What are the advantages and disadvantages of using a Logarithmic model?
The advantage is that a logarithmic model predicts that sprint times won’t just drop to 0 seconds. There is a limit in other words to how fast a human can go. Which, by as far as we can tell is true. A disadvantage of the log model is that it is much more difficult to fit than a straight line. A way around this difficulty is to plot the data on a log-scale graph (log y = mx+b) and fit a line to the data. Then you just transform the linear equation (10^(log y) = 10^(mx+b)) to get your model in rectilinear form.
