NCAAursinus   <- Talks   <-  Sean Forman   <-  You Are Here

   / /

Modelling Individual Games

• Instead of 50-50, I need to find the probability of SJU beating Liberty.

• Given $T_1$ and $T_2$'s ratings and a location, what is the probability of $T_1$ defeating $T_2$, $f(T_1, T_2, loc) =$?

• What to do next?

• ?

• ?

• ?

• ?

• ?

• Gather data!

• In 2004, 4,400+ NCAA Division games between 326 teams.

• We know the outcome, the location and the difference in their Sagarin ratings.

/