A couple of questions that you may have

Some exercises use 2 MSS as the initial valuefor the congestion window, why is that?
It's ok. In the book and in the classroom an initial value of 1 MSS is used instead. Modern TCP implementations are commonly using 2 MSS instead. That is why you can see this value sometimes.


What happens after a loss event in TCP congestion control?
In the book and in the classroom we have mentioned the new threshold value is half of current congestion window[RFC2001]. However, once you factor in that flow-control does limit the sending capability of the sender, what we call the transmission window is the number of segments that can actually be transmitted and it is equal to min(receiverWindow, congestionWindow). Modern TCP implementations use half of the used transmission window as the new value for the threshold after a loss event[RFC2581]. That means that the new threshold could be lower than just half of the congestion window.

----
The only purpose of this information is for your own peace of mind, given that some of you have detected some contradiction between the book/lectures and some of the solved exercises out there.