In order to have a more accurate measure of just how many trains are cancelled and how many are delayed, I've started to routinely gather the information and I'm putting the raw data on the web site.

Once I've got a good chunk of data available, I'll run some further analysis.

All trains have run so far today, so I've not been able to test the new code as well as I would like - please let me know if you spot any errors or strange things on my page when we've had something to wrong with the service.

As the data gathers, we'll be able to quantify what's happening.  It could be simply that the 17:46 is more noticed because it's used by more people - because of our customer base, I would lalk about the 17:02 from Melksham as the one I notice ... these days, I always check if it's running before dropping anyone off, and make other arrangements for them if I have to.

What the statistics are never going to be able to tell me is about how many people are on trains connection on to the line that are delayed, so that the connections miss.