Do I really need to do this? This is like when Lefty34 demanded that Warner2Bruce provide empirical proof that the phrase "you win some, you lose some" is applicable to baseball.
But okay... Here goes...
From May 2009:
From October 2010:
From November 2010:
Meanwhile...
Don't worry though... NASCAR ratings have been doing better this year:
They are now at a three year high, and still only averaging half as many viewers as the NBA Finals, and almost as much as the 9 million that watched the NBA All Star game.
And here's the head to head numbers from the week of the conference finals/Coca Cola 600:
Yes it is. NASCAR fans are more likely to select that sport as their favorite sport than NBA fans are to select that as their favorite sport.
Let's say this poll interviewed 1000 people. Let's say 200 are NASCAR fans. They watch the races every weekend, know the drivers, etc.
Now let's say the other 800 are your average sports fans who don't give a shit about NASCAR and only know the big drivers.
All 200 of those NASCAR fans are going to choose that as their favorite sport. But only a segment of the other 800 is going to choose the NBA.
So you have 200 people who love NASCAR, so that's 20%. Then you have the other 80%, half of which choose football but really also like basketball and don't give a shit about NASCAR. Ask the same respondents to order them 1 through 5 and NASCAR gets 200 first place votes and 800 last place votes. Meanwhile the NBA gets a few first place votes and an even distribution of 2nd, 3rd and 4th place votes.
True. I was possibly thinking of response bias but while not paying attention in statistics classes I gleaned that most surveys are screwy in some way. This one provides an incomplete answer to the question of which sport is more popular between the NBA and NASCAR.
I venture to guess that if you asked 1000 pollsters to choose between the two, the NBA would have the majority. But if you ask the 1000 pollsters to choose one out of all of the pro sports, you're bound to have the hardcore NASCAR fans stand out as outliers. They are more of a niche community and likely to be vocal about their patronage of the sport in a one-choice survey.
But okay... Here goes...
From May 2009:
A double-digit decline in ratings for NASCAR on FOX has officials researching why the audience is shrinking and what, if anything, can be done to reverse the trend.
Ratings for the first four Chase races in 2010 were down 27 percent compared with last year. The decline is enough of a concern that top ESPN and NASCAR executives met in Charlotte last weekend to come up with ways to reverse the trend.
Four years after signing a record $4.48 billion media deal with Fox, ESPN and Turner, NASCAR has lost nearly a quarter of its TV viewership base, a four-year trend of massive viewer defections that has been punctuated by the erosion of the young male demographic.
the 2011 NBA Finals on ABC – with the Miami Heat leading the Dallas Mavericks 2-1 in the best-of-seven series – are averaging 15,311,000 viewers, 10,543,000 household impressions and a 9.1 rating, according to Nielsen.
The network’s presentation of the first 13 races of the NASCAR Sprint Cup series hit a three-year high averaging 8.6 million viewers per race, up +9% over last year (7.9 million viewers). That is the largest one-year audience increase in the 11-year history of NASCAR on FOX.
And here's the head to head numbers from the week of the conference finals/Coca Cola 600:
As far as the poll goes, asking a sports fan to give their favorite sport is not selection bias.
Let's say this poll interviewed 1000 people. Let's say 200 are NASCAR fans. They watch the races every weekend, know the drivers, etc.
Now let's say the other 800 are your average sports fans who don't give a shit about NASCAR and only know the big drivers.
All 200 of those NASCAR fans are going to choose that as their favorite sport. But only a segment of the other 800 is going to choose the NBA.
So you have 200 people who love NASCAR, so that's 20%. Then you have the other 80%, half of which choose football but really also like basketball and don't give a shit about NASCAR. Ask the same respondents to order them 1 through 5 and NASCAR gets 200 first place votes and 800 last place votes. Meanwhile the NBA gets a few first place votes and an even distribution of 2nd, 3rd and 4th place votes.
And what you described isn't even the correct use of the term selection bias because selection bias has to do with sample selection that isn't random. It has nothing to do with how someone from that sample responds to the question.
I venture to guess that if you asked 1000 pollsters to choose between the two, the NBA would have the majority. But if you ask the 1000 pollsters to choose one out of all of the pro sports, you're bound to have the hardcore NASCAR fans stand out as outliers. They are more of a niche community and likely to be vocal about their patronage of the sport in a one-choice survey.
Comment