Saturday, June 24, 2006

Bookies Offer Longer Odds On Pete Post News Of New Housemates;

Bookies responded to news of 5 new housemates entering a section of the BB house late last night by lengthening Pete's win odds - offering up to 1.71 on the nation's favourite tourette's sufferer. Below is the best price available chart over the last 3 days - click through if you'd like to see who's offering the highest odds (and how much they'll let you have on)

Meanwhile, Pete is still riding high in the Special Bets poll.. 44% of the group have Pete as their first choice favourite;

and he's also a very popular 2nd choice favourite - which bodes well for him as when other contestants leave, he'll pick up further votes;

Here's the current state of affairs in the win market on our odds comparison site (click through for live prices);

Have a free £25 bet on Big Brother (or the World Cup!) with our partners Blue Square

Find the best prices on Big Brother with Special Bets' Odds Comparison Site
special bets specialbets big brother reality tv betting odds eviction bet pete davina aisleyne vote big brother polls who goes? eviction % prediction

Friday, June 23, 2006

5 New Housemates Are Coming In - See Their First Polls HERE!

Five new housemates are "incoming" on Friday say channel 4 . Find out what the survey group think of the newbies here first! PS - The Special Bets survey group were only 0.1% wrong on Lisa's eviction % - 60.4% vs 60.3% - impressive eh?

Special Bets 5th Eviction Prediction!

We've read through the spoiler this evening (posted below) and there's nothing much in there to cause a big swing away from Lisa in the polling in our view, so what's our call? This is Maths Geek (or Statman's) big moment so we'll leave the analysis to him - over to you MG - PS did you really send stalker -style letters to Carol Vorderman?

Hi, thanks Special Bets, the probability of you accusing me of stalking Carol Vorderman on here is so high it crashed my Apple Mac.. anyway.. I've taken a long hard look at the data you've presented me with and the numbers are telling me that your responders this week have placed the eviction % almost "bang on" 60%, with a 4% confidence interval either side. The key factor right now is how the swing voters will act, and I suspect that over 50% of the voters do not act until eviction night.

Looking at the data you've presented me with and at the most recent trend it would appear that initial enthusiasm for Lisa is giving way to a late run for the other housemates.

On that basis I would expect that the actual result (currently 60.4%) will fall within the -4% rather than the +4% confidence range. Of course, whilst the numbers are our friend we can only work with what we have and as Vanessa Williams famously once said "sometimes the snow comes down in June" Thanks! Statman

Spoiler For Tonight's Show

moved to here;

Thursday, June 22, 2006

Special Bets 5th Eviction Poll - Results So Far!

We've finally got our Oxford University Maths Grad Geek to have a look at our poll for us this week to do some statistical analysis- he's put together some well, interesting stuff but maybe completely useless - more about him later.

We're having a record response to this week's poll from our ever-increasing survey group and we should be well over 1000 respondents by tomorrow - the biggest non-anonymous Big Brother poll in the world? Let us know if there's a bigger one!

OK, so on with the data...

Asking all 844 responders (Weds - Thurs)

A whopping 89% of you think Lisa is a gonner...

And you would seem to be right;

What about Lisa's eviction %?

Well those of you who told us you were actually going to vote were more in favour of evicting Lisa..

So, 61.6% seems to be a better figure.. so over to the Maths Geek... what have you to add?

Thanks! I prefer to be known as "Statman" but whatever... We can deduce how large the margin of error between your survey and the the actual population vote is likely to be using a measure called standard error.

Put simply, the standard error is a measurement of the standard deviation of the sampling distribution of the sample statistic (which in your case is eviction percentage).

Using this, and the size of your sample, I calculate that you can be 95% confident that the actual eviction percentage will fall within +/- 5.44% of your survey result. It's worth pointing out that as your sample size increases, your margin of error shrinks, albeit at a marginally decreasing rate. For example, if you doubled the amount of responses and yielded the same result, the margin would fall to +/- 3.01%

Isn't maths fun.

We hope the Maths Geek has something erm... more usable for tomorrow so we can get our bets on the right evictee %

We will update you on the final results tomorrow and will unveil out first ever spoiler survey to try to work out what effect the highlight show will have on the final evictee %. The Geek is supposed to be making sense of it all for us - so we'll see. Thanks for all the feedback poll fans!

Special Bets

**Spoiler** for tonight's show;

moved to here;

Wednesday, June 21, 2006

**SPOILER*** For Tonight's Show;