Now that every Ivy League football team has played a single league game, we at the Yale Undergraduate Sports Analytics Group (YUSAG) present our updated predictions for the league standings. For more on our methodology please see our earlier article in which we introduced the models, BY THE NUMBERS: Previewing the Ivy League football season.
To note a few salient results, the updated predictions show the particular importance of each single game of Ivy League play. For example, notice that both ranking systems still rate Dartmouth as a better team than Penn, even after Penn beat Dartmouth on Saturday. However, Penn has a higher chance of winning the league in both models. This is a direct result of Penn having one more win than Dartmouth in league play and further shows that each win is one step closer to the title, while each loss makes it a much harder road to the top of the league.
It is also noteworthy that our preliminary ELO ratings model does not take into account out of league games. Therefore, it is likely that Yale is rated higher in our model than it perhaps should be, given the two lopsided defeats to non-Ivy opponents. Meanwhile, these same defeats have caused Yale’s Sagarin rating to plummet, especially because that model is highly score-based. Nevertheless, since both models give more points for wins over highly ranked opponents, Yale has a chance to greatly increase its rating this weekend against Dartmouth.