i would like a data feed of all the BBC Have Your Say responses which have a) been most recommended and b) feature a large amount of capital letters (eg BOYCOTT CHINA). That way, you could launch a very low cost newspaper to rival the Daily Express/Mail without any real journalism. And thank-you newsbiscuit.com for "Pisswizard". Great name for a cat or a dog.
On 28/09/2007, Brian Butterworth <[EMAIL PROTECTED]> wrote: > On 28/09/2007, Richard Lockwood <[EMAIL PROTECTED]> wrote: > > That's possibly a bit of a sledgehammer / nut scenario for a poll like > > "What should we call the Tikkabilla ("Play School" for you other > > oldies...) hamster?". > > > > :-) > > > > Cheers, > > > > Rich. > > > You tell that to Socks and Cookie! > > I can't see how you can look for trends in the data and come up with a > confidence level for a vote without all the data. > > > > > > On 9/28/07, Brian Butterworth <[EMAIL PROTECTED]> wrote: > > > Martin, > > > > > > On 26/09/2007, Martin Belam <[EMAIL PROTECTED]> wrote: > > > > > > > > > > > > No I hadn't, thanks for pointing it out. I used to be Senior Producer > on > > > Online Voting at the BBC for a couple of years, and so I have some quite > > > strong opinions about when it is right to run an online vote and when > the > > > correct reaction is "You did *what*?" - most of those views are probably > > > more suited to the pub than this mailing list ;-) > > > > > > I guess one reason you might actually want to release a version of the > > > original data is that it could be put out to external scrutiny, that is > to > > > say *our* scrutiny. Perhaps it might be possible to detect the problems > > > with each poll and potentially correct for them. For example: > > > > > > - if the referring page is listed, any links from other sites to the > vote > > > could be detected; > > > - some form of IP-address-to-region (with a confidence-level perhaps) to > > > detect location problems; > > > - matching against botnet lists; > > > - hashed reference to the bbc.co.uk login name; > > > - hashed version of the IP address; > > > - date and time of the vote; > > > - hashed version of the bbc.co.uk tracking cookie; and > > > - the vote data > > > > > > What you could possibly get back is a confidence level (as a percentage) > as > > > to the reliablity of the result... > > - > > Sent via the backstage.bbc.co.uk discussion group. To unsubscribe, please > visit > http://backstage.bbc.co.uk/archives/2005/01/mailing_list.html. Unofficial > list archive: > http://www.mail-archive.com/backstage@lists.bbc.co.uk/ > > > > > > -- > Please email me back if you need any more help. > > Brian Butterworth > www.ukfree.tv - Sent via the backstage.bbc.co.uk discussion group. To unsubscribe, please visit http://backstage.bbc.co.uk/archives/2005/01/mailing_list.html. Unofficial list archive: http://www.mail-archive.com/backstage@lists.bbc.co.uk/