Originally Posted by derek500:
“I don't disagree with any of that.
But, if a household turns to Sky Sports Mix to watch the Netball, Sky using their data will have a more accurate idea of the number of homes watching than BARB with their 2,000 or so homes with Sky.”
“I don't disagree with any of that.
But, if a household turns to Sky Sports Mix to watch the Netball, Sky using their data will have a more accurate idea of the number of homes watching than BARB with their 2,000 or so homes with Sky.”
So having said you don't disagree with it, you then spend the next paragraph disagreeing with it. While Sky will know how many of the connected boxes were tuned the the correct channel, that's all they know. The BARB figures will still be more accurate, even for niche sports on low-reach channels, because the sample method is better.
And it is the method that is key.
As any statistician will tell you, increasing the sample size beyond a certain point doesn't increase the accuracy of the data significantly, it just gives you more data.
BARB samples around 6,500 households. If it sampled 65,000 the data wouldn't be ten times more accurate, it would only be around three times more accurate.
If the data from the Sky boxes was sampled in as accurate a way as BARB data, then it would be fifty times more accurate (from a sample size more than 1,000 times greater).
But most of the data from Sky boxes is "bad data". It tells you nothing at all.
Quote:
“For a niche sport such as Netball it would be almost impossible to get one Netball fan on the panel, and BARB wouldn't actively try. I assume it gets a zero rating from BARB?
I expect it's like Ladies Golf, that I'm involved with, the audience is mainly female golfers, club members and their families. Something Sky can monitor, but BARB is far too generic.”
“For a niche sport such as Netball it would be almost impossible to get one Netball fan on the panel, and BARB wouldn't actively try. I assume it gets a zero rating from BARB?
I expect it's like Ladies Golf, that I'm involved with, the audience is mainly female golfers, club members and their families. Something Sky can monitor, but BARB is far too generic.”
Actively trying to get netball fans on the panel would also be wholly unscientific. But statistically the panel would already include 15-20 households where someone plays netball regularly, because the panel is representative of the greater population.
I haven't seen any netball ratings, but I'd guess that BARB probably reports figures of 15-40,000 for a live game.



