MNF

[Drink IRON City]

B b b b ut Ben heldid the ball long time in beginning of career.

****, Ben was held back in the passing yardage due to HIGH RUSH OFFENSE, imagine his yard total had he been turned lose from the get-go.

Rivers and Manning are not in Ben's league. Ben by far the best QB of that class.

Now if I or someone falls out of the tree, then I could see a head injury detouring their reasoning.


Salute the nation

 

[Drink IRON City]

What I love about this argument is that I have said multiple times that I think Ben is better than Rivers. But, I can see the claim or argument one might make for Rivers. That is all I am saying. But some of the people on this board are so rigid in their beliefs that they can't even fathom the thought of someone disagreeing with them. And this is the root problem with so many on this board.

**** yeah, this ^^^^^^^^^^^^


Salute the nation

 

[slashsteel]

**** yeah, this ^^^^^^^^^^^^


Salute the nation

Sounds like he forgot he was the pot calling the kettle...

 

[Drink IRON City]

Sounds like he forgot he was the pot calling the kettle...

I felt like I had to keep him in CHECK.


Salute the nation

 

[SteelerFan448]

I’m a statics guy so a sample size like that has a margin of error of probably 5%.

That still doesn't change that people will vote against Roethlisberger for just about anything because they hate the guy.

 

[madinsomniac]

I’m a statics guy so a sample size like that has a margin of error of probably 5%.

How do you figure? Its a non scientific poll, attached to an internet article.
There were 478 responses. There are over 150 million nfl fans… if those 150 million fans were more homogeneous than you could male an argument of maybe a 5% moe,, but with 32 separate fanbases interjecting varied bias you can’t set any sort of confidence because you dont have a known sample bias breakdown..

Its the equivalent of randomly polling 500 people about how they feel about the president without knowing their political party or if they are from varied locations or not…

Ask 500 democrats in Washington and your results would vary greatly from asking 500 republicans in texas, which is why surveys target diversified interests to accurately reflect the target market.

 

[FordFairLane]

How do you figure? Its a non scientific poll, attached to an internet article.
There were 478 responses. There are over 150 million nfl fans… if those 150 million fans were more homogeneous than you could male an argument of maybe a 5% moe,, but with 32 separate fanbases interjecting varied bias you can’t set any sort of confidence because you dont have a known sample bias breakdown..

Its the equivalent of randomly polling 500 people about how they feel about the president without knowing their political party or if they are from varied locations or not…

Ask 500 democrats in Washington and your results would vary greatly from asking 500 republicans in texas, which is why surveys target diversified interests to accurately reflect the target market.

So all 150 million fans saw that survey.

In my statics class in college it was said a sample size of 1000 would have a 3% margin of error. And that’s speaking about the entire country. So, yes. A was being liberal by saying 5%.

 

[FordFairLane]

That still doesn't change that people will vote against Roethlisberger for just about anything because they hate the guy.

Oh, golly. I guess that means Roethlisberger is the undisputed best QB of all time. Because people might hate him…

 

[SteelerFan448]

Oh, golly. I guess that means Roethlisberger is the undisputed best QB of all time. Because people might hate him…

How do you get this from a fan poll is likely skewed toward Rivers simply because they don't like Ben?

 

[FordFairLane]

How do you get this from a fan poll is likely skewed toward Rivers simply because they don't like Ben?

What about the people who don’t Rivers? Every liberal I know hates Rivers because he’s a devout Christian. Maybe he would have won by even more if not for that hatred?

You are trying to argue for unknowns.

 

[slashsteel]

How do you get this from a fan poll is likely skewed toward Rivers simply because they don't like Ben?

That is how he acts when people don't agree with his assessment. His pompous prick meter starts going off.

 

[slashsteel]

What about the people who don’t Rivers? Every liberal I know hates Rivers because he’s a devout Christian. Maybe he would have won by even more if not for that hatred?

You are trying to argue for unknowns.

Now you are just pulling crazy **** out of your ***, sounds like you are making **** up on the fly.

Just ridiculous crap.

 

[Drink IRON City]

What about the people who don’t Rivers? Every liberal I know hates Rivers because he’s a devout Christian. Maybe he would have won by even more if not for that hatred?

