Posted by Alex Moscoso (@AlexPMoscoso) on November 18th, 2013
It’s been over a week since the season started and all 12 teams have at least three games under their belts. Michigan State is as good as advertised after beating #1 Kentucky (even if they followed it up with a lackluster performance against Columbia). But what can we conclude from the other teams’ performances, where most games have been lopsided victories against inferior opponents? This makes it difficult to gauge which team has under- or overperformed so early in the year, but we here at the RTC Big Ten microsite are always up for a challenge. Prior to the first tip, we recorded each team’s predicted outcome using KenPom.com. To measure how teams have performed thus far, we will now compare their season performances against their preseason expected outcomes.
The table below illustrates each team’s performance in games already played against what they were expected to do, helping us evaluate their consistency and long-term projections.
The table above displays each team’s performance for each game relative to their expected preseason expected outcome. For example, if a team was expected to win by 10 points, but ended up winning by only five points, then that team underperformed by five points (shown as -5 in the table). If that same team had won by 20 points, then that team would have overperformed by 10 points. Underperformances are marked in red and overtperformances are marked in green. The average and standard deviation of each teams’ differential performances are calculated to measure their overall consistency so far. Finally, the far-right column in the table shows the change in total wins for the season that KenPom is projecting. For example, if a team was initially expected to win 18 games, but is now expected to win 21 games, their record difference is shown as +3. This metric not only takes into account each individual team’s season performance thus far, but also the performance of all its opponents.
Here are our five takeaways from this analysis:
Read the rest of this entry »