A new and formal paper of mine describing how one can use machine learn methodologies to help determine which NBA teams will make the playoffs is now online:
Have a look!
It is another New York Knicks season where fans have to wait until next year to see if the Knicks will make the playoffs or not.
Yesterday, there was a lot buzz around the idea that Phil Jackson may want to keep Kurt Rambis on as head coach, and as usual, there were numerous people that were very vocal in their criticism.
However, in actuality, the Knicks were much closer to the playoffs than people realize. A previous post of mine described in detail using data science methodologies the criteria a team must meet to have a high probability of making the playoffs.
Using the decision tree generated in that post, I evaluated the Knicks playoffs chances this season based on possible playoff criteria scenarios, and found the following:
One sees that a big problem was the Knicks margin of victory, which was too negative. However, even in this case, there are possibilities that existed that would have allowed the Knicks to make the playoffs. For example, a slight increase in the Knicks’ opponent’s field goal attempts or a very slight decrease in the Knicks’ field goal attempts per game would have greatly impacted their playoff chances.
These metrics can easily be adjusted for the upcoming season which will likely require a more organized execution of the triangle offense and discipline on both ends of the floor. They really are almost there!
What goes into making a cosmological model? Here is a presentation (that was part of my Ph.D. dissertation) that I have reproduced and embedded here to describe what actually goes into the making of a cosmological model. After describing some general properties, I describe specifically a early-universe model that contains a viscous fluid and a magnetic field.
The background mathematics can be found in this old presentation of mine here: