meatiershower]

N7 means squat. I got to 1080 before I stopped bothering, and run consistantly 3rd on Gold. A fairly large percentage of the 1st and 2nd place players are below me. I played buttloads of Bronze and Silver, promoted every time I had any chance too, and promoted before purchasing packs to utilize character cards.

N7 means nothing as far as actual playing skill goes. An individual player will reach his skill peak and stay there. This does not prevent him from playing at that skill level and promoting, thus ranking up. His skill did not increase.

[Kuriiiiiii]

My N7 is 300 something, I play better then many of my friends who are N7 2000+

N7 unless it's below 100 doens't mean anything.

[elPrimoFilipino]

Omega2079 wrote...

FirroSeranel wrote...

For me to prove myself correct, all I have to do is cite one single reference where score or N7 rating were good predictors of skill. That's it. One reference, and I'm done. It matters.

For -you- to prove that they don't matter on the other hand? One reference doesn't do it. You'd have to systematically disprove every single instance in the history of the game in which a high-N7 played better than a low N7, and show that in every single instance in the history of the game, the top-scorer was actually the weakest link while the guy who scored 20k on Gold was actually the best player. Not once. Every single time, in every single group of players, in every single match, ever.

You can't. It's literally impossible. So stop trying!

LOL, no. It doesn't work like that. Your statements show you don't understand what you're talking about.

You can't have it both ways. One example does not show correlation. Correlation does not imply causation.

I don't think N7 means squat, but I was willingly to listen to him.  Now, he has taken an infallible stance, which makes him just another raving forum poster who obviously knows how it all works.

Modifié par elPrimoFilipino, 18 juillet 2012 - 06:43 .

[Thoragoros]

Lets face it, it is just not encouraging to be prepping in the lobby for Gold or Platinum and seeing that one of your teammates has a 150 N7 Rank.

Typically, you'll be think "I'm going to spend a lot of time reviving that guy" or "We're not making it past Wave 5."

[Zero132132]

Omega2079 wrote...

LOL, no. It doesn't work like that. Your statements show you don't understand what you're talking about.

You can't have it both ways. One example does not show correlation. Correlation does not imply causation.

Actually, correlation does imply that there's SOME causal relationship. Regardless of that, if correlation could be proven (which I doubt), you could actually rationally assume, without regards to causation, that if high N7 level is presented, a high score will present itself as well.

Of course, you actually can't claim correlation from one sample. Or even two. A two-sample OLS regression will pretty much always yield a 100% correlation on a single predictor variable. The only way it wouldn't is if the two samples included someone of the exact same N7 level or score, but not the same on both.

I really doubt that N7 rankings or scores are normally distributed, too, so we'd need a LOT of samples to invoke the central limit theorem, since one of the assumptions behind OLS regression is the normality of residuals. Any individual person's experience can't even be shown to imply correlation.

[Feauce]

Thoragoros wrote...

Lets face it, it is just not encouraging to be prepping in the lobby for Gold or Platinum and seeing that one of your teammates has a 150 N7 Rank.

Typically, you'll be think "I'm going to spend a lot of time reviving that guy" or "We're not making it past Wave 5."

If the person is using Common weapons, depending on class, I could maybe see that, but otherwise, not really. Is it inconceivable that they could've only promoted a few times, reached Lv 20 on the classes they use, and stayed there? (Resisting the temptation to post a Princess Bride picture with this one...)

[Doc Magnus]

Omega2079 wrote...

xp is shared, so your n7 rank could be built entirely from quality, top-tier gold performance, or worst-of-the-worst leaching on bronze. That's why N7 rank means little, esp. above 120.

Well said. 

[FirroSeranel]

Feauce wrote...

Are you trying to justify something like booting people you arbitrarily judge as "dead weight" or a false sense of superiority? What's the purpose here? Neither score or N7 rating are accurate portrayals of how someone played during a match. There are a number of things that aren't scored, there's ways that someone will get points that should either be someone else's or split with them, and then there's the fact that the killshot is often times worth more than wearing an enemy down.

No, I'm trying to debunk the people who say that score and N7 are meaningless.

[JxLegend]

Good equip and teamwork > N7 rating. I boot people for several reasons but not low N7 I let them run with us and get some credits and get better stuff after all this is a co-op game

[FirroSeranel]

Zero132132 wrote...

Omega2079 wrote...

LOL, no. It doesn't work like that. Your statements show you don't understand what you're talking about.

You can't have it both ways. One example does not show correlation. Correlation does not imply causation.

Actually, correlation does imply that there's SOME causal relationship. Regardless of that, if correlation could be proven (which I doubt), you could actually rationally assume, without regards to causation, that if high N7 level is presented, a high score will present itself as well.

Of course, you actually can't claim correlation from one sample. Or even two. A two-sample OLS regression will pretty much always yield a 100% correlation on a single predictor variable. The only way it wouldn't is if the two samples included someone of the exact same N7 level or score, but not the same on both.

I really doubt that N7 rankings or scores are normally distributed, too, so we'd need a LOT of samples to invoke the central limit theorem, since one of the assumptions behind OLS regression is the normality of residuals. Any individual person's experience can't even be shown to imply correlation.

But I'm not trying to prove a correlation.  I'm just saying that -others- can't prove there -isn't- any.  Which is true.  And yes, it does work like that.  Not because I'm a "raving forum poster"  Because that's how -logic- works.