All the times given by witnesses are estimates of the time, and they are estimates made based upon their recollection of events that, for the most part, were just mundane things (i.e. saw a woman and a man talking in the street). So witnesses have to reconstruct the times of things. Add that to the fact they are all referencing different clocks, which may not be synchronized, the times stated needed to be viewed as rough estimates only. I found that the only way to really put things together was to splice things more in terms of references to other events by durations (i.e. I was home for approximately 15 minutes when I heard men shouting and running along Fairclough type thing) ... and by fitting things together using those durations, guided in part by their estimation of the time, a more or less coherent story can be formed. That was how I put together the Stride simulation. There's room for improvement in that one, of course, as it's a very complicated data set given how many people are involved, but overall I think it hangs together. I'm sure there are some mistakes in it, and I'm sure some of the assumptions I had to make are wrong, but the gist seems ok in my opinion.
- Jeff
- Jeff
Comment