FB: Liberty League

[Bombers798891]

Boy, I wish we got paid more than 58 cents per 1,000 page views with you two going at it.

I'll give you an address where you can send me my cut 

[dlippiel]

Honestly it's been a good discussion and dlip has enjoyed talking Union/Ithaca. It really is a great rivalry and one that hopefully continues. That was one of the reasons dlip was so disappointed with last's year Dutchmen performance out in Ithaca. For such a big game, one that many of us went into with some high hopes, dlip felt like he got punched in the face after. Not only from Ithaca's victory but from Union's lack of execution on defense. dlip thinks Audino learned a lot from that game regarding first game preparation, especially when playing against a quality opponenet like the Bombers. Even in cam this year, the bit dlip has seen, Audino seems like he is dotting his i's and crossing his t's a bit more. Plus dlip noticed the intensity and focus seemed to be really top notch. We shall see. Either way it has been enjoyable discussing this issue with you two.

[maxpower]

Don't work as a pollster, Bombers.

Yeah Bombers, if you worked with polls, you'd have to act with a modicum of professionalism and not be a homer and not pick fights on the message boards while trumpeting your position . Because pollsters can't do that, right?

[Mr. Ypsi]

I'll weigh in belatedly here.  Frank, I AM a statistician (and occasional pollster), and you are mis-using the 'margin of error' term.  That refers to sampling error (the possibility that due to luck of the draw in random sampling you have overrepresented a particular position).  Since you are presenting your own opinion, you are a sample of one.  There is no sampling error in your opinion; you either think team A will win, or lose, or are unsure ('a toss-up').  It is impossible to establish (statistically) a 'margin of error'.

(There, Pat, that oughta generate a few more cents in page views! )

[Pat Coleman]

Don't work as a pollster, Bombers.

Yeah Bombers, if you worked with polls, you'd have to act with a modicum of professionalism and not be a homer and not pick fights on the message boards while trumpeting your position . Because pollsters can't do that, right?

Maxpower raises an excellent point, Frank.

[Frank Rossi]

Don't work as a pollster, Bombers.

Yeah Bombers, if you worked with polls, you'd have to act with a modicum of professionalism and not be a homer and not pick fights on the message boards while trumpeting your position . Because pollsters can't do that, right?

Maxpower raises an excellent point, Frank.

Except, I:

1) stated that Ithaca is the 4-point favorite in my book;

2) set up criteria so that I didn't make misleading or irrational predictions in most or all of the EFr8 games;

3) still proved my point with that modicum of professionalism of NOT being a homer in the first place (since I actually believe the toss-up games may break worse than 50/50 for the EFr8 personally, but haven't stated that until now since you're going to actually question my professionalism when I took extra time putting that list together to treat it with the utmost professionalism); and

4) asked someone who kept beating a drum about my safe approach to at least quantify their own feelings to see that their prediction was for a one-touchdown game for a game 3 weeks away.

Yeah, where should I send my check for $5.80 for the 10,000 page views you won't be getting while I go to the Travers tomorrow instead of wasting my time actually putting together thought-provoking posts that actually challenge people.  And from now on, I'll just put out my statements with no factual or historical backup of my opinions or positions since apparently you'd prefer that.

[Frank Rossi]

I'll weigh in belatedly here.  Frank, I AM a statistician (and occasional pollster), and you are mis-using the 'margin of error' term.  That refers to sampling error (the possibility that due to luck of the draw in random sampling you have overrepresented a particular position).  Since you are presenting your own opinion, you are a sample of one.  There is no sampling error in your opinion; you either think team A will win, or lose, or are unsure ('a toss-up').  It is impossible to establish (statistically) a 'margin of error'.

(There, Pat, that oughta generate a few more cents in page views! )

And if I used historical results in those cases, there would be no margin of error and no sampling, Ypsi?

[Pat Coleman]

I don't know who you're addressing, Frank, with the last point, but it sounds like me. I think if you divide the number of page views (clearly posted on the East Region index board) by the six years this particular software has been running, you'd find you vastly overestimate the number of page views this is likely to get.

You've publicly put your voter status into this conversation. Maxpower was right to call you on it. I am pointing that out.

[Mr. Ypsi]

I'll weigh in belatedly here.  Frank, I AM a statistician (and occasional pollster), and you are mis-using the 'margin of error' term.  That refers to sampling error (the possibility that due to luck of the draw in random sampling you have overrepresented a particular position).  Since you are presenting your own opinion, you are a sample of one.  There is no sampling error in your opinion; you either think team A will win, or lose, or are unsure ('a toss-up').  It is impossible to establish (statistically) a 'margin of error'.

(There, Pat, that oughta generate a few more cents in page views! )

And if I used historical results in those cases, there would be no margin of error and no sampling, Ypsi?

There is ALWAYS the likelihood of error in predictions.  But YOU introduced the 'margin of error' term.  I merely observed that (since you referred to polling) you were mis-using the term.

You may be right; you may be wrong; but as a sample of one, the probability of error cannot be assessed.  Using historical results may be right; it may be wrong; but the probability ('margin') of error cannot be (statistically) assessed.

That is all I posted.  Having now earned Pat an extra 3-5 cents, I will withdraw from a fight that doesn't involve me!

[Frank Rossi]

Where did I do that exactly?  My comments were regarding polling methodologies in things like politics, NOT your poll, Pat.

[Pat Coleman]

+10 karma points to the first person who quotes the line I'm talking about.

[pg04]

Don't work as a pollster, Bombers.  Ever see how they define toss-ups with sampling data?

Do I win? LOL

P.S.  I think Ypsi and Max do make excellent points, whether or not anyone wants to believe it! 

[Frank Rossi]

If that's the comment, Pat, you're dead wrong.  I'm not a pollster.  I'm a voter.  I was referring to the use of "toss-up states" in election polling even when the result isn't tied.

[pg04]

Do you have any pull on the Conference rankings?  Maybe he meant your comment about the conference dropping to 6 or 7 during the season.  Regardless, I just want to see if he can actually dish out 10 karma points at once. 

[Pat Coleman]

I can't! Looks like I can give two, so I owe you eight.

Frank -- regardless, you do talk about your voterness far more than anyone else I can think of, maybe even Keith. Just tone it down.