http://farm3.static.flickr.com/2652/3808547741_92e41315f1.jpg

Birmingham City Council’s leisure centre website isn’t an easy place to look for information, harder still to try to take data, but that’s just what I’ve spent some considerable time trying to do.

As part of my investigation, I decided to explore how easy it would be to ‘scrape’ the website to see if I could start to compare opening times from pool to pool in the city. My interest was piqued by reading a report prepared for the Amateur Swimming Association, which has been published on the web.

The report, which seeks to assess provision of swimming pools in the UK, is a long, rather difficult read, but brilliantly researched and full of fascinating detail. I’ve prepared a post about it here, which demonstrates that Birmingham lags behind other places in the UK for swimming pool provision.

It does, however, also suggests at one point that there is the possibility that the data it’s based on might not always be completely accurate. Why? Well, it all comes from a database called Active Places. Active Places contains information for every leisure centre in the country – and the opening times, size of facilities, etc. However, as the report notes, it relies on councils to update the information in a timely and accurate fashion.

With that in mind I thought I’d try to see how easy it is for me, on my own, to get hold of the information from the website and then make some similar calculations.

What I did
As I’ve already established, it’s possible to scrape data from the Birmigham City Council’s website, because – weirdly – it’s all held in tables. By using the ‘importhtml’ formula in Google Spreadsheets I lifted opening times onto a spreadsheet for every swimming pool that Birmingham City Council owns and runs. Have a look at the spreadsheet here.

Or look at this cut-down version with the headline figures. NB: Hours are decimalised – so 2.5 would be two and a half hours.

Counting

I’ve been reading a very interesting book recently, called The Tiger That Isn’t. It’s about statistics – and points out many of the very common errors journalists tend to make when they’re using numbers in stories. One of those errors, as it happens, is not realising that counting is more complicated than you might immediately imagine. It says that often you’re not counting nice, simple units, but mushy peas. It turns out that’s just the case with swimming pools. Firstly, pools are different lengths and different widths. Some pools are open some of the time and some are open a lot of the time, while others are only partially open.

For the purposes, therefore, of simplicity and to avoid what might have been a great many sore heads and long nights, I chose to keep my comparison fairly simple. I decided to count the hours for all main pools (since all pools have at least one pool) and ignore size. This means I’m not yet able to say with confidence what the swimming pool’s capacity for swims is, but I can at least give an impression of how much of a week a pool is being used for.

I put the details on a map, so you can get an idea of how Birmingham’s distribution of pools looks. Remember: there are two pools missing from the visualisation, with Harborne being redeveloped and with no figures for Small Heath:

View Larger Map

And here’s Many Eyes visualisation of my efforts:

Is it useful?
Comparing different facilities, as The Tiger That Isn’t points out, is fraught with danger. But understanding how many hours a swimming pool is open to the public for is very helpful. If one facility is open for a particularly short period of time it’s worth asking why. It might be that I’ve made a miscalculation, or the information I’m using isn’t accurate (more about that in a minute), or it might be because there’s something wrong with the way the facility is being used. Since I’m not yet in the position to comment on that I won’t, but I hope this analysis might serve as the introduction to a more informed analysis. In the case of Shenley Court, for example, the pool is open for only a few hours because it’s only 18 metres long and is used most of the time for swimming lessons.

Fluctuations
What is clear, however, is that there are very significant variations in opening times for different pools. That’s something that I’m going to try to find out more about, because one has to ask why. I’d also like to compare it to other data to see if it has any impact on swimming and, indeed, on health.

Is it fair?
Earlier I said that doubt has been cast on the accuracy of Active Places, because it relies on local authorities to update it. I don’t know whether there’s much cause for concern, but I wonder about the value of the opening hours as they exist on the Birmingham City Council’s website. Some are accompanied by a warning: ‘Please check with the pool for more information on opening hours,’ for example. That means I can’t say that what I’ve collected is er, water tight, but hopefully it can still be the start of a more in-depth examination.