## Probablity

For the discussion of math. Duh.

Moderators: gmalivuk, Moderators General, Prelates

kozzoz
Posts: 8
Joined: Fri Jan 29, 2010 4:10 am UTC

### Probablity

I got seconded to London last week for a project and am amazed at the rent-a-bike system in place. For those of you who don't know about it, peppered throughout the city are Barclays-branded rent-a-bikes where you swipe a card, grab a bike from a bike rack, and toodle off to the bike station nearest your destination. The first 30 minutes of the bike hire are free, although registration and annual fee is £45. Great system as you'd easily rack up £45 using public transport.

From a marketing perspective it's great for Barclays. I don't know what they paid for the rights, but it's a coup. The term 'barclays bike' is a proper noun now, and indelibly associated with this initiative. Plus from a corporate social responsibility angle, they can pull the usage data, make a few assumptions and with the help of their friendly neighbourhood actuary work out how much carbon they're offsetting. If they really get creative, they should be able to demonstrate the benefit to people's healths as measured in life years. Clever.

Anyway, onto the problem. I was trying to work out how many bikes London would need to service such a huge population, and noticed that each bike has a registration number on it (17340 on the above pic). On the walk to the office, I looked for the highest registration number on a bike, and (assuming that each bike has a unique, sequential registration number) tried to work out how many Barclays Bikes there are in London. The highest number I saw was 19566.

What are the steps that I need to take to figure out how many bikes there are likely to be? I recall from long-dead stats classes that distributions can be used to give a probability range, but have no clue as to how / why to choose a particular distribution based on the data that I have.

ps: Google tells me there are 6,000 in total so the registration number has *nothing* to do with the actual number of bikes, but I'd still like to know how to solve.

Torn Apart By Dingos
Posts: 817
Joined: Thu Aug 03, 2006 2:27 am UTC

### Re: Probablity

kozzoz
Posts: 8
Joined: Fri Jan 29, 2010 4:10 am UTC

### Re: Probablity

Brilliant! Thank you.

WarDaft
Posts: 1583
Joined: Thu Jul 30, 2009 3:16 pm UTC

### Re: Probablity

Given a sufficient sample size, we could also make an estimation of how many of the registration numbers are in use, to see if we can back up Google's claim.
All Shadow priest spells that deal Fire damage now appear green.
Big freaky cereal boxes of death.