Thursday, February 07, 2008
Is the NHL More Popular than the NBA?
by Jes
Over at The FanHouse, we were sent a link to something called "The Harris Poll", which asks a bunch of (American) people what their favourite sport is, and then ranks them, plays around with the data, and shows us just how popular the boring NFL brand of football really is.
What made our eyes pop out was the fact that 5% of respondents claimed hockey as their favourite sport, while only 3% claimed professional basketball, aka The NBA, as their favourite sport.
'Sup, bitches?
The NHL is more popular than the NBA?? WTF??
Of course, this news is far too good to be true, as any set of TV ratings will tell you.
As with any poll, this instantly screams "SMALL SAMPLE SIZE!". 2032 adults is hardly going to capture the true tastes of the general American public.
Given how regional NHL interest is, the figures could wildly fluctuate if you hit a good patch (Minnesota) or bad patch (West Virginia) when polling people.
It's always been a pet peeve of mine that the media, and the pollsters themselves, seem to forget a simple little concept called SAMPLING ERROR.
In statistics, sampling error is the error caused by observing a sample instead of the whole population.
An estimate of a quantity of interest, such as an average or percentage, will generally be subject to sample-to-sample variation. These variations in the possible sample values of a statistic can theoretically be expressed as sampling errors, although in practice the exact sampling error is typically unknown. Sampling error also refers more broadly to this phenomenon of random sampling variation.
For example, say a recent poll asked 2,000 Canadians which flavour of ice cream they like the most. The first poll had a +/- margin of error of 3%, with 38% choosing Vanilla and 23% choosing Chocolate.
Now, the second poll shows 35% choosing Vanilla, and 26% choosing chocolate.
If the media were covering the poll, the headline would be like "Support for Chocolate growing!" or "Vanilla Losing Popularity".
Nevermind the fact that the growth/drop was 3% for each, which was the exactly sampling error, the media seems to think that any such movement is indicative of the mindset of the population, rather than the expected variation from one poll to the next.
Just remember, kiddies, that any movement that is within, or very close to, a margin of error on any poll does not necessarily indicate a change in public opinion, as a whole.
Of course, the Harris Poll doesn't even bother publishing a NUMBER for sampling error, but at least they acknowledge there is a such thing and give their reasons for not putting a hard number out there.
Still, I did enjoy the poll, and would love to see a similar poll done with a far large sample size. Perhaps ESPN, with all of their resources, could conduct a poll through their website.
Anyway, back to the original topic at hand ... Big Mac took the reigns of this story for us at FanHouse, and made a good point about how the NHL really is gaining ground to the NBA.
The numbers for the NHL are pretty much flat throughout the survey. It's the NBA's numbers that have really dwindles, as the American audience seems to be turned off by the new generation of American 'ballers.
Over at The FanHouse, we were sent a link to something called "The Harris Poll", which asks a bunch of (American) people what their favourite sport is, and then ranks them, plays around with the data, and shows us just how popular the boring NFL brand of football really is.
What made our eyes pop out was the fact that 5% of respondents claimed hockey as their favourite sport, while only 3% claimed professional basketball, aka The NBA, as their favourite sport.
The NHL is more popular than the NBA?? WTF??
Of course, this news is far too good to be true, as any set of TV ratings will tell you.
As with any poll, this instantly screams "SMALL SAMPLE SIZE!". 2032 adults is hardly going to capture the true tastes of the general American public.
Given how regional NHL interest is, the figures could wildly fluctuate if you hit a good patch (Minnesota) or bad patch (West Virginia) when polling people.
It's always been a pet peeve of mine that the media, and the pollsters themselves, seem to forget a simple little concept called SAMPLING ERROR.
In statistics, sampling error is the error caused by observing a sample instead of the whole population.
An estimate of a quantity of interest, such as an average or percentage, will generally be subject to sample-to-sample variation. These variations in the possible sample values of a statistic can theoretically be expressed as sampling errors, although in practice the exact sampling error is typically unknown. Sampling error also refers more broadly to this phenomenon of random sampling variation.
For example, say a recent poll asked 2,000 Canadians which flavour of ice cream they like the most. The first poll had a +/- margin of error of 3%, with 38% choosing Vanilla and 23% choosing Chocolate.
Now, the second poll shows 35% choosing Vanilla, and 26% choosing chocolate.
If the media were covering the poll, the headline would be like "Support for Chocolate growing!" or "Vanilla Losing Popularity".
Nevermind the fact that the growth/drop was 3% for each, which was the exactly sampling error, the media seems to think that any such movement is indicative of the mindset of the population, rather than the expected variation from one poll to the next.
Just remember, kiddies, that any movement that is within, or very close to, a margin of error on any poll does not necessarily indicate a change in public opinion, as a whole.
Of course, the Harris Poll doesn't even bother publishing a NUMBER for sampling error, but at least they acknowledge there is a such thing and give their reasons for not putting a hard number out there.
Still, I did enjoy the poll, and would love to see a similar poll done with a far large sample size. Perhaps ESPN, with all of their resources, could conduct a poll through their website.
Anyway, back to the original topic at hand ... Big Mac took the reigns of this story for us at FanHouse, and made a good point about how the NHL really is gaining ground to the NBA.
This survey does dovetail well with some other data that we've seen about the relative popularity of the two sports. Back in November, Sports Business Daily published a study about the most popular sports sites on the Web, and that survey pegged the total traffic of NBA.com and NHL.com as pretty much even. Later, I called the league office in New York, and they told me that had the numbers from Canada been included, that NHL.com would have been far more popular.
Of course, once you add in those numbers from China, the NHL suffers a bit. The global village and all that.
If I was working for the NHL, I might take some time to send these numbers to a couple of sports pages around the country to make a simple case: You might be ignoring the NHL, but there's a market of people out there that you aren't necessarily serving. You'd think there might be some potential in that.
The numbers for the NHL are pretty much flat throughout the survey. It's the NBA's numbers that have really dwindles, as the American audience seems to be turned off by the new generation of American 'ballers.
Comments:
<< Home
Economists have a good approach to polls: they're meaningless because they cost the respondent nothing.
Eg., a pollster calls me at dinner. Asks me whether I prefer hockey or basketball. I suspect it's a polling firm hired by the Committee to Build a Free Arena for the Edmonton Oilers. I say, "basketball."
But when it comes to spending my money or time I'll watch hockey over basketball 100 times out of 100. That's an economic choice and it's far more indicative than answering a poll question.
Check out Slate.com for its running feature on betting for the presidential candidates. Same concept.
Eg., a pollster calls me at dinner. Asks me whether I prefer hockey or basketball. I suspect it's a polling firm hired by the Committee to Build a Free Arena for the Edmonton Oilers. I say, "basketball."
But when it comes to spending my money or time I'll watch hockey over basketball 100 times out of 100. That's an economic choice and it's far more indicative than answering a poll question.
Check out Slate.com for its running feature on betting for the presidential candidates. Same concept.
# posted by 
: 1:15 pm
: 1:15 pm
Great point: Economics of polls. TV Ratings and $ speak volumes.
Another point brought up is how the NHL is unlikely to be the 2nd, 3rd, or 4th choice for many people while the NBA would.
Another point brought up is how the NHL is unlikely to be the 2nd, 3rd, or 4th choice for many people while the NBA would.
With 2,032 respondents, you would actually expect somewhere in the range of a 2% variation.
Not a big difference, but still meaningful when we're comparing 5% to 3%.
Economists have a good approach to polls: they're meaningless because they cost the respondent nothing.
Except that polls aren't meaningless. :)
Maybe in that example you mentioned above, where you are intentionally providing an incorrect response because you have the foreknowledge to manipulate the ultimate result you'd like to see. But beyond that situation being far from the norm, there are measures in place to prevent and account for those cases.
Not a big difference, but still meaningful when we're comparing 5% to 3%.
Economists have a good approach to polls: they're meaningless because they cost the respondent nothing.
Except that polls aren't meaningless. :)
Maybe in that example you mentioned above, where you are intentionally providing an incorrect response because you have the foreknowledge to manipulate the ultimate result you'd like to see. But beyond that situation being far from the norm, there are measures in place to prevent and account for those cases.
This is funny cause i work at a polling place and we have a similar question (which i dont feel there is so called economic polling factor because the survey is lengthy and random). What are your favorite sports. I think once someone said the nhl, and only a couple more times, 4? Someone said nba. Not exactly scientific, but i thought it represented public opinon. Weird/cool to hear otherwise with this article.
# posted by : 9:16 pm
: 9:16 pm
Hockey may have more die-hard fans, but the NBA in general has more casual fans...
Case for the former: look at how many hockey blogs are on the net vs. NBA blogs.
Case for the latter: outside of the two cities in the NBA finals, Atlanta delivers the highest ratings for the finals. And pro basketball in Atlanta has sucked for decades.
Case for the former: look at how many hockey blogs are on the net vs. NBA blogs.
Case for the latter: outside of the two cities in the NBA finals, Atlanta delivers the highest ratings for the finals. And pro basketball in Atlanta has sucked for decades.
# posted by : 6:33 am
: 6:33 am
Hockey may have more die-hard fans, but the NBA in general has more casual fans...
That's a very good point. Most people I know who like hockey love hockey. Most people I know who like basketball tend to follow baseball and football as well, often at the collegiate level.
It's anecdotal, but I would put some weight to it. The casual market for basketball is much larger; I imagine the size of a casual market is in some way related to how ingrained that sport is in a culture.
That's a very good point. Most people I know who like hockey love hockey. Most people I know who like basketball tend to follow baseball and football as well, often at the collegiate level.
It's anecdotal, but I would put some weight to it. The casual market for basketball is much larger; I imagine the size of a casual market is in some way related to how ingrained that sport is in a culture.
I hate the freaking NBA with a passion but for the NHL to compete with the NBA hockey players need to wear short sleeves and get a buttload of unreadable tattoos then fans with switch over.
Post a Comment
# posted by : 11:03 am
: 11:03 am << Home
