> Amitava mentioned in the lecture that insertion sort is faster than merge sort up until 6000 elements. However, the graph of both algorithms show that the intercept is much smaller, and insertion sort is only faster for less than 100 elements approximately. As such, what is the correct approximate input size for which insertion sort is faster than merge sort so that I may decide which to implement. > > Thanks.

> I was thinking about this myself, but computers are so fast that for sorting sub 100 elements it makes hardly any practical difference, so I went with merge sort only. Also, different sources on the internet say different numbers of `n` items are the equilibrium; I have seen n stated as 55, 38, 100 etc. > > I came across this link in my research which was quite informative: https://w3.cs.jmu.edu/lam2mo/cs240_2014_08/pa04-sorting.html#:~:text=Switching%20Strategies&text=As%20you%20can%20see%2C%20binary,faster%20for%20all%20larger%20inputs. > > Although I could be entirely wrong so if someone more knowledgable replies feel free to ignore this :)