YOU are trying to argue for unknowns.

Quite the mirror I’sd say.

Rivers isn’t as good a QB as Ben ROETHLISBERGER, regardless of how you spin it. Your opinion is your’s, you do NOT have to try and make it everybody else's. This is where you fail.


Salute the nation

 

[madinsomniac]

So all 150 million fans saw that survey.

In my statics class in college it was said a sample size of 1000 would have a 3% margin of error. And that’s speaking about the entire country. So, yes. A was being liberal by saying 5%.

I deal with relative accuracy on a weekly basis at work… it only works if the sampling is correctly represented… people think that the math equation is the only part of that, but it only works when what you are sampling has generally similar representation…

Misunderstanding that concept is the first step to bad data.

Stratification in a test market requires either general representative sampling or correction for bias.

Its not just the number of responses that gives you that, its the number of representative data sets…

So in this case, you wouldn’t want 478 people from one city or 239 from 2… you would probably want 15 from the 32 nfl cities… but then the results aren’t 480 people, its 15 data sets.

A portion of what I do is directly built from the same mathematical theories that drive statistical polling, only instead of just doing half assed math and throwing it out into the public, I have to follow rigid scientific methods as regulated by the government…
So when I have to verify how accurate a sampling system is, I often cannot simply test a single point over and over or test sporadically in areas… I have to scientifically determine the level of potential stratification, then use a series of logical steps to come up with a representative sample plan.

In some cases, the difference in a test source is going to be statistically negligible and representative sampling could just be one sample point.. but in other cases it could be something much higher…

But lets say I sample 48 points… that’s mathematically considered 1 data set for relative accuracy purposes… so if i do that 12 times.. that is 576 samples, but each set of 48 is averaged together then counted as 1 when we calculate standard deviation and confidence coefficients.. we don’t use 576 as the number to set the T value, we would use 12….

A lot of people can’t grasp this concept… even some professional pollsters… its a personal pet peeve of mine..

 

[FordFairLane]

I deal with relative accuracy on a weekly basis at work… it only works if the sampling is correctly represented… people think that the math equation is the only part of that, but it only works when what you are sampling has generally similar representation…

Misunderstanding that concept is the first step to bad data.

Stratification in a test market requires either general representative sampling or correction for bias.

Its not just the number of responses that gives you that, its the number of representative data sets…

So in this case, you wouldn’t want 478 people from one city or 239 from 2… you would probably want 15 from the 32 nfl cities… but then the results aren’t 480 people, its 15 data sets.

This survey was on the world wide web. It wasn't on a Chargers website or a Browns fan page. It was on SBNation which is a sports page that covers baseball to UFC. Saying that there's bias because people hate Ben is asinine.

https://www.similarweb.com/website/sbnation.com/#geography

Here you can break down the demographic in greater detail if you like. But as you can see the majority of the traffic is from American and spread throughout the country.

A portion of what I do is directly built from the same mathematical theories that drive statistical polling, only instead of just doing half assed math and throwing it out into the public, I have to follow rigid scientific methods as regulated by the government…
So when I have to verify how accurate a sampling system is, I often cannot simply test a single point over and over or test sporadically in areas… I have to scientifically determine the level of potential stratification, then use a series of logical steps to come up with a representative sample plan.

In some cases, the difference in a test source is going to be statistically negligible and representative sampling could just be one sample point.. but in other cases it could be something much higher…

But lets say I sample 48 points… that’s mathematically considered 1 data set for relative accuracy purposes… so if i do that 12 times.. that is 576 samples, but each set of 48 is averaged together then counted as 1 when we calculate standard deviation and confidence coefficients.. we don’t use 576 as the number to set the T value, we would use 12….

A lot of people can’t grasp this concept… even some professional pollsters… its a personal pet peeve of mine..

So SBnation gets 7.3m hits monthly or 243k daily. The sample size is 478. Proportion 50%. What is the margin of error with the standard 95% confidence level? I'll give you a hint. Less than 5%.

 

[Drink IRON City]

BACON


Salute the nation

 

[madinsomniac]

This survey was on the world wide web. It wasn't on a Chargers website or a Browns fan page. It was on SBNation which is a sports page that covers baseball to UFC. Saying that there's bias because people hate Ben is asinine.

https://www.similarweb.com/website/sbnation.com/#geography

Here you can break down the demographic in greater detail if you like. But as you can see the majority of the traffic is from American and spread throughout the country.



So SBnation gets 7.3m hits monthly or 243k daily. The sample size is 478. Proportion 50%. What is the margin of error with the standard 95% confidence level? I'll give you a hint. Less than 5%.

None of that addresses the potential issue of polling bias. In fact its the basis of the reason national polls more often than not fail worldwide..

Margin of error should be repeatable and reflective of real world results… fan blog polls aren’t going to be accurate at all.. the deviation of perception from fanbase to fanbase is far greater than the 5 or 10% that is required to calculate MOE the way you are doing it … youd have to ask some demographic questions and tgen bias adjust the data accordingly

 

[FordFairLane]

None of that addresses the potential issue of polling bias. In fact its the basis of the reason national polls more often than not fail worldwide..

Margin of error should be repeatable and reflective of real world results… fan blog polls aren’t going to be accurate at all.. the deviation of perception from fanbase to fanbase is far greater than the 5 or 10% that is required to calculate MOE the way you are doing it … youd have to ask some demographic questions and tgen bias adjust the data accordingly

Yes, it does. It is called the margin of error... You are just messing with me now? Right?

The margin of error is a statistic expressing the amount of random sampling error in the results of a survey. The larger the margin of error, the less confidence one should have that a poll result would reflect the result of a census of the entire population. The margin of error will be positive whenever a population is incompletely sampled and the outcome measure has positive variance, which is to say, whenever the measure varies.

 

[madinsomniac]

Yes, it does. It is called the margin of error... You are just messing with me now? Right?

The margin of error is a statistic expressing the amount of random sampling error in the results of a survey. The larger the margin of error, the less confidence one should have that a poll result would reflect the result of a census of the entire population. The margin of error will be positive whenever a population is incompletely sampled and the outcome measure has positive variance, which is to say, whenever the measure varies.

And I’m telling you that the margin of error cannot be accurately determined unless your sample is known to be representative of the population…

The generic way that you learn it in high school is based on ideal conditions… thats why national polls rarely accurately reflect real world turnout either in politics or marketing… and those actually try to use some representative sampling, at least far more than an online poll would…

If those MOE were correct, then for instance the end result of a vote would fall within the margin of error of the most recent poll more often than not.
This is famously not the case in national polls across the world…

Scientific polling is based on scientific sampling, but unless the first step of representative sampling is done then all the results are potentially biased and unrepeatable… pollsters typically ignore that… and here you are using a totally unscientific and relatively small poll to try to prove a national opinion

 

[Hypocycloids]

https://www.sbnation.com/nfl/2018/11/30/18119674/ben-roethlisberger-vs-philip-rivers-steelers-chargers-super-bowl

61% Picked Rivers. But I am sure they were all Chargers fans.

What I am saying is it isn't as cut and dry as it would be on this board. I expect the majority to pick Ben. Hell, I would pick Ben. But I recognize my own homerism. I am going to pick my guy. But you can't see the forest from the tree. I can step back and look at the argument for Rivers and there is a solid argument. He had less talented offensive weapons, less talented offensive lines, and less talented defenses. Stat line is pretty similar. Ben had 64.4% 64,088 yards 7.6 y/a 418/211 to Rivers 64.9% 63,440 7.8 y/a 421/209. Considering Rivers played 5 less games I would give Rivers a slight nod on stats but not a major difference. So at the end of the day I would say the playoff stats and record would trump Rivers.

I would pick Ben as well. I am surprised at the results. Due to the similar stats, I would have anticipated a much closer margin with one player edging out the other. This was a blowout imo.
Could it be that the majority of respondents were Seahawk fans still disgruntled over the SB loss to Ben?

 

[slashsteel]

And I’m telling you that the margin of error cannot be accurately determined unless your sample is known to be representative of the population…

The generic way that you learn it in high school is based on ideal conditions… thats why national polls rarely accurately reflect real world turnout either in politics or marketing… and those actually try to use some representative sampling, at least far more than an online poll would…

If those MOE were correct, then for instance the end result of a vote would fall within the margin of error of the most recent poll more often than not.
This is famously not the case in national polls across the world…

Scientific polling is based on scientific sampling, but unless the first step of representative sampling is done then all the results are potentially biased and unrepeatable… pollsters typically ignore that… and here you are using a totally unscientific and relatively small poll to try to prove a national opinion

You are wasting your time his quest to be right far out weighs common sense.

People do hate Ben so it isn't assinine to think anything with his name involved wouldn't be tainted.

When I first moved to Pittsburgh I was surprised at the level of dislike for BigBen. In the bars downtown I heard plenty of stories confirming the topic. And that was in Pittsburgh, so you know outside of it wouldn't be any different.

He is judged for allegations from the women that accused him and he is judged from his earlier behavior in and out of establishments.

Over they years I heard plenty of bias towards him.

So for any poster to not realize it is either purposely being naive or lives under a rock.

 

[madinsomniac]

You are wasting your time his quest to be right far out weighs common sense.

People do hate Ben so it isn't assinine to think anything with his name involved wouldn't be tainted.

When I first moved to Pittsburgh I was surprised at the level of dislike for BigBen. In the bars downtown I heard plenty of stories confirming the topic. And that was in Pittsburgh, so you know outside of it wouldn't be any different.

He is judged for allegations from the women that accused him and he is judged from his earlier behavior in and out of establishments.

Over they years I heard plenty of bias towards him.

So for any poster to not realize it is either purposely being naive or lives under a rock.

Nah the Polling MOE thing is a decades long pet peeve of mine… people using small polls to predict the opinions of huge populations is a long time junk science that already irks me… then the generic non scientific aspect of that particular one make it even less accurate than a more scientific one … less to do with this particular topic than something I have been irked about for a very long time just bubbling out

 

[slashsteel]

Nah the Polling MOE thing is a decades long pet peeve of mine… people using small polls to predict the opinions of huge populations is a long time junk science that already irks me… then the generic non scientific aspect of that particular one make it even less accurate than a more scientific one … less to do with this particular topic than something I have been irked about for a very long time just bubbling out

Yeah but what you described makes common sense. But I got ya.

 

[topseed]

I have compared Roethlisberger to Matt Ryan and Phillip Rivers before in the past. If either of those two players were on the TEAM of Ben Roethlisberger they would have more wins than they currently have.

Well, just because you decided to compare them doesn't mean that there's any comparison.

Palmer played in a different era. And Vinny threw 127 more picks as Joe.

Oh. So, as usual, you want to qualify statitics when they don't support your argument.

Yes, you wear rose colored glasses. Go ask a non-Steelers fan about Rivers and Roethlisberger if they were comparable. I am sure they will give Ben the nod because of the wins and super bowls but if you ask would Rivers have won a super bowl or two with the teams Roethlisberger played on you are probably going to get pissed at them too.

Rivers played on some pretty good Chargers teams, too. Like 2006, when they were 14-2 and went one and done in the playoffs with PR compiling a 55.5 rating. Or 2009, when they were 13-3, and went one and done in the playoffs with good buddy compiling a 76.9 rating.

Ask most knowledgeable football fans about Philip Rivers, and they'll describe him as a pretty good regular season quarterback. Not in the same class as Ben.

 

[FordFairLane]

And I’m telling you that the margin of error cannot be accurately determined unless your sample is known to be representative of the population…

The generic way that you learn it in high school is based on ideal conditions… thats why national polls rarely accurately reflect real world turnout either in politics or marketing… and those actually try to use some representative sampling, at least far more than an online poll would…

If those MOE were correct, then for instance the end result of a vote would fall within the margin of error of the most recent poll more often than not.
This is famously not the case in national polls across the world…

Scientific polling is based on scientific sampling, but unless the first step of representative sampling is done then all the results are potentially biased and unrepeatable… pollsters typically ignore that… and here you are using a totally unscientific and relatively small poll to try to prove a national opinion

Go look at my last post. Every needed measurement is there. Except all the Ben haters who “obviously” skew any survey that deals with him.

Dude. You are wrong. I know you can’t admit it but I know you know. That’s why you keep digging deeper. There’s two types of humans. The ones who can admit they are wrong. And slashsteel. Don’t be slashsteel.