The first three weeks the Ivy League men’s basketball season have brought a little bit of everything: suspense, surprise and a whole lot of fun. This weekend’s slate of eight games, however, should top anything we’ve seen thus far. Friday and Saturday will mark the first weekend that all of the Ancient Eight teams play back-to-back games, an idiosyncrasy seen in few Division I conferences outside of the Ivy League.
There’s no possible way that one could watch all eight games this weekend, but choosing the right game(s) to watch is harder than it seems. Using my NCAA hoops prediction model, I identified which games are most important in determining which teams make “Ivy Madness”, the conference’s postseason tournament.
For each game being played this weekend, I assumed first that the home team won the game and calculated the resulting playoff probabilities for each team. I then switched the hypothetical winner of the game to the visiting team, and examined the subsequent changes in each team’s playoff odds. Summing up the differences in each team’s playoff odd’s between the two scenarios allowed me to create a metric, called Playoff Swing Factor (PSF), to evaluate the importance of each game.
By this method, Brown’s Saturday game at Columbia is the most important in deciding the Ancient Eight playoff race. A Columbia win really boosts the Lions’ own playoff chances while all but ending those of Brown, while a Bear upset in Morningside Heights would put them squarely back in the race for the conference’s No. 4 seed. We must also consider the residual effects that this game has on the other six teams. For example, a Lions win Saturday night would be very damaging to Penn’s chances of catching Columbia for the fourth seed.
Note that when I claim Brown at Columbia is a must watch game this weekend, I only mean to suggest that it the most critical in determining playoff outcomes. The Princeton-at-Harvard game Saturday night is arguably the best game to watch, but the outcome of the game has little effect on the playoff race compared to games between middle of the conference teams.
Last week, we were 4–1 in our picks, missing the score differentials by a combined twelve points in five games. Here are our picks for this weekend’s epic conference slate:
— Luke Benz, Yale Undergraduate Sports Analytics Group President