Are You Interested in Pushing the Model Further?

The data set that was available spanned 20 14 to 20 18, when there was a major change point event in the pandemic. We will be testing against newer networks, seeing forms and finding ways to improve it. I don't want to shy away from seeing whether this is a poor model when we actually find the right model to fit. It won't be on track with the best line estimate, of course, because there was a 30 % drop in crash rates. However, i would be interested to see where there is a fell betwien the confidence interval bounds that were included in the model